Forum Discussion

allaboutunix's avatar
10 years ago

DIMM replacement on VCS active node

Hi Team,

We have to replace DIMM on the active  node in which VCS services is currently running.

Kindly guide step by step procedure  for this activity,also please suggest the prerequisities before starting this activity.

This is very very crucial activity as Class A application is running on the active node.

Do we move the SG to passive node(Jupiter) and then perform the  activity so that apps will be online.What will be its steps do to it.

Currently, SG is running in Venus server.in which We have to perform activity.

-- SYSTEM STATE
-- System               State                Frozen

A  Venus              RUNNING              0
A   Jupiter            RUNNING              0

-- GROUP STATE
-- Group           System               Probed     AutoDisabled    State

B  ClusterService  Venus    Y          N               ONLINE
B  ClusterService   Jupiter              Y          N               OFFLINE
B  ORA_SG_Group   Venus             Y          N               ONLINE
B  ORA_SG_Group   Jupiter         Y          N               OFFLINE

Kindly suggest as soon as possible.

  • Hi Allaboutunix,

     

    For the cluster states you described, the steps are:

     

    1.    Switchover the ORA_SG_Group from active node(Venus) to passive node(Jupiter). This will ensure uptime of ORA_SG_Group group while DIMM replacement takes place on Venus.

    # hagrp -switch ORA_SG_Group -to Jupiter

     

    2.    You need to also switchover ClusterService from Venus to Jupiter.

    # hagrp -switch ClusterService -to Jupiter

     

    3.    Freeze the Venus node. Freezing Venus will prevent any accidental failover(either user initiated or automated) of service groups.

    # hasys -freeze -persistent Venus

    "hasys -freeze -persistent -evacuate Venus" can be used instead of step #1, #2, and #3. However, as class A applications are involved, we would suggest smaller atomic operation instead of composite operation. It is easier to monitor progress for smaller atomic operation v/s composite operations.

     

    4.    Stop VCS on Venus

    # hastop -sys Venus

     

    5.    After successful completion of DIMM, start VCS on Venus. You should also unfreeze it.

    # hasys -unfreeze -persistent Venus

     

    6.    Switchback ORA_SG_Group and ClusterService from passive node(Jupiter) to active node(Venus). Repeat step #1 and #2 for Venus node.

     

    Thanks & Regards,
    Sunil Y

  • Hi Allaboutunix,

     

    For the cluster states you described, the steps are:

     

    1.    Switchover the ORA_SG_Group from active node(Venus) to passive node(Jupiter). This will ensure uptime of ORA_SG_Group group while DIMM replacement takes place on Venus.

    # hagrp -switch ORA_SG_Group -to Jupiter

     

    2.    You need to also switchover ClusterService from Venus to Jupiter.

    # hagrp -switch ClusterService -to Jupiter

     

    3.    Freeze the Venus node. Freezing Venus will prevent any accidental failover(either user initiated or automated) of service groups.

    # hasys -freeze -persistent Venus

    "hasys -freeze -persistent -evacuate Venus" can be used instead of step #1, #2, and #3. However, as class A applications are involved, we would suggest smaller atomic operation instead of composite operation. It is easier to monitor progress for smaller atomic operation v/s composite operations.

     

    4.    Stop VCS on Venus

    # hastop -sys Venus

     

    5.    After successful completion of DIMM, start VCS on Venus. You should also unfreeze it.

    # hasys -unfreeze -persistent Venus

     

    6.    Switchback ORA_SG_Group and ClusterService from passive node(Jupiter) to active node(Venus). Repeat step #1 and #2 for Venus node.

     

    Thanks & Regards,
    Sunil Y