Starting with Oracle Database 18c, Oracle have made a few changes. The biggest one perhaps is the intention to make quarterly releases each year from now on. Hence, Oracle Database 18c Release 1 in Q1 of 2018 (18.1), Release 2 in Q2 (18.2) and so on. Not all of these releases will be or are available for download for on premise installation. At the time of writing, 18.3 is only available for download for 4 different platforms and 18.4 is only available for a further 3 platforms. We’ll be using 18.3 for Linux x86-64 for a series of Oracle Database 18c articles. Enjoy!
Also starting with Oracle Database 18c, Oracle have changed the way the database software is installed. Previously, you’d grab the zip file(s) from OTN or wherever, copy them to some staging area on your target server, unzip them then run the installer to actually install the code set into your OFA designated path (if you were doing things properly that is). That’s now changed.
Nowadays, you still grab the zip file(s) from Oracle, but now you have to copy them directly to your pre-created OFA designated path, unzip them and run the installer to install the code set in that directory. In other words, the installer will not create an OFA path for you like it has previously. I’m sure this will lead to many installations being in the wrong place, so let’s start off my running a deinstall so you can get it right the second time.
Given it’s 18.3, following OFA rules the path should be /u01/app/oracle/product/18.3.0/dbhome_1. So let’s uninstall it. These steps assume you haven’t created a database already. If you have, the deinstall process will detect it (and any associated Listener) and remove those too:
[oracle@orasvr01 dbhome_1]$ cd deinstall
[oracle@orasvr01 deinstall]$ pwd
/u01/app/oracle/product/18.0.0/dbhome_1/deinstall
[oracle@orasvr01 deinstall]$ ./deinstall
Checking for required files and bootstrapping …
Please wait …
Location of logs /u01/app/oraInventory/logs/
############ ORACLE DECONFIG TOOL START ############
################### DECONFIG CHECK OPERATION START ###################
## [START] Install check configuration ##
Checking for existence of the Oracle home location /u01/app/oracle/produc/18.0.0/dbhome_1
Oracle Home type selected for deinstall is: Oracle Single Instance Database
Oracle Base selected for deinstall is: /u01/app/oracle
Checking for existence of central inventory location /u01/app/oraInventory
## [END] Install check configuration
Network Configuration check config START
Network de-configuration trace file location: /u01/app/oraInventory/logs/netdc_check2019-07-12_11-27-40AM.log
Network Configuration check config END
Database Check Configuration START
Database de-configuration trace file location: /u01/app/oraInventory/logs/databasedc_check2019-07-12_11-27-40AM.log
Use comma as separator when specifying list of values as input
Specify the list of database names that are configured in this Oracle home []:
Database Check Configuration END
################### DECONFIG CHECK OPERATION END ###################
################# DECONFIG CHECK OPERATION SUMMARY #################
Oracle Home selected for deinstall is: /u01/app/oracle/product/18.0.0/dbhome_1
Inventory Location where the Oracle home registered is: /u01/app/oraInventory
Do you want to continue (y - yes, n - no)? [n]: y
A log of this session will be written to: '/u01/app/oraInventory/logs/deinstall_deconfig2019-07-12_11-27-39-AM.out'
Any error messages from this session will be written to: '/u01/app/oraInventory/logs/deinstall_deconfig2019-07-12_11-27-39-AM.err'
################## DECONFIG CLEAN OPERATION START ##################
Database de-configuration trace file location: /u01/app/oraInventory/logs/databasedc_clean2019-07-12_11-27-40AM.log
Network Configuration clean config START
Network de-configuration trace file location: /u01/app/oraInventory/logs/netdc_clean2019-07-12_11-27-40AM.log
De-configuring Naming Methods configuration file…
Naming Methods configuration file de-configured successfully.
De-configuring Local Net Service Names configuration file…
Local Net Service Names configuration file de-configured successfully.
De-configuring backup files…
Backup files de-configured successfully.
The network configuration has been cleaned up successfully.
Network Configuration clean config END
################### DECONFIG CLEAN OPERATION END ###################
################# DECONFIG CLEAN OPERATION SUMMARY #################
#############################################################
####### ORACLE DECONFIG TOOL END #######
Using properties file /tmp/deinstall2019-07-12_11-27-30AM/response/deinstall_2019-07-12_11-27-39-AM.rsp
Location of logs /u01/app/oraInventory/logs/
###### ORACLE DEINSTALL TOOL START ######
################# DEINSTALL CHECK OPERATION SUMMARY #################
A log of this session will be written to: '/u01/app/oraInventory/logs/deinstall_deconfig2019-07-12_11-27-39-AM.out'
Any error messages from this session will be written to: '/u01/app/oraInventory/logs/deinstall_deconfig2019-07-12_11-27-39-AM.err'
################## DEINSTALL CLEAN OPERATION START ##################
## [START] Preparing for Deinstall ##
Setting LOCAL_NODE to orasvr01
Setting CRS_HOME to false
Setting oracle.installer.invPtrLoc to /tmp/deinstall2019-07-12_11-27-30AM/oraInst.loc
Setting oracle.installer.local to false
## [END] Preparing for Deinstall ##
Setting the force flag to false
Setting the force flag to cleanup the Oracle Base
Oracle Universal Installer clean START
Detach Oracle home '/u01/app/oracle/product/18.0.0/dbhome_1' from the central inventory on the local node : Done
Delete directory '/u01/app/oracle/product/18.0.0/dbhome_1' on the local node : Done
The Oracle Base directory '/u01/app/oracle' will not be removed on local node. The directory is in use by Oracle Home '/u01/app/oracle/product/19.3.0/dbhome_1'.
Oracle Universal Installer cleanup was successful.
Oracle Universal Installer clean END
## [START] Oracle install clean ##
## [END] Oracle install clean ##
################### DEINSTALL CLEAN OPERATION END ###################
################# DEINSTALL CLEAN OPERATION SUMMARY #################
Successfully detached Oracle home '/u01/app/oracle/product/18.0.0/dbhome_1' from the central inventory on the local node.
Successfully deleted directory '/u01/app/oracle/product/18.0.0/dbhome_1' on the local node.
Oracle Universal Installer cleanup was successful.
Review the permissions and contents of '/u01/app/oracle' on nodes(s) 'orasvr01'.
If there are no Oracle home(s) associated with '/u01/app/oracle', manually delete '/u01/app/oracle' and its contents.
Oracle deinstall tool successfully cleaned up temporary directories.
#######################################################################
####### ORACLE DEINSTALL TOOL END #######
Task #2: Installing Oracle Database 18c (GUI).
There’s really not much to write home about with regards to installing 18c compared to previous versions. It has changed just a little so let’s quickly run through what the GUI installation looks like:
[oracle@orasvr01 dbhome_1]$ pwd
/u01/app/oracle/product/18.3.0/dbhome_1
[oracle@orasvr01 dbhome_1]$ ls -l
-rw-r--r-- 1 oracle oinstall 4564649047 Jul 12 11:40 LINUX.X64_180000_db_home.zip
[oracle@orasvr01 dbhome_1]$ unzip LINUX.X64_180000_db_home.zip
(output is boring so I didn't include it)
[oracle@orasvr01 dbhome_1]$ export ORACLE_BASE=/u01/app/oracle
[oracle@orasvr01 dbhome_1]$ export DISPLAY=<your_workstation_or_IP>:0.0
[oracle@orasvr01 dbhome_1]$ ./runInstaller
Select Set Up Software Only then click NextSelect Single instance database installation then click NextSelect Enterprise Edition then click NextThis screen tells you where the installer will install the software. Click Next.These groups are created beforehand here. Click Next.The installer runs the pre-req checks. It only takes a few seconds.Correct any pre-req check failures then click Next.
In my case, I haven’t configured enough swap space. Since this is a non-production system this doesn’t matter so I can safely check the Ignore All check box, click Next and move on. Incidentally, if you need to configure more swap via a swap file, check this out.
You see this only if you chose to ignore some pre-req check failures.
Review your settings then click Install.Tick tock tick tock…
If all goes well, the installer tells you to run the root.sh script.
[root@orasvr01 ~]# /u01/app/oracle/product/18.3.0/dbhome_1/root.sh
Performing root user operation.
The following environment variables are set as:
ORACLE_OWNER= oracle
ORACLE_HOME= /u01/app/oracle/product/18.3.0/dbhome_1
Enter the full pathname of the local bin directory: [/usr/local/bin]:
The contents of "dbhome" have not changed. No need to overwrite.
The contents of "oraenv" have not changed. No need to overwrite.
The contents of "coraenv" have not changed. No need to overwrite.
Entries will be added to the /etc/oratab file as needed by
Database Configuration Assistant when a database is created
Finished running generic part of root script.
Now product-specific root actions will be performed.
Do you want to setup Oracle Trace File Analyzer (TFA) now ? yes|[no] : no
Oracle Trace File Analyzer (TFA - Non Daemon Mode) is available at :
/u01/app/oracle/product/18.3.0/dbhome_1/suptools/tfa/release/tfa_home/bin/tfactl
Note :
1. tfactl will use TFA Daemon Mode if TFA already running in Daemon Mode and user has access to TFA
2. tfactl will configure TFA Non Daemon Mode only if user has no access to TFA Daemon mode or TFA Daemon mode is not installed
OR
Oracle Trace File Analyzer (TFA - Daemon Mode) can be installed by running this script :
/u01/app/oracle/product/18.3.0/dbhome_1/suptools/tfa/release/tfa_home/install/roottfa.sh
Click OK in the Execute Configuration Scripts window.
Although it doesn’t take that long to install the database software, sometimes you just haven’t got the time (or patience) to sit there and click-click-click. Silent (non-interactive) installations have been around for a while and Oracle Database 18c also supports that method. Rather useful if you’ve got a load of installations to do and need to (mostly) automate them.
Silent installations reply upon a response file which contains most or all of the values you would be prompted for if you were to run an interactive installation. Response files have a specific format which you must adhere to. A database software installation response file template can be found in $ORACLE_HOME/install/response/db_install.rsp. A more effective way to generate a populated response file is to run the GUI installation and save the response file towards the end of that process. That’s what I did to demonstrate the silent installation coming up shortly.
I always use a staging area to store the database media zip file. To speed up the copying, unzipping and installation of the software, I use this simple script. The significant line of code which runs the silent installation is this:
Here is the screen output from running the script as the oracle user:
[oracle@orasvr01 shell]$ ./silent_db_install.sh -help
Usage: silent_db_install.sh DB_Release Zip_File_Path Response_File_Path DELETE|NODELETE
where: DB_Release = Database_Version.Database_Release.0 (e.g. 18.3.0) Zip_File_Path = Full path to the database binaries zip file Response_File_Path = Full path to the silent install response file DELETE|NODELETE = Delete the zip file in the new ORACLE_HOME (DELETE) or not (NODELETE)
[oracle@orasvr01 shell]$ ./silent_db_install.sh 18.3.0 /u01/MEDIA/database/18c/LINUX.X64_180000_db_home.zip /home/oracle/scripts/shell/install_183.rsp DELETE *** Checking parameters *** Copying zip file: /u01/MEDIA/database/18c/LINUX.X64_180000_db_home.zip to the new ORACLE_HOME: /u01/app/oracle/product/18.3.0/dbhome_1 *** Unzipping LINUX.X64_180000_db_home.zip into /u01/app/oracle/product/18.3.0/dbhome_1 *** Deleting /u01/app/oracle/product/18.3.0/dbhome_1/LINUX.X64_180000_db_home.zip *** Passing execution to the Oracle Installer in silent mode
Launching Oracle Database Setup Wizard…
The response file for this session can be found at:
/u01/app/oracle/product/18.3.0/dbhome_1/install/response/db_2019-08-16_12-10-14PM.rsp
You can find the log of this install session at:
/tmp/InstallActions2019-08-16_12-10-14PM/installActions2019-08-16_12-10-14PM.log
As a root user, execute the following script(s):
1. /u01/app/oraInventory/orainstRoot.sh
2. /u01/app/oracle/product/18.3.0/dbhome_1/root.sh
Execute /u01/app/oraInventory/orainstRoot.sh on the following nodes:
[orasvr01]
Execute /u01/app/oracle/product/18.3.0/dbhome_1/root.sh on the following nodes:
[orasvr01]
Successfully Setup Software.
Moved the install session logs to:
/u01/app/oraInventory/logs/InstallActions2019-08-16_12-10-14PM
*** Please check the Oracle Installer output for additional root scripts to run.
Note, the first database software installation does take into account the absence of an Oracle inventory. Subsequent installations on the same server leverage a preexisting inventory. Also note, I had to fix the swap space issue otherwise the silent installation complained. Fixing a low swap space issue is documented here.
For completeness, here’s the output from the two root scripts referenced in the above output:
[root@orasvr01 ~]# /u01/app/oraInventory/orainstRoot.sh
Changing permissions of /u01/app/oraInventory.
Adding read,write permissions for group.
Removing read,write,execute permissions for world.
Changing groupname of /u01/app/oraInventory to oinstall.
The execution of the script is complete.
[root@orasvr01 app]# /u01/app/oracle/product/18.3.0/dbhome_1/root.sh
Check /u01/app/oracle/product/18.3.0/dbhome_1/install/root_orasvr01.mynet.com_2019-08-16_12-12-24-830643173.log for the output of root script
The script can be run multiple times on the same server. Each time it will create a new OFA compliant ORACLE_HOME, copy the zip file to it, unzip it, run a silent installation and optionally delete the zip file from the new ORACLE_HOME (but not from the staging area). Hopefully it will save you some time.
With the recent release of Oracle Enterprise Manager Cloud Control 13c Release 2 Plug-in Update 1, I felt it was time to finally say goodbye to 12c Release 5.
The Oracle documentation does a pretty decent job of explaining how to uninstall Oracle Enterprise Manager (OEM) Cloud Control (CC) 12c Release 5 which can be accessed here.
The only weirdness I encountered was while uninstalling the Oracle Management Agents. Everything worked successfully, but the final output of the AgentDeinstall.pl script was this:
Can't locate Carp.pm in @INC (@INC contains:
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/5.10.0/x86_64-linux-thread-multi
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/5.10.0
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/site_perl/5.10.0/x86_64-linux-thread-multi
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/site_perl/5.10.0
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/5.10.0/x86_64-linux-thread-multi
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/5.10.0/x86_64-linux-thread-multi
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/5.10.0
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/site_perl/5.10.0/x86_64-linux-thread-multi
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/site_perl/5.10.0
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/site_perl .) at
/u01/oracle/agent/core/12.1.0.5.0/perl/lib/5.10.0/File/Path.pm line 32.
The uninstall worked OK, so I’m guessing this message is just a ‘special feature’ which we can ignore. 😏
To install OEM CC 13c R2, we will need a server. I’m going to use the same server (oraemcc) to run the Oracle Management Service (OMS) and the Management Repository Database (MRDB). Since we’ll need a little more space than we did for the OEM CC 12c R5 installation, I increased the size of the /u01 and /u02 file systems on oraemcc accordingly:
File System
Old Size
New Size
Usage
/u01
30 GB
50 GB
Oracle Software Installations
/u02
30 GB
80 GB
Oracle Database Files
There are 4 main tasks we need to complete to get OEM CC 13c R2 up and running. Click the link you need:
We will be using a CDB with a PDB containing the Management Repository. You don’t have to set it up this way, but I’m going to use another PDB within the same CDB to host an RMAN Catalog. See Part 12 of my Build Your Own Oracle Infrastructure series for more details.
Task #2: Install the Oracle Database Software (12.1.0.2).
This particular task is covered here so I won’t dwell on it again. Once the database software is installed, copy the template zip file to $ORACLE_HOME/assistants/dbca/templates and unzip it:
Again, I have already covered this task here for 12c Release 5. It’s essentially the same process and is reasonably documented by Oracle here. However, a few things are worth noting.
Note #1: Pick the Correct Template.
The names of the templates are so long now that they extend beyond the width of Template field as shown in the screen shot. That’s fine except you can’t increase the width of the field to see which template is which. I could increase the size of the window, but not the width of the field. Genius!
Click to Open Full Size
The template selected in the screen shot is the one for a small repository. Clicking any of the radio buttons in the Select column and then clicking the Show Details button will show you more information about that template. How about putting “small”, “medium” or “large” at the beginning of the template name? Thanks Captain Obvious! 😉
Note #2: Do Not Configure Enterprise Manager.
The documentation does mention this, but it’s worth repeating because not following this instruction will lead to bad things happening. When you get to Step #5 in the dbca dialogue, make sure both Enterprise Manager options are UNCHECKED, as shown in the screen shot below.
Click to Open Full Size
Note #3: Run the shpool SQL Script.
The documentation still includes this warning:
“When you run Oracle Database Configuration Assistant to create the database, on the Database Content screen, by default the Run the following scripts option is selected to run the shpool script. The script sets the shared pool in the database. Make sure you RETAIN THE SELECTION. Do not deselect it.”
There is no Database Content screen as the left panel of the screen shots above clearly show. The script still isn’t called “shpool” either. It’s actually called:
Unlike with the 12c Release 5 procedure, there is no option to specify a custom script on the Step 9 screen. This is what the Step 9 screen looks like when using the 13c Release 2 database template:
Click to Open Full Size
As you can see, no option to specify the shpool SQL script. Instead, I let the dbca run to completion, then ran the shpool SQL script from a SQL*Plus session once the database had been created:
[oracle@oraemcc templates]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/assistants/dbca/templates
[oracle@oraemcc templates]$ . oraenv
ORACLE_SID = [oracle] ? PADMIN
The Oracle base remains unchanged with value /u01/app/oracle
[oracle@oraemcc templates]$ sqlplus / as sysdba
SQL*Plus: Release 12.1.0.2.0 Production on Mon Feb 26 13:59:34 2018
Copyright (c) 1982, 2014, Oracle. All rights reserved.
Connected to:
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Partitioning, OLAP, Advanced Analytics and Real Application Testing options
SQL> @shpool_12.1.0.2.0_Database_SQL_for_EM13_2_0_0_0.sql
System altered.
System altered.
System altered.
I did a little research on this and it would appear not having the option to specify a SQL script on on the Step 9 screen may have something to do with the template I used. YMMV.
Note #4: The PDB Containing the Management Repository.
Using this template creates a CDB (called PADMIN in my case) and a PDB called EMPDBREPOS. You can call the CDB what you like, but the name of the PDB is fixed. You can tell the PDB contains the Management Repository tablespaces by using these queries:
SQL> show con_name
CON_NAME
------------------------------
CDB$ROOT
SQL> select file_name from dba_data_files;
FILE_NAME
--------------------------------------------------------------------------------
/u02/oradata/PADMIN/datafile/o1_mf_users_f96sgg98_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_undotbs1_f96sghfp_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_system_f96sbzxo_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_sysaux_f96s8dcb_.dbf
SQL> alter session set container=EMPDBREPOS;
Session altered.
SQL> show con_name
CON_NAME
------------------------------
EMPDBREPOS
SQL> select file_name from dba_data_files;
FILE_NAME
--------------------------------------------------------------------------------
/u02/oradata/PADMIN/datafile/o1_mf_users_f96slktd_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_system_f96sfc35_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_sysaux_f96sllyg_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_mgmt_tab_f96smp4m_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_mgmt_ecm_f96sbwrj_.dbf
/u02/oradata/PADMIN/datafile/o1_mf_mgmt_ad4_f96so3by_.dbf
Task #4: Install OEM CC 13.2.
There are different hardware and software requirements for installing Release 13.2 depending upon the size and complexity of your infrastructure. We will be performing the simple/small installation. Here are the links to the relevant sections of Oracle’s documentation:
Despite performing a simple/small installation, it makes sense to go through the main items in the pre-installation/prerequisite check list. These are documented in full here, but here are a selection of just 9 to keep you amused:
Check #1: White Space.
Ensure there is no white space in the name of the directory from which you will run the installer:
[oracle@oraemcc ~]$ cd /u01/MEDIA/oem_13cr2p1
[oracle@oraemcc oem_13cr2p1]$ ls -l
-rwxr-xr-x 1 oracle oinstall 2123211088 Feb 24 18:45 em13200p1_linux64-2.zip
-rwxr-xr-x 1 oracle oinstall 741526563 Feb 24 18:45 em13200p1_linux64-3.zip
-rwxr-xr-x 1 oracle oinstall 2084231936 Feb 24 18:46 em13200p1_linux64-4.zip
-rwxr-xr-x 1 oracle oinstall 109191154 Feb 24 18:45 em13200p1_linux64-5.zip
-rwxr-xr-x 1 oracle oinstall 2146696423 Feb 24 18:47 em13200p1_linux64-6.zip
-rwxr-xr-x 1 oracle oinstall 771426157 Feb 24 18:47 em13200p1_linux64-7.zip
-rwxr-xr-x 1 oracle oinstall 554606940 Feb 24 18:47 em13200p1_linux64.bin
Check #2: DISPLAY Variable.
Ensure your DISPLAY environment variable is set accordingly (i.e. to the hostname or IP address of the server from where you’ll run the installation):
If you used the database template, then this should have been taken care of already, but let’s check:
SQL> show con_name
CON_NAME
------------------------------
CDB$ROOT
SQL> show parameter optimizer_adaptive_features
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
optimizer_adaptive_features boolean FALSE
SQL> alter session set container=empdbrepos;
Session altered.
SQL> show con_name
CON_NAME
------------------------------
EMPDBREPOS
SQL> show parameter optimizer_adaptive_features
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
optimizer_adaptive_features boolean FALSE
Check #4: Middleware Home Path.
Ensure the full path to the Middleware home directory does not exceed 70 characters:
In a production environment, I have come across situations where the required ports are being blocked by the corporate firewall. This provides an excellent opportunity to makes friends with the Network Administrator. If you’re really lucky and work in a bureaucratic nightmare, you’ll get to fill out lots and lots of forms to ask really nicely if your ports can be opened up, pretty please. 😀
The port requirements are fully documented here, but essentially boil down to these (YMMV):
Usage
Port#
Enterprise Manager Upload HTTP Port
4889
Enterprise Manager Upload HTTP SSL Port
4903
OHS HTTP Port
9788
OHS HTTP SSL Port
9851
Oracle BI Publisher HTTP Port
9701
Oracle BI Publisher HTTP SSL Port
9803
OEM Central Console HTTP Port
7788
OEM Central Console HTTP SSL Port
7802
Node Manager HTTP SSL Port
7403
Managed Server HTTP Port
7202
Managed Server HTTP SSL Port
7301
Management Agent Port
3872
Admin Server HTTP SSL Port
7102
Use the netstat command to figure out if a port is being used. If the command returns no output, then it’s free. Let’s use port #1521 as an example because we know the TNS Listener process is using that port:
Gone are the days when we’d invoke the runInstaller installer. Now it’s called a Wizard. Coloring books and crayons are optional. A couple of things to note here. First, you’ll be executing the .bin file and second, you don’t unzip any of the other files. The ‘wizard’ will do that for us since we can’t be trusted, apparently:
[oracle@oraemcc ~]$ cd /u01/MEDIA/oem_13cr2p1
[oracle@oraemcc oem_13cr2p1]$ ls -l
-rw-r--r-- 1 oracle oinstall 2123211088 Feb 24 18:45 em13200p1_linux64-2.zip
-rw-r--r-- 1 oracle oinstall 741526563 Feb 24 18:45 em13200p1_linux64-3.zip
-rw-r--r-- 1 oracle oinstall 2084231936 Feb 24 18:46 em13200p1_linux64-4.zip
-rw-r--r-- 1 oracle oinstall 109191154 Feb 24 18:45 em13200p1_linux64-5.zip
-rw-r--r-- 1 oracle oinstall 2146696423 Feb 24 18:47 em13200p1_linux64-6.zip
-rw-r--r-- 1 oracle oinstall 771426157 Feb 24 18:47 em13200p1_linux64-7.zip
-rw-r--r-- 1 oracle oinstall 554606940 Feb 24 18:47 em13200p1_linux64.bin
I had to add execute permissions before the .bin file would run:
0%...............................................................................100%
Launcher log file is /tmp/OraInstall2018-02-26_07-08-06PM/launcher2018-02-26_07-08-06PM.log.
Starting Oracle Universal Installer
Checking if CPU speed is above 300 MHz. Actual 2665.986 MHz Passed
Checking monitor: must be configured to display at least 256 colors. Actual 16777216 Passed
Checking swap space: must be greater than 512 MB. Actual 16383 MB Passed
Checking if this platform requires a 64-bit JVM. Actual 64 Passed (64-bit not required)
Preparing to launch the Oracle Universal Installer from /tmp/OraInstall2018-02-26_07-08-06PM
====Prereq Config Location main===
/tmp/OraInstall2018-02-26_07-08-06PM/stage/prereq
EMGCInstaller args -scratchPath
EMGCInstaller args /tmp/OraInstall2018-02-26_07-08-06PM
EMGCInstaller args -sourceType
EMGCInstaller args network
EMGCInstaller args -timestamp
EMGCInstaller args 2018-02-26_07-08-06PM
EMGCInstaller args -paramFile
EMGCInstaller args /tmp/sfx_ArNIXh/Disk1/install/linux64/oraparam.ini
EMGCInstaller args -nocleanUpOnExit
DiskLoc inside SourceLoc/u01/MEDIA/oem_13cr2p1
EMFileLoc:/tmp/OraInstall2018-02-26_07-08-06PM/oui/em/
ScratchPathValue :/tmp/OraInstall2018-02-26_07-08-06PM
Then what looks suspiciously like an installer screen appears and we’re off to the races. Uncheck the option to receive security updates and click Next:
Click to Open Full Size
Ignore the patronizing warning message by clicking Yes:
Click to Open Full Size
Select the Skip radio button and click Next:
Click to Open Full Size
So here we have a couple of prereq check failures. I left these in to drawn your attention to a ‘gotcha’.
Click to Open Full Size
We’re using the same server to run both the OMS and the MRDB. Consequently, we needed to install the Database 12c Release 1 database and before that, we installed the Oracle Database 12c Release 1 pre-install package into the OS. That installation is documented here.
The softnofiles issue is slightly more devious. The soft and hard limits for the number of files a process can have open is (usually) set in the file, /etc/security/limits.conf:
[root@oraemcc ~]# vi /etc/security/limits.conf
...
oracle soft nofile 8192
oracle hard nofile 65536
...
However, the error is reporting the soft limit is currently set to 16384 which, according to the value in /etc/security/limits.conf, is NOT the current soft limit. Trust me, setting either the hard or soft limit in this file will have no effect on the error the OEM 13.2 installer is reporting. So what’s going on?
Installing the pre-install RDBMS package creates a configuration file which overrides the limits.conf file:
[root@oraemcc ~]# cd /etc/security/limits.d
[root@oraemcc limits.d]# ls -l
-rw-r--r--. 1 root root 191 Aug 18 2015 90-nproc.conf
-rw-r--r-- 1 root root 1093 Feb 26 21:52 oracle-rdbms-server-12cR1-preinstall.conf
[root@oraemcc limits.d]# view oracle-rdbms-server-12cR1-preinstall.conf
# oracle-rdbms-server-12cR1-preinstall setting for nofile soft limit is 1024
oracle soft nofile 1024
# oracle-rdbms-server-12cR1-preinstall setting for nofile hard limit is 65536
oracle hard nofile 16384
...
This is where the OEM 13.2 installer is getting its information from and is actually using the hard limit, not the soft limit. Editing this file will fix the problem, but you will need to reboot afterwards, re-start the MRDB and start the OEM 13.2 installation again.
[root@oraemcc limits.d]# vi oracle-rdbms-server-12cR1-preinstall.conf
# oracle-rdbms-server-12cR1-preinstall setting for nofile soft limit is 1024
oracle soft nofile 30000
# oracle-rdbms-server-12cR1-preinstall setting for nofile hard limit is 65536
oracle hard nofile 65536
Notice it’s the hard limit value the OEM 13.2 installer references. With all the prereq checks successfully completed, click Next:
Click to Open Full Size
Select the Simple option of Create a new Enterprise Manager system, then click Next:
Click to Open Full Size
Use these values to populate this next screen, then click Next:
Field
Value
Middleware Home Location
/u01/app/oracle/product/middleware
Agent Base Directory
/u01/app/oracle/product/agent
Home Name
oraemcc.mynet.com
Click to Open Full Size
Dream up a suitable password for all the accounts mentioned in the next screen. Write it down now. You will need it shortly. Also, use these values to populate this next screen, then click Next:
Field
Value
Database Host Name
oraemcc.mynet.com
Port
1521
Service/SID
emdbrepos.mynet.com
SYS Password
The SYS password you used to create the MRDB
Click to Open Full Size
Now we hit another little gotcha. Clicking Next on the previous screen attempted to make a connection to the PDB containing the Management Repository. However, when we rebooted earlier, the CDB probably came back up, but the PDB probably did not. If that was the case, you see this rather misleading error message:
Click to Open Full SIze
Let’s login to the CDB and find out what’s going on:
SQL> select name, open_mode from v$pdbs;
NAME OPEN_MODE
------------------------------ ----------
PDB$SEED READ ONLY
EMPDBREPOS MOUNTED
SQL> alter pluggable database EMPDBREPOS open;
Pluggable database altered.
SQL> select name, open_mode from v$pdbs;
NAME OPEN_MODE
------------------------------ ----------
PDB$SEED READ ONLY
EMPDBREPOS READ WRITE
Click the OK button on the error window, then click Next again. That will land you on this next screen. I’m not interested in Oracle BI Publisher, so I unchecked those options and left the Software Library Location at its default value (/u01/app/oracle/product/swlib):
Click to Open Full Size
Click Next and we end up on a summary screen. Here’s the top half:
Click to Open Full Size
Here’s the bottom half showing all the default ports:
Click to Open Full Size
Click the Install button and then wait. Your next task will be to run a root script. It took close to 2 hours to reach that stage on my system. YMMV:
Click to Open Full Size
Time passes:
Click to Open Full Size
Eventually you’ll see this window appear:
Click to Open Full Size
In a separate root shell session, run the allroot.sh script then click the OK button:
[root@oraemcc middleware]# ./allroot.sh
Starting to execute allroot.sh .........
Starting to execute /u01/app/oracle/product/middleware/root.sh ......
/etc exist
/u01/app/oracle/product/middleware
Finished product-specific root actions.
/etc exist
Finished execution of /u01/app/oracle/product/middleware/root.sh ......
Starting to execute /u01/app/oracle/product/agent/agent_13.2.0.0.0/root.sh ......
Finished product-specific root actions.
/etc exist
Finished execution of /u01/app/oracle/product/agent/agent_13.2.0.0.0/root.sh ......
Finally, you’ll see an installation summary screen. Click Close and you’re done!
Click to Open Full Size
All that remains is to scoot over to your OEM CC 13.2 URL and login as SYSMAN:
Click to Open Full Size
If you have any comments or questions about this post, please use the Contact form here.
Sometimes you need to do things as the root user. To gain access to the root account is a multi-step process. Here’s what I do:
Fire up Putty and login as the admin user. The password will be the same as the admin user you setup when you initially configured your NAS. In order to get a Putty session established, you will need to ensure SSH is enabled. To do that, open up Control Panel, click Terminal & SNMP and check the Enable SSH service box.
Once you’re logged in as admin, enter this command:
admin@YOURNAS:/$ sudo su -
Password: <your admin password>
root@YOURNAS:~#
Adding Entries to /etc/hosts.
One thing you’ll need root access for is adding entries to the /etc/hosts file. This is useful if you want to take advantage of hostname resolution when setting up NFS client access on your NAS. You can use the IP address of NFS clients, so it’s not strictly necessary to add entries to /etc/hosts. Note, your custom entries to /etc/hosts do survive a DiskStation Manager (DSM) upgrade. You can use vi to edit /etc/hosts.
Setting Up NFS Access from Linux.
To conserve space on your Linux servers, it’s often a good idea to store Oracle media on an NFS drive rather than directly on the server itself. Setting up NFS access is a two step process. Step one is to setup NFS permissions via the DSM. Step two is to configure an NFS share on your Linux server.
Step #1: Setup NFS Permissions using Disk Station Manager (DSM).
After creating a volume using the DSM, you’ll need to configure NFS client access to that volume.
In DSM open up Control Panel. Click on Shared Folder. Right click the relevant volume, then click Edit. Once you’re in the Edit window, click on the NFS Permissions tab. Click the Create button. This will enable you to create an NFS rule. I use these entries:
Field
Value
Hostname or IP*:
NFS client IP or hostname (from /etc/hosts)
Privilege:
Read/Write
Squash:
No mapping (allows file ownership changes)
Enable asynchronous
Check
Allow connections from non-privileged ports
Leave Unchecked
Allow users to access mounted subfolders
Check
Click OK and the NAS will process your changes. That’s the NAS side of things completed.
Step #2: Configure an NFS Share on Linux.
The second stage is to add some configuration to the Linux server.
Add a mount point directory in the root file system. Let’s call it nas (how original!):
[root@orasvr01 ~]$ cd /
[root@orasvr01 ~]$ mkdir /nas
Make the mount point directory accessible to everyone:
[root@orasvr01 ~]# chmod 777 /nas
Edit the /etc/fstab. This is where it gets a bit tricky. The following entry would be just fine to access files in the NAS from Linux:
However, things get a little more complicated when it comes to accessing Oracle files in the NAS. The My Oracle Support (MOS) document ID 359515.1 describes the NFS mount parameters needed to access certain types of files.
We’re now ready to create an Oracle RAC database. It’s mostly quite similar to creating a single instance Oracle database. The major differences come with how the various infrastructure components are administered. We’ll get to that in good time.
So far we have the +CRS ASM Diskgroup which provides storage for the Votedisk, the OCR and the GIMR database (_MGMTDB). Our own RAC database will also use ASM storage, so we’ll need to create a couple more ASM Diskgroups before we create the database.
On racnode1, login as the grid user and fire up the ASM Configuration Assistant (asmca):
[grid@racnode1 ~]$ . oraenv
ORACLE_SID = [grid] ? +ASM1
The Oracle base has been set to /u01/app/grid
Set your DISPLAY environment variable so it points back to your workstation:
Click on Create. Enter DATA for the Disk Group Name, select External for the Redundancy level and select ASMDISK02 and ASMDISK03. Then Click OK:
Click OK:
Click Create:
Enter FRA for the Disk Group Name, select External for the Redundancy level and select ASMDISK04 and ASMDISK05. Then click OK:
Click OK in the Disk Group:Creation pop-up window:
Then click Exit to close the ASM Configuration Assistant.
Et voila! Two shiny new ASM Diskgroups ready for our RAC database to use. It will be pleased.
Task #2: Create an Oracle 12c RAC Database.
A quick way to set your oracle environment correctly when no databases currently exist on the server is to edit the /etc/oratab file to include a dummy line:
[oracle@racnode1 ~]$ vi /etc/oratab
dummy:/u01/app/oracle/product/12.1.0/dbhome_1:N
Use the standard oraenv script to set your environment, then unset the ORACLE_SID:
[oracle@racnode1 ~]$ . oraenv
ORACLE_SID = [oracle] ? dummy
The Oracle base has been set to /u01/app/oracle
[oracle@racnode1 ~]$ unset ORACLE_SID
You can tell you’re pointing to the correct set of Oracle executables by using the which command:
[oracle@racnode1 database]$ which dbca
/u01/app/oracle/product/12.1.0/dbhome_1/bin/dbca
Finally, set your DISPLAY variable correctly and fire up the Database Configuration Assistant (dbca):
Select Use the Same Administrative Password for All Accounts. Enter the SYSDBA password twice then click Next:
Complete the next screen using the values shown in the screen shot, then click Next:
Create a password for the ASMSNMP ASM instance account, then click OK:
Just click Next:
On the Memory tab, select Use Automatic Memory Management and enter 2048 MB for the Memory Size, then click the Character Sets tab:
On the Character Sets tab, select Use Unicode (AL32UTF8) then click Next:
Ensure Create Database is checked, then click Next:
The Prerequisite Checks screen highlights the swap size failure which we can safely ignore. Click Next:
On the Summary screen, click Finish:
The familiar progress bar tells you how things are going:
Eventually you’ll see the Finish screen. Click Close and you’re done.
You now have an Oracle 12c RAC database up and running with 2 instances. In the next section we’ll run through a series of tests to find out if all is well.
Task #3: Test Oracle 12c GI and RAC Database.
After you build out your cluster and RAC database, it’s important that you test everything is working properly, especially when it will become a production system. Testing really falls into two categories, functionality testing and workload testing. Functionality testing focuses on finding out if the configuration behaves in the way it should. Workload testing focuses on finding out if the configuration can handle the expected workload in terms of response time and throughput.
You can view an outline test plan here. It’s a decent start at putting together a comprehensive test plan, but it does omit three key areas. It does not contain any workload testing, network failure testing or any destructive testing. Let’s get testing and see what happens!
Section #1: Check & Test Clusterware.
The first set of checks and tests focuses on Clusterware. It gets us comfortable with using the various Clusterware administration commands and helps with identifying where certain resources are located.
Test #1B. Check: Resource TARGET and STATE columns.
The crs_stat command is supposed to be deprecated, but still works in 12c.
What we’re looking for here is any line where the Target and State column values are different. The Target value is what the status should be and the State value is what the status actually is. So, a Target value of ONLINE and a State value of OFFLINE could mean one of two things. Either the resource is coming online, but hadn’t quite made it when you ran the command or that resource has a problem and cannot come online. In which case, it would be time to dive into the log files. As we can see from this output, all is well.
The 3 main services are online on all nodes. Good news.
[grid@racnode1 ~]$ crsctl check cluster -all
**************************************************************
racnode1:
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
**************************************************************
racnode2:
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
**************************************************************
Test #1D. Check: CRS Verification.
Similar to Test 1C, this output shows the HA services are all OK too.
[grid@racnode1 ~]$ crsctl check crs
CRS-4638: Oracle High Availability Services is online
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
[grid@racnode2 ~]$ crsctl check crs
CRS-4638: Oracle High Availability Services is online
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
Test #1E. Check: CTSS.
Remember we’re not using NTP for server time synchronization, so Oracle’s CTSS should be doing the work for us which it seems to be.
[grid@racnode1 ~]$ crsctl check ctss
CRS-4701: The Cluster Time Synchronization Service is in Active mode.
CRS-4702: Offset (in msec): 0
[grid@racnode2 ~]$ crsctl check ctss
CRS-4701: The Cluster Time Synchronization Service is in Active mode.
CRS-4702: Offset (in msec): 0
Test #1F. Check: DNS.
Remember we configured DNS on the OEM Cloud Control server (oraemcc.mynet.com, 200.200.10.16). It seems to be working for both RAC nodes.
[grid@racnode1 ~]$ crsctl query dns -servers
CRS-10018: the following configuration was found on the system:
CRS-10019: There are 1 domains in search order. They are: mynet.com
CRS-10022: There are 1 name servers. They are: 200.200.10.16
CRS-10020: number of retry attempts for name lookup is: 4
CRS-10021: timeout for each name lookup is: 5
[grid@racnode2 ~]$ crsctl query dns -servers
CRS-10018: the following configuration was found on the system:
CRS-10019: There are 1 domains in search order. They are: mynet.com
CRS-10022: There are 1 name servers. They are: 200.200.10.16
CRS-10020: number of retry attempts for name lookup is: 4
CRS-10021: timeout for each name lookup is: 5
Test #1G. Check: Votedisk(s).
The Votedisk seems fine, but best practice says we should have more than just one and an odd number of them. We might want to fix this later.
[grid@racnode1 ~]$ crsctl query css votedisk
## STATE File Universal Id File Name Disk group
-- ----- ----------------- --------- ---------
1. ONLINE 6623c2c7b3a24fc5bffd28b93c053000 (/dev/oracleasm/disks/ASMDISK01) [CRS]
Located 1 voting disk(s).
Test #1H. Check: OCR File(s).
We can have up to 5 OCR files and we currently only have one. We might want to fix that later. Also, to perform the logical corruption check, we need to run this ocrcheck command as root.
[grid@racnode1 ~]$ cat /etc/oracle/ocr.loc
#Device/file +CRS getting replaced by device +CRS/cluster1/OCRFILE/registry.255.898728137
ocrconfig_loc=+CRS/cluster1/OCRFILE/registry.255.898728137
[grid@racnode1 ~]$ ocrcheck -details
Status of Oracle Cluster Registry is as follows :
Version : 4
Total space (kbytes) : 409568
Used space (kbytes) : 1588
Available space (kbytes) : 407980
ID : 1641944593
Device/File Name : +CRS/cluster1/OCRFILE/registry.255.898728137
Device/File integrity check succeeded
Device/File not configured
Device/File not configured
Device/File not configured
Device/File not configured
Cluster registry integrity check succeeded
Logical corruption check bypassed due to non-privileged user
[root@racnode1 ~]# /u01/app/12.1.0/grid/bin/ocrcheck -details
Status of Oracle Cluster Registry is as follows :
Version : 4
Total space (kbytes) : 409568
Used space (kbytes) : 1588
Available space (kbytes) : 407980
ID : 1641944593
Device/File Name : +CRS/cluster1/OCRFILE/registry.255.898728137
Device/File integrity check succeeded
Device/File not configured
Device/File not configured
Device/File not configured
Device/File not configured
Cluster registry integrity check succeeded
Logical corruption check succeeded
Test #1I. Check: OCR Backups.
The OCR is backed up automatically every 4 hours. If the cluster has been up and running long enough you can see the backup files. Running this command on any node should produce the same output.
[grid@racnode1 ~]$ ocrconfig -showbackup
racnode1 2016/01/27 10:15:35 /u01/app/12.1.0/grid/cdata/cluster1/backup00.ocr 0
racnode1 2016/01/27 06:14:48 /u01/app/12.1.0/grid/cdata/cluster1/backup01.ocr 0
racnode1 2016/01/27 02:14:11 /u01/app/12.1.0/grid/cdata/cluster1/backup02.ocr 0
racnode1 2016/01/26 18:13:01 /u01/app/12.1.0/grid/cdata/cluster1/day.ocr 0
racnode1 2016/01/15 04:31:01 /u01/app/12.1.0/grid/cdata/cluster1/week.ocr 0
PROT-25: Manual backups for the Oracle Cluster Registry are not available
Test #1J. Check: OLR.
The OLR plays in important role in starting the cluster if the OCR is stored in ASM. Better make sure each node’s OLR is alive and well.
[grid@racnode1 ~]$ cat /etc/oracle/olr.loc
olrconfig_loc=/u01/app/12.1.0/grid/cdata/racnode1.olr
crs_home=/u01/app/12.1.0/grid
[grid@racnode1 ~]$ ocrcheck -local -config
Oracle Local Registry configuration is :
Device/File Name : /u01/app/12.1.0/grid/cdata/racnode1.olr
[grid@racnode1 ~]$ ls -l /u01/app/12.1.0/grid/cdata/racnode1.olr
-rw------- 1 root oinstall 503484416 Feb 6 14:22 /u01/app/12.1.0/grid/cdata/racnode1.olr
[grid@racnode2 ~]$ cat /etc/oracle/olr.loc
olrconfig_loc=/u01/app/12.1.0/grid/cdata/racnode2.olr
crs_home=/u01/app/12.1.0/grid
[grid@racnode2 ~]$ ocrcheck -local -config
Oracle Local Registry configuration is :
Device/File Name : /u01/app/12.1.0/grid/cdata/racnode2.olr
[grid@racnode2 ~]$ ls -l /u01/app/12.1.0/grid/cdata/racnode2.olr
-rw------- 1 root oinstall 503484416 Feb 6 18:24 /u01/app/12.1.0/grid/cdata/racnode2.olr
Test #1K. Check: GPnP.
The Grid Plug ‘n’ Play profile is an XML file which also plays an important role in cluster startup. It contains important node configuration information including the storage location of some GI resources, the node’s networking configuration and so on. Make sure it’s where it ought to be and eye ball its contents to make sure they appear accurate. They should be as they’re maintained by the GPnP daemon (gpnpd.bin).
[grid@racnode1 ~]$ cd /u01/app/12.1.0/grid/gpnp/racnode1/profiles/peer
[grid@racnode1 peer]$ ls -l
-rw-r--r-- 1 grid oinstall 1908 Dec 17 22:44 profile.old
-rw-r--r-- 1 grid oinstall 1850 Dec 17 22:36 profile_orig.xml
-rw-r--r-- 1 grid oinstall 1908 Dec 17 22:44 profile.xml
[grid@racnode2 ~]$ cd /u01/app/12.1.0/grid/gpnp/racnode2/profiles/peer
[grid@racnode2 peer]$ ls -l
-rw-r--r-- 1 grid oinstall 1850 Dec 17 23:05 profile_orig.xml
-rw-r--r-- 1 grid oinstall 1908 Dec 17 23:05 profile.xml
Test #1L. Check: SCAN VIPs.
The SCAN Listeners each need a VIP on which to operate. These commands show use they’re up and running and open for business.
[grid@racnode1 ~]$ srvctl status scan
SCAN VIP scan1 is enabled
SCAN VIP scan1 is running on node racnode2
SCAN VIP scan2 is enabled
SCAN VIP scan2 is running on node racnode2
SCAN VIP scan3 is enabled
SCAN VIP scan3 is running on node racnode1
[grid@racnode1 ~]$ srvctl config scan
SCAN name: cluster1-scan.mynet.com, Network: 1
Subnet IPv4: 200.200.10.0/255.255.255.0/eth0, static
Subnet IPv6:
SCAN 0 IPv4 VIP: 200.200.10.122
SCAN VIP is enabled.
SCAN VIP is individually enabled on nodes:
SCAN VIP is individually disabled on nodes:
SCAN 1 IPv4 VIP: 200.200.10.120
SCAN VIP is enabled.
SCAN VIP is individually enabled on nodes:
SCAN VIP is individually disabled on nodes:
SCAN 2 IPv4 VIP: 200.200.10.121
SCAN VIP is enabled.
SCAN VIP is individually enabled on nodes:
SCAN VIP is individually disabled on nodes:
[grid@racnode1 ~]$ ps -ef | grep -v grep | grep SCAN
grid 7739 1 0 Jan26 ? 00:00:41 /u01/app/12.1.0/grid/bin/tnslsnr
LISTENER_SCAN3 -no_crs_notify -inherit
[grid@racnode2 ~]$ ps -ef | grep -v grep | grep SCAN
grid 16190 1 0 Jan26 ? 00:00:43 /u01/app/12.1.0/grid/bin/tnslsnr
LISTENER_SCAN1 -no_crs_notify -inherit
grid 7736 1 0 Jan26 ? 00:00:41 /u01/app/12.1.0/grid/bin/tnslsnr
LISTENER_SCAN2 -no_crs_notify -inherit
[grid@racnode1 ~]$ ping 200.200.10.120
PING 200.200.10.120 (200.200.10.120) 56(84) bytes of data.
64 bytes from 192.168.0.120: icmp_seq=1 ttl=64 time=0.288 ms
64 bytes from 192.168.0.120: icmp_seq=2 ttl=64 time=0.152 ms
--- 192.168.0.120 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1362ms
rtt min/avg/max/mdev = 0.152/0.220/0.288/0.068 ms
[grid@racnode1 ~]$ ping 200.200.10.121
PING 200.200.10.121 (200.200.10.121) 56(84) bytes of data.
64 bytes from 192.168.0.121: icmp_seq=1 ttl=64 time=0.072 ms
64 bytes from 192.168.0.121: icmp_seq=2 ttl=64 time=0.028 ms
--- 192.168.0.121 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1311ms
rtt min/avg/max/mdev = 0.028/0.050/0.072/0.022 ms
[grid@racnode1 ~]$ ping 200.200.10.122
PING 200.200.10.122 (200.200.10.122) 56(84) bytes of data.
64 bytes from 192.168.0.122: icmp_seq=1 ttl=64 time=0.310 ms
64 bytes from 192.168.0.122: icmp_seq=2 ttl=64 time=0.186 ms
--- 192.168.0.122 ping statistics ---
2 packets transmitted, 2 received, 0% packet loss, time 1314ms
rtt min/avg/max/mdev = 0.186/0.248/0.310/0.062 ms
Test #1M. Check: Node VIPs.
To avoid a lengthy TCP timeout in the event of a node failure, its VIP needs to failover to a surviving node. For it to do that, it needs to be enabled and pingable.
[grid@racnode1 ~]$ srvctl status vip -node racnode1
VIP racnode1-vip.mynet.com is enabled
VIP racnode1-vip.mynet.com is running on node: racnode1
[grid@racnode1 ~]$ srvctl status vip -node racnode2
VIP racnode2-vip.mynet.com is enabled
VIP racnode2-vip.mynet.com is running on node: racnode2
[grid@racnode1 ~]$ ping racnode1-vip.mynet.com
PING racnode1-vip.mynet.com (200.200.10.111) 56(84) bytes of data.
64 bytes from racnode1-vip.mynet.com (200.200.10.111): icmp_seq=1 ttl=64 time=0.024 ms
64 bytes from racnode1-vip.mynet.com (200.200.10.111): icmp_seq=2 ttl=64 time=0.031 ms
64 bytes from racnode1-vip.mynet.com (200.200.10.111): icmp_seq=3 ttl=64 time=0.026 ms
--- racnode1-vip.mynet.com ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2441ms
rtt min/avg/max/mdev = 0.024/0.027/0.031/0.003 ms
[grid@racnode1 ~]$ ping racnode2-vip.mynet.com
PING racnode2-vip.mynet.com (200.200.10.112) 56(84) bytes of data.
64 bytes from racnode2-vip.mynet.com (200.200.10.112): icmp_seq=1 ttl=64 time=0.308 ms
64 bytes from racnode2-vip.mynet.com (200.200.10.112): icmp_seq=2 ttl=64 time=0.174 ms
64 bytes from racnode2-vip.mynet.com (200.200.10.112): icmp_seq=3 ttl=64 time=0.163 ms
--- racnode2-vip.mynet.com ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2513ms
rtt min/avg/max/mdev = 0.163/0.215/0.308/0.065 ms
Test #1N. Check: Nodeapps.
Nodeapps comprise a node’s VIP, network and ONS daemon. Check to see if they’re all up and running.
[grid@racnode1 ~]$ srvctl status nodeapps
VIP racnode1-vip.mynet.com is enabled
VIP racnode1-vip.mynet.com is running on node: racnode1
VIP racnode2-vip.mynet.com is enabled
VIP racnode2-vip.mynet.com is running on node: racnode2
Network is enabled
Network is running on node: racnode1
Network is running on node: racnode2
ONS is enabled
ONS daemon is running on node: racnode1
ONS daemon is running on node: racnode2
Test #1O. Check: Node Participation.
You might think you know which nodes make up your cluster, but does the cluster agree? In our case, yes it does!
[grid@racnode1 ~]$ olsnodes -n -i -s -t -a
racnode1 1 racnode1-vip.mynet.com Active Hub Unpinned
racnode2 2 racnode2-vip.mynet.com Active Hub Unpinned
Section #2: Check & Test ASM.
ASM has become the de facto storage standard for Oracle databases, so we’d better check the clustered version is healthy.
Test #2B. Check: ASM Instances.
Ensure the ASM instances are up, running and can accept logins.
[grid@racnode1 ~]$ srvctl status asm
ASM is running on racnode2,racnode1
[grid@racnode1 ~]$ sqlplus / as sysasm
SQL*Plus: Release 12.1.0.2.0 Production on Sat Feb 6 17:14:59 2016
Copyright (c) 1982, 2014, Oracle. All rights reserved.
Connected to:
Oracle Database 12c Enterprise Edition Release 12.1.0.2.0 - 64bit Production
With the Real Application Clusters and Automatic Storage Management options
SQL>
[grid@racnode1 ~]$ asmcmd
ASMCMD> ls -l
State Type Rebal Name
MOUNTED EXTERN N CRS/
MOUNTED EXTERN N DATA/
MOUNTED EXTERN N FRA/
Test #2C. Check: ASM Diskgroups.
Next, check that all the ASM Diskgroups are OK.
[grid@racnode1 ~]$ echo "select name from v\$asm_diskgroup;" | sqlplus -s / as sysasm
NAME
------------------------------
FRA
DATA
CRS
[grid@racnode1 ~]$ for dg in FRA DATA CRS> do> srvctl status diskgroup -diskgroup ${dg} -detail> done
Disk Group FRA is running on racnode2,racnode1
Disk Group FRA is enabled
Disk Group DATA is running on racnode2,racnode1
Disk Group DATA is enabled
Disk Group CRS is running on racnode2,racnode1
Disk Group CRS is enabled
Test #2D. Check: ASM Diskgroup Metadata.
Check the ASM Diskgroup metadata is present and correct. No error messages means it is.
[grid@racnode1 ~]$ sqlplus / as sysasm
SQL> alter diskgroup crs check all;
Diskgroup altered.
SQL> alter diskgroup data check all;
Diskgroup altered.
[grid@racnode1 ~]$ asmcmd
ASMCMD> chkdg FRA
Diskgroup altered.
Test #2E. Check: ASM Disks.
Ensure all the ASM Disks you have configured are visible.
[grid@racnode1 ~]$ sqlplus / as sysasm
SQL> select dg.name DG, d.name Disk, decode(d.GROUP_NUMBER, 0,'Unallocated', 'In Use') State, d.path from v$asm_disk d left outer join v$asm_diskgroup dg on dg.group_number = d.group_number order by dg.name, d.path;
DG DISK STATE PATH
---------- ---------- ----------- ------------------------------
CRS CRS_0000 In Use /dev/oracleasm/disks/ASMDISK01
DATA DATA_0000 In Use /dev/oracleasm/disks/ASMDISK02
DATA DATA_0001 In Use /dev/oracleasm/disks/ASMDISK03
DATA DATA_0002 In Use /dev/oracleasm/disks/ASMDISK06
FRA FRA_0000 In Use /dev/oracleasm/disks/ASMDISK04
FRA FRA_0001 In Use /dev/oracleasm/disks/ASMDISK05
Unallocated /dev/oracleasm/disks/ASMDISK07
Unallocated /dev/oracleasm/disks/ASMDISK08
Unallocated /dev/oracleasm/disks/ASMDISK09
Unallocated /dev/oracleasm/disks/ASMDISK10
ASMCMD> lsdsk --discovery
Path
/dev/oracleasm/disks/ASMDISK01
/dev/oracleasm/disks/ASMDISK02
/dev/oracleasm/disks/ASMDISK03
/dev/oracleasm/disks/ASMDISK04
/dev/oracleasm/disks/ASMDISK05
/dev/oracleasm/disks/ASMDISK06
/dev/oracleasm/disks/ASMDISK07
/dev/oracleasm/disks/ASMDISK08
/dev/oracleasm/disks/ASMDISK09
/dev/oracleasm/disks/ASMDISK10
Test #2F. Check: ASM Clients.
Verify what is connected to each ASM instance.
[grid@racnode1 ~]$ sqlplus / as sysasm
SQL> select dg.NAME, c.INSTANCE_NAME, c.DB_NAME, c.CLUSTER_NAME, c.STATUS from v$asm_client c, v$asm_diskgroup dg where c.GROUP_NUMBER = dg.GROUP_NUMBER order by c.DB_NAME;
NAME INSTANCE_N DB_NAME CLUSTER_NAME STATUS
---------- ---------- -------- ------------ ------------
DATA +ASM1 +ASM cluster1 CONNECTED
CRS +ASM1 +ASM cluster1 CONNECTED
FRA RAC1DB1 RAC1DB cluster1 CONNECTED
DATA RAC1DB1 RAC1DB cluster1 CONNECTED
CRS -MGMTDB _mgmtdb cluster1 CONNECTED
[grid@racnode2 ~]$ sqlplus / as sysasm
NAME INSTANCE_N DB_NAME CLUSTER_NAME STATUS
---------- ---------- -------- ------------ ------------
CRS +ASM2 +ASM cluster1 CONNECTED
DATA +ASM2 +ASM cluster1 CONNECTED
DATA RAC1DB2 RAC1DB cluster1 CONNECTED
FRA RAC1DB2 RAC1DB cluster1 CONNECTED
Section #3: Check & Test Databases & Instances.
Test #3C. Check: User RAC Database & Instance.
Confirm where the database instances are running.
[oracle@racnode1 ~]$ srvctl status database -d RAC1DB
Instance RAC1DB1 is running on node racnode1
Instance RAC1DB2 is running on node racnode2
Test #3D. Check: User RAC Database DBVERIFY.
For the sake of brevity, we’ll run the DBVERIFY executable (dbv) for only one datafile.
[oracle@racnode1 ~]$ sqlplus / as sysdba
SQL> select name from v$datafile;
NAME
--------------------------------------------------------------------------------
+DATA/RAC1DB/DATAFILE/system.258.898779945
+DATA/RAC1DB/DATAFILE/sysaux.257.898779761
+DATA/RAC1DB/DATAFILE/undotbs1.260.898780207
+DATA/RAC1DB/DATAFILE/undotbs2.265.898783783
+DATA/RAC1DB/DATAFILE/users.259.898780201
[oracle@racnode1 scripts]$ cat dbv_parfile.txt
FILE='+DATA/RAC1DB/DATAFILE/users.259.898780201'
LOGFILE='users_datafile.out'
FEEDBACK=10
USERID=system/<removed>
[oracle@racnode1 scripts]$ dbv parfile=dbv_parfile.txt
DBVERIFY: Release 12.1.0.2.0 - Production on Mon Feb 15 11:54:52 2016
Copyright (c) 1982, 2014, Oracle and/or its affiliates. All rights reserved.
[oracle@racnode1 scripts]$ cat users_datafile.out
DBVERIFY: Release 12.1.0.2.0 - Production on Mon Feb 15 11:54:52 2016
Copyright (c) 1982, 2014, Oracle and/or its affiliates. All rights reserved.
DBVERIFY - Verification starting : FILE = +DATA/RAC1DB/DATAFILE/users.259.898780201
DBVERIFY - Verification complete
Total Pages Examined : 640
Total Pages Processed (Data) : 30
Total Pages Failing (Data) : 0
Total Pages Processed (Index): 5
Total Pages Failing (Index): 0
Total Pages Processed (Other): 588
Total Pages Processed (Seg) : 0
Total Pages Failing (Seg) : 0
Total Pages Empty : 17
Total Pages Marked Corrupt : 0
Total Pages Influx : 0
Total Pages Encrypted : 0
Highest block SCN : 0 (0.0)
SQL> select instance_name from v$instance;
select instance_name from v$instance
*
ERROR at line 1:
ORA-25408: can not safely replay call
SQL> /
INSTANCE_NAME
----------------
RAC1DB2
Section #4: Test System & Cluster.
Test #4C. Test: Cluster Restart.
[root@racnode1 ~]# cd /u01/app/12.1.0/grid/bin
[root@racnode1 ~]# ./crsctl stop cluster -all
The output from this command is quite verbose, but it can be viewed here.
[root@racnode1 bin]# ./crsctl check cluster -all
**************************************************************
racnode1:
CRS-4535: Cannot communicate with Cluster Ready Services
CRS-4530: Communications failure contacting Cluster Synchronization Services daemon
CRS-4534: Cannot communicate with Event Manager
**************************************************************
racnode2:
CRS-4535: Cannot communicate with Cluster Ready Services
CRS-4530: Communications failure contacting Cluster Synchronization Services daemon
CRS-4534: Cannot communicate with Event Manager
**************************************************************
[grid@racnode2 ~]$ crsctl status resource -t
CRS-4535: Cannot communicate with Cluster Ready Services
CRS-4000: Command Status failed, or completed with errors.
[root@racnode1 bin]# ./crsctl start cluster -all
The output from the start cluster command is also a little verbose. It can be viewed here.
Note, despite the cluster having been shutdown and re-started, the instance RAC1DB1 on racnode1 remained shutdown. This is because that instance was shutdown in Test 3J. Re-starting the cluster does NOT automatically re-start cluster resources which had been previously shutdown. This is expected behavior.
The HTML version is very useful because some of the references it contains are hyperlinks to Oracle’s technical support documentation. How convenient is that?
Note, running the cluvfy command with the -html option will overwrite any existing cvucheckreport.html file.
An example of this command’s output can be seen here. Looks like a couple of things could do with some attention.
Test #5C. Check: Run ORAchk.
ORAchk is a utility provided by Oracle Support which analyzes your operating system setup, Oracle software installation(s) and configurations and compares them with Oracle best practices, recommendations and available patches. The objective is to identify known problem areas before they cause disruption to database services. ORAchk can only be downloaded via MOS, Doc ID 1268927.2. As the oracle software installation user (usually oracle), you unzip the ORAchk zip file, then run the orachk command with the required parameters. For example:
[oracle@racnode1 ORAchk]$ ./orachk -a -o verbose
The great thing about ORAchk is it only analyzes the things it needs to or you want it to. It’s output is saved to a time stamped directory, the main component being a large HTML report file. When opened in a browser, the report shows an overall System Health Score out of 100. For each component checked, the report provides individual test feedback in the form of a Pass, Fail, Warning or Info message. You can view the details of each test which provides recommendations and even links to technical support documentation which show how to fix the problem. In some cases, the details contain actual SQL code which will fix the problem.
The not so great thing about ORAchk is that it is very fussy and as a result does not always generate correct or accurate results. For example, some perfectly valid unalias commands in the .bash_profile of the grid user caused ORAchk to fail multiple tests and even time out running others. Code timing out is especially problematic to the extent there are two environment variables you can set to allow a longer timeout periods. These are documented in the User Guide. The values I have used which (mostly) worked are:
RAT_ROOT_TIMEOUT=800
RAT_TIMEOUT=280
Suffice it to say, ORAchk isn’t perfect but is still worth trying. Just make sure you verify its results.
Task #4: Common Administration Tasks.
This task will see us perform a few corrective actions based upon the output of the health checks and ORAchk. We’ll also complete some admin which includes some of the things you’d typically do to a cluster and RAC database. We having fun yet? 🙂
ORAchk highlighted our RAC database is not running in archivelog mode. So let’s fix that.
Task #4.1a: Review Parameters.
The normal parameters used to control the archiving of the Online Redo Logs have no effect when using ASM and a Fast Recovery Area. The parameters which do matter are these:
SQL> show parameter db_recovery_file_dest
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
db_recovery_file_dest string +FRA
db_recovery_file_dest_size big integer 4785M
The parameter db_recovery_file_dest effectively trumps the traditional log_archive_dest_1 parameter and ensures the archived Online Redo Logs are stored in the Fast Recovery Area. In our case, that happens to be implemented as an ASM Diskgroup called +FRA. Also, the traditional parameter log_archive_format, which defaults to %t_%s_%r.dbf, has no effect since the archived Online Redo Log file names will follow the OMF standard naming convention.
[oracle@racnode1 ~]$ . oraenv
ORACLE_SID = [RAC1DB1] ?
The Oracle base remains unchanged with value /u01/app/oracle
[oracle@racnode1 ~]$ sqlplus / as sysdba
SQL*Plus: Release 12.1.0.2.0 Production on Tue Feb 16 21:02:46 2016
Copyright (c) 1982, 2014, Oracle. All rights reserved.
Connected to an idle instance.
SQL> startup mount
ORACLE instance started.
Total System Global Area 2147483648 bytes
Fixed Size 2926472 bytes
Variable Size 1275070584 bytes
Database Buffers 855638016 bytes
Redo Buffers 13848576 bytes
Database mounted.
Task #4.1d: Enable Archiving & Shutdown the Instance.
SQL> alter database archivelog;
Database altered.
SQL> shutdown immediate
ORA-01109: database not open
Database dismounted.
ORACLE instance shut down.
SQL> exit
SQL> archive log list
Database log mode Archive Mode
Automatic archival Enabled
Archive destination USE_DB_RECOVERY_FILE_DEST
Oldest online log sequence 316
Next log sequence to archive 317
Current log sequence 317
SQL> alter system switch logfile;
System altered.
SQL> select name from v$archived_log;
NAME
--------------------------------------------------------------------------------
+FRA/RAC1DB/ARCHIVELOG/2016_02_16/thread_1_seq_317.261.903993279
From racnode2 (Thread #2 – RAC1DB2):
SQL> archive log list
Database log mode Archive Mode
Automatic archival Enabled
Archive destination USE_DB_RECOVERY_FILE_DEST
Oldest online log sequence 265
Next log sequence to archive 265
Current log sequence 266
SQL> alter system switch logfile;
System altered.
SQL> select name from v$archived_log;
NAME
--------------------------------------------------------------------------------
+FRA/RAC1DB/ARCHIVELOG/2016_02_16/thread_1_seq_317.261.903993279
+FRA/RAC1DB/ARCHIVELOG/2016_02_16/thread_2_seq_265.262.903993427
We can also see the archived Online Redo Logs via ASMCMD:
ASMCMD> pwd
+FRA/RAC1DB/ARCHIVELOG/2016_02_16
ASMCMD> ls -l
Type Redund Striped Time Sys Name
ARCHIVELOG UNPROT COARSE FEB 16 21:00:00 Y thread_1_seq_317.261.903993279
ARCHIVELOG UNPROT COARSE FEB 16 21:00:00 Y thread_2_seq_265.262.903993427
Task #4.2: Add an Additional OCR File.
We can have up to 5 OCR files and the fact Oracle allows for that suggests we ought to have more than just one. So let’s fix that.
Task #4.2a: Review Current OCR Configuration.
[grid@racnode2 ~]$ ocrcheck
Status of Oracle Cluster Registry is as follows :
Version : 4
Total space (kbytes) : 409568
Used space (kbytes) : 1588
Available space (kbytes) : 407980
ID : 1641944593
Device/File Name : +CRS
Device/File integrity check succeeded
Device/File not configured
Device/File not configured
Device/File not configured
Device/File not configured
Cluster registry integrity check succeeded
Logical corruption check bypassed due to non-privileged user
So, we know we have just one OCR file located in the +CRS ASM Diskgroup. Let’s take a look at that file using ASMCMD:
ASMCMD> pwd
+crs/cluster1/OCRFILE
ASMCMD> ls -l
Type Redund Striped Time Sys Name
OCRFILE UNPROT COARSE FEB 16 22:00:00 Y REGISTRY.255.898728137
Task #4.2b: Add an OCR Location.
I created an additional ASM Diskgroup called CRS2. The ocrconfig command must be run as root.
[root@racnode1 ~]# . oraenv
ORACLE_SID = [root] ? +ASM1
The Oracle base has been set to /u01/app/grid
[root@racnode1 ~]# ocrconfig -add +CRS2
Note, initial attempts to run this command repeatedly generated PROT-16 and PROC-23 errors. CRS was running fine on both nodes, the CRS log file contained no errors and the ASM instance alert log did not tell me anything of interest. A reboot of both nodes made this problem go away. Go figure. YMMV.
Task #4.2c: Review Updated OCR Configuration.
[grid@racnode1 ~]$ ocrcheck
Status of Oracle Cluster Registry is as follows :
Version : 4
Total space (kbytes) : 409568
Used space (kbytes) : 1604
Available space (kbytes) : 407964
ID : 1641944593
Device/File Name : +CRS
Device/File integrity check succeeded
Device/File Name : +CRS2
Device/File integrity check succeeded
Device/File not configured
Device/File not configured
Device/File not configured
Cluster registry integrity check succeeded
Logical corruption check bypassed due to non-privileged user
[grid@racnode1 ~]$ cat /etc/oracle/ocr.loc
#Device/file getting replaced by device +CRS2/cluster1/OCRFILE/REGISTRY.255.904309547
ocrconfig_loc=+CRS/cluster1/OCRFILE/registry.255.898728137
ocrmirrorconfig_loc=+CRS2/cluster1/OCRFILE/REGISTRY.255.904309547
ASMCMD> pwd
+crs2/cluster1/OCRFILE
ASMCMD> ls -l
Type Redund Striped Time Sys Name
OCRFILE UNPROT COARSE FEB 20 14:00:00 Y REGISTRY.255.904309547
Task #4.3: Add an Additional Votedisk.
Oracle best practice dictates we should have an odd number of Votedisks. Yes, I know 1 is an odd number but we should probably have more than one. Let’s fix that.
Task #4.3a: Review Current Votedisk Configuration.
[grid@racnode2 ~]$ crsctl query css votedisk
## STATE File Universal Id File Name Disk group
-- ----- ----------------- --------- ---------
1. ONLINE 6623c2c7b3a24fc5bffd28b93c053000 (/dev/oracleasm/disks/ASMDISK01) [CRS]
Located 1 voting disk(s).
Note, there is currently only ONE Votedisk because it resides in the +CRS ASM Diskgroup which is configured with external redundancy and a single ASM Disk. Hence, only ONE Votedisk.
Task #4.3b: Add Additional Votedisks.
Since we’re using ASM to store the Votedisk, we can’t just add more Votedisks to other ASM Diskgroups. It doesn’t work that way. Instead,the Votedisk redundancy (i.e. the number of Votedisks you have) is dependent upon the redundancy level of the ASM Diskgroup where the Votedisk(s) are stored. However, there is a mismatch between ASM Diskgroup redundancy and Oracle Clusterware Votedisk redundancy.
Redundancy Level
Minimum # of ASM Disks
# of Clusterware Votedisk Files
External
1
1
Normal
2
3
High
3
5
What this means is, if you create an ASM Diskgroup with NORMAL redundancy and 2 ASM Disks, trying to create Votedisks in that ASM Diskgroup would fail. This is because NORMAL redundancy for Votedisks would need 3 ASM Disks, not 2. Using a NORMAL redundancy ASM Diskgroup with 3 ASM Disks could contain Votedisks as all 3 disks would be used to store Votedisks.
To further make the point, I created a 4 disk NORMAL redundancy ASM Diskgroup called DATA2:
SQL> select dg.name, dg.type Redundancy, d.name, d.pathfrom v$asm_disk d, v$asm_diskgroup dgwhere dg.group_number = d.group_numberand dg.name = 'DATA2'order by d.name;
NAME REDUNDANCY NAME PATH
---------- ---------- ---------- ------------------------------
DATA2 NORMAL DATA2_0000 /dev/oracleasm/disks/ASMDISK08
DATA2 NORMAL DATA2_0001 /dev/oracleasm/disks/ASMDISK09
DATA2 NORMAL DATA2_0002 /dev/oracleasm/disks/ASMDISK10
DATA2 NORMAL DATA2_0003 /dev/oracleasm/disks/ASMDISK06
Now we’ll move the Votedisk from the +CRS ASM Diskgroup to the +DATA2 ASM Diskgroup:
[grid@racnode2 ~]$ crsctl replace votedisk +DATA2
Successful addition of voting disk fd93674940d24f72bfcf07b3e1ac6d14.
Successful addition of voting disk fc9912809b3f4f3fbf64add0b80b34e4.
Successful addition of voting disk 5e840512f2904f41bfbc5dd468b4051b.
Successful deletion of voting disk 6623c2c7b3a24fc5bffd28b93c053000.
Successfully replaced voting disk group with +DATA2.
CRS-4266: Voting file(s) successfully replaced
As you can see, the NORMAL redundancy setting of the ASM Diskgroup created 3 Votedisks:
[grid@racnode2 ~]$ crsctl query css votedisk
## STATE File Universal Id File Name Disk group
-- ----- ----------------- --------- ---------
1. ONLINE fd93674940d24f72bfcf07b3e1ac6d14 (/dev/oracleasm/disks/ASMDISK08) [DATA2]
2. ONLINE fc9912809b3f4f3fbf64add0b80b34e4 (/dev/oracleasm/disks/ASMDISK09) [DATA2]
3. ONLINE 5e840512f2904f41bfbc5dd468b4051b (/dev/oracleasm/disks/ASMDISK10) [DATA2]
Located 3 voting disk(s).
Note, only 3 of the 4 ASM Disks in the +DATA2 ASM Diskgroup are used for Votedisks.
Task #4.4: Implement oswatcher.
ORAchk highlighted our RAC nodes are not running oswatcher. Let’s fix that.
The oswatcher utility available via Oracle Support, Doc ID 301137.1. It captures and analyzes performance data from the underlying operating system. The capture aspect is documented in MOS Doc ID 1531223.1 and the analyze aspect is documented in MOS Doc ID 461053.1.
Task #4.4a: Download.
Within MOS, navigate to Doc ID 301137.1, then click the OSWatcher Download link. This will take you to the OSWatcher Download section of the document. There you’ll see another link called:
Click here to download OSWatcher for Solaris, Linux and HP-UX. (AIX users see link below)
Click that link and you’ll be prompted to download a file called oswbb733.tar (where the 733 means version 7.3.3). Once this file is downloaded, copy it to an appropriate location on ALL the nodes in your cluster.
Task #4.4b: Install.
Installing OSWatcher is simply a case of untarring the tar file, then changing the file permissions:
[oracle@racnode1 ~]$ cd ./media/oswatcher
[oracle@racnode1 OSWatcher]$ tar -xvf oswbb733.tar
[oracle@racnode1 oswatcher]$ ls -l
drwxr-xr-x 4 oracle oinstall 4096 Feb 27 2015 oswbb
-rw-r--r-- 1 oracle oinstall 6318080 Feb 21 15:10 oswbb733.tar
[oracle@racnode1 oswatcher]$ cd oswbb
[oracle@racnode1 oswbb]$ chmod 744 *
[oracle@racnode1 oswbb]$ ls -l
-rwxr--r-- 1 oracle oinstall 67 Jan 15 2014 call_du.sh
-rwxr--r-- 1 oracle oinstall 68 Oct 7 2013 call_sar.sh
-rwxr--r-- 1 oracle oinstall 71 Jan 7 2014 call_uptime.sh
drwxr--r-- 4 oracle oinstall 4096 Sep 8 2014 docs
-rwxr--r-- 1 oracle oinstall 626 Jan 15 2014 Example_extras.txt
-rwxr--r-- 1 oracle oinstall 1864 Oct 7 2013 Exampleprivate.net
-rwxr--r-- 1 oracle oinstall 772 May 8 2014 ifconfigsub.sh
-rwxr--r-- 1 oracle oinstall 743 Oct 7 2013 iosub.sh
-rwxr--r-- 1 oracle oinstall 1486 Jan 8 2014 ltop.sh
-rwxr--r-- 1 oracle oinstall 542 Oct 7 2013 mpsub.sh
-rwxr--r-- 1 oracle oinstall 740 Oct 7 2013 nfssub.sh
-rwxr--r-- 1 oracle oinstall 5062 Sep 17 2014 OSWatcherFM.sh
-rwxr--r-- 1 oracle oinstall 35108 Feb 27 2015 OSWatcher.sh
-rwxr--r-- 1 oracle oinstall 233897 Feb 27 2015 oswbba.jar
-rwxr--r-- 1 oracle oinstall 414 Oct 7 2013 oswib.sh
-rwxr--r-- 1 oracle oinstall 435 Jan 8 2014 oswnet.sh
-rwxr--r-- 1 oracle oinstall 825 Oct 7 2013 oswrds.sh
-rwxr--r-- 1 oracle oinstall 524 Oct 7 2013 oswsub.sh
-rwxr--r-- 1 oracle oinstall 1445 Oct 17 2013 psmemsub.sh
drwxr--r-- 2 oracle oinstall 4096 May 2 2014 src
-rwxr--r-- 1 oracle oinstall 2574 Feb 26 2015 startOSWbb.sh
-rwxr--r-- 1 oracle oinstall 558 Apr 17 2014 stopOSWbb.sh
-rwxr--r-- 1 oracle oinstall 746 Nov 6 2013 tarupfiles.sh
-rwxr--r-- 1 oracle oinstall 4219 Nov 6 2013 tar_up_partial_archive.sh
-rwxr--r-- 1 oracle oinstall 545 Feb 23 2015 vmsub.sh
-rwxr--r-- 1 oracle oinstall 1486 Feb 26 2015 xtop.sh
User guides and README files are in the docs subdirectory.
Task #4.4c: Configure.
OSWatcher basically runs a series of OS monitoring commands at a predetermined frequency, stores the output of those commands for a predetermined amount of time in a predetermined location.
The defaults are a frequency of 30 seconds with output data kept for the last 48 hours in an archive directory where you untarred the software. These can all be customized.
In addition, Oracle Support recommends you use OSWatcher to monitor Grid Infrastructure’s private interconnect networks. This is done by creating a special file called private.net with the appropriate entries:
[oracle@racnode1 oswbb]$ cat private.net
######################################################################
#Linux Example
######################################################################
echo "zzz ***"`date`
traceroute -r -F racnode1-priv
traceroute -r -F racnode2-priv
######################################################################
# DO NOT DELETE THE FOLLOWING LINE!!!!!!!!!!!!!!!!!!!!!
######################################################################
rm locks/lock.file
Task #4.4d: Run OSWatcher.
The script which starts OSWatcher is called startOSWbb.sh. It can take up to 4 parameters:
Parameter Position
Meaning
1
Sample Frequency (Seconds)
2
Archive History (Hours)
3
OS Compression Utility to Use
4
Archive Directory Path
For example, to capture stats once per minute, keep those stats for 12 hours, use gzip to compress the archive data and store the archive data in /nas1/OSWATCHER/racnode1, use this command:
[oracle@racnode1 oswbb]$ ./startOSWbb.sh 60 12 gzip /nas1/OSWATCHER/racnode1
Info...Zip option IS specified.
Info...OSW will use gzip to compress files.
Testing for discovery of OS Utilities...
VMSTAT found on your system.
IOSTAT found on your system.
MPSTAT found on your system.
IFCONFIG found on your system.
NETSTAT found on your system.
TOP found on your system.
Warning... /proc/slabinfo not found on your system.
Testing for discovery of OS CPU COUNT
oswbb is looking for the CPU COUNT on your system
CPU COUNT will be used by oswbba to automatically look for cpu problems
CPU COUNT found on your system.
CPU COUNT = 1
Discovery completed.
Starting OSWatcher v7.3.3 on Sun Feb 21 17:32:10 CST 2016
With SnapshotInterval = 60
With ArchiveInterval = 12
OSWatcher - Written by Carl Davis, Center of Expertise,
Oracle Corporation
For questions on install/usage please go to MOS (Note:301137.1)
If you need further assistance or have comments or enhancement
requests you can email me Carl.Davis@Oracle.com
Data is stored in directory: /nas1/OSWATCHER/racnode1
Starting Data Collection...
Once OSWatcher is underway, it will echo heartbeat information to your screen at the frequency you specified (in our case every 60 seconds):
oswbb heartbeat:Sun Feb 21 17:40:16 CST 2016
It’s OK to log out of the session which initiated OSWatcher. The processes continue to run in the background:
Analyzing the OSWatcher data might be a topic of a future update. You can’t have everything all in one go.
Task #4.5: Check I/O Calibration.
The RAC Database Health Check reported this failure:
Verification Check : FILESYSTEMIO_OPTIONS
Verification Description : Checks FILESYSTEMIO_OPTIONS parameter
Verification Result : WARNING
Verification Summary : Check for FILESYSTEMIO_OPTIONS failed
Additional Details : FILESYSTEMIO_OPTIONS=setall supports both direct I/O and
asynchronus I/O which helps to achieve optimal performance
with database data files
Database(Instance) Status Expected Value Actual Value
---------------------------------------------------------------------------------------
rac1db FAILED filesystemio_options = SETALL filesystemio_options = none
We know our RAC database is using ASM for its storage which should optimize I/O for us. One of the prerequisites of running Oracle’s I/O Calibration is to ensure asynchronous I/O is enabled for all the database datafiles, so let’s check that now:
Asynchronous I/O seems to be in play. We know all these files are in the +DATA ASM Diskgroup, which contains 2 ASM Disks:
SQL> select dg.name DG,
d.name Disk,
decode(d.GROUP_NUMBER,
0,'Unallocated',
'In Use') State,
d.path
from v$asm_disk d left outer join v$asm_diskgroup dg
on dg.group_number = d.group_number
order by
dg.name,
d.path;
DG DISK STATE PATH
---------- ------------ ----------- ------------------------------
CRS CRS_0000 In Use /dev/oracleasm/disks/ASMDISK01
CRS2 CRS2_0000 In Use /dev/oracleasm/disks/ASMDISK07
DATA DATA_0000 In Use /dev/oracleasm/disks/ASMDISK02
DATA DATA_0001 In Use /dev/oracleasm/disks/ASMDISK03
DATA2 DATA2_0003 In Use /dev/oracleasm/disks/ASMDISK06
DATA2 DATA2_0000 In Use /dev/oracleasm/disks/ASMDISK08
DATA2 DATA2_0001 In Use /dev/oracleasm/disks/ASMDISK09
DATA2 DATA2_0002 In Use /dev/oracleasm/disks/ASMDISK10
FRA FRA_0000 In Use /dev/oracleasm/disks/ASMDISK04
FRA FRA_0001 In Use /dev/oracleasm/disks/ASMDISK05
I/O calibration is discussed in the Oracle Database Performance Tuning Guide and can be reviewed here. Let’s run the calibration code multiple times to see what results we get with FILESYSTEMIO_OPTIONS set to NONE and SETALL. Here’s the calibration code:
Having run the code multiple times with FILESYSTEMIO_OPTIONS set to NONE, let’s change it to SETALL. It’s not a dynamic parameter so we’ll need to bounce the database:
SQL> show parameter filesystemio_options
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
filesystemio_options string none
SQL> alter system set filesystemio_options = SETALL scope=spfile;
System altered.
[oracle@racnode1 racnode1]$ srvctl stop database -d RAC1DB
[oracle@racnode1 racnode1]$ srvctl start database -d RAC1DB
[oracle@racnode1 racnode1]$ srvctl status database -d RAC1DB
Instance RAC1DB1 is running on node racnode1
Instance RAC1DB2 is running on node racnode2
SQL> show parameter filesystemio_options
NAME TYPE VALUE
------------------------------------ ----------- ------------------------------
filesystemio_options string SETALL
These are the results averaged for multiple runs of the code:
FILESYSTEMIO_OPTIONS
Max IOPS
Latency
Max MBPS
NONE
9,631
9.67
105.67
SETALL
9,796
8
105.67
As you can see, the Max IOPS improved slightly with SETALL which also performed with slightly lower latency. That said, the two sets of results are so close that it makes little difference. However, things didn’t get worse by setting FILESYSTEMIO_OPTIONS to SETALL, so let’s call that a win. Sort of.
Task #4.6: Apply a Patch (22191349).
ORAchk recommended applying the Grid Infrastructure Patch Set Update 12.1.0.2.160119 (JAN2016). There are multiple steps involved, so let’s get started.
Note, this is a rolling patch. It can be applied to a running system. Clusterware will be shutdown and patched as you apply the patch one node at a time.
Task #4.6a: Update opatch.
To apply this patch, the opatch utility must be version 12.1.0.1.7 or later. So let’s check the version in the GI and database $ORACLE_HOME directories:
[grid@racnode1 OPatch]$ ./opatch version
OPatch Version: 12.1.0.1.3
OPatch succeeded.
[oracle@racnode1 OPatch]$ ./opatch version
OPatch Version: 12.1.0.1.3
OPatch succeeded.
We need to update opatch which is done by downloading the latest version from MOS, patch #6880880.
We’ll update the GI and database $ORACLE_HOME directories on racnode1 first, taking a backup of the original OPatch directories as we go.
Check the new opatch version then remove the zip file:
[grid@racnode1 OPatch]$ ./opatch version
OPatch Version: 12.1.0.1.10
OPatch succeeded.
[grid@racnode1 OPatch]$ cd ..
[grid@racnode1 grid]$ ls -l *.zip
-rw-r--r-- 1 grid oinstall 52774823 Feb 29 11:48 p6880880_121010_Linux-x86-64.zip
[grid@racnode1 grid]$ rm -f p6880880_121010_Linux-x86-64.zip
Note, don’t forget to change the GI $ORACLE_HOME permissions back to 755.
Now update opatch in the database $ORACLE_HOME:
[oracle@racnode1 dbhome_1]$ mv OPatch OPatch_12.1.0.1.3
[oracle@racnode1 dbhome_1]$ cp /nas1/PATCHES/6880880/p6880880_121010_Linux-x86-64.zip .
[oracle@racnode1 dbhome_1]$ unzip p6880880_121010_Linux-x86-64.zip
[oracle@racnode1 dbhome_1]$ cd OPatch
[oracle@racnode1 OPatch]$ ./opatch version
OPatch Version: 12.1.0.1.10
OPatch succeeded.
[oracle@racnode1 dbhome_1]$ cd ..
[oracle@racnode1 dbhome_1]$ ls -l *.zip
-rw-r--r-- 1 oracle oinstall 52774823 Feb 29 11:59 p6880880_121010_Linux-x86-64.zip
[oracle@racnode1 dbhome_1]$ rm -f p6880880_121010_Linux-x86-64.zip
Now repeat the same tasks for the GI and database $ORACLE_HOME directories on racnode2.
Task #4.6b: Create an Oracle Configuration Manager (OCM) Response File.
You must have an OCM response file with file permissions of 775.
[grid@racnode1 bin]$ pwd
/u01/app/12.1.0/grid/OPatch/ocm/bin
[grid@racnode1 bin]$ ./emocmrsp -no_banner -output /u01/app/12.1.0/grid/OPatch/ocm/ocm.rsp
Provide your email address to be informed of security issues, install and
initiate Oracle Configuration Manager. Easier for you if you use your My
Oracle Support Email address/User Name.
Visit http://www.oracle.com/support/policies.html for details.
Email address/User Name:
You have not provided an email address for notification of security issues.
Do you wish to remain uninformed of security issues ([Y]es, [N]o) [N]: Y
The OCM configuration response file (/u01/app/12.1.0/grid/OPatch/ocm/ocm.rsp) was
successfully created.
[grid@racnode1 bin]$ cd ..
[grid@racnode1 ocm]$ chmod 775 ocm.rsp
[grid@racnode1 ocm]$ ls -l
drwxr-x--- 2 grid oinstall 4096 Nov 13 14:15 bin
drwxr-x--- 2 grid oinstall 4096 Nov 13 14:15 doc
drwxr-x--- 2 grid oinstall 4096 Nov 13 14:15 lib
-rw-r----- 1 grid oinstall 638 Nov 13 14:15 ocm_platforms.txt
-rwxrwxr-x 1 grid oinstall 621 Feb 29 12:44 ocm.rsp
-rwxr-x--- 1 grid oinstall 42560960 Nov 13 14:15 ocm.zip
Repeat this step for the GI $ORACLE_HOME on racnode2.
Task #4.6c: Check the $ORACLE_HOME Inventories.
On racnode1, we need to check the GI and database $ORACLE_HOME inventories.
Again, the output is quite verbose, but it can be viewed here.
Checking the inventories tells us there are no interim patches applied to either the GI or database $ORACLE_HOME directories.
Task #4.6d: Unzip the Patch.
The patch needs to be unzipped and accessible to all nodes in the cluster. It makes it easier that way. Fortunately I have some NAS storage which is mounted on both RAC nodes.
[grid@racnode1 grid]$ cd /nas1/PATCHES/22191349/grid
[grid@racnode1 grid]$ unzip p22191349_121020_Linux-x86-64.zip
Task #4.6e: Check For One-Off Patch Conflicts.
We already know that neither $ORACLE_HOME directories contain any patches at this point, so there can’t be any conflicts by definition. Still, we’ll run the conflict check anyway which must be done as the root user:
[root@racnode1 grid]# pwd
/nas1/PATCHES/22191349/grid
[root@racnode1 grid]# export PATH=$PATH:/u01/app/12.1.0/grid/OPatch
[root@racnode1 grid]# which opatchauto
/u01/app/12.1.0/grid/OPatch/opatchauto
[root@racnode1 grid]# /u01/app/12.1.0/grid/OPatch/opatchauto apply /nas1/PATCHES/22191349/grid/22191349 -analyze -ocmrf /u01/app/12.1.0/grid/OPatch/ocm/ocm.rsp
OPatch Automation Tool
Copyright (c)2014, Oracle Corporation. All rights reserved.
OPatchauto Version : 12.1.0.1.10
OUI Version : 12.1.0.2.0
Running from : /u01/app/12.1.0/grid
opatchauto log file: /u01/app/12.1.0/grid/cfgtoollogs/opatchauto/22191349/
opatch_gi_2016-02-29_14-02-15_analyze.log
NOTE: opatchauto is running in ANALYZE mode. There will be no change to your system.
OCM RSP file has been ignored in analyze mode.
Parameter Validation: Successful
Configuration Validation: Successful
Patch Location: /nas1/PATCHES/22191349/grid/22191349
Grid Infrastructure Patch(es): 21436941 21948341 21948344 21948354
DB Patch(es): 21948344 21948354
Patch Validation: Successful
Grid Infrastructure home: /u01/app/12.1.0/grid
DB home(s): /u01/app/oracle/product/12.1.0/dbhome_1
Analyzing patch(es) on "/u01/app/oracle/product/12.1.0/dbhome_1" ...
Patch "/nas1/PATCHES/22191349/grid/22191349/21948344" successfully analyzed on
"/u01/app/oracle/product/12.1.0/dbhome_1" for apply.
Patch "/nas1/PATCHES/22191349/grid/22191349/21948354" successfully analyzed on
"/u01/app/oracle/product/12.1.0/dbhome_1" for apply.
Analyzing patch(es) on "/u01/app/12.1.0/grid" ...
Patch "/nas1/PATCHES/22191349/grid/22191349/21436941" successfully analyzed on
"/u01/app/12.1.0/grid" for apply.
Patch "/nas1/PATCHES/22191349/grid/22191349/21948341" successfully analyzed on
"/u01/app/12.1.0/grid" for apply.
Patch "/nas1/PATCHES/22191349/grid/22191349/21948344" successfully analyzed on
"/u01/app/12.1.0/grid" for apply.
Patch "/nas1/PATCHES/22191349/grid/22191349/21948354" successfully analyzed on
"/u01/app/12.1.0/grid" for apply.
SQL changes, if any, are analyzed successfully on the following database(s): RAC1DB
Apply Summary:
Following patch(es) are successfully analyzed:
GI Home: /u01/app/12.1.0/grid: 21436941,21948341,21948344,21948354
DB Home: /u01/app/oracle/product/12.1.0/dbhome_1: 21948344,21948354
opatchauto succeeded.
As we suspected, no conflicts so we can proceed and apply the patch.
Task #4.6f: Apply the Patch.
There are some options for applying the patch. We’ll choose the ‘non-shared GI and database $ORACLE_HOME and ACFS is not configured’ option. Each node has to be patched separately on a running system, i.e. the cluster is up. Let’s patch racnode1 first and enjoy the nightmare together. 🙂
The patch is applied by the root user using this command template:
This had the effect of installing patches 21948344 and 21948354 into the database $ORACLE_HOME, but it failed to install any patches into the GI $ORACLE_HOME. The reason was there was insufficient disk space. The patch log file contained these messages:
...
[Feb 29, 2016 2:27:40 PM] Running prerequisite checks...
[Feb 29, 2016 2:27:41 PM] Space Needed : 8180.049MB
[Feb 29, 2016 2:27:41 PM] Space Usable : 7786.043MB
[Feb 29, 2016 2:27:41 PM] Required amount of space(8180.049MB) is not available.
[Feb 29, 2016 2:27:41 PM] Prerequisite check "CheckSystemSpace" failed.
The details are:
Required amount of space(8180.049MB) is not available.
...
The /u01 file system did not have the 8 GB of free disk it needed. It would have been nice if the patch conflict command had told me this. Thanks Oracle! As is customary in patch README files, the command to rollback the patch is also documented. However, attempts to rollback the patch also failed because Clusterware wasn’t running and my attempts to re-start it on racnode1 repeatedly failed. A node reboot solved that problem and I tried to rollback the patch for a second time. That failed as well. This time the rollback command reported these errors in the log file:
COMMAND EXECUTION FAILURE :
Using configuration parameter file: /u01/app/12.1.0/grid/OPatch/opatchautotemp_racnode1/
patchwork/crs/install/crsconfig_params
Oracle Clusterware active version on the cluster is [12.1.0.2.0]. The cluster upgrade
state is [ROLLING PATCH]. The cluster active patch level is [0].
PROC-28: Oracle Cluster Registry already in current version
CRS-1153: There was an error setting Oracle Clusterware to rolling patch mode.
CRS-4000: Command Start failed, or completed with errors.
2016/02/29 16:01:23 CLSRSC-430: Failed to start rolling patch mode
ERROR:
Died at /u01/app/12.1.0/grid/OPatch/opatchautotemp_racnode1/patchwork/crs/
install/crspatch.pm line 776.
Clusterware was already in ‘rolling patch mode’, so you’d think trying to set it to that mode wouldn’t be a problem, right? Wrong! Unfortunately, this is an unpublished pub. Fortunately, Oracle already knows about this and has a workaround documented in MOS, Doc ID 1943498.1. It boils down to running this command:
[root@racnode1 ~]# crsctl stop rollingpatch
This command appears to switch Clusterware back into NORMAL mode, so it can be switched into ROLLING PATCH mode when you attempt to rollback the patch or apply it. The log file also contains this rather useful tidbit of information:
--------------After fixing the cause of failure you have two options shown below:
Run 'opatchauto resume'
or
Manually run the commands listed below
---------------------------------------------------------------------------------
...
I wasn’t about to trust ‘opatchauto resume’, so I opted for running the 17 fully documented commands in sequential order. Some needed to be run as root, some as oracle and some as grid. Unbelievably, they all worked! We can see the patches which were applied by running opatch lsinventory commands:
Note, opatch stores patch rollback information and files in a sub-directory within each $ORACLE_HOME called .patch_storage. This has the potential to consume a large quantity of disk space and thus reduce the free space available to the filesystem which contains the Oracle code installations. Running out of disk space in this filesystem is bad news. You should NOT just delete/archive the .patch_storage directory, so Oracle recommends one of 3 courses of action in the event disk space becomes an issue:
Increase the size of the filesystem. (I know – genius right? 😂)
Use symbolic links and relocate .patch_storage some place else. (This is what I did – worked fine)
Run the command: opatch util cleanup. (Documented in MOS Doc ID 550522.1)
So, with the $ORACLE_HOME directories patched on racnode1, it’s time to attempt applying the patch on racnode2. Yes, I did check I had more than 8 GB of free space in the /u01 filesystem before I started. Thanks for asking. When applying the patch actually works as intended, it looks like this:
[root@racnode2 ~]# . oraenv
ORACLE_SID = [+ASM2] ?
The Oracle base remains unchanged with value /u01/app/grid
[root@racnode2 trace]# export PATH=$PATH:/u01/app/12.1.0/grid/OPatch
[root@racnode2 trace]# which opatchauto
/u01/app/12.1.0/grid/OPatch/opatchauto
[root@racnode2 ~]# /u01/app/12.1.0/grid/OPatch/opatchauto apply /nas1/PATCHES/22191349/grid/22191349 -ocmrf /u01/app/12.1.0/grid/OPatch/ocm/ocm.rsp
OPatch Automation Tool
Copyright (c)2014, Oracle Corporation. All rights reserved.
OPatchauto Version : 12.1.0.1.10
OUI Version : 12.1.0.2.0
Running from : /u01/app/12.1.0/grid
opatchauto log file : /u01/app/12.1.0/grid/cfgtoollogs/opatchauto/22191349/
opatch_gi_2016-03-01_13-17-42_deploy.log
Parameter Validation: Successful
Configuration Validation: Successful
Patch Location: /nas1/PATCHES/22191349/grid/22191349
Grid Infrastructure Patch(es): 21436941 21948341 21948344 21948354
DB Patch(es): 21948344 21948354
Patch Validation: Successful
Grid Infrastructure home: /u01/app/12.1.0/grid
DB home(s): /u01/app/oracle/product/12.1.0/dbhome_1
Performing prepatch operations on CRS Home... Successful
Applying patch(es) to "/u01/app/oracle/product/12.1.0/dbhome_1" ...
Patch "/nas1/PATCHES/22191349/grid/22191349/21948344" successfully applied
to "/u01/app/oracle/product/12.1.0/dbhome_1".
Patch "/nas1/PATCHES/22191349/grid/22191349/21948354" successfully applied
to "/u01/app/oracle/product/12.1.0/dbhome_1".
Applying patch(es) to "/u01/app/12.1.0/grid" ...
Patch "/nas1/PATCHES/22191349/grid/22191349/21436941" successfully applied
to "/u01/app/12.1.0/grid".
Patch "/nas1/PATCHES/22191349/grid/22191349/21948341" successfully applied
to "/u01/app/12.1.0/grid".
Patch "/nas1/PATCHES/22191349/grid/22191349/21948344" successfully applied
to "/u01/app/12.1.0/grid".
Patch "/nas1/PATCHES/22191349/grid/22191349/21948354" successfully applied
to "/u01/app/12.1.0/grid".
Performing postpatch operations on CRS Home... Successful
SQL changes, if any, are applied successfully on the following database(s): RAC1DB
Apply Summary:
Following patch(es) are successfully installed:
GI Home: /u01/app/12.1.0/grid: 21436941,21948341,21948344,21948354
DB Home: /u01/app/oracle/product/12.1.0/dbhome_1: 21948344,21948354
opatchauto succeeded.
Again, running the opatch lsinventory command in each $ORACLE_HOME will verify the patches have been applied.
Task #4.6g: Run datapatch.
The final step involves running a script called datapatch which is located in the OPatch directory under $ORACLE_HOME. The documentation says for a RAC environment, run this from only one node. We’re left to assume we do this after all the nodes have been patched. There are instructions for both a “Standalone DB” and “Single/Multitenant (CDB/PDB) DB”. We know our user RAC database is not a CDB, but the GIMR database (_MGMTDB) is. Since the GI $ORACLE_HOME was patched and that codeset runs the _MGMTDB database, again we have to assume we need to run the datapatch script for that database as well. It would be nice if Oracle’s documentation made that clear.
The RAC1DB instances are already up and running, so we’ll dive straight in and run the datapatch script from racnode1:
[oracle@racnode1 OPatch]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/OPatch
[oracle@racnode1 OPatch]$ ./datapatch -verbose
SQL Patching tool version 12.1.0.2.0 on Wed Mar 2 10:16:57 2016
Copyright (c) 2015, Oracle. All rights reserved.
Log file for this invocation: /u01/app/oracle/cfgtoollogs/sqlpatch/
sqlpatch_2324_2016_03_02_10_16_57/sqlpatch_invocation.log
Connecting to database...OK
Bootstrapping registry and package to current versions...done
Determining current state...done
Current state of SQL patches:
Bundle series PSU:
ID 160119 in the binary registry and ID 160119 in the SQL registry
Adding patches to installation queue and performing prereq checks...
Installation queue:
Nothing to roll back
Nothing to apply
SQL Patching tool complete on Wed Mar 2 10:19:11 2016
Finally, we need to verify everything worked:
SQL> select patch_id, action, status from dba_registry_sqlpatch;
PATCH_ID ACTION STATUS
--------- --------------- ---------------
21948354 APPLY SUCCESS
Next, we’ll run the datapatch script in the _MGMTDB database. The script requires that all pluggable databases be open, so let’s double check that before we run the script:
[grid@racnode1 ~]$ . oraenv
ORACLE_SID = [-MGMTDB] ?
The Oracle base remains unchanged with value /u01/app/grid
[grid@racnode1 ~]$ sqlplus / as sysdba
SQL> select name, open_mode from v$pdbs;
NAME OPEN_MODE
------------------------------ ----------
PDB$SEED READ ONLY
CLUSTER1 READ WRITE
[grid@racnode1 ~]$ cd $ORACLE_HOME/OPatch
[grid@racnode1 OPatch]$ pwd
/u01/app/12.1.0/grid/OPatch
[grid@racnode1 OPatch]$ ./datapatch -verbose
SQL Patching tool version 12.1.0.2.0 on Wed Mar 2 10:48:36 2016
Copyright (c) 2015, Oracle. All rights reserved.
Log file for this invocation: /u01/app/grid/cfgtoollogs/sqlpatch/
sqlpatch_14626_2016_03_02_10_48_36/sqlpatch_invocation.log
Connecting to database...OK
Note: Datapatch will only apply or rollback SQL fixes for PDBs
that are in an open state, no patches will be applied to closed PDBs.
Please refer to Note: Datapatch: Database 12c Post Patch SQL Automation
(Doc ID 1585822.1)
Bootstrapping registry and package to current versions...done
Determining current state...done
Current state of SQL patches:
Bundle series PSU:
ID 160119 in the binary registry and ID 160119 in PDB CDB$ROOT, ID 160119
in PDB PDB$SEED, ID 160119 in PDB CLUSTER1
Adding patches to installation queue and performing prereq checks...
Installation queue:
For the following PDBs: CDB$ROOT PDB$SEED CLUSTER1
Nothing to roll back
Nothing to apply
SQL Patching tool complete on Wed Mar 2 10:50:57 2016
Finally, we need to verify everything worked:
SQL> select patch_id, action, status from dba_registry_sqlpatch;
PATCH_ID ACTION STATUS
-------- --------------- ---------------
21948354 APPLY SUCCESS
During the writing of Part 11, the January 2016 PSUs were released by Oracle. These are cumulative and bundle together many patches which were released previously. As always, we read the README file which comes with the patch very carefully to see what we’re getting into.
Note, this is a non-rolling patch and requires downtime to apply it.
The January 2016 PSU contains these patches:
Patch #
Description
Applicable Homes
21948354
DB PSU 12.1.0.2.160119
Both DB Homes and Grid Home
21948344
OCW PSU 12.1.0.2.160119
Both DB Homes and Grid Home
22139226
OJVM 12.1.0.2.160119
Only DB Homes
21948341
ACFS PSU 12.1.0.2.160119
Only Grid Home
21436941
DBWLM PSU 12.1.0.2.5
Only Grid Home
Since we’ve already applied Patch #22191349, we can check to see if there’s any overlap. These are the results of running opatch lsinventory for the GI and database $ORACLE_HOME directories:
Patch #
Applied to Homes
21948354
DB & GI
21948344
DB & GI
21948341
GI
21436941
GI
So there is a significant overlap with only Patch #22139226 adding anything new. We could apply just that one patch or we could test the application of the whole patch to make sure it’s smart enough to skip already applied patches and apply just the one we want (#22139226). So let’s do that. We’ll follow the same approach as before, even though we know some of the ground work has already been done.
Task #4.7a: Update opatch.
We must use opatch version 12.1.0.1.7 or later. We already have version 12.1.0.1.10.
Task #4.7b: Create an Oracle Configuration Manager (OCM) Response File.
Since things have changed, we’ll go ahead and re-generate the ocm.rsp file. Just for fun, I ran the diff command to compare the old ocm.rsp file and the new one and it reported they were different. So we’ll use a new one just to be on the safe side.
[grid@racnode1 bin]$ pwd
/u01/app/12.1.0/grid/OPatch/ocm/bin
[grid@racnode1 bin]$ ./emocmrsp -no_banner -output /u01/app/12.1.0/grid/OPatch/ocm/ocm.rsp
Provide your email address to be informed of security issues, install and
initiate Oracle Configuration Manager. Easier for you if you use your My
Oracle Support Email address/User Name.
Visit http://www.oracle.com/support/policies.html for details.
Email address/User Name: <just press Enter here>
You have not provided an email address for notification of security issues.
Do you wish to remain uninformed of security issues ([Y]es, [N]o) [N]: Y
The OCM configuration response file (/u01/app/12.1.0/grid/OPatch/ocm/ocm.rsp) was
successfully created.
[grid@racnode1 bin]$ cd ..
[grid@racnode1 ocm]$ chmod 775 ocm.rsp
Repeat this step on racnode2.
Task #4.7c: Check the $ORACLE_HOME Inventories.
As before, the point of this step is to generate documentation of what the inventories looked like before the patch was applied. The documentation also says if this command fails, you should contact Oracle Support so presumably a successful run must mean the inventories are healthy.
Task #4.7d: Unzip the Patch.
[grid@racnode1 ~]$ cd /nas1/PATCHES/22191676/grid
[grid@racnode1 grid]$ unzip p22191676_121020_Linux-x86-64.zip
Task #4.7e: Check For One-Off Patch Conflicts.
This is where the patch software should detect the presence of the existing patches and report that they’ll be skipped. Thus leaving only Patch #22191349 to be applied to the database $ORACLE_HOME. Here’s the command to check for conflicts:
So what did this command report? Well, the main opatchauto log file reported this:
...
The following patch(es) are duplicate patches with patches installed in the
Oracle Home.
[ 21948354]
You have already installed same patch(es) with same UPI(s) or same version(s).
These patch(es) will be skipped.
Opatchauto skipped installing the above patch in /u01/app/12.1.0/grid
...
The GI $ORACLE_HOME log file (referenced in the main opatchauto log file) reported this:
...
The following patch(es) are duplicate patches with patches installed in the
Oracle Home.
[ 21436941 21948341 21948344]
You have already installed same patch(es) with same UPI(s) or same version(s).
These patch(es) will be skipped.
...
So, for the GI $ORACLE_HOME, Patches #21948354, #21948344, #21948341 and #21436941 will all be skipped. This is correct!
The main opatchauto log file also reported this:
...
The following patch(es) are duplicate patches with patches installed in the
Oracle Home.
[ 21948354]
You have already installed same patch(es) with same UPI(s) or same version(s).
These patch(es) will be skipped.
Opatchauto skipped installing the above patch in
/u01/app/oracle/product/12.1.0/dbhome_1
...
The database $ORACLE_HOME log file (referenced in the main opatchauto log file) reported this:
...
Patches to apply -> [ 21948344 22139226 ]
Duplicate patches to filter -> [ 21948344 ]
The following patches are duplicate and are skipped:
[ 21948344 ]
OPatch found that the following patches are not required.
All of the constituents of the composite patch [ 21948354 ] are already installed
in the Oracle Home.
...
So, for the database $ORACLE_HOME, Patches #21948354 and 21948344 will be skipped leaving Patch #22139226 to be applied. Again, this is correct!
Task #4.7f: Apply the Patch.
Since we know there are no patches to apply to the GI $ORACLE_HOME, we can choose the option to only apply the patch to the database $ORACLE_HOME. The command template for that is:
Since we’ll be using the opatchauto command from the database $ORACLE_HOME, we should probably generate an OCM response file there:
[oracle@racnode1 bin]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/OPatch/ocm/bin
[oracle@racnode1 bin]$ ./emocmrsp -no_banner -output /u01/app/oracle/product/12.1.0/dbhome_1/OPatch/ocm/ocm.rsp
Provide your email address to be informed of security issues, install and
initiate Oracle Configuration Manager. Easier for you if you use your My
Oracle Support Email address/User Name.
Visit http://www.oracle.com/support/policies.html for details.
Email address/User Name: <just press Enter here>
You have not provided an email address for notification of security issues.
Do you wish to remain uninformed of security issues ([Y]es, [N]o) [N]: Y
The OCM configuration response file (/u01/app/oracle/product/12.1.0/dbhome_1/
OPatch/ocm/ocm.rsp) was successfully created.
[oracle@racnode1 bin]$ cd ..
[oracle@racnode1 ocm]$ chmod 775 ocm.rsp
First, we need to stop the RAC database:
[oracle@racnode1 ~]$ . oraenv
ORACLE_SID = [RAC1DB1] ?
The Oracle base remains unchanged with value /u01/app/oracle
[oracle@racnode1 ~]$ srvctl stop database -d RAC1DB
Next, stop Clusterware on the remote node (racnode2):
[root@racnode2 ~]# . oraenv
ORACLE_SID = [+ASM2] ?
The Oracle base remains unchanged with value /u01/app/grid
[root@racnode2 ~]# crsctl stop crs
Next, ensure Clusterware is still running on the local node (racnode1):
[root@racnode1 ~]# . oraenv
ORACLE_SID = [+ASM1] ?
The Oracle base remains unchanged with value /u01/app/grid
[root@racnode1 ~]# crsctl check crs
CRS-4638: Oracle High Availability Services is online
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
Next, run the opatchauto command according to the template referenced earlier:
The screen output can be viewed here. Overall, it says Patch #22139226 was applied, but that some SQL was not run because the RAC database was down. The good news is the SQL commands/scripts which were not executed are documented in the log file. All they do is query the existence of the SPFILE, startup an instance in upgrade mode, set CLUSTER_DATABASE=FALSE, run the datapatch script (in the $ORACLE_HOME/OPatch directory), then set CLUSTER_DATABASE=TRUE. We’ll apply the patch to racnode2, then run these SQL commands/scripts.
Following the same sequence of steps for racnode2, first generate a new OCM response file:
[oracle@racnode2 bin]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/OPatch/ocm/bin
[oracle@racnode2 bin]$ ./emocmrsp -no_banner -output /u01/app/oracle/product/12.1.0/dbhome_1/OPatch/ocm/ocm.rsp
Provide your email address to be informed of security issues, install and
initiate Oracle Configuration Manager. Easier for you if you use your My
Oracle Support Email address/User Name.
Visit http://www.oracle.com/support/policies.html for details.
Email address/User Name: <just press Enter here>
You have not provided an email address for notification of security issues.
Do you wish to remain uninformed of security issues ([Y]es, [N]o) [N]: Y
The OCM configuration response file (/u01/app/oracle/product/12.1.0/dbhome_1/
OPatch/ocm/ocm.rsp) was successfully created.
[oracle@racnode2 bin]$ cd ..
[oracle@racnode2 ocm]$ chmod 775 ocm.rsp
Next, re-start Clusterware on racnode2:
[root@racnode2 22191676]# . oraenv
ORACLE_SID = [+ASM2] ?
The Oracle base remains unchanged with value /u01/app/grid
[root@racnode2 22191676]# crsctl start crs
After a few moments, check the Clusterware software stack has re-started:
[root@racnode2 22191676]# crsctl check crs
CRS-4638: Oracle High Availability Services is online
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
Next, shutdown Clusterware on the remote node (racnode1):
[root@racnode1 ~]# . oraenv
ORACLE_SID = [RAC1DB1] ? +ASM1
The Oracle base has been changed from /u01/app/oracle to /u01/app/grid
[root@racnode1 ~]# crsctl stop crs
Next, set your environment and run the opatchauto command:
[root@racnode2 ~]# . oraenv
ORACLE_SID = [root] ? RAC1DB2
The Oracle base has been set to /u01/app/oracle
[root@racnode2 ~]# export PATH=$PATH:$ORACLE_HOME/OPatch
[root@racnode2 ~]# which opatchauto
/u01/app/oracle/product/12.1.0/dbhome_1/OPatch/opatchauto
The screen output was the same as the opatchauto run on racnode1 and Patch #22139226 was successfully applied. Now we’ll run the SQL commands which were missed due to the RAC database not being up and running.
The first SQL script just checks for the existence of an SPFILE, so we’ll skip that.
The second SQL script alters the CLUSTER_DATABASE parameter FALSE in the SPFILE:
[oracle@racnode2 ~]$ srvctl start instance -d RAC1DB -n racnode2
[oracle@racnode2 ~]$ sqlplus / as sysdba
SQL> @/u01/app/oracle/product/12.1.0/dbhome_1/cfgtoollogs/oplan/dataPatchfileForSetClusterFalse.sql
System altered.
The third SQL script starts up the instance in upgrade mode:
[oracle@racnode2 ~]$ srvctl stop instance -d RAC1DB -n racnode2
[oracle@racnode2 ~]$ sqlplus / as sysdba
Connected to an idle instance.
SQL> @/u01/app/oracle/product/12.1.0/dbhome_1/cfgtoollogs/oplan/dataPatchfileForStartupUpgrade.sql
ORACLE instance started.
Total System Global Area 2147483648 bytes
Fixed Size 2926472 bytes
Variable Size 1207961720 bytes
Database Buffers 922746880 bytes
Redo Buffers 13848576 bytes
Database mounted.
Database opened.
The fourth script runs the datapatch script:
[oracle@racnode2 ~]$ /bin/bash -c 'ORACLE_HOME=/u01/app/oracle/product/12.1.0/dbhome_1 ORACLE_SID=RAC1DB2 /u01/app/oracle/product/12.1.0/dbhome_1/OPatch/datapatch'
SQL Patching tool version 12.1.0.2.0 on Thu Mar 3 19:30:35 2016
Copyright (c) 2015, Oracle. All rights reserved.
Connecting to database...OK
Determining current state...done
Adding patches to installation queue and performing prereq checks...done
Installation queue:
Nothing to roll back
The following patches will be applied:
22139226 (Database PSU 12.1.0.2.160119, Oracle JavaVM Component (Jan2016))
Installing patches...
Patch installation complete. Total patches installed: 1
Validating logfiles...done
SQL Patching tool complete on Thu Mar 3 19:34:52 2016
The fifth SQL script alters the CLUSTER_DATABASE parameter TRUE in the SPFILE and shuts down the instance:
[oracle@racnode2 ~]$ sqlplus / as sysdba
SQL> @/u01/app/oracle/product/12.1.0/dbhome_1/cfgtoollogs/oplan/dataPatchfileForSetClusterTrue.sql
System altered.
Database closed.
Database dismounted.
ORACLE instance shut down.
We can now re-start the Clusterware on racnode1:
[root@racnode1 ~]# crsctl start crs
CRS-4123: Oracle High Availability Services has been started.
After waiting for a few moments, we can check the status of CRS:
[root@racnode1 ~]# crsctl check crs
CRS-4638: Oracle High Availability Services is online
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
Finally, we can re-start the RAC database:
[oracle@racnode2 ~]$ srvctl start database -d RAC1DB
[oracle@racnode2 ~]$ srvctl status database -d RAC1DB
Instance RAC1DB1 is running on node racnode1
Instance RAC1DB2 is running on node racnode2
Note:
Always read and understand the patch README file before doing anything else.
The README file will not always tell you exactly what to do. Sometimes you have to use common sense and your DBA gene to figure stuff out.
Always test applying a patch to a sandbox or test system for 3 main reasons:
To develop your own procedure so running it in production becomes a mechanical process with little to no thinking required.
To verify that it actually works.
To determine how long it takes so you can schedule an appropriate production outage if necessary.
Task #4.8: Add a Node to the Cluster.
We have a 2 node RAC cluster called cluster1 comprised of servers racnode1 and racnode2. Now let’s expand that cluster to 3 nodes by adding racnode3. After starting up racnode3 using Oracle VM Manager and connecting to the console, there are multiple steps to prepare racnode3 to join the cluster. This mostly involves changes to the OS, networking and shared storage.
Task #4.8a: Install the Cloud Control Management Agent.
Let’s get this out of the way first. Installation instructions for the Management Agent can be found here.
Task #4.8b: Fix the Networking.
Remember that racnode3 is a clone of racnode1, so it’s networking will be identical to racnode1. We need to change that so racnode3 has its own networking setup. An example of how to do that is documented here.
Task #4.8c: Remove IPv6.
If you’re not using IPv6 addressing, its presence on the vNICs will only confuse matters. So it’s best to remove it. Instructions on how to do that can be found here.
Task #4.8d: Fix the Hostname.
When racnode3 powers up, you’ll notice it’s called racnode1. We need to change its identity to racnode3. The procedure to do that is documented here.
Task #4.8e: Edit /etc/hosts.
On racnode3, update the /etc/hosts file so it’s identical to the other RAC nodes. You know it makes sense. A copy of this file can be found here. We will be configuring shared storage on the Openfiler server a little later, so in preparation for that add this entry to the /etc/hosts file on openfiler:
200.200.20.13 racnode3-storage
Task #4.8f: Edit /etc/resolv.conf.
Copy the /etc/resolv.conf file from either of the other 2 RAC nodes. It should contain just these entries:
# Generated by NetworkManager
search mynet.com
nameserver 200.200.10.16
Task #4.8g: Edit /etc/security/limits.conf.
Ensure this file has these entries:
####################################
# for oracle user
####################################
oracle soft nofile 8192
oracle hard nofile 65536
oracle soft nproc 2048
oracle hard nproc 16384
oracle soft stack 10240
oracle hard stack 32768
oracle soft core unlimited
oracle hard core unlimited
oracle hard memlock 5500631
####################################
# for grid user
####################################
grid soft nofile 8192
grid hard nofile 65536
grid soft nproc 2048
grid hard nproc 16384
grid soft stack 10240
grid hard stack 32768
grid soft core unlimited
grid hard core unlimited
Task #4.8h: Rename /etc/ntp.conf.
We’re using Oracle’s Cluster Time Synchronization Service rather than NTP to keep time within the cluster. However, the mere presence of this file can cause problems, so we need to rename it. Here are the steps to do that:
[root@racnode3 ~]# ls -l /etc/ntp.conf
-rw-r--r--. 1 root root 1770 Oct 26 10:38 /etc/ntp.conf
[root@racnode3 ~]# service ntpd status
ntpd is stopped
[root@racnode3 ~]# chkconfig ntpd off
[root@racnode3 ~]# mv /etc/ntp.conf /etc/ntp.conf.ORIG
Task #4.8i: Configure User Equivalency.
User equivalency or passwordless SSH needs to be setup for the racnode3 users oracle and grid. Instructions for configuring SSH can be found here.
Task #4.8j: Configure Shared Storage Within Openfiler.
First, we need to add racnode3-storage to the Network Access Configuration section. Open up the Openfiler management console, click on System, then add the entry for racnode3-storage using these values:
Name
Network/Host
Netmask
Type
racnode3-storage
200.200.20.13
255.255.255.255
Share
The screen should look similar to this:
Next, Click Volumes. The screen should change to this:
Then click iSCSI Targets over on the right hand side. The screen should change to this:
The following procedure should be repeated for iSCSI targets iqn.2006-01.com.openfiler:c1vg-vol01 through iqn.2006-01.com.openfiler:c1vg-vol10.
Use the Select iSCSI Target pull down menu to select the iqn.2006-01.com.openfiler:c1vg-vol01 (it should already be selected by default), then click the Change button. Click on Network ACL and use the Access pull down menu to select Allow for racnode3-storage. The screen should look similar to this:
Click the Update button, then click on Target Configuration and select the next iSCS target.
Task #4.8k: Configure Shared Storage at the OS Level.
First, verify the iSCSI client rpm is installed, then configure the iSCSI client to start on boot up of the node:
[root@racnode3 ~]# rpm -qa | grep iscsi
iscsi-initiator-utils-6.2.0.873-14.0.1.el6.x86_64
[root@racnode3 ~]# chkconfig iscsid on
[root@racnode3 ~]# chkconfig iscsi on
Next, discover the iSCSI Targets coming from the openfiler server:
[root@racnode3 ~]# oracleasm configure -i
Configuring the Oracle ASM library driver.
This will configure the on-boot properties of the Oracle ASM library
driver. The following questions will determine whether the driver is
loaded on boot and what permissions it will have. The current values
will be shown in brackets ('[]'). Hitting <ENTER> without typing an
answer will keep that current value. Ctrl-C will abort.
Default user to own the driver interface []: grid
Default group to own the driver interface []: asmadmin
Start Oracle ASM library driver on boot (y/n) [n]: y
Scan for Oracle ASM disks on boot (y/n) [y]: y
Writing Oracle ASM library driver configuration: done
Next, initialize Oracle ASM:
[root@racnode1 ~]# oracleasm init
Creating /dev/oracleasm mount point: /dev/oracleasm
Loading module "oracleasm": oracleasm
Configuring "oracleasm" to use device physical block size
Mounting ASMlib driver filesystem: /dev/oracleasm
Finally, instantiate the ASM Disks:
[root@racnode3 ~]# oracleasm scandisks
Reloading disk partitions: done
Cleaning any stale ASM disks...
Scanning system for ASM disks...
Instantiating disk "ASMDISK01"
Instantiating disk "ASMDISK02"
Instantiating disk "ASMDISK03"
Instantiating disk "ASMDISK04"
Instantiating disk "ASMDISK05"
Instantiating disk "ASMDISK06"
Instantiating disk "ASMDISK07"
Instantiating disk "ASMDISK08"
Instantiating disk "ASMDISK09"
Instantiating disk "ASMDISK10"
Task #4.8m: Run the cluvfy Utility.
As the grid user, run the following cluvfy utility command from the original software installation node (racnode1):
[grid@racnode3 ~]$ olsnodes -n -i -s -t -a
racnode1 1 racnode1-vip.mynet.com Active Hub Unpinned
racnode2 2 racnode2-vip.mynet.com Active Hub Unpinned
racnode3 3 racnode3-vip.mynet.com Active Hub Unpinned
Next, check CRS is up and running:
[grid@racnode3 ~]$ crsctl check crs
CRS-4638: Oracle High Availability Services is online
CRS-4537: Cluster Ready Services is online
CRS-4529: Cluster Synchronization Services is online
CRS-4533: Event Manager is online
Next, check the status of the SCAN VIPs:
[grid@racnode3 ~]$ srvctl status scan
SCAN VIP scan1 is enabled
SCAN VIP scan1 is running on node racnode1
SCAN VIP scan2 is enabled
SCAN VIP scan2 is running on node racnode3
SCAN VIP scan3 is enabled
SCAN VIP scan3 is running on node racnode2
Next, check the status of the nodeapps:
[grid@racnode3 ~]$ srvctl status nodeapps
VIP racnode1-vip.mynet.com is enabled
VIP racnode1-vip.mynet.com is running on node: racnode1
VIP racnode2-vip.mynet.com is enabled
VIP racnode2-vip.mynet.com is running on node: racnode2
VIP racnode3-vip.mynet.com is enabled
VIP racnode3-vip.mynet.com is running on node: racnode3
Network is enabled
Network is running on node: racnode1
Network is running on node: racnode2
Network is running on node: racnode3
ONS is enabled
ONS daemon is running on node: racnode1
ONS daemon is running on node: racnode2
ONS daemon is running on node: racnode3
Finally, check the status of ASM:
[grid@racnode3 ~]$ srvctl status asm
ASM is running on racnode2,racnode3,racnode1
Task #4.8o: Run the Database addnode.sh
With racnode3 now officially a member of the cluster, we now need to make it a database server. This is done in two stages. The first is to clone the database software using an existing cluster node, followed by adding a RAC database instance to racnode3.
Note, a non-domain qualified hostname is sufficient, i.e. racnode3 rather than racnode3.mynet.com.
The addnode script invokes a database installer GUI. Click the checkbox for racnode3:
Click Next:
As usual, it’s complaining about swap space which we know we can ignore. Click the Ignore All checkbox, then click Next:
Click Yes:
On the Summary screen, click Install:
The database $ORACLE_HOME on racnode1 is cloned to racnode3:
Towards the end of the cloning process, you’ll be asked to run a root.sh script on racnode3:
[root@racnode3 dbhome_1]# pwd
/u01/app/oracle/product/12.1.0/dbhome_1
[root@racnode3 dbhome_1]# ./root.sh
Performing root user operation.
The following environment variables are set as:
ORACLE_OWNER= oracle
ORACLE_HOME= /u01/app/oracle/product/12.1.0/dbhome_1
Enter the full pathname of the local bin directory: [/usr/local/bin]:
The contents of "dbhome" have not changed. No need to overwrite.
The contents of "oraenv" have not changed. No need to overwrite.
The contents of "coraenv" have not changed. No need to overwrite.
Entries will be added to the /etc/oratab file as needed by
Database Configuration Assistant when a database is created
Finished running generic part of root script.
Now product-specific root actions will be performed.
Once the root.sh script completes, return to the GUI, click OK to confirm the completion of the root.sh script, then click Close:
Task #4.8p: Run the Database Configuration Assistant (dbca).
All that’s happened is the Oracle database software has been installed onto racnode3. To extend the RAC database to a third instance, we need to run the Database Configuration Assistant from a node where an instance is already up and running, e.g. racnode1:
[oracle@racnode1 ~]$ . oraenv
ORACLE_SID = [RAC1DB1] ?
The Oracle base remains unchanged with value /u01/app/oracle
[oracle@racnode1 ~]$ which dbca
/u01/app/oracle/product/12.1.0/dbhome_1/bin/dbca
[oracle@racnode1 ~]$ dbca
Select Instance Management, then click Next:
Select Add an Instance, then click Next:
Select RAC1DB, then click Next:
Click Next:
Accept the default instance name of RAC1DB3 and click Next:
On the Summary screen, click Finish:
The GUI does its thing:
Finally, click Close:
We should now have a new instance up and running on racnode3. Let’s find out:
After adding a RAC1DB3 entry to the local /etc/oratab file:
[oracle@racnode3 dbhome_1]$ . oraenv
ORACLE_SID = [oracle] ? RAC1DB3
The Oracle base has been set to /u01/app/oracle
[oracle@racnode3 dbhome_1]$ sqlplus / as sysdba
SQL> select instance_name from v$instance;
INSTANCE_NAME
----------------
RAC1DB3
We know we’re using ASM, but we had no direct control over where the new instance’s files were created. So let’s make sure they exist in the correct place:
SQL> select lf.member from v$logfile lf where lf.group# in ( select l.group# from v$log l where l.thread# = 3 );
MEMBER
--------------------------------------------------------------------------------
+DATA/RAC1DB/ONLINELOG/group_6.270.905859405
+FRA/RAC1DB/ONLINELOG/group_6.481.905859405
+DATA/RAC1DB/ONLINELOG/group_5.271.905859405
+FRA/RAC1DB/ONLINELOG/group_5.482.905859407
SQL> select file_name from dba_data_files where tablespace_name = 'UNDOTBS3';
FILE_NAME
--------------------------------------------------------------------------------
+DATA/RAC1DB/DATAFILE/undotbs3.269.905859401
Finally, let’s make sure the cluster agrees the RAC1DB3 instance is part of our setup:
[oracle@racnode3 ~]$ srvctl status database -d RAC1DB
Instance RAC1DB1 is running on node racnode1
Instance RAC1DB2 is running on node racnode2
Instance RAC1DB3 is running on node racnode3
And for those of you who are paranoid like myself, let’s just see the process running:
Note, before we dive into deleting the node we just added, there are a few items worth mentioning so we’re all on the same page:
Our RAC database is admin-managed and we will be removing the instance and the Oracle Database software. The procedure is different for a policy-managed database.
Before removing an instance, ensure any services associated with that instance are relocated to other instances.
Ensure the services are configured such that the deleted instance is neither a preferred nor available instance.
Task #4.9a: Run the Database Configuration Assistant (dbca).
From a node other than the one which runs the instance you wish to delete, fire up dbca:
[oracle@racnode1 ~]$ . oraenv
ORACLE_SID = [RAC1DB1] ?
The Oracle base remains unchanged with value /u01/app/oracle
[oracle@racnode1 ~]$ dbca
Select Instance Management, then click Next:
Select Delete an Instance, then click Next:
Select RAC1DB, then click Next:
Select RAC1DB3, then click Next:
Click Finish:
Click OK to proceed:
The familiar progress bar tracks progress, then rather bizarrely a dialog box appears with a red cross, the digit 1 and an OK button. Clicking the OK button closes the dbca. Question is, did the removal of the RAC1DB3 instance succeed? Let’s find out:
SQL> select tablespace_name from dba_tablespaces where tablespace_name like 'UNDO%';
TABLESPACE_NAME
------------------------------
UNDOTBS1
UNDOTBS2
Looks like the instance RAC1DB3 has been deleted.
The Oracle Database software is still installed and the racnode3 node is still part of the cluster:
[oracle@racnode3 dbhome_1]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1
[grid@racnode3 ~]$ olsnodes -n -i -s -t -a
racnode1 1 racnode1-vip.mynet.com Active Hub Unpinned
racnode2 2 racnode2-vip.mynet.com Active Hub Unpinned
racnode3 3 racnode3-vip.mynet.com Active Hub Unpinned
Task #4.9b: Deinstall the Oracle Database Software.
To update the Oracle Inventory on racnode3, run this:
[oracle@racnode3 bin]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/oui/bin
[oracle@racnode3 bin]$ ./runInstaller -updateNodeList ORACLE_HOME=/u01/app/oracle/product/12.1.0/dbhome_1 "CLUSTER_NODES={racnode3}" -local
Starting Oracle Universal Installer...
Checking swap space: must be greater than 500 MB. Actual 4044 MB Passed
The inventory pointer is located at /etc/oraInst.loc
'UpdateNodeList' was successful.
To delete the database $ORACLE_HOME on racnode3, run this:
The output is quite verbose and can be viewed here. There was one failure. The script could not delete the directory /u01/app/oracle because another session was in that directory. Lesson learned! Apart from that, the Oracle Database software was completely removed from racnode3.
Finally, update the Oracle Inventories on the remaining nodes by running this:
[oracle@racnode1 bin]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/oui/bin
[oracle@racnode1 bin]$ ./runInstaller -updateNodeList ORACLE_HOME=/u01/app/oracle/product/12.1.0/dbhome_1 "CLUSTER_NODES={racnode1,racnode2}" LOCAL_NODE=racnode1
Starting Oracle Universal Installer...
Checking swap space: must be greater than 500 MB. Actual 4032 MB Passed
The inventory pointer is located at /etc/oraInst.loc
'UpdateNodeList' was successful.
The Oracle Database software installation has been deleted, but racnode3 is still seen as part of the cluster:
[oracle@racnode3 oracle]$ pwd
/u01/app/oracle
[oracle@racnode3 oracle]$ ls -la
drwxrwxr-x 2 oracle oinstall 4096 Mar 7 21:09 .
drwxr-xr-x 6 root oinstall 4096 Mar 4 10:59 ..
[grid@racnode1 ~]$ olsnodes -n -i -s -t -a
racnode1 1 racnode1-vip.mynet.com Active Hub Unpinned
racnode2 2 racnode2-vip.mynet.com Active Hub Unpinned
racnode3 3 racnode3-vip.mynet.com Active Hub Unpinned
Task #4.9c: Deinstall the Oracle Grid Infrastructure Software.
To update the Oracle Inventory on racnode3, run this:
[grid@racnode3 bin]$ pwd
/u01/app/12.1.0/grid/oui/bin
[grid@racnode3 bin]$ ./runInstaller -updateNodeList ORACLE_HOME=/u01/app/12.1.0/grid "CLUSTER_NODES={racnode3}" CRS=TRUE -silent -local
Starting Oracle Universal Installer...
Checking swap space: must be greater than 500 MB. Actual 4044 MB Passed
The inventory pointer is located at /etc/oraInst.loc
'UpdateNodeList' was successful.
To delete the GI $ORACLE_HOME on racnode3, run the following command.
Note, this will shutdown Clusterware and ASM on racnode3 and relocate the LISTENER_SCANn to another node.
The output is quite verbose and can be viewed here. In addition, part way through a generated script needs to be run as the root user. Its output can be viewed here.
To update the Oracle Inventories of the remaining nodes, run these commands:
[grid@racnode1 bin]$ pwd
/u01/app/12.1.0/grid/oui/bin
[grid@racnode1 bin]$ ./runInstaller -updateNodeList ORACLE_HOME=/u01/app/12.1.0/grid "CLUSTER_NODES={racnode1,racnode2}" CRS=TRUE -silent
Starting Oracle Universal Installer...
Checking swap space: must be greater than 500 MB. Actual 4028 MB Passed
The inventory pointer is located at /etc/oraInst.loc
'UpdateNodeList' was successful.
[oracle@racnode1 bin]$ pwd
/u01/app/oracle/product/12.1.0/dbhome_1/oui/bin
[oracle@racnode1 bin]$ ./runInstaller -updateNodeList ORACLE_HOME=/u01/app/oracle/product/12.1.0/dbhome_1 "CLUSTER_NODES={racnode1,racnode2}"
Starting Oracle Universal Installer...
Checking swap space: must be greater than 500 MB. Actual 4028 MB Passed
The inventory pointer is located at /etc/oraInst.loc
'UpdateNodeList' was successful.
[root@racnode1 ~]# crsctl delete node -n racnode3
CRS-4661: Node racnode3 successfully deleted.
Finally, let’s verify racnode3 is no longer part of the cluster:
[grid@racnode1 ~]$ cluvfy stage -post nodedel -n racnode3 -verbose
Performing post-checks for node removal
Checking CRS integrity...
The Oracle Clusterware is healthy on node "racnode1"
CRS integrity check passed
Clusterware version consistency passed.
Result: Node removal check passed
Post-check for node removal was successful.
[grid@racnode1 ~]$ olsnodes -n -i -s -t -a
racnode1 1 racnode1-vip.mynet.com Active Hub Unpinned
racnode2 2 racnode2-vip.mynet.com Active Hub Unpinned
So there it is, racnode3 successfully removed from the cluster.
That about wraps it up for Part 11. In Part 12 we tackle Oracle Recovery Manager. You never know, you might want to back something up and restore it again. See you next time.
If you have any comments or questions about this post, please use the Contact form here.
When starting up VMs from OVM Manager, you may encounter this error:
Error: Device 1 (vif) could not be connected. Hotplug scripts not working
Note, the Device could be 0, 2, etc. depending upon the number of vNICs you have configured.
What’s happening is the vNICs are not being initialized by the Hotplug scripts within the specified time limit. It could also be the side effect of a Xen memory leak which is a known bug. If the VM can be started without any vNICs, then you’ll know you have this issue. Here are 3 things you can try if you’re in this situation.
Solution #1: Add the vNICs one at a time.
RACNODE2_VM has an ID of 0004fb00000600000703eab3e0c76af5.
We know from creating the VM storage repositories earlier, that the VM_Filesystems_Repo has an ID of 0004fb000003000059416081b6e25e36. If we log onto the Oracle VM server itself, we can locate this VM’s configuration file called vm.cfg:
[root@ovmsvr]# cd /
[root@ovmsvr]# cd OVS
[root@ovmsvr]# cd Repositories
[root@ovmsvr]# cd 0004fb000003000059416081b6e25e36
[root@ovmsvr]# cd VirtualMachines
[root@ovmsvr]# cd 0004fb00000600000703eab3e0c76af5
[root@ovmsvr]# pwd
/OVS/Repositories/0004fb000003000059416081b6e25e36/VirtualMachines/
0004fb00000600000703eab3e0c76af5
[root@ovmsvr]# ls -l
total 0
-rw------- 1 root root 981 Dec 17 22:03 vm.cfg
The vm.cfg file contains the VM configuration data. The first line in vm.cfg specifies the vNIC MAC addresses. For example:
Make a copy of this line and comment it out so you have a record of what it looked like originally. Then edit the original line, removing all but the first mac and bridge pair. Then start the VM. If that works, stop the VM, edit this line to include the first 2 mac and bridge pairs, then start the VM again. If that works, try starting the VM with this line containing all 3 mac and bridge pairs (i.e. as the line looked originally). This has worked for me, but not always.
Solution #2: Restart the ovs-agent service.
The problem may be caused by a memory leak which is a known bug. This is documented on MOS, Doc ID 2077214.1. What that note effectively tells you to do is restart the ovs-agent service:
[root@ovmsvr ~]# service ovs-agent restart
Stopping Oracle VM Agent: [ OK ]
Starting Oracle VM Agent: [ OK ]
This is a quick and easy workaround and has worked for me.
Solution #3: Edit and re-start.
The problem could be the value of DEVICE_CREATE_TIMEOUT which defaults to 100. According to MOS, Doc ID 1089604.1, this timeout value can be increased. Thus giving the Hotplug scripts more time to complete their tasks. This is done by editing this file, /etc/xen/xend-config.sxp.
Find this line:
#(device-create-timeout 100)
and change it to this:
(device-create-timeout 150)
I have had some success with this solution, but VMs have refused to start after making this change. I got around that by using solution #2.
Note, to make this change effective restart of the OVM Server, which takes ALL your VMs down!
Error Logging Into Oracle VM Manager:
If you’ve not logged into OVM Manager for a while, you may find it’s taken an extended vacation. This can take the form of having to wait a really long time after you’ve entered your username and password, then it throws an “unexpected” error with this accompanying text:
java.lang.NullPointerException
I know, right? How unusual for Java to have had a problem. I mean it’s so rock solid and stable!
Now, you’re probably wanting to log in because you have something to do and don’t have a week to trawl through countless log files trying to find out why Java has had a hissy fit. So the quickest and simplest way to get logged into OVM Manager after seeing this error is to stop and re-start it:
[root@ovmmgr ~]# service ovmm stop
Stopping Oracle VM Manager [ OK ]
[root@ovmmgr ~]# service ovmm start
Starting Oracle VM Manager [ OK ]
Once OVM Manager is up and running, trying logging in again.
VM Stuck on ‘Stopping’:
Every now and then stopping a VM puts the VM into a ‘Stopping’ state in OVM Manager and that’s where it stays. Aborting the stop or restart and trying again has no effect other than to wind you up. Brilliant! Fortunately there’s a back door method to kick the VM to death, then resurrect it safely. For this you will need the ID for the VM. This can be obtained via OVM Manager.
Click the arrowhead to the left of the VM name. This opens up the Configuration tab. Make a note of the long VM ID string.
Next, log into the OVM server as root and locate the directory where the VM configuration file is located (vm.cfg). The path will be something like this:
Next, run two xm commands. One to destroy the VM (it doesn’t actually destroy it – rather it destroys the processes running the VM) and the other to create the VM using its configuration file: Here’s an example (where the string ending in 8f4 is the ID of the VM):
After building and configuring the hardware infrastructure, it’s time to install some Oracle software. If all the prior tasks have been completed successfully, the software installations should be easy. In Part 10, we’ll install the OEM Management Agent, Oracle 12c Grid Infrastructure and the database software.
Task #1: Install the OEM Cloud Control Management Agent.
To avoid the hassle of adding targets manually to Cloud Control later, we’ll install the Management Agent now so it’ll just pick up targets as they materialize. Just 4 steps should do it.
Task #1a: Create a Management Agent Software Directory.
We need to create a base directory for the Management Agent software. It’s required that the base directory and all its parent directories are owned by the installing user (oracle). Therefore, we can’t install the Management Agent into /u01/app/oracle because /u01/app is owned by the grid user. So, on both racnode1 and racnode2, create a directory for the Management Agent software installation using these commands:
Since the Management Agent software will be the first Oracle software installed on the RAC server nodes, there won’t be a pre-existing Oracle Inventory it can write to. That causes a problem when Cloud Control tries to deploy the Management Agent. To get around that, do this:
A fully documented /etc/hosts file can be viewed here.
Task 1d: Use Cloud Control.
For this walkthrough we’ll use racnode2 as the example. You could deploy Management Agents to both racnode servers at the same time or install them one at a time. Your call.
Log into Cloud Control, click on Setup ⇒ Add Target ⇒ Add Targets Manually:
Click Add Host:
Click on the Add menu, choose Manually then enter hostname (racnode2.mynet.com) and platform (Linux x86-64):
Click Next, then use these values to populate the screen:
Field
Value
Installation Base Directory
/u01/oracle/agent
Instance Directory
/u01/oracle/agent/agent_inst
In addition, create a new Named Credential for the oracle OS user then click OK:
With all the fields populated, the screen should look similar to this:
Click Next:
Click Deploy Agent:
The deploy proceeds until it starts whining about running a root script. This can be ignored for now. Click the Continue menu then choose Continue, All hosts:
Now you can run the root.sh script as specified on the screen (/u01/oracle/agent/core/12.1.0.5.0/root.sh):
Then click the Done button and you’re done. Management Agent Deployed!
You can check all is well with the Management Agent by running these commands as the oracle user:
[oracle@racnode2 ~]$ cd /u01/oracle/agent/agent_inst/bin
[oracle@racnode2 bin]$ ./emctl status agent
Oracle Enterprise Manager Cloud Control 12c Release 5
Copyright (c) 1996, 2015 Oracle Corporation. All rights reserved.
---------------------------------------------------------------
Agent Version : 12.1.0.5.0
OMS Version : 12.1.0.5.0
Protocol Version : 12.1.0.1.0
Agent Home : /u01/oracle/agent/agent_inst
Agent Log Directory : /u01/oracle/agent/agent_inst/sysman/log
Agent Binaries : /u01/oracle/agent/core/12.1.0.5.0
Agent Process ID : 4519
Parent Process ID : 4258
Agent URL : https://racnode2.mynet.com:3872/emd/main/
Local Agent URL in NAT : https://racnode2.mynet.com:3872/emd/main/
Repository URL : https://oraemcc.mynet.com:4903/empbs/upload
Started at : 2016-01-26 13:55:19
Started by user : oracle
Operating System : Linux version 3.8.13-118.2.2.el6uek.x86_64 (amd64)
Last Reload : (none)
Last successful upload : 2016-02-02 21:50:25
Last attempted upload : 2016-02-02 21:50:25
Total Megabytes of XML files uploaded so far : 8.58
Number of XML files pending upload : 0
Size of XML files pending upload(MB) : 0
Available disk space on upload filesystem : 32.51%
Collection Status : Collections enabled
Heartbeat Status : Ok
Last attempted heartbeat to OMS : 2016-02-02 21:52:44
Last successful heartbeat to OMS : 2016-02-02 21:52:44
Next scheduled heartbeat to OMS : 2016-02-02 21:53:44
---------------------------------------------------------------
Agent is Running and Ready
Task #2: Run the Grid Infrastructure Installer.
With the Grid Infrastructure software copied to racnode1 and unzipped, ensure your DISPLAY environment variable is set to the IP address or hostname of your workstation:
Note, the Grid Infrastructure software is installed by the grid user.
Test this is working by running an X Windows program, making sure the display opens on your workstation. Most people tend to use xclock as do I from time to time. I prefer xeyes though because it’s funnier:
[grid@racnode1 grid]$ xeyes
With the display set correctly, invoke the Grid Infrastructure installer:
[grid@racnode1 grid]$ ./runInstaller
Select Install and Configure Oracle Grid Infrastructure for a Cluster, then click Next:
Select Configure a Standard cluster, then click Next:
Select Advanced Installation, then click Next:
Use these values to populate the next screen, then click Next:
Field
Value
Cluster Name
cluster1
SCAN Name
cluster1-scan.mynet.com
SCAN Port
1521
Configure GNS
Uncheck this box
Use these values to populate the next screen, then click SSH connectivity:
Public Hostname
Virtual Hostname
racnode1.mynet.com
racnode1-vip.mynet.com
racnode2.mynet.com
racnode2-vip.mynet.com
After clicking SSH connectivity, enter the OS password of the grid user, then click Setup. Let the installer re-do the SSH setup. It uses RSA keys. Again, let Oracle take care of this. If the installation goes wrong due to a user equivalency problem, it won’t be your fault! Once SSH is re-configured, click Next.
Use these values to populate the next screen, then click Next:
Interface Name
Subnet
Use for
eth0
200.200.10.0
Public
eth1
200.200.20.0
Do Not Use
eth2
200.200.30.0
Private
Note, although we are using eth1 to access shared storage, Oracle Grid Infrastructure has no business on this network so it must be flagged as “Do Not Use”.
Select Use Standard ASM for storage, then click Next:
Click the Change Discovery Path button:
Change the Disk Discovery Path to /dev/oracleasm/disks/* then click OK:
Use these values to populate the next screen, then click Next:
Field
Value
Disk group name
CRS
Redundancy
External
Disk Path
Check /dev/oracleasm/disks/ASMDISK01
Click Use same password for these accounts, enter a password for SYSASM twice, then click Next:
Select Do not use Intelligent Platform Management Interface (IPMI):
Click the check box to Register with Enterprise Manager (EM) Cloud Control. Use these values to populate the screen, then click Next:
The 3 OS groups should be pre-populated with these values. Click Next:
OS Group
OS Group Name
Oracle ASM Administrator (OSASM) Group
asmadmin
Oracle ASM DBA (OSDBA for ASM) Group
asmdba
Oracle ASM Operator (OSOPER for ASM) Group (Optional)
asmoper
Use these values to populate the next screen, then click Next:
Field
Value
Oracle base
/u01/app/grid
Software location
/u01/app/12.1.0/grid
Make sure the Automatically run configuration scripts check box is unchecked. We’ll be running root scripts as the root user from the command line later. Click Next:
This next screen shows the results of the prerequisite checks. As you can see, the swap space issue came up which we know we can ignore. The screen also shows a Device Checks for ASM failure. What this boils down to is the installer’s inability to detect shared storage on both the RAC nodes.
Here is the full text of the failure:
Device Checks for ASM - This is a prerequisite check to verify that the specified
devices meet the requirements for ASM.
Check Failed on Nodes: [racnode2, racnode1]
Verification result of failed node: racnode2
Expected Value : cvuqdisk-1.0.9-1
Actual Value : cvuqdisk-1.0.9-1
Details:
-
PRVF-9802 : Attempt to get 'udev' information from node "racnode2" failed No UDEV rule
found for device(s) specified - Cause: An attempt to read the ''udev'' permission or
rule file failed, or the permission or rule file did not contain any rules for the
specified device or devices.
- Action: Make sure that the ''udev'' permission or rule directory is created, the
''udev'' permission or rule file is available and accessible by the user running the
check and that the permission or rule file contains the correct ''udev'' rule for the
device or devices being checked.
Back to Top
Verification result of failed node: racnode1
Expected Value : cvuqdisk-1.0.9-1
Actual Value : cvuqdisk-1.0.9-1
Details:
-
PRVF-9802 : Attempt to get 'udev' information from node "racnode1" failed No UDEV rule
found for device(s) specified - Cause: An attempt to read the ''udev'' permission or
rule file failed, or the permission or rule file did not contain any rules for the
specified device or devices.
- Action: Make sure that the ''udev'' permission or rule directory is created, the
''udev'' permission or rule file is available and accessible by the user running the
check and that the permission or rule file contains the correct ''udev'' rule for the
device or devices being checked.
What this essentially means is a check was made for the presence of the CVUQDISK package on both racnode1 and racnode2 (remember we installed that here). The installer is looking for version 1.0.9-1 and it found version 1.0.9-1, but apparently that wasn’t good enough. Further checks were made for udev rules regarding the shared disk selected earlier (ASMDISK01) and the installer couldn’t find those either. I actually tried this installation both with and without udev rules and it didn’t make any difference to the reporting of this ‘failure’. I even tried re-installing the CVUQDISK package on both nodes and that also made no difference. My thought was if this was such a significant problem we’d find out soon enough because the GI installation would fail. Hint – it doesn’t fail so ignore this ‘failure’ if you see it.
Click Yes:
On the Summary screen, click Install:
The installation thunders along:
Until a window pops up inviting you to run a root.sh script on both nodes.
Note, it is important that you run this script to completion in the order of the nodes listed, i.e. racnode1 first and racnode2 second.
The root.sh scripts take quite a while to run and their output is quite verbose. You can see what it looks like here. Once the root.sh script completes on racnode2, click OK. The installation will continue. Again, this part of the installation takes some time:
When the installation completes, you’ll see this screen:
That’s Grid Infrastructure installed. Next, install the database software.
Task #3: Run the Database Installer.
Using FileZilla or your favorite file copy utility, copy the 2 Oracle Database software zip files to racnode1 and unzip them.
Note, the database software is installed by the oracle user.
With the display setup correctly, invoke the database installer:
[oracle@racnode1 database]$ ./runInstaller
Uncheck the I wish to receive security updates via My Oracle Support option, then click Next:
Choose to remain uninformed by clicking Yes:
Select the Install database software only option, then click Next:
Select the Oracle Real Application Clusters database installation option, then click Next:
Make sure both nodes are selected. If you want the installer to re-do the SSH (user equivalency) setup, click the SSH connectivity button. When complete, click Next:
Select your preferred language:
Select Enterprise Edition, then click Next:
Use the following values to populate the next screen, then click Next:
Field
Value
Oracle base
/u01/app/oracle
Software location
/u01/app/oracle/product/12.1.0/dbhome_1
The next screen should pre-populate with the correct OS groups. If so, click Next:
OS Group Privilege
OS Group
Database Administrator (OSDBA) group
dba
Database Operator (OSOPER) group (Optional)
oper
Database Backup and Recovery (OSBACKUPDBA) group
backupdba
Data Guard administrative (OSDGDBA) group
dgdba
Encryption Key Management administrative (OSKMDBA) group
kmdba
As with the Grid Infrastructure installation, the installer complains about swap space. We can safely ignore this prerequisite check failure by checking Ignore All and clicking Next:
Click Yes:
On the Summary screen, click Install:
The familiar progress bar screen appears:
After a short while, this window appears inviting you to run the root.sh script on racnode1 and racnode2. Always run the root.sh script to completion in the order the nodes are listed. When complete, click OK:
The root.sh script output looks like this:
[root@racnode1 ~] cd /u01/app/oracle/product/12.1.0/dbhome_1
[root@racnode1 dbhome_1]# ./root.sh
Performing root user operation.
The following environment variables are set as:
ORACLE_OWNER= oracle
ORACLE_HOME= /u01/app/oracle/product/12.1.0/dbhome_1
Enter the full pathname of the local bin directory: [/usr/local/bin]:
The contents of "dbhome" have not changed. No need to overwrite.
The contents of "oraenv" have not changed. No need to overwrite.
The contents of "coraenv" have not changed. No need to overwrite.
Entries will be added to the /etc/oratab file as needed by
Database Configuration Assistant when a database is created
Finished running generic part of root script.
Now product-specific root actions will be performed.
[root@racnode2 ~] cd /u01/app/oracle/product/12.1.0/dbhome_1
[root@racnode2 dbhome_1]# ./root.sh
Performing root user operation.
The following environment variables are set as:
ORACLE_OWNER= oracle
ORACLE_HOME= /u01/app/oracle/product/12.1.0/dbhome_1
Enter the full pathname of the local bin directory: [/usr/local/bin]:
The contents of "dbhome" have not changed. No need to overwrite.
The contents of "oraenv" have not changed. No need to overwrite.
The contents of "coraenv" have not changed. No need to overwrite.
Entries will be added to the /etc/oratab file as needed by
Database Configuration Assistant when a database is created
Finished running generic part of root script.
Now product-specific root actions will be performed.
Finally, the successful completion screen is displayed:
We now have everything we need in place to create an Oracle RAC database. Obviously it’s important to test and verify that what you have installed is actually up and running correctly. We’ll run a series of tests after we create a RAC database in Part 11. See you there!
If you have any comments or questions about this post, please use the Contact form here.
The last major hurdle we need to overcome before installing some Oracle software and building a RAC database is to configure shared storage.
Shared storage is the key to making Oracle RAC work. The same storage concurrently visible to and updatable from multiple servers really is the heart of the system.
Enterprise class dedicated SAN or NAS filers are very cool but very expensive. What we need is a scaled down economical solution which does the same thing. The answer? Openfiler!
Download the Openfiler 2.99.1 ISO and burn it to a DVD using ImgBurn. Then use the DVD to boot the shared storage server computer. Installing Openfiler is very simple and similar to installing Oracle Linux. Since we’re not installing Openfiler in a virtual environment, it’s a bit tricky to capture screen shots of a bare metal install. So I’ll have to rely on a text description. Fear not, this install is ridiculously easy and completes in 8 simple steps.
Task #1a: Press Enter to install in graphical mode.
After the server boots from DVD, simply hit the Enter key to get the install started.
Task #1b: Click Next.
Once the GUI comes up, just hit the Enter key to continue.
Task #1c: Select keyboard.
Select your desired keyboard layout from the menu and click Next.
Task #1d: Select installation drive.
The shared storage server has 4 drives in total, identified as sda, sdb, sdc and sdd:
Device
Physical Drive
Usage
sda
Samsung 850 EVO 120GB 2.5-Inch SATA III Internal SSD
Openfiler OS
sdb
Western Digital 300 GB VelociRaptor SATA 3 Internal HD
Data
sdc
Western Digital 300 GB VelociRaptor SATA 3 Internal HD
Data
sdd
Samsung 850 EVO 120GB 2.5-Inch SATA III Internal SSD
Data
Select the option to Remove all partitions on selected drives and create default layout. Leave the first drive, sda, checked and uncheck all the other drives. Finally, select sda as the boot installation drive and click Next.
Task #1e: Configure network devices.
The shared storage server has two NICs. Click Edit and configure them with these values:
Interface
Field
Value
eth0
IP Address
200.200.10.6
Prefix (Netmask)
255.255.255.0
eth1
IP Address
200.200.20.6
Prefix (Netmask)
255.255.255.0
Note, do not forget to disable IPv6 support and ensure both NICs are Active on Boot.
Set the remaining networking fields manually with these values, then click Next:
Field
Value
Hostname
openfiler.mynet.com
Gateway
200.200.10.1
Primary DNS
200.200.10.1
Secondary DNS
8.8.8.8
Task #1f: Choose timezone.
Select your preferred timezone from the menu. Leave System clock uses UTC checked. Click Next.
Task #1g: Choose root password.
Enter a root password twice. Write it down somewhere safe. Now! Don’t argue, just do it. Click Next.
Task #1h: Install the software.
A familiar looking left to right progress bar shows the installation’s progress. It only takes about a minute, then you’ll see a congratulations screen. Click Reboot.
Note, this reboot sometimes sticks at a completely blank screen. Don’t be alarmed. If this happens hard power off the machine, then power it back on and all will be well.
When the machine comes back up, it will display the URL for the Openfiler console, which in our case will be, https://200.200.10.6:446/. The default username is “openfiler” and the default password is, wait for it, “password”. No quotes of course.
Once you get logged in, you’ll see a status screen similar to this:
Told you it was easy! 😀
Task #2: Configure Openfiler.
Installing Openfiler was quite easy. Configuring the storage it manages is a little involved. There are 8 steps to this. Pay close attention. Maximum attention to detail is needed here!!!
Task #2a: Edit /etc/hosts.
Add the following to the openfiler server’s /etc/hosts file:
On the status screen, click Services to go to the Manage Services page:
For iSCSI target, click Enable and Start. For iSCSI Initiator, click Enable and start. The screen should now look like this:
Task #2c: Network Access Configuration.
Click on System to display the networking configuration. Use the following values to to populate the Network Access Configuration section:
Name
Network/Host
Netmask
Type
racnode1-storage
200.200.20.11
255.255.255.255
Share
racnode2-storage
200.200.20.12
255.255.255.255
Share
The screen should now look similar to this:
Note, it’s tempting to put the hostname in the Network/Host column especially if you’ve added the correct entries in the Openfiler server’s /etc/hosts file. However, don’t! It doesn’t work when you come to discover the storage from the Oracle RAC nodes. You MUST use the IP address!
Task #2d: Partition the Disk.
Click Volumes to display the Volume Group Management screen which currently shows no volume groups having been created:
Click Block Devices which shows the physical disks attached to the server:
The /dev/sda device is where Openfiler is installed. For the first Oracle RAC cluster, we’ll be using the other Samsung SSD, /dev/sdd. Click on /dev/sdd:
In the Create a partition in /dev/sdd section, use these values then click Create:
Field
Value
Mode
Primary
Partition Type
Physical Volume
Starting Cylinder
1
Ending Cylinder
14593
The screen should now look like this:
Note, despite the values you used, Openfiler insists on keeping 5% of the disk free.
Task #2e: Create a Volume Group.
Click Volume Groups. Enter a new volume group name, cluster1-vg and check the box to select the /dev/sdd1 partition:
Then click Add volume group and the screen display changes to this:
Task #2f: Create Logical Volumes.
The logical volumes we create will eventually end up as ASM disks visible to the Oracle RAC node servers.Click Add Volume.
We can now use this screen to create volumes within the cluster1-vg volume group. Use the following values to create 10 x 10 GB volumes:
Volume Name
Volume Description
Required Space (MB)
Filesystem / Volume type
c1vg-vol01
Cluster1 VG Volume #1
10240
block (iSCSI,FC,etc)
c1vg-vol02
Cluster1 VG Volume #2
10240
block (iSCSI,FC,etc)
c1vg-vol03
Cluster1 VG Volume #3
10240
block (iSCSI,FC,etc)
c1vg-vol04
Cluster1 VG Volume #4
10240
block (iSCSI,FC,etc)
c1vg-vol05
Cluster1 VG Volume #5
10240
block (iSCSI,FC,etc)
c1vg-vol06
Cluster1 VG Volume #6
10240
block (iSCSI,FC,etc)
c1vg-vol07
Cluster1 VG Volume #7
10240
block (iSCSI,FC,etc)
c1vg-vol08
Cluster1 VG Volume #8
10240
block (iSCSI,FC,etc)
c1vg-vol09
Cluster1 VG Volume #9
10240
block (iSCSI,FC,etc)
c1vg-vol10
Cluster1 VG Volume #10
10240
block (iSCSI,FC,etc)
When you’re done, your screen should look similar to this:
Task #2g: Create iSCSI Targets for Logical Volumes.
An iSCSI target is a way of identifying the logical volumes. Since we have 10 logical volumes, we’ll have 10 iSCSI targets. An iSCSI target is actually identified by a Target IQN or iSCSI Qualified Name. The following procedure will need to be repeated for each of the 10 logical volumes.
Click Volumes, then iSCSI Targets.
The default Target IQNs are not very meaningful, so we’ll add some meaning by changing their names according to the values below:
Logical Volume Name
Target IQN
c1vg-vol01
iqn.2006-01.com.openfiler:c1vg-vol01
c1vg-vol02
iqn.2006-01.com.openfiler:c1vg-vol02
c1vg-vol03
iqn.2006-01.com.openfiler:c1vg-vol03
c1vg-vol04
iqn.2006-01.com.openfiler:c1vg-vol04
c1vg-vol05
iqn.2006-01.com.openfiler:c1vg-vol05
c1vg-vol06
iqn.2006-01.com.openfiler:c1vg-vol06
c1vg-vol07
iqn.2006-01.com.openfiler:c1vg-vol07
c1vg-vol08
iqn.2006-01.com.openfiler:c1vg-vol08
c1vg-vol09
iqn.2006-01.com.openfiler:c1vg-vol09
c1vg-vol10
iqn.2006-01.com.openfiler:c1vg-vol10
After changing the Target IQN, click the Add button. The screen display changes to this:
Ignore the Settings section as we won’t be changing any of those values.
Context is extremely important at this stage and you need to make sure your focus is on the correct object. Therefore, make sure you select the Target IQN you just created from the Select iSCSI Target menu. If it isn’t, as will be the case after you create the first one, select the Target IQN you need, then click the Change button. Selecting the Target IQN from the menu isn’t sufficient. You MUST click the Change button to make it the ‘active’ Target IQN.
Click LUN Mapping. The screen changes to this:
Click the Map button next to the Logical Volume you wish to map the currently active Target IQN to. First time through, the Map New LUN to Target will show “iqn.2006-01.com.openfiler:c1vg-vol01” and click the Map button on the line beginning, Cluster 1 VG Volume #1. The screen changes to this:
Next, click Network ACL (Access Control List). Here you define which hosts (Oracle RAC nodes) can have access to this Target IQN. Since Openfiler only knows about two nodes on the storage network (200.200.20.x), racnode1-storage and racnode2-storage are listed.
Use the Access menu to change Deny to Allow. Click Update. The screen will refresh but nothing else happens which can be a little deceiving. As long as you hit Update and the screen refreshes (don’t blink or you’ll miss it), you’re fine. Click Target Configuration to go back to this screen and repeat the procedure for c1vg-vol02 through c1vg-vol10.
Task #2h: Review iSCSI targets.
Once you’re done creating a Target IQN, LUN Mapping and Network ACL for all 10 logical volumes, you can check your handiwork. Return to the home page by clicking Status, then click iSCSI Targets:
Task #3: Configure iSCSI Targets in Oracle Linux.
Our attention now turns to making the iSCSI targets visible and usable on the Oracle RAC node servers. To do that, we use the iSCSI initiator client which we installed back in Part 7. There are 6 steps in all, so here we go.
Task #3a: Verify iSCSI client (ALL nodes).
Verify the iSCSI client is installed on both Oracle RAC nodes:
Then configure the iSCSI client to start on boot up on both Oracle RAC nodes:
[root@racnode1 ~]# chkconfig iscsid on
[root@racnode1 ~]# chkconfig iscsi on
[root@racnode2 ~]# chkconfig iscsid on
[root@racnode2 ~]# chkconfig iscsi on
Trying to start the services has little effect at this stage, but running the iscsiadm discovery command makes everything spring into life. You’ll see.
Perform this step on all nodes in the cluster.
Task #3b: Discover iSCSI targets (ALL nodes).
The iscsiadm discovery command starts the iscsid service and then discovers the iSCSI targets coming from Openfiler.
Note, we’re using the portal openfiler-storage (the -p option) which is on the storage network (200.200.20.x). However, the discovery output lists targets on the storage network AND the public network (200.200.10.x). I think this is an Openfiler quirk/bug. Needless to say, we’re only interested in the targets on the storage network.
Repeat this step on ALL nodes in the cluster.
Task #3c: Login to iSCSI targets (ALL nodes).
Using the iSCSI target name returned by the discovery command, use the iscsiadm command to log into each iSCSI target, making sure you reference the storage network IP addresses (200.200.20.x):
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol01 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol01, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol01, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol02 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol02, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol02, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol03 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol03, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol03, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol04 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol04, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol04, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol05 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol05, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol05, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol06 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol06, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol06, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol07 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol07, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol07, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol08 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol08, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol08, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol09 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol09, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol09, portal: 200.200.20.6,3260] successful.
[root@racnode1 scripts]# iscsiadm -m node -T iqn.2006-01.com.openfiler:c1vg-vol10 -p 200.200.20.6 -l
Logging in to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol10, portal: 200.200.20.6,3260] (multiple)
Login to [iface: default, target: iqn.2006-01.com.openfiler:c1vg-vol10, portal: 200.200.20.6,3260] successful.
This has the effect of Oracle Linux recognizing the iSCSI targets as disk devices and assigns them a device name to each one. A partial fdisk -l listing confirms what has happened:
To ensure the iscsi client on each node logs into the iSCSI targets each time the system is booted or the iscsid service is restarted, automatic login must be configured. To do that, run these commands, again making sure you reference the storage network IP addresses (200.200.20.x):
Note, these commands do not produce any output. Which is a little unfriendly, but don’t be scared!
Repeat this step on ALL nodes in the cluster.
Task #3e: Partition the disks (ONE node only).
Now that the iSCSI targets coming from the Openfiler shared storage server are visible as disk devices, we need to partition them. Each drive will have a single partition. Repeat the following command for /dev/sda through /dev/sdj:
[root@racnode1 scripts]# fdisk /dev/sda
Device contains neither a valid DOS partition table, nor Sun, SGI or OSF disklabel
Building a new DOS disklabel with disk identifier 0xdce55f18.
Changes will remain in memory only, until you decide to write them.
After that, of course, the previous content won't be recoverable.
Warning: invalid flag 0x0000 of partition table 4 will be corrected by w(rite)
WARNING: DOS-compatible mode is deprecated. It's strongly recommended to
switch off the mode (command 'c') and change display units to
sectors (command 'u').
Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-10240, default 1):
Using default value 1
Last cylinder, +cylinders or +size{K,M,G} (1-10240, default 10240):
Using default value 10240
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
Note, partitioning the disk devices is only done from ONE node.
The discovery of the iSCSI targets by Oracle Linux just happened to map c1vg-vol01-lun-0 to /dev/sda, c1vg-vol02-lun-0 to /dev/sdb and so on. After the next reboot, these mappings may change. We need to avoid that and instead have a consistent iSCSI target to disk device mapping each time the server starts. This is achieved using an Oracle Linux device mapping utility called udev.
Note, using Oracle’s ASM library driver (Task #4 Steps #1 & #2) can also achieve the same thing udev does, i.e. a consistent mapping between iSCSI targets and disk devices.
Thus, configuring udev is, strictly speaking, unnecessary.
That said, you can use udev or ASMlib or both if you want to. This procedure uses both, just to demonstrate both ways of doing the same thing, one with Oracle’s help (ASMlib) and one without (udev).
Fortunately, each iSCSI target has a unique ID which remains consistent across reboots. So we need to find out what that is. Repeat the following command for /dev/sda through /dev/sdj, making a note of the output for each disk device:
Next, create a udev rules script called 99-oracle-iscsi.rules and place it in the /etc/udev/rules.d directory where udev can find it.
Note, the 99 signifies the order in which the script will run. Lower numbered scripts run first. The “oracle-iscsi” part is just our way of identifying the script as being related to Oracle. The “.rules” part makes it a rules script.
The rules script will contain 10 lines, one for each disk device. Each line will look like this:
Only two entries need to be changed for each disk device. The RESULT parameter needs to be set to the ID output by the scsi_id command for each disk device. The NAME parameter needs to be set to the naming convention you wish to follow. In the above example, I’m using iscsi/asm-disk01 for /dev/sda. The NAME parameter for /dev/sdb will therefore be, iscsi/asm-disk02 and so on.
Once the file is saved, it should have this ownership and permissions:
[root@racnode1 rules.d]# pwd
/etc/udev/rules.d
[root@racnode1 rules.d]# ls -l | grep oracle
-rw-r--r-- 1 root root 2030 Dec 16 17:06 99-oracle-iscsi.rules
To test the udev rules script, run this command for each disk device (/block/sda/sda1 through /block/sdj/sdj1):
[root@racnode1 ~]# udevadm test /block/sda/sda1
Then check the /dev/iscsi directory and the persistent iscsi disk devices should be listed:
[root@racnode1 iscsi]# pwd
/dev/iscsi
[root@racnode1 iscsi]# ls -l
brw-rw---- 1 root disk 8, 1 Dec 16 17:12 asm-disk01
brw-rw---- 1 root disk 8, 17 Dec 16 17:16 asm-disk02
brw-rw---- 1 root disk 8, 33 Dec 16 17:19 asm-disk03
brw-rw---- 1 root disk 8, 49 Dec 16 17:20 asm-disk04
brw-rw---- 1 root disk 8, 65 Dec 16 17:20 asm-disk05
brw-rw---- 1 root disk 8, 81 Dec 16 17:20 asm-disk06
brw-rw---- 1 root disk 8, 97 Dec 16 17:20 asm-disk07
brw-rw---- 1 root disk 8, 113 Dec 16 17:21 asm-disk08
brw-rw---- 1 root disk 8, 129 Dec 16 17:21 asm-disk09
brw-rw---- 1 root disk 8, 145 Dec 16 17:21 asm-disk10
Repeat this step for ALL nodes in the cluster.
Now reboot ALL nodes.
Task #4: Configure Oracle ASMlib.
Only a couple of steps to follow here to configure ASMLib.
[root@racnode1 ~]# oracleasm configure -i
Configuring the Oracle ASM library driver.
This will configure the on-boot properties of the Oracle ASM library
driver. The following questions will determine whether the driver is
loaded on boot and what permissions it will have. The current values
will be shown in brackets ('[]'). Hitting <ENTER> without typing an
answer will keep that current value. Ctrl-C will abort.
Default user to own the driver interface []: grid
Default group to own the driver interface []: asmadmin
Start Oracle ASM library driver on boot (y/n) [n]: y
Scan for Oracle ASM disks on boot (y/n) [y]: y
Writing Oracle ASM library driver configuration: done
[root@racnode1 ~]# oracleasm init
Creating /dev/oracleasm mount point: /dev/oracleasm
Loading module "oracleasm": oracleasm
Configuring "oracleasm" to use device physical block size
Mounting ASMlib driver filesystem: /dev/oracleasm
Repeat this step for ALL nodes in the cluster.
Task #4b: Create ASM disks (ONE node only).
Using the oracleasm createdisk command, create an ASM disk for each iSCSI device created in Task #3 Step #6 (/dev/iscsi/asm-disk01 through /dev/iscsi/asm-disk10):
Note, the remnants of ASM disk headers can be left behind even when a drive has been re-partitioned. This can prevent the new ASM disk from being instantiated. You will see this when that happens:
[root@racnode1 disks]# oracleasm createdisk ASMDISK01 /dev/iscsi/asm-disk01
Device "/dev/iscsi/asm-disk01" is already labeled for ASM disk "ASMDISK01"
If you see this, you need to destroy the old disk header. A simple way to do that is documented here. Once the header is wiped, the oracleasm createdisk command can create a new one. Once all the ASM disks have been created, they can be listed using this command:
Note, the user and group ownership is grid:asmadmin:
[root@racnode1 disks]# pwd
/dev/oracleasm/disks
[root@racnode1 disks]# ls -l
brw-rw---- 1 grid asmadmin 8, 1 Dec 16 18:31 ASMDISK01
brw-rw---- 1 grid asmadmin 8, 17 Dec 16 18:23 ASMDISK02
brw-rw---- 1 grid asmadmin 8, 33 Dec 16 18:23 ASMDISK03
brw-rw---- 1 grid asmadmin 8, 49 Dec 16 18:23 ASMDISK04
brw-rw---- 1 grid asmadmin 8, 65 Dec 16 18:24 ASMDISK05
brw-rw---- 1 grid asmadmin 8, 81 Dec 16 18:26 ASMDISK06
brw-rw---- 1 grid asmadmin 8, 97 Dec 16 18:26 ASMDISK07
brw-rw---- 1 grid asmadmin 8, 113 Dec 16 18:26 ASMDISK08
brw-rw---- 1 grid asmadmin 8, 129 Dec 16 18:26 ASMDISK09
brw-rw---- 1 grid asmadmin 8, 145 Dec 16 18:26 ASMDISK10
Note, you can determine the mapping between the ASM disks and the block devices by examining the major and minor device numbers. For example, the major number of ASMDISK01 is 8 and its minor number is 1. Similarly, the major number for ASMDISK02 is also 8 and its minor number is 17. Compare the major and minor numbers with a directory listing of the block devices:
[root@racnode1 disks]# ls -l /dev/sd*1
brw-rw---- 1 root disk 8, 1 Jan 26 13:55 /dev/sda1
brw-rw---- 1 root disk 8, 17 Jan 26 13:55 /dev/sdb1
brw-rw---- 1 root disk 8, 33 Jan 26 13:55 /dev/sdc1
brw-rw---- 1 root disk 8, 49 Jan 26 13:55 /dev/sdd1
brw-rw---- 1 root disk 8, 65 Jan 26 13:55 /dev/sde1
brw-rw---- 1 root disk 8, 81 Jan 26 13:55 /dev/sdf1
brw-rw---- 1 root disk 8, 97 Jan 26 13:55 /dev/sdg1
brw-rw---- 1 root disk 8, 113 Jan 26 13:55 /dev/sdh1
brw-rw---- 1 root disk 8, 129 Jan 31 14:05 /dev/sdi1
brw-rw---- 1 root disk 8, 145 Jan 26 13:55 /dev/sdj1
So we can see that ASMDISK01 maps to /dev/sda1, ASMDISK02 maps to /dev/sdb1 and so on.
The other node will not see the ASM disks immediately. Use this command to instantiate the disks on racnode2:
[root@racnode2 ~]# oracleasm scandisks
Reloading disk partitions: done
Cleaning any stale ASM disks...
Scanning system for ASM disks...
Instantiating disk "ASMDISK08"
Instantiating disk "ASMDISK10"
Instantiating disk "ASMDISK05"
Instantiating disk "ASMDISK09"
Instantiating disk "ASMDISK07"
Instantiating disk "ASMDISK01"
Instantiating disk "ASMDISK04"
Instantiating disk "ASMDISK02"
Instantiating disk "ASMDISK03"
Instantiating disk "ASMDISK06"
Note, for Oracle Database 12c there is a new mechanism called the ASM Filter Driver. Obviously we haven’t used that here. Maybe when we re-visit this section in the future we can migrate to using that.
So there you have it. Shared storage configured and ready to go. Everything is now in place for a successful install of the Grid Infrastructure and database software. See you in Part 10 for that.
If you have any comments or questions about this post, please use the Contact form here.
It’s acronym time boys and girls! Before we get to the fun part of configuring the shared storage in Part 9, we need setup a few more slightly less fun items.
Secure Shell or SSH and Domain Name System or DNS.
After that we take the thrill ride of running the Cluster Verification Utility or CVU for the first time. This stuff is pretty simple, so let’s crack on.
The RAC node from where you install the Grid Infrastructure and Oracle Database software effectively becomes the node from which you perform ALL subsequent installs. Hence, the software only needs to reside on that one server and is copied from that server to all the other nodes in the cluster as part of the installation process. Consequently, this node needs remote access to the same user account on all the other nodes without being prompted for a password.
User equivalency is effectively the setup of identical user accounts on all nodes. Identical means the same username, UID, GIDs and password. The accounts which require user equivalency are oracle and grid.
The installation of the Grid Infrastructure and database software can setup SSH for you, but you’ll need SSH configured before you get to that stage in order for the CVU to weave its magic on remote nodes. That’s why we’re setting it up ahead of the software installs. There are 8 steps in total.
Task #1a: Login as oracle and create an ssh directory (all nodes).
[oracle@racnode1 ~]$ ssh-keygen -t rsa
Generating public/private rsa key pair.
Enter file in which to save the key (/home/oracle/.ssh/id_rsa):
Enter passphrase (empty for no passphrase): <just press the Enter key>
Enter same passphrase again: <just press the Enter key>
Your identification has been saved in /home/oracle/.ssh/id_rsa.
Your public key has been saved in /home/oracle/.ssh/id_rsa.pub.
The key fingerprint is:
3a:46:e5:58:c8:c1:f9:de:63:4f:d7:e4:29:d9:aa:b7 oracle@racnode1.mynet.com
The key's randomart image is:
<snipped>
*** Repeat for all nodes in the cluster. ***
Task #1c: Generate DSA Keys (all nodes).
[oracle@racnode1 ~]$ ssh-keygen -t dsa
Generating public/private dsa key pair.
Enter file in which to save the key (/home/oracle/.ssh/id_dsa):
Enter passphrase (empty for no passphrase): <just press the Enter key>
Enter same passphrase again: <just press the Enter key>
Your identification has been saved in /home/oracle/.ssh/id_dsa.
Your public key has been saved in /home/oracle/.ssh/id_dsa.pub.
The key fingerprint is:
32:40:a2:1f:e6:85:23:2f:80:e6:a0:15:9f:5e:01:9e oracle@racnode1.mynet.com
The key's randomart image is:
<snipped>
*** Repeat for all nodes in the cluster. ***
Task #1d: Create the authorized_keys file (all nodes).
[oracle@racnode1 ~]$ touch ~/.ssh/authorized_keys
[oracle@racnode1 ~]$ ls -l .ssh
-rw-r--r-- 1 oracle oinstall 0 Dec 13 15:55 authorized_keys
-rw------- 1 oracle oinstall 668 Dec 13 15:47 id_dsa
-rw-r--r-- 1 oracle oinstall 617 Dec 13 15:47 id_dsa.pub
-rw------- 1 oracle oinstall 1675 Dec 13 15:46 id_rsa
-rw-r--r-- 1 oracle oinstall 409 Dec 13 15:46 id_rsa.pub
*** Repeat for all nodes in the cluster. ***
Task #1e: Capture Key Fingerprints to the authorized_keys file (first node).
First, ssh into the node you’re logged into and capture the output to authorized_keys.
[oracle@racnode1 ~]$ cd .ssh
[oracle@racnode1 .ssh]$ ssh racnode1 cat /home/oracle/.ssh/id_rsa.pub >> authorized_keys
The authenticity of host 'racnode1 (200.200.10.11)' can't be established.
RSA key fingerprint is 35:a5:78:68:36:c3:c2:42:f5:df:da:5f:2c:56:2b:a7.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'racnode1,200.200.10.11' (RSA) to the list of known hosts.
oracle@racnode1's password: <enter the oracle password>
[oracle@racnode1 .ssh]$ ssh racnode1 cat /home/oracle/.ssh/id_dsa.pub >> authorized_keys
*** Repeat these steps by using ssh to connect to every other node in the cluster, again capturing the output into authorized_keys. ***
Task #1f: Copy the authorized_keys file to every node.
The authorized_keys file on the first node now contains all the data which all the other nodes need. Use scp to copy it to the other nodes.
*** Repeat the scp to each node in the cluster. ***
Task #1g: Secure the authorized_keys file (all nodes).
[oracle@racnode1 .ssh]$ chmod 600 authorized_keys
*** Repeat for all nodes in the cluster. ***
Task #1h: Test passwordless connectivity both ways between all nodes.
[oracle@racnode1 .ssh]$ ssh racnode2 date
Sun Dec 13 16:30:13 CST 2015
[oracle@racnode2 .ssh]$ ssh racnode1 date
The authenticity of host 'racnode1 (200.200.10.11)' can't be established.
RSA key fingerprint is 35:a5:78:68:36:c3:c2:42:f5:df:da:5f:2c:56:2b:a7.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'racnode1,200.200.10.11' (RSA) to the list of known hosts.
Sun Dec 13 16:30:32 CST 2015
[oracle@racnode2 .ssh]$ ssh racnode1 date
Sun Dec 13 16:30:35 CST 2015
Once you have SSH setup (also known as “user equivalency”) for the oracle user, repeat the same 8 steps for the grid user.
Task #2: Configure DNS.
For the most part you can get away with using /etc/hosts files to resolve hostnames. However, with the introduction of Single Client Access Name (SCAN) in Oracle Database 11g Release 2, things got a little more complicated. A cluster SCAN must now resolve to 3 IP addresses, so a /etc/hosts files won’t help.
There is a ‘smoke and mirrors’ way to get a tweaked nslookup script to return 3 IP addresses, but why not just configure DNS properly? In a production environment, there would already be a dedicated DNS server, but in our environment using a whole VM just for DNS would be excessive to say the least. The smart way to go would be to use a server that’s used for something else and is always on. Top of the list is our Oracle Enterprise Manager Cloud Control server, oraemcc.mynet.com (IP: 200.200.10.16). It won’t mind and DNS won’t interfere with its other duties.
Whenever I’ve asked a production Network Administrator to set this up for me, I’ve sometimes (not always) got the eye roll and look of exasperation. Like I’m asking them to split the atom or something. In truth, setting up DNS isn’t that difficult, but it is fiddly. One semi colon, bracket or period out of place and the whole thing doesn’t work and more infuriatingly, won’t tell you why. So, with your keenest eye on the prize, let’s configure DNS using these 7 steps.
Task #2a: Install the bind package.
DNS functionality is provided by installing the bind package. This can be done using yum:
The installation of bind will create the file /etc/named.conf. It will contain multiple lines, most of which you will not need to edit. Before you start to edit this file, take a copy of it in case you need to start over. You will also need to have several pieces of information to hand to correctly edit this and subsequent files. They are:
Information
Value
Hostname of your DNS Server
oraemcc.mynet.com
IP Address of your DNS Server
200.200.10.16
IP Address of your ISP’s DNS Servers
8.8.8.8 & 8.8.4.4
Internal network domain name
mynet.com
The items which need to be edited are shown in red in this example /etc/named.conf file. Copy this file to your own system and make any necessary changes. When you edit your file be VERY CAREFUL with the double quotes and periods. Missing something seemingly so innocent will make you wish you hadn’t. Three important points about this file.
The forwarders line is there so server names which are not on your internal network can still be resolved. These servers would likely be on the internet, so your ISP’s DNS servers handle those lookups instead.
The zone “10.200.200.in-addr.arpa.” line is there to facilitate reverse lookups. The 10.200.200 part describes the first three octets of the IP address your DNS server is on, but in reverse order. Our DNS server has the IP address 200.200.10.16, hence 10.200.200.
The ownership and file permissions are crucial. Get these wrong and DNS won’t startup.
[root@oraemcc ~]# ls -l /etc/named.conf
-rw-r----- 1 root named 1114 Aug 19 17:15 /etc/named.conf
Notice the file is owned by root, with named as the group. The file permissions are 640. If your file does not match this, then run these commands:
[root@oraemcc ~]# chgrp named /etc/named.conf
[root@orasemcc ~]# chmod 640 /etc/named.conf
Task #2c: Create /var/named/mynet.com.zone file.
The next step is to create your /var/named/mynet.com.zone file which will resolve hostnames on your domain, mynet.com.
The items which need to be edited are shown in red in this example /var/named/mynet.com.zone file. Copy this file to your own system and make any necessary changes. Ensure this file has the correct group ownership and file permissions:
[root@oraemcc ~]# chgrp named /var/named/mynet.com.zone
[root@oraemcc ~]# chmod 640 /var/named/mynet.com.zone
[root@oraemcc ~]# ls -l /var/named/mynet.com.zone
-rw-r----- 1 root named 1133 Dec 21 12:44 /var/named/mynet.com.zone
This step creates the /var/named/10.200.200.in-addr.arpa file which does reverse lookups for addresses on the 200.200.10.x network. It’s not strictly necessary to implement reverse lookups, but since we’re in the neighborhood we may as well.
The items which need to be edited are shown in red in this example /var/named/10.200.200.in-addr.arpa file. Copy this file to your own system and make any necessary changes. Again, ensure the file has the correct group and file permissions.
[root@oraemcc ~]# chgrp named /var/named/10.200.200.in-addr.arpa
[root@oraemcc ~]# chmod 640 /var/named/10.200.200.in-addr.arpa
[root@oraemcc ~]# ls -l /var/named/10.200.200.in-addr.arpa
-rw-r----- 1 root named 916 Dec 21 12:26 10.200.200.in-addr.arpa
Task #2e: Edit /etc/resolv.conf.
Create or edit a /etc/resolv.conf file on each server which will use your DNS setup. These are the only lines which should be in this file:
search mynet.com
nameserver 200.200.10.16
It is important that is file be identical across all the RAC server nodes as it’s one of the things the CVU checks for. This is the file which Network Manager trashes each time networking starts up on a server. So it’s important that this file remain intact and consistent.
Task #2f: Start the DNS service.
[root@oraemcc ~]# service named start
Starting named: [ OK ]
Checking the status of DNS generates some interesting output:
[root@oraemcc ~]# service named status
version: 9.8.2rc1-RedHat-9.8.2-0.37.rc1.el6_7.5
CPUs found: 2
worker threads: 2
number of zones: 21
debug level: 0
xfers running: 0
xfers deferred: 0
soa queries in progress: 0
query logging is OFF
recursive clients: 0/0/1000
tcp clients: 0/100
server is up and running
named (pid 9931) is running...
Task #2g: Test DNS is working.
We have configured DNS and have started the service, so now’s the time to make sure it actually works.
The Cluster Verification Utility (CVU) script, runcluvfy.sh, is included with the 12c Grid Infrastructure software. You should have already downloaded and unzipped these files. If not, refer to this prior step.
Login to racnode1 as the grid user, locate the CVU script, then run the script with the parameters shown:
This will run the CVU, echo its output to the screen and copy the output to a file called cvu.out.
The vast majority of the checks passed. Therefore, it only really makes sense to highlight the three which failed and more importantly what to do about them.
Failure #1: IPv6 Addressing.
The first error has to do with IPv6 addressing. Since we’re not using IPv6 (one day perhaps), we could ignore this error. However, IPv6 interface information will come up in the Grid Infrastructure installation which only serves to confuse matters. I therefore prefer to remove the IPv6 references. These are the errors:
Check: TCP connectivity of subnet "2002:4b49:1933::"
Source Destination Connected?
------------------------------ ------------------------------ ----------------
racnode1 : 2002:4b49:1933:0:221:f6ff:fe04:4298 racnode1 : 2002:4b49:1933:0:221:f6ff:fe04:4298 passed
racnode2 : 2002:4b49:1933:0:221:f6ff:fed2:45a0 racnode1 : 2002:4b49:1933:0:221:f6ff:fe04:4298 failed
ERROR:
PRVG-11850 : The system call "connect" failed with error "13" while executing exectask on node
"racnode2" Permission denied
racnode1 : 2002:4b49:1933:0:221:f6ff:fe04:4298 racnode2 : 2002:4b49:1933:0:221:f6ff:fed2:45a0 failed
ERROR:
PRVG-11850 : The system call "connect" failed with error "13" while executing exectask on node
"racnode1" Permission denied
racnode2 : 2002:4b49:1933:0:221:f6ff:fed2:45a0 racnode2 : 2002:4b49:1933:0:221:f6ff:fed2:45a0 passed
Result: TCP connectivity check failed for subnet "2002:4b49:1933::"
You can confirm the presence of IPv6 addressing by checking the output of the ifconfig command. For example:
The second failure reported in the CVU output had to do with Network Time Protocol (NTP). You will recall from the installation of Oracle Linux on racnode1, we chose not to synchronize the date and time over the network. That’s because I prefer to use Oracle’s Cluster Time Synchronization Service instead. These are the NTP errors reported by the CVU:
Starting Clock synchronization checks using Network Time Protocol(NTP)...
Checking existence of NTP configuration file "/etc/ntp.conf" across nodes
Node Name File exists?
------------------------------------ ------------------------
racnode2 yes
racnode1 yes
The NTP configuration file "/etc/ntp.conf" is available on all nodes
NTP configuration file "/etc/ntp.conf" existence check passed
No NTP Daemons or Services were found to be running
PRVF-5507 : NTP daemon or service is not running on any node but NTP
configuration file exists on the following node(s):
racnode2,racnode1
Result: Clock synchronization check using Network Time Protocol(NTP) failed
When Grid Infrastructure is installed, it will detect if NTP is running. If it is, then Cluster Time Synchronization Service is started in “observer mode”. If NTP is not running, then Cluster Time Synchronization Service is started in “active mode” and that’s what we want. However, if NTP is down but the installer sees the NTP configuration file /etc/ntp.conf, then it assumes NTP will spontaneously start by itself and the world will end as we know it. To avoid that from happening, we need to ensure NTP is down, stays down and can’t find its configuration file. We do that by renaming it:
[root@racnode1 ~]# service ntpd status
ntpd is stopped
[root@racnode1 ~]# chkconfig ntpd off
[root@racnode1 ~]# mv /etc/ntp.conf /etc/ntp.conf.ORIG
Thus the world is saved. Hurrah! ☑
Failure #3: Insufficient Swap Space.
I originally built the racnodes with 4 GB of RAM, but things ran a bit slow so I upped the RAM to 6 GB. One of the great things about virtualization is the ability to add more resource to a VM through software. Assuming the resources are physically present and available.
Note, Oracle VM cannot over allocate physical server memory. You can only allocate what you physically have available in the Oracle VM Server.
Since I originally had 4 GB of RAM allocated, I also configured 4 GB of swap space. Increasing the amount of RAM means you should also increase the amount of swap space. Here’s the official Oracle sizing table:
RAM
Swap Space
Between 1 GB & 2 GB
1.5x RAM
Between 2 GB & 16 GB
Same size as RAM
Greater than 16 GB
16 GB
Here is what the CVU reported regarding memory and swap space:
Check: Total memory
Node Name Available Required Status
------------ ------------------------ ------------------------ ----------
racnode2 5.8369GB (6120432.0KB) 4GB (4194304.0KB) passed
racnode1 5.8369GB (6120432.0KB) 4GB (4194304.0KB) passed
Result: Total memory check passed
Check: Swap space
Node Name Available Required Status
------------ ------------------------ ------------------------ ----------
racnode2 4GB (4194300.0KB) 5.8369GB (6120432.0KB) failed
racnode1 4GB (4194300.0KB) 5.8369GB (6120432.0KB) failed
Result: Swap space check failed
There are a few ways to increase the amount of swap space on a Linux server. If you have sufficient free disk you can add an additional swap partition. If the current swap space is in a logical volume, you could potentially re-size it to increase the amount of swap space. The quickest and simplest way to increase the available swap space is to add a swap file. The commands to do that are here. Since this is a non-production system, swapping won’t be a problem so we can safely ignore this failure. If it were a production system, we’d obviously do the right thing….blame the Systems Administrator! 👺
Well, that about wraps it up for SSH, DNS and CVU.
Now it’s time to start the process of building our first Oracle RAC system. To do that, we will need two Oracle Database 12c server nodes running Oracle Linux. This installment shows how to build that rather efficiently using a cool capability of Oracle VM.
In Oracle VM Manager, click on the Servers and VMs tab, highlight RACNODE1_VM, then click on the green arrow head Start icon to start the VM:
This will start the Oracle Linux 6 installation because this VM has the V52218-01.iso file loaded in its virtual CD/DVD drive and that device is listed first in the Boot Order. With the Install or Upgrade and existing system option highlighted, press Enter:
Skip the media test:
Click Next:
Choose your language, then click Next:
Choose your keyboard layout, then click Next:
Choose Basic Storage Devices, then click Next:
Click the ‘Yes, discard any data’ button, then click Next:
Enter the hostname “racnode1.mynet.com”, then click Configure Network:
If you’re following this series, then you should be getting good at configuring the network settings. So just use these values to configure the 3 network interfaces:
Field
eth0 (Public)
eth1 (Storage)
eth2 (Private)
IP Address
200.200.10.11
200.200.20.11
200.200.30.11
Netmask (auto-populates)
24
24
24
Gateway
200.200.10.1
200.200.20.1
N/A
Primary DNS
200.200.10.1
N/A
N/A
Secondary DNS
8.8.8.8
N/A
N/A
With the network configured, choose your timezone then click Next:
Choose a root password, then click Next:
Choose Use All Space, check Review and modify partitioning layout, then click Next:
Using the right pointing arrow key, move the Xen Virtual Block Device (40960 MB) over to the Install Target Devices pane, then click Next:
Edit the following screen using these values, then click Next:
Logical Volume
Size (MB)
lv_root
34312
lv_swap
6144
Click Format:
Click Write changes to disk:
Check Install boot loader on /dev/xvdb. Select Oracle Linux Server 6 (Label) and /dev/mapper/vg_racnode1-lv_root (Device). Click Next:
Select Database Server and Customize now, then click Next:
Use the following to complete the next screen, then click Next:
Category
Option
Base System
Leave the default options
Servers
Leave the default options
Web Services
Leave the defaults options
Databases
Uncheck everything
System Management
Leave the default options
Virtualization
Leave the default options
Desktops
Check everything
Applications
Check Internet Browser
Development
Leave the default options
UEK3 kernel repo
Leave the default options
Lanugages
Leave the default options
The Oracle Linux installation will start installing 1249 packages:
Eventually the installation will complete and you’ll see a rather splendid congratulations screen. Do NOT click Reboot yet:
Go back to Oracle VM Manager. Click on the Servers and VMs tab, highlight RACNODE1_VM, then click on the pencil icon to edit RACNODE1_VM. Click on the Disks tab. Click on the Eject a CDROM icon:
This will remove the V52218-01.iso from the CD/DVD drive, so when you reboot the Oracle Linux installation won’t start over. Click OK. Then return to the ‘congratulations’ screen and click Reboot. You may need to close the console session and open a new one. When the system comes back up, you’ll see this ‘welcome’ screen. Click Forward:
Accept the License Agreement and click Forward:
Choose the option to register at a later time and click Forward:
Click the ‘No thanks’ button, then click Forward:
Rather annoyingly, the harassment continues. Ignore the scaremongering and click Forward:
Don’t create a user here. We’ll be doing that later. Click Forward:
Set the system date and time if necessary. Uncheck the option to synchronize date and time over the network. We’ll be using Oracle’s Cluster Time Synchronization Service later on. If the time is wrong, then it’ll be Oracle’s fault not NTP’s fault. See what I did there? 🙂 Click Forward:
Uncheck Enable kdump, then click Finish:
Click the Yes to agree to a reboot:
Click OK to start the system reboot:
When the system comes back up, you’ll see the familiar login screen:
That’s Oracle Linux 6.6 installed. Time to configure it for Oracle Database 12c RAC duties.
Task #2: Configure Oracle Linux on racnode1.
For reference, the official Oracle Database 12c Release 1 RAC installation guide can be found here. Our installation has been simplified into the following 12 steps. That’s right, our very own 12 step programme. 🍺
Task #2a: Update Oracle Linux.
Let’s grab the latest and greatest packages and update the Linux kernel. To do that we need to run a yum update. To do that, we need to create a yum repository first. With the yum repository in place, run the update:
[root@racnode1 ~]# yum update
This will take a while. It will ask you to confirm progress in a couple of places, so don’t go to lunch or it’ll be waiting for you when you get back.
Note, this update will upgrade the version of Oracle Linux from 6.6 to 6.7.
Task #2b: Install the Oracle 12c Release 1 Pre-Installation Package.
This package makes a number of changes to Linux which aids in setting things up the way you want them. It’s actually on the Oracle Linux 6 DVD in /server/Packages and can be installed using this command:
A number of changes are made to the /etc/password and /etc/group files. The UIDs and GIDs it creates may not be what you want. After you have installed the package, you can use this script to correct the setup and add the additional users and groups you’ll need later.
As a final post package installation step, add these entries to the /etc/security/limits.conf file:
####################################
# for oracle user
####################################
oracle soft nofile 8192
oracle hard nofile 65536
oracle soft nproc 2048
oracle hard nproc 16384
oracle soft stack 10240
oracle hard stack 32768
oracle soft core unlimited
oracle hard core unlimited
oracle hard memlock 5500631
####################################
# for grid user
####################################
grid soft nofile 8192
grid hard nofile 65536
grid soft nproc 2048
grid hard nproc 16384
grid soft stack 10240
grid hard stack 32768
grid soft core unlimited
grid hard core unlimited
Now that the oracle and grid users exist, change their passwords.
Task #2c: Edit User Profiles.
Add this line to the .bash_profile for both oracle and grid:
umask 022
This next change is a personal preference thing, but I always unalias the ls and vi commands. I don’t need vim in my life and I find the different colors for different types of file completely unnecessary. Sanity can be returned to your default bash shell by adding the following lines to the .bash_profile for both oracle and grid:
unalias ls
unalias vi
Task #2d: Disable SELinux.
SELinux can cause havoc when trying to instantiate ASM disks and it needs to be disabled. Instructions for doing so can be found here.
Task #2e: Disable the Linux Firewall.
Generally speaking, firewalls are a good thing.
In a production environment, you’d want to work with your Network Administrator and have them open up the ports necessary for everything to work.
However, in our environment we can trust that we won’t hack ourselves and we’ll go ahead and disable the Linux firewall.
Task #2f: Disable Network Manager.
Network Manager has an annoying habit of jumping all over your /etc/resolv.conf file. Thus destroying your neatly crafted configuration. You need to disable Network Manager to prevent this from happening.
Task #2g: Install the iSCSI Initiator.
This can be selected as an installation option or you can add it now using this yum command:
Installing cvuqdisk will enable the Cluster Verification Utility to detect shared storage. Allegedly. It’s included in the 12c Grid Infrastructure downloads referenced here.
You’ll need to copy the two Grid Infrastructure zip files to racnode1. FileZilla does the job. I located these files within the home directory of the grid user:
[root@racnode1 ~]# cd ~grid/media/gi_12102
[root@racnode1 gi_12102]# ls -l
drwxr-xr-x 7 grid oinstall 4096 Dec 16 22:10 grid
-rw-r--r-- 1 grid oinstall 1747043545 Dec 12 15:33 linuxamd64_12102_grid_1of2.zip
-rw-r--r-- 1 grid oinstall 646972897 Dec 12 15:32 linuxamd64_12102_grid_2of2.zip
Unzipping these files created the grid directory. Within the grid directory is an rpm directory and that’s where you’ll find the cvuqdisk rpm:
[root@racnode1 gi_12102]# cd grid/rpm
[root@racnode1 rpm]# ls -l
-rwxr-xr-x 1 grid oinstall 8976 Jun 30 2014 cvuqdisk-1.0.9-1.rpm
RACNODE1_VM had 2 virtual disks allocated to it. The /dev/xvdb disk was used for the installation of Oracle Linux. The /dev/xvdc disk will be used to build the /u01 file system. That’s where all the Oracle software will be installed. The following series of commands will partition the /dev/xvdc disk, build and mount the /u01 file system:
[root@racnode1 ~]# fdisk /dev/xvdc
WARNING: DOS-compatible mode is deprecated. It's strongly recommended to
switch off the mode (command 'c') and change display units to
sectors (command 'u').
Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First cylinder (1-3916, default 1):
Using default value 1
Last cylinder, +cylinders or +size{K,M,G} (1-3916, default 3916):
Using default value 3916
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
[root@racnode1 ~]# mkfs -t ext4 -m 0 /dev/xvdc1
[root@racnode1 ~]# cd /
[root@racnode1 /]# mkdir /u01
[root@racnode1 /]# mount /dev/xvdc1 /u01
[root@racnode1 /]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/vg_racnode1-lv_root
35G 11G 23G 31% /
tmpfs 2.0G 72K 2.0G 1% /dev/shm
/dev/xvdb1 477M 132M 316M 30% /boot
/dev/xvdc1 30G 44M 30G 1% /u01
Finally, edit the /etc/fstab file so the /u01 file system will be mounted each time the system restarts:
[root@racnode1 ~]# vi /etc/fstab
Add this line:
/dev/xvdc1 /u01 ext4 defaults 0 0
Task #2k: Add Oracle software directories.
Now the /u01 file system is mounted, we can create the directories we’ll use to install the 12c Grid Infrastructure and Oracle Database 12c software:
Although we will be using DNS to resolve most things within our various networks, it’s very helpful to have all the addressing documented in each RAC node’s /etc/hosts file. All the entries required for every server’s /etc/hosts file can be found here.
Task #3: Clone RACNODE1_VM.
Now that we have the racnode1 server fully configured, we need a second server configured in a similar way. The hard way would be to create another VM then repeat all the steps up to this point to configure a second server. The easy way would be to clone RACNODE1_VM. Let’s go that route!
In Oracle VM Manager, click on the Servers and VMs tab, highlight RACNODE1_VM, then click the Stop icon (the red square). This will shutdown racnode1 and stop the VM.
Highlight RACNODE1_VM, right click then select the Clone or Move option:
Select Create a clone of this VM, then click Next:
Use the following values to complete the next screen, then click OK:
Field
Value
Clone to a
Virtual Machine
Clone Count
4
Name Index
2
Clone Name
RACNODE_VM
Target Server Pool
ServerPool1
This has the effect of creating 4 more VMs starting with RACNODE_VM.2 and ending with RACNODE_VM.5, all identical to RACNODE1_VM:
Note, the screen shows a 4096 MB memory allocation. This was increased to 6144 MB on a subsequent re-build Click here for further details.
Highlight RACNODE_VM.2 and click the pencil icon to edit the VM. Change the Name to RACNODE2_VM:
Click the Disks tab and change the names of the 2 virtual disks to RACNODE2_OS and RACNODE2_u01, then click OK:
Click the arrow head to the left of RACNODE2_VM to display its properties:
Task #4: Modify racnode2.
At this point, it’s time to start the VMs. In Oracle VM Manager, click on the Servers and VMs tab, highlight RACNODE1_VM, then click on the green arrow head Start icon to start the VM. Once RACNODE1_VM is up and running, repeat this procedure to start RACNODE2_VM. Connect to the console of RACNODE2_VM by clicking on the Launch Console icon.
Note, you may run into problems starting VMs. Click here for potential solutions/workarounds if you see an error similar to this:
Error: Device 1 (vif) could not be connected. Hotplug scripts not working
After RACNODE2_VM has started up, the first thing you’ll notice is the login screen announces you’re looking at racnode1.mynet.com. RACNODE2_VM is an identical clone of RACNODE1_VM, so it’s hardly surprising the server running inside RACNODE2_VM thinks it’s racnode1. Fortunately this is easy to fix by following the next couple of steps.
Task #4a: Fix the networking.
When cloning RACNODE1_VM, Oracle VM Manager was smart enough to allocate 3 new MAC addresses to the 3 vNICs allocated to each clone. However, the NIC configuration files in Linux don’t know that and will have to be edited manually.
Login to the ‘fake’ racnode1 and locate the NIC configuration files:
[root@racnode1 ~]# cd /etc/sysconfig/network-scripts
[root@racnode1 network-scripts]# ls -l ifcfg-eth*
-rw-r--r--. 3 root root 320 Dec 11 17:36 ifcfg-eth0
-rw-r--r-- 3 root root 269 Dec 13 16:13 ifcfg-eth1
-rw-r--r-- 3 root root 248 Dec 13 16:13 ifcfg-eth2
ifcfg-eth0 is the configuration file for the public network interface.
ifcfg-eth1 is the configuration file for the storage network interface.
ifcfg-eth2 is the configuration file for the private interconnect network interface.
Each file contains the exact same entries as the equivalent file on the real racnode1. Some of these parameters are consistent across all the servers (e.g. gateways and DNS) and some are specific to an individual server. We need to change the ones which are specific to an individual server. Those parameters are UUID, IPADDR and HWADDR.
The UUID is a Universal Unique IDentifier. It’s a number that Linux gives to each NIC. The entry in the ifcfg-eth* files looks like this:
UUID=eaba99ea-c88e-4cf2-b990-bee55e752e91
To get a new UUID, use this command:
[root@racnode1 ~]# uuidgen eth0
Substitute eth0 for eth1 and eth2 to generate new UUIDs for those interfaces. Once you have 3 new UUIDs, one for each NIC, update the UUID entry in the ifcfg-eth* files.
The IPADDR is the IP address assigned to a given NIC. The entry in the ifcfg-eth* files looks like this:
IPADDR=200.200.10.11
You can get the correct IP addresses for racnode2’s eth0, eth1 and eth2 interfaces from the /etc/hosts file. If clicking the link is too challenging, here’s the cheat sheet:
Server
NIC
IP Address
racnode2
eth0
200.200.10.12
racnode2
eth1
200.200.20.12
racnode2
eth2
200.200.30.12
Using these IP addresses, update the IPADDR entry in the ifcfg-eth* files.
The HWADDR is basically the MAC address which Oracle VM Manager allocates to the VM’s vNICs. You can see the MAC addresses allocated to RACNODE2_VM in a few different ways.
The vm.cfg contains the MAC addresses for each vNIC in eth0, eth1 and eth2 order. The location of the vm.cfg file is referenced here.
Alternatively, go back to this screen in Oracle VM Manager and click on the Networks tab:
Once again, the Ethernet Network name maps to vNICs like this:
Ethernet Network
Linux NIC
MAC Address
Management_Public
eth0
00:21:F6:D2:45:A0
Shared_Storage_Public
eth1
00:21:F6:E6:F6:71
GI_Interconnect_Private
eth2
00:21:F6:2F:DE:C5
Using the MAC addresses you obtain from your environment, update the HWADDR entry in the ifcfg-eth* files.
To bring up the network interfaces using the updated and correct values for UUID, IPADDR and HWADDR, each interface must be stopped and re-started. Using eth0 as an example, stopping and re-starting the interface is done with these commands:
Once all 3 interfaces have been re-started, you can view their status using this command:
[root@racnode1 ~]# ifconfig -a
Task #4b: Change the hostname.
The method for changing an Oracle Linux hostname is documented here.
Once the networking and hostname issues have been fixed, I strongly recommend a reboot. If the system comes back up with the correct name (racnode2.mynet.com) and the correct network configuration, you’re ready to move onto the dark arts discussed in Part 8 – SSH, DNS and CVU. See you there!
If you have any comments or questions about this post, please use the Contact form here.
