Oracle 11gR2 Grid Infrastructure Upgrade Bug

I’ve still been working away on migrating a 10gR2 RAC cluster running on a RHEL4 2 node cluster up to 11gR2. However I seem to have hit an install issue right at the first hurdle.

The system involved is an x86-64 system running a 2.6.9-22 kernel. The grid infrastructure install seemed to be passing by without incident, I had chosen the “Upgrade Grid Infrastructure” option at the Installation Type dialog box during the install, and had ran through the install, until the very end when a box pops up asking for the script to be run:

This was not a worry, so I went off and attempted to run it, but then the following occurred:

[root@linuxrac1 11.2.0]# ./ 
Running Oracle 11g script...

The following environment variables are set as:
    ORACLE_OWNER= oracle
    ORACLE_HOME=  /var/opt/grid/11.2.0

Enter the full pathname of the local bin directory: [/usr/local/bin]: 
The file "dbhome" already exists in /usr/local/bin.  Overwrite it? (y/n) 
The file "oraenv" already exists in /usr/local/bin.  Overwrite it? (y/n) 
The file "coraenv" already exists in /usr/local/bin.  Overwrite it? (y/n) 

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 script.
Now product-specific root actions will be performed.
2009-12-21 15:45:59: Parsing the host name
2009-12-21 15:45:59: Checking for super user privileges
2009-12-21 15:45:59: User has super user privileges
Using configuration parameter file: /var/opt/grid/11.2.0/crs/install/crsconfig_params
Creating trace directory
Cannot get node network interfaces (/var/opt/grid/11.2.0/bin/oifcfg iflist -p -n failed.) at /var/opt/grid/11.2.0/crs/install/ line 1786.

Running the oifcfg command by itself, seem to actually produce some sensible output:

[root@linuxrac1 11.2.0]# /var/opt/grid/11.2.0/bin/oifcfg iflist -p -n

This is the same output if you run oifcfg from the 10gR2 install:

[root@linuxrac1 11.2.0]# /var/opt/oracle/product/crs/bin/oifcfg iflist -p -n

However, it seems like some commands may be running, incorrectly, from the 11gR2 when they should be running from the old 10gR2 home:

The 11.2 command produces the following

[root@linuxrac1 log]# /var/opt/grid/11.2.0/bin/oifcfg getif -global

Failed to initialize GPnP

While the 10.2 command produces following output:

[root@linuxrac1 log]# /var/opt/oracle/product/crs/bin/oifcfg getif -global
eth0 global public
eth1 global cluster_interconnect

I took out a Service Request with Oracle and after a couple of weeks, they are conceding the possibility that this might be a bug, 9126737 which is as yet unpublished. They are muttering about needing a backport for this fix. This of course will take an as yet indeterminate quantity of time.

This service request has been on the go for 3 weeks, and I don’t feel any closer to a solution.

I can’t believe that such a mainstream platform can have a fundamental bug, that is preventing from installing the 11.2 Grid Infrastructure.

I’d very much like to hear from anyone if you have managed to upgrade a 10gR2 RAC system to the new 11.2 Grid Infrastructure, particularly on Linux.


4 thoughts on “Oracle 11gR2 Grid Infrastructure Upgrade Bug

  1. Hi.

    I was able to successfully upgrade RAC to 11.2 (during beta-testing period) on
    Red Hat Enterprise Linux AS release 4 (Nahant Update 8)

    So your issue may be related to x86-64, or something changed in wrong way before going production 🙂


  2. Hi Dude

    I upgrade successfully the CRS and ASM however , I was getting errors not being published by metalink yet looking deeper I looked into my interfaces registered in the CR and I shocked when I see
    -(0)$ oifcfg getif
    bond1 global cluster_interconnect

    I proceed to fix it however I got the following error

    -(0)$ oifcfg setif -global bond0/xxxx:public

    PRIF-26: Error in update the profiles in the cluster

    I brought up the CRS in exclusive mode no success same problem trying to delete the private address , also trying to update the diskstring from the ASM I got the following error, I have filed those cases to Oracle Support no fix yet =(
    the scary thing was I didnt have problems upgrading the CRS

    SQL> alter system set asm_diskstring=’ORCL:*’;
    alter system set asm_diskstring=’ORCL:*’
    ERROR at line 1:
    ORA-29780: unable to connect to GPnP daemon [CLSGPNP_ERR]

    SQL> show parameter spfile;

    ———————————— ———– ——————————
    spfile string /u01/app/11.2.0/grid/dbs/spfil
    SQL> create pfile=’/tmp/pfile’ from spfile;

    File created.

    SQL> show parameter diskgroup

    ———————————— ———– ——————————
    asm_diskgroups string DATA01DG, RECOVERDEST01DG
    SQL> create spfile=’+DATA01DG’ from pfile=’/tmp/pfile’;
    create spfile=’+DATA01DG’ from pfile=’/tmp/pfile’
    ERROR at line 1:
    ORA-29780: unable to connect to GPnP daemon [CLSGPNP_ERR]
    ORA-29780: unable to connect to GPnP daemon [CLSGPNP_ERR]
    ORA-29780: unable to connect to GPnP daemon [CLSGPNP_ERR]

    same error when I tr

  3. We faced same issue during install and worked with Oracle and found that if any CRS related variables are set (CRS_HOME,ORA_CRS_HOME) may cause this issue. Please check this if it is helpfull to u.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s