SUMMARY: Freshening Solaris 7/E4500/Cluster 2.2

From: Angelo McComis (sunmanagers_at_mccomis.com)
Date: 08/27/04

  • Next message: Elaine .: "disksuite query" 
    To: "Sun Managers Mailing List" <sunmanagers@sunmanagers.org>
Date: Fri, 27 Aug 2004 08:47:14 -0400

    Thanks to all who responded.
     
    Special Kudos to Darren E for doing almost the exact same exercise recently
    and sending me the steps he took including the commands. (I had to translate
    them into Cluster 2.2's commands, but that was relatively painless).
     
    Antonio Marongiu also pointed out that Cluster 2.2 is not supported on
    Solaris 9 (I missed that when I was going over all the various compatibility
    matrices).
     
    Daniel Denes shared that ran into something similar a few weeks back, Sun
    said to take everything down, patch it all, then bring it back up. Of
    course, that goes against the reason we all buy clustering software. He said
    he did it basically the way I had outlined, which was to freeze the cluster,
    patch the inactive node, bring it back up, unfreeze and migrate to the newly
    patched node, then patch the other. When he got to the unfreeze/migrate
    point, both boxes panicked, but after the reset, everything was fine.
     
    Thanks everyone.
     
    Summary Follows:
     
    Step 1: Swing processes over to Node 1 (away from the node you're patching)
    Begin on Node 2:
    Step 2: Ensure all processes are up on the other node
    Step 3: Create/Adjust /etc/vfstab files to not see any shared storage
    Step 4: Bring node down to OK prompt (single user mode)
    Step 5: Boot node without the cluster (boot -xs)
    Step 6: Ensure node booted without the cluster and shared storage not
    mounted
    Step 7: Apply Sun Recommended Patches to node and Reboot (multi-reboots
    and/or OBP patches if needed too)
    Step 8: Ensure node boots ok into multi-user mode
    Step 9: Add node into Cluster Membership
    Step 10: Swing process over to Node 2 (away from Node 1)
    Step 11: Verify nothing running on Node 1
    Step 12: Verify everything is running ok on Node 2
    Step 13: Follow Steps 2 through 9 on Node 1

    Thanks all,
     
    Angelo
      

    --
Angelo McComis
Senior IT Consultant
sunmanagers at mccomis dot com
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
  • Next message: Elaine .: "disksuite query"

    Relevant Pages

    • Re: Cannot use GPO Deployment from Cluster Volume
      ... I use commands similar to the bottom of your link. ... I do this because of the following error while using the GUI. ... Please remove any local users or groups from the permissions." ... Any ideas why the Cluster Administrator would have this issue? ...
      (microsoft.public.windows.server.clustering)
    • Re: cluadmin Hung
      ... Or use cluster.exe commands. ... It happens pretty randomly, I generally get a call from Helpdesk, that they are not able to access cluster. ... I get a issue where in cluadmin.exe is hung. ... However I can see all resources are online through cluster.exe command line. ...
      (microsoft.public.windows.server.clustering)
    • Re: Using cluster.exe to create physical disks
      ... You shouldn't be trying to mount the disk on the other node ... For your cluster.exe script, you could try using the command syntax "cluster ... System error 2 has occurred. ... These are the commands I am running ...
      (microsoft.public.windows.server.clustering)
    • Re: Add VCS resource using CLI
      ... but best done with the ha CLI commands rather than ... modifying main.cf. ... to restart the cluster using a hastop -force, but I do not recommend this. ...
      (comp.unix.solaris)
    • Re: HACMP question
      ... Boot and service must be on the same VLAN but seperate subnets. ... cluster services is started. ... to aixoradb2 in the event of a failover. ... You have the vg which is on the disk imported, ...
      (AIX-L)