Sunday, September 18, 2011

How to restore Oracle Grid Infrastructure OCR and vote disk on ASM

After reading the book about Oracle RAC (see previous post) I decided to test recovery of OCR (Oracle Cluster Registry) file and vote disk placed on ASM disk groups. This is an operation I never did before. I was lucky enough and my ASM based RAC installation never had similar issue but I did it before when OCR and vote disk were placed on raw or block devices.
According to Oracle doc it should be straight forward and it really is but you need to careful otherwise you can hit strange errors, like "segmentation fault" for ocrconfig when Grid Infrastructure is down.
Here is step by step instruction:
  1. Check if you can access OCR files
    [root@node1 ~]# /u01/app/11.2.0/grid/bin/ocrcheck
    PROT-602: Failed to retrieve data from the cluster registry
    PROC-26: Error while accessing the physical storage
    
  2. Stop CRS on all nodes
    [root@node1 ~]# export CRS_HOME=/u01/app/11.2.0/grid
    [root@node1 ~]# $CRS_HOME/bin/crsctl stop crs
    CRS-2796: The command may not proceed when Cluster Ready Services is not running
    CRS-4687: Shutdown command has completed with errors.
    CRS-4000: Command Stop failed, or completed with errors.
    
    If you got error like that try force option
    [root@node1 ~]# $CRS_HOME/bin/crsctl stop crs -f
    CRS-2791: Starting shutdown of Oracle High Availability Services-managed resources on 'node1'
    CRS-2673: Attempting to stop 'ora.ctssd' on 'node1'
    CRS-2673: Attempting to stop 'ora.evmd' on 'node1'
    CRS-2673: Attempting to stop 'ora.asm' on 'node1'
    CRS-2673: Attempting to stop 'ora.mdnsd' on 'node1'
    CRS-2677: Stop of 'ora.asm' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.cluster_interconnect.haip' on 'node1'
    CRS-2677: Stop of 'ora.mdnsd' on 'node1' succeeded
    CRS-2677: Stop of 'ora.cluster_interconnect.haip' on 'node1' succeeded
    CRS-2677: Stop of 'ora.evmd' on 'node1' succeeded
    CRS-2677: Stop of 'ora.ctssd' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.cssd' on 'node1'
    CRS-2677: Stop of 'ora.cssd' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.diskmon' on 'node1'
    CRS-2673: Attempting to stop 'ora.crf' on 'node1'
    CRS-2677: Stop of 'ora.crf' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.gipcd' on 'node1'
    CRS-2677: Stop of 'ora.diskmon' on 'node1' succeeded
    CRS-2677: Stop of 'ora.gipcd' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.gpnpd' on 'node1'
    CRS-2677: Stop of 'ora.gpnpd' on 'node1' succeeded
    CRS-2793: Shutdown of Oracle High Availability Services-managed resources on 'node1' has completed
    CRS-4133: Oracle High Availability Services has been stopped.
    [root@node1 ~]#
    
  3. When OCR file and vote disk are located on ASM disk group running ASM instance is required for recovery. Current version of Oracle Grid Infrastructure (GI) can be started in exclusive mode on one node to allow checks and restore.
    There is a slight change between version 11.2.0.1 and 11.2.0.2 - in first release we need to start GI in exclusive mode and then stop CRS service (as showed below). In 11.2.0.2 we can start GI without CRS but old steps are still working.
    Steps for 11.2.0.1
    [root@node1 ~]# $CRS_HOME/bin/crsctl start crs -excl
    CRS-4123: Oracle High Availability Services has been started.
    CRS-2672: Attempting to start 'ora.mdnsd' on 'node1'
    CRS-2676: Start of 'ora.mdnsd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.gpnpd' on 'node1'
    CRS-2676: Start of 'ora.gpnpd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.cssdmonitor' on 'node1'
    CRS-2672: Attempting to start 'ora.gipcd' on 'node1'
    CRS-2676: Start of 'ora.cssdmonitor' on 'node1' succeeded
    CRS-2676: Start of 'ora.gipcd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.cssd' on 'node1'
    CRS-2672: Attempting to start 'ora.diskmon' on 'node1'
    CRS-2676: Start of 'ora.diskmon' on 'node1' succeeded
    CRS-2676: Start of 'ora.cssd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.ctssd' on 'node1'
    CRS-2672: Attempting to start 'ora.cluster_interconnect.haip' on 'node1'
    CRS-2676: Start of 'ora.ctssd' on 'node1' succeeded
    CRS-5017: The resource action "ora.cluster_interconnect.haip start" encountered the following error:
    Start action for HAIP aborted
    CRS-2674: Start of 'ora.cluster_interconnect.haip' on 'node1' failed
    CRS-2679: Attempting to clean 'ora.cluster_interconnect.haip' on 'node1'
    CRS-2681: Clean of 'ora.cluster_interconnect.haip' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.asm' on 'node1'
    CRS-2674: Start of 'ora.asm' on 'node1' failed
    CRS-2673: Attempting to stop 'ora.ctssd' on 'node1'
    CRS-2677: Stop of 'ora.ctssd' on 'node1' succeeded
    CRS-4000: Command Start failed, or completed with errors.
    [root@node1 ~]#
    
    Errors are expected so now we need to stop CRS (if it has been started)
    [root@node1 ~]# $CRS_HOME/bin/crsctl stop resource ora.crsd -init
    [root@node1 ~]# $CRS_HOME/bin/crsctl stop resource ora.crsd -init
    CRS-2500: Cannot stop resource 'ora.crsd' as it is not running
    CRS-4000: Command Stop failed, or completed with errors.
    
    Steps for 11.2.0.2 - this allow us to avoid starting CRS which is trying to start resources. Issues related to CRS resource will bring down HAIP and ASM - and this is root cause of errors shown in 11.2.0.1 example.
    [root@node1 ~]# $CRS_HOME/bin/crsctl start crs -excl -nocrs
    CRS-4123: Oracle High Availability Services has been started.
    CRS-2672: Attempting to start 'ora.mdnsd' on 'node1'
    CRS-2676: Start of 'ora.mdnsd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.gpnpd' on 'node1'
    CRS-2676: Start of 'ora.gpnpd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.cssdmonitor' on 'node1'
    CRS-2672: Attempting to start 'ora.gipcd' on 'node1'
    CRS-2676: Start of 'ora.cssdmonitor' on 'node1' succeeded
    CRS-2676: Start of 'ora.gipcd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.cssd' on 'node1'
    CRS-2672: Attempting to start 'ora.diskmon' on 'node1'
    CRS-2676: Start of 'ora.diskmon' on 'node1' succeeded
    CRS-2676: Start of 'ora.cssd' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.cluster_interconnect.haip' on 'node1'
    CRS-2672: Attempting to start 'ora.ctssd' on 'node1'
    CRS-2676: Start of 'ora.ctssd' on 'node1' succeeded
    CRS-2676: Start of 'ora.cluster_interconnect.haip' on 'node1' succeeded
    CRS-2672: Attempting to start 'ora.asm' on 'node1'
    CRS-2676: Start of 'ora.asm' on 'node1' succeeded
    
    No errors has been displayed and GI has been started in exclusive mode.
  4. Now it is time to check if ASM instance is running and recreate ASM disk groups.
    [oracle@node1 ~]$ ps aux | grep pmon
    oracle    8009  0.0  0.9 495804 18564 ?        Ss   20:37   0:00 asm_pmon_+ASM1
    oracle    8260  0.0  0.0  61184   760 pts/3    S+   20:38   0:00 grep pmon
    
    ASM is running. Now we can connect and create a disk group for OCR and vote disk. Be sure that your new group will have proper COMPATIBLE.ASM value. You can put OCR and vote on groups with COMPATIBLE.ASM greater than 11.2.
    [oracle@node1 ~]$ export ORACLE_HOME=/u01/app/11.2.0/grid
    [oracle@node1 ~]$ export PATH=$PATH:$ORACLE_HOME/bin
    [oracle@node1 ~]$ export ORACLE_SID=+ASM1
    [oracle@node1 ~]$ sqlplus / as sysasm
    SQL*Plus: Release 11.2.0.2.0 Production on Tue Aug 30 20:39:14 2011
    Copyright (c) 1982, 2010, Oracle.  All rights reserved.
    Connected to:
    Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
    With the Real Application Clusters and Automatic Storage Management options
    SQL> show parameter asm_di
    NAME                                 TYPE        VALUE
    ------------------------------------ ----------- ------------------------------
    asm_diskgroups                       string
    asm_diskstring                       string
    
    SQL> alter system set asm_diskstring = '/dev/asmdisk*';
    System altered.
    
    SQL> create diskgroup DATA external redundancy disk '/dev/asmdisk11' attribute 'COMPATIBLE.ASM' = '11.2';
    Diskgroup created.
    
  5. In next step we need to create spfile for ASM instance as it will be used for rest of cluster to find out location of ASM disks using 'asm_diskstring'.
    First text init.ora file has been created and then new binary spfile has been created on ASM disk group.
    [oracle@node1 rac-cluster]$ cat /tmp/init.ora
    *.asm_power_limit=1
    *.diagnostic_dest='/u01/app/oracle'
    *.instance_type='asm'
    *.large_pool_size=12M
    *.remote_login_passwordfile='EXCLUSIVE'
    *.asm_diskstring = '/dev/asmdisk*'
    [oracle@node1 rac-cluster]$ sqlplus / as sysasm
    SQL*Plus: Release 11.2.0.2.0 Production on Tue Aug 30 21:23:36 2011
    Copyright (c) 1982, 2010, Oracle.  All rights reserved.
    Connected to:
    Oracle Database 11g Enterprise Edition Release 11.2.0.2.0 - 64bit Production
    With the Real Application Clusters and Automatic Storage Management options
    
    SQL> create spfile='+DATA' from pfile='/tmp/init.ora';
    File created.
    
  6. When disk group is ready it is time now to restore OCR from backup file.
    Oracle Grid Infrastructure is doing OCR backup automatically into following location $GRID_HOME/cdata/. Let's restore and check if OCR file is recognized properly after it.
    [root@node1 ~]# $CRS_HOME/bin/ocrconfig -restore /u01/app/11.2.0/grid/cdata/rac-cluster/backup_20110830_201118.ocr
    
    [root@node1 ~]# $CRS_HOME/bin/ocrcheck
    Status of Oracle Cluster Registry is as follows :
             Version                  :          3
             Total space (kbytes)     :     262120
             Used space (kbytes)      :       3132
             Available space (kbytes) :     258988
             ID                       :  660203047
             Device/File Name         :      +DATA
                                        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
    
  7. Now it is time to restore vote disk. This process will read asm_diskstring from ASM instance and will place vote files on these disks. See what happen when asm_diskstring is empty.
    [root@node1 ~]# $CRS_HOME/bin/crsctl replace votedisk +DATA
    CRS-4602: Failed 27 to add voting file 28652f742fc44f28bfc6d12d1412a604.
    Failed to replace voting disk group with +DATA.
    CRS-4000: Command Replace failed, or completed with errors.
    
    Error message from log file.
    [cssd(7894)]CRS-1638:Unable to locate voting file with ID 1b37b25b-686c4fb4-bfb82eac-357f48df that is being added to the list of configured voting files; details at (:CSSNM00022:) in /u01/app/11.2.0/grid/log/node1/cssd/ocssd.log
    2011-08-30 20:44:12.256
    
    When asm_diskstring is set up properly it is looking much better.
    [root@node1 ~]# $CRS_HOME/bin/crsctl replace votedisk +DATA
    Successful addition of voting disk 4ca8c2b58d394ff1bf7a9b88dd9f5fc3.
    Successfully replaced voting disk group with +DATA.
    CRS-4266: Voting file(s) successfully replaced
    [root@node1 ~]#  $CRS_HOME/bin/crsctl query css votedisk
    ##  STATE    File Universal Id                File Name Disk group
    --  -----    -----------------                --------- ---------
     1. ONLINE   4ca8c2b58d394ff1bf7a9b88dd9f5fc3 (/dev/asmdisk11) [DATA]
    Located 1 voting disk(s).
    
  8. In last step Grid Infrastructure need to be restarted on all nodes.
    [root@node1 ~]#  $CRS_HOME/bin/crsctl stop crs -f
    CRS-2791: Starting shutdown of Oracle High Availability Services-managed resources on 'node1'
    CRS-2673: Attempting to stop 'ora.mdnsd' on 'node1'
    CRS-2673: Attempting to stop 'ora.ctssd' on 'node1'
    CRS-2673: Attempting to stop 'ora.asm' on 'node1'
    CRS-2677: Stop of 'ora.asm' on 'node1' succeeded
    CRS-2677: Stop of 'ora.mdnsd' on 'node1' succeeded
    CRS-2677: Stop of 'ora.ctssd' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.cssd' on 'node1'
    CRS-2677: Stop of 'ora.cssd' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.gipcd' on 'node1'
    CRS-2673: Attempting to stop 'ora.diskmon' on 'node1'
    CRS-2677: Stop of 'ora.diskmon' on 'node1' succeeded
    CRS-2677: Stop of 'ora.gipcd' on 'node1' succeeded
    CRS-2673: Attempting to stop 'ora.gpnpd' on 'node1'
    CRS-2677: Stop of 'ora.gpnpd' on 'node1' succeeded
    CRS-2793: Shutdown of Oracle High Availability Services-managed resources on 'node1' has completed
    CRS-4133: Oracle High Availability Services has been stopped.
    
    [root@node1 ~]#  $CRS_HOME/bin/crsctl start crs
    CRS-4123: Oracle High Availability Services has been started.
    
    If everything went well similar result is expected
    [root@node1 ~]# $CRS_HOME/bin/crsctl check cluster -all
    **************************************************************
    node1:
    CRS-4537: Cluster Ready Services is online
    CRS-4529: Cluster Synchronization Services is online
    CRS-4533: Event Manager is online
    **************************************************************
    node2:
    CRS-4537: Cluster Ready Services is online
    CRS-4529: Cluster Synchronization Services is online
    CRS-4533: Event Manager is online
    **************************************************************
    node3:
    CRS-4537: Cluster Ready Services is online
    CRS-4529: Cluster Synchronization Services is online
    CRS-4533: Event Manager is online
    **************************************************************
    
Now Grid Infrastructure is working and if database ASM disk groups has been destroyed as well or we have everything in one ASM disk groups (like in my example) it is time to restore database from backup using Oracle Recovery Manager. But this is a different story.

Hope it help someone to restore Grid Infrastructure after disk crash.
regards,
Marcin