* * README - VERITAS Cluster Server 5.0MP1 (Solaris/sparc) * *
* * Patches 139349-01, 139350-01, 139351-01 * *
Patch Names/Numbers : 139349-01, 139350-01, 139351-01
-----------------------------------------------------
Date:
-----
2008-10-13
OS:
---
Solaris
OS Version:
-----------
All supported Solaris versions for VCS5.0MP1 on the sparc architecture (8, 9 and 10)
139349-01 (Solaris 8)
139350-01 (Solaris 9)
139351-01 (Solaris 10)
Packages:
---------
VRTSvcs
---------------------
Etrack Incidents: 1216309, 1282209, 1285122, 1280144, 643021, 1061056, 1225536, 631644, 1187581, 1161339, 1133171, 597530
-----------------
1216309 SxRT5.0MP3:I18N: VCS: hares -display - attribute - date format doesn't respect locale settings on JPN
1282209 Output from action EP is not always visible to the caller.
1285122 Race condition for MSG_RES_DELETE followed by MSG_RES_ADD
1280144 Subroutines required to set and reset locale in ag_i18n_inc module
643021 VCSAgExecInContainer() should not create the requested child process if the container is down
1061056 Child process of Agents' should not make calls to logging functions of VCSAgLog object
1225536 Reset OnlineRetryCount, OnlineWaitCount and OfflineWaitCount when resource is flushed
631644 Cookie APIs don't provide atomic clearing of cookie
1187581 ActionTimeout attribute issues
1161339 Application agent does not inherit user defined LANG parameter
1133171 haconf -dump does not change the SRDF NumThreads variable
597530 Need for an OfflineWaitLimit attribute that works the same way as the OnlineWaitLimit
Fixes Applied for Products:
---------------------------
Veritas Cluster Server for Solaris (sparc)
Install/Uninstall Instructions:
-------------------------------
BEFORE GETTING STARTED
----------------------
This patch is for VCS 5.0MP1 on Solaris sparc systems. The patch fixes the list of problems with the VCS agent framework given by the list above.
If you are running any VCS version below VCS 5.0MP1, you will have to upgrade
to VCS 5.0MP1 before applying this patch.
If you are already running VCS 5.0MP1, you can apply this patch directly
(guidelines below).
NOTE: This patch should only be applied on top of VCS 5.0MP1. VCS 5.0MP3
obsoletes this patch. VCS 5.0MP3 can be applied on top of this patch as
VCS5.0MP3 would contain either the same or a newer version of the binary files
contained in this patch.
If VCS5.0MP3 is available, one can and should directly apply VCS5.0MP3 on top
of VCS5.0GA or VCS5.0MP1 and does not need to apply this point-patch.
PATCH'S BASE VERSION
--------------------
VCS 5.0MP1
INSTALLING THE VCS PATCH
------------------------
To install the patch:
1. Copy the point patch to a directory on your system:
# cp <patch.gz> .
2. Unzip the compressed patch files that you want to install:
# gunzip *.gz
3. Extract the compressed patch files from the tar files:
# tar -xvf <filename.tar>
For example, for patch 139349-01, enter:
# tar -xvf 139349-01.tar
4. Make the VCS configuration writable. On any system, type:
# haconf -makerw
5. Freeze all service groups. On any system, type:
# hagrp -freeze <service_group> -persistent
6. Verify that service groups are frozen. On any system, type:
# hastatus -sum
7. Save this configuration change to the main.cf file with the groups frozen.
On any system, type:
# haconf -dump -makero
8. Stop VCS. On any system, type:
# hastop -all -force
To apply the patch one system at-a-time, stop VCS on a single system using
the following command:
# hastop -local -force
9. Check that VCS has shut down. On each system, type:
# ps -ef | grep vcs
# ps -ef | grep -i vrts
If the VCS CmdServer process is running, it can be left running.
10. Change directory to the patch location.
On each node in the cluster, install the patch:
# cd <patch_location>
# patchadd 139349-01
11. After installation is complete, verify that the patch has been installed.
On each system, type:
# showrev -p | grep 139349-01
12. Restart VCS. On each system in the cluster, type:
# hastart
If you chose to apply the patch one system at-a-time, after restarting VCS
on the system that was patched, go to step 8 and repeat the patching
procedure for each individual system.
13. After VCS has been restarted on all the systems, verify all resources have
been probed. On any system, type:
# hastatus -summary
14. Make the configuration writable and unfreeze all service groups.
On any system, type:
# haconf -makerw
# hagrp -unfreeze <service_group> -persistent
15. Save the changes to the configuration file (main.cf). On any system, type:
# haconf -dump -makero
REMOVING THE PATCH
-------------------
To remove the patch from a running VCS cluster:
1. Make the VCS configuration writable. On any system, type:
# haconf -makerw
2. Freeze the service groups. On any system, type:
# hagrp -freeze <service_group> -persistent
3. Save the configuration file (main.cf) with the groups frozen.
On any system, type:
# haconf -dump -makero
4. Stop VCS. On any system, type:
# hastop -all -force
To remove the patch one system at-a-time, stop VCS on a single system using
the following command:
# hastop -local -force
5. On each node in the cluster, type the following command:
# patchrm 139349-01
6. Verify that the patch has been removed from each system.
For example on a Solaris 8 machine, type:
# showrev -p | grep 139349-01
It should not list the patch on this system.
7. Restart VCS. On each system in the cluster, type:
# hastart
If you chose to remove the patch one system at-a-time, after restarting VCS
on the system where the patch was removed, go to step 4 and repeat the
procedure for each individual system.
8. After restarting VCS on all the systems, verify all resources have been probed.
On any system, type:
# hastatus -summary
9. Make the configuration writable and unfreeze all service groups.
On any system, type:
# haconf -makerw
# hagrp -unfreeze <service_group> -persistent
10. Save this change to the configuration file (main.cf). On any system, type:
# haconf -dump -makero