4 Replies Latest reply: Apr 30, 2013 4:18 AM by 1005406 RSS

    ORA-15490: not all instances could add/drop the volume

    1005406
      I have a 4-nodes cluster on RedHat6.

      I created a diskgroup DG_ORA_DIR2.

      I tried to create a volume, but got the message above:

      [grid@CLU ~]$ asmcmd volcreate -G DG_ORA_DIR2 V_DIR_SIG2 -s 2136M
      ORA-15032: not all alterations performed
      ORA-15460: volume name 'V_DIR_SIG2' is already in use (DBD ERROR: OCIStmtExecute)

      However the volume was created anyway. But it is DISABLED:

      [grid@CLU ~]$ asmcmd volinfo -G DG_ORA_DIR2 -a
      Diskgroup Name: DG_ORA_DIR2

      Volume Name: V_DIR_SIG2
      Volume Device: /dev/asm/v_dir_sig2-133
      State: DISABLED
      Size (MB): 2240
      Resize Unit (MB): 32
      Redundancy: MIRROR
      Stripe Columns: 4
      Stripe Width (K): 128
      Usage:
      Mountpath:

      I cannot delete it:

      [grid@CLU ~]$ asmcmd voldelete -G DG_ORA_DIR2 V_DIR_SIG2
      ORA-15032: not all alterations performed
      ORA-15490: not all instances could add/drop the volume (DBD ERROR: OCIStmtExecute)

      I cannot do anything actually:

      [grid@CLU ~]$ asmcmd voldisable  -G DG_ORA_DIR2 -a
      ORA-15032: not all alterations performed
      ORA-15480: error reported in the ASM volume driver (DBD ERROR: OCIStmtExecute)
      [grid@CLU ~]$ asmcmd volenable  -G DG_ORA_DIR2 -a
      ORA-15032: not all alterations performed
      ORA-15480: error reported in the ASM volume driver (DBD ERROR: OCIStmtExecute)

      My diskgroup has these parameters:
      SQL> ALTER DISKGROUP DG_ORA_DIR2 SET ATTRIBUTE 'compatible.rdbms' = '11.2.0.0.0';
      SQL> ALTER DISKGROUP DG_ORA_DIR2 SET ATTRIBUTE 'compatible.asm' = '11.2.0.3.0';
      SQL> ALTER DISKGROUP DG_ORA_DIR2 SET ATTRIBUTE 'compatible.advm' = '11.2.0.0.0';

      I checked the alertlog: nothing.
      I have no idea left... Anyone can help please?
        • 1. Re: ORA-15490: not all instances could add/drop the volume
          User719416-Oracle
          Hi,
          This can occur if the volume is open from a previously unclean dismount or shutdown.
          The volume needs to be closed before it can be deleted in the driver.

          Could you please check each of your nodes and first see if the volume devices exists?
          If so, you can verify that the volume is open by running:
          lsof /dev/asm/v_dir_sig2-133
          Close all open volume references and retry the volume delete command.
          • 2. Re: ORA-15490: not all instances could add/drop the volume
            1005406
            Hi, thanks for the reply.

            I checked on every node and got this:

            [grid@CLU ~]$ ls /dev/asm/v_dir_sig2-133
            ls: impossible d'accéder à /dev/asm/v_dir_sig2-133: Aucun fichier ou dossier de ce type
            [grid@CLU ~]$ lsof /dev/asm/v_dir_sig2-133
            lsof: status error on /dev/asm/v_dir_sig2-133: No such file or directory
            lsof 4.82

            It seems that the volume doesn't exist.


            FYI: When I tried to create the volume, another volume was not really clean (its diskgroup had been dropped). ASMCMD didn't want to remove this other volume. I had to recreate a diskgroup with the same name and I was abble to remove the volume.
            But i'm not sure this has something to do with my issue.

            Edited by: Eska on 25 avr. 2013 23:58
            • 3. Re: ORA-15490: not all instances could add/drop the volume
              User719416-Oracle
              Hi,
              What version of ASM are you using?

              There should be more information in the alert.log file on the node where you attempted to add/drop the volume. Specifically, it should indicate the node number where the volume
              failed to be created/dropped. On the node indicated in the alert.log file, there should be more information in the CKPT trace file about exact error seen.
              • 4. Re: ORA-15490: not all instances could add/drop the volume
                1005406
                Hi,

                Thanks for the answer.

                First I've to say that the problem is solved. This morning I tried to execute the same command again and this time it succeed. I really don't know how and why.

                When the problem what still there I check the alertlog and nothing was in there.

                My version is:
                Oracle Database 11g Enterprise Edition Release 11.2.0.3.0 - 64bit Production
                PL/SQL Release 11.2.0.3.0 - Production
                CORE 11.2.0.3.0 Production
                TNS for Linux: Version 11.2.0.3.0 - Production
                NLSRTL Version 11.2.0.3.0 - Production

                Thanks anyway :) Sorry I can't tell you what solved the issue :(