The purpose of this document is to provide an example of the steps required to update a 5.0 or 5.0MP1 2-node cluster to 5.0MP3 with minimal downtime.


Because VxVM and VxFS are upgraded on the nodes from single-user mode, while the service groups are running on the other node (s), this can be done while the root devices are encapsulated.

During the update process, the Volume Manager and VxFS Filesystem components will remain at a lower level than VCS as the process is completed.
 

However, this process does not apply to CFS or RAC installations, since Storage Foundation with Cluster File System (CFS) and/or Oracle RAC require all Veritas product components to be at the same level

The process may be modified for use with clusters that have more than 2 nodes.

1.) Upgrade VCS on both nodes to5.0Mp3
  -   Freeze the service groups.
a) Make the cluster configuration read/write.
  # haconf -makerw
b) Persistent freeze each and all cluster systems (perform this step for each cluster node).
  # hasys -freeze -persistent
c) Save and dump the clusterconfiguration.
  # haconf -dump-makero
  -   Upgrade the VCS portion of the product on all systems.  
a) Stop VCS while leaving the servicegroups online, running outside VCS.
  #hastop -all -force
b) Verify on each node that port h (HAD -high availability daemon) is no longer running.
This is an example of an expected output:
  # gabconfig -a
GAB PortMemberships
===============================================================
Porta gen   22590b membership 01

c) Update VCS only on all nodes of the cluster ( the update must be done for all cluster nodes at the same time, or the cluster nodes will not join due to version mismatch errors).
  # cd/cluster_server
  # ./installvcs  
(Enter all cluster node names when prompted)

Note that the installation script may indicate a reboot upon completion, but no reboot for any of the clustered systems is required at this point in time.
-   Start VCS on all nodes to verify functionality and switch service groups to one node
a)Start VCS on each node.
  #hastart
b) Verify that the cluster nodes joined successfully by running the command on each node (ports a, and h should have a single line in the output).
This is an example of an expected output:
  # gabconfig-a      
GAB PortMemberships
===============================================================
Porta gen   22590b membership 01
Port h gen   225916 membership01

c) Verify the status of the service groups in the cluster.
  # hastatus-sum

2.) Switch all the service groups to one node(call this node) and persistent freeze the systems.
a) Make the cluster configuration read/write.
  #haconf -makerw
b) Persistent unfreeze each and all clustersystems (perform this step for each cluster node).
  # hasys -unfreeze -persistent
c) Switch all service groups from one node (to be upgraded in the next step) to another node.
  #hagrp -switch -to
d)Persistent freeze each and all cluster systems (perform this step for each cluster node).
  # hasys -freeze-persistent
e) Save and dump the clusterconfiguration.
  # haconf -dump-makero
.
3.) Bring other nodes down to single-user mode -and upgrade it with SUN and Veritas patches.
   - First stophad, gab, andvxfen
            #  hastop-local
     #  vxfenconfig-U
     #  gabconfig-U

-  Now bring the system down to single-user mode

      # init s

- Log onto the console to run the upgrade from single-user mode

-     Apply any SUN patches at this time.

    At this time, VCS is updated on all nodes.
   The product installation scripts will only update the lower versions of the installed products.
   There is a known issue with the 5.0MP3 installation script.
   Executing any of the installation scripts more than one time (as is required for a minimal downtime update) may trigger
   the issue.  There is a hotfix to resolve this issue.
   The link for the hotfix is in the RelatedDocuments section near the bottom of this document.
   The Perl script in this hotfix may be utilized with any of the installations scripts included with the product (for example:
   installer, installmp,installsf, installvcs, and installvm).
   Please download the CPIhotfix, copy the Perl script in the hotfix to /var/tmp and utilize the Perl script from the hotfix as
   required by the command below.
         a) Update VxFS and VxVM onnode.
            # ./installmp -require /var/tmp/rem_drv_patch.pl
         b) Upon completion of the script, reboot is indicated.
            # reboot

4.) Confirm the cluster join after reboot of the upgraded node
-  Following reboot of the upgraded node, check gabconfig -a output to ensure had, gab, and vxfen (if applicable) are running on the cluster nodes.   The following is an example of what the output of the command should look like:
     # gabconfig-a      
         GABPortMemberships
         ===============================================================
         Port a gen   22590b membership 01
         Port h gen   225916 membership 01

5.) Unfreeze the systems and then switch the service groups to upgraded nodes. Then persistent freeze the system for the next node upgrade.
     a) Make the cluster configuration read /write.
        # haconf-makerw
     b) Unfreeze the system to be upgraded in the next step.
        # hasys-unfreeze -persistent  
     c)Switch all groups to the upgraded node.
        # hagrp -switch -to
     d) Freeze persistent all nodes.
        # hasys-freeze -persistent
     e) Save and dump the cluster configuration.
      # haconf-dump -makero

6.) Bring the other node down to a single user-mode - and upgrade it with the SUN and Veritas patches.
   - First stop had, gab, andvxfen
        #  hastop-local
        #  vxfenconfig-U
        #  gabconfig -U

-  Now bring the system down to single-user mode

      #init s

- Log onto the console to run the upgrade from single-user mode

-     Apply any SUN patches at this time.

-     Upgrade VxVM andVxFS

       a) Update VxFS and VxVM onnode.
            #./installmp -require /var/tmp/rem_drv_patch.pl
      b) Upon completion of the script, reboot is indicated.
          #reboot

7.) Confirm the cluster join after reboot of an upgraded node
-  Following the reboot of the upgraded node, check gabconfig -a output to ensure had, gab, and vxfen (if applicable) are running on the cluster nodes.   The following is an example of what the output of the command should look like:
   #gabconfig -a      
     GAB PortMemberships
     ===============================================================
   Port a gen   22590b membership 01
    Port h gen  225916 membership 01

8.) Unfreeze persistent the first node, Andre-distribute the service groups between the two nodes.
       a) Make the cluster configuration read/ write.
          #haconf -makerw
       b) Unfreeze the systems.
       # hasys-unfreeze -persistent
       c)Switch any desired groups.
          #hagrp -switch -to
      d) Save and dump the clusterconfiguration.
          #haconf -dump -makero
       e) Verify all service groups are online.
          #hastatus -sum

The upgrade is complete.

If there are any questions in relation to this process, or an attempt of this process results in any unexpected error or messages, call Technical Support for additional assistance.