Etrack Incidents: 1294441

Last Modified: October 29, 2008

POINT PATCH FOR VCS 5.0MP1 or 5.0MP3
------------------------------------

SUMMARY
-------
The deliverables for this point patch contain:

   README           (this file)
   NFSRestartAgent (replacement for $VCS_HOME/bin/NFSRestart/NFSRestartAgent)


Packages affected
-----------------
    VRTSvcsag


Base Version
------------
    VCS 5.0MP1 or 5.0MP3 (For Solaris Sparc)


Supported OS
------------
    Solaris 8, 9 and 10


This patch is issued to fix the following incidents:
===================================================
1294441: [VCS 5.0][NFS resources stuck in OFFLINE_WAITING-PROPAGATE]


PATCHED EXECUTABLES DETAILS
---------------------------
$ cksum NFSRestartAgent
564559242      67656   NFSRestartAgent


To install the patch:
---------------------

If VCS is running:

A) On any node in the cluster:

1. Make the VCS configuration writable:

      # haconf -makerw

2. Freeze the service groups where the application resource is running. 

      # hagrp -freeze <service_group> -persistent
3. Save the configuration file (main.cf) with the groups frozen. 

      # haconf -dump -makero     

B) One node at a time, stop VCS (leaving the applications running),
and update the agent:

1. Stop VCS

      # hastop -local -force

2. Remove/rename the following file
       VCS_HOME/bin/NFSRestart/NFSRestartAgent

3. Copy the NFSRestartAgent binary from this patch to
       $VCS_HOME/bin/NFSRestart/NFSRestartAgent

4. Start VCS

      # hastart

5. Repeat steps B1-B4 for every node in the cluster.

C) After updating the agent on each node in the cluster,
on any node in the cluster:

1. Make the configuration writable and unfreeze the service groups.  
       # haconf -makerw

       # hagrp -unfreeze <service_group> -persistent
2. Save the configuration file (main.cf). On any system, type:

       # haconf -dump -makero


If VCS is not running:

A) One node at a time, update the agent:

1. Remove/rename the following file
       $VCS_HOME/bin/NFSRestart/NFSRestartAgent

2. Copy the NFSRestartAgent binary from this patch to
       $VCS_HOME/bin/NFSRestart/NFSRestartAgent

3. Repeat steps A1-A2 for every node in the cluster.