errors from cmcluster

[sameer_4]

I am getting the following error in syslog
cmclconfd[1391]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory

and i have
HPUX11i-OE-MC -> B.11.23 HP-UX Mission Critical

Does this causing the errors .As i have checkeed everything otherwise like /etc/rc.config.d/cmcluster etc .I do not have MCSG . Do i need to uninstall HPUX11i-OE-MC.How to do it , will it have any impact on the system .

[David Child_1]

Sameer,

Since you have the mission critical OE it automatically comes with MC/SG. Since you are not using MC/SG I wonder if you really need the Mission Critical OE.

The only two major products you get with HPUX11i-OE-MC that you don't get with the Enterprise OE are MC/SG and Workload Manager.

Even though you have the Mission Critical OE you should be able to safely remove MC/SG. This should take care of your issue.

David

[Scot Bean]

Yes, this is a normal msg when ServiceGuard is installed, but not configured / setup.

ServiceGuard is packaged with the MC OE.

If you want to turn off the SG daemons, couple of things to do:

- set AUTOSTART_CMCLD to 0 in /etc/rc.config.d/cmcluster

- 'swremove' just the ServiceGuard product

[Fred Ruffet]

Scot, there are two threads open for this problem. Please refer to http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=690092.

Sameer, please close this one.

Regards,

Fred

[Geoff Wild]

Scot has the answer - just set AUTOSTART_CMCLD to 0 in /etc/rc.config.d/cmcluster

Which is good practise even if you want it configured - because if you have had a problem - you might not want the cluster to start right away...

Rgds...Geoff

[Stephen Doud]

When does the message occur? Right after bootup, or later on?
If just after bootup - it may be that someone set AUTOSTART_CMCLD=1, the issue stated before.
But if it occurs sometime after the server has been booted, then someone or some process/cron may be issuing a Serviceguard command.

Example:
# logger ==================================================
# logger ___________ TEST STARTED __________
# cmviewconf | more
cmviewconf: Binary file does not exist.

... and in syslog.log:
Sep 7 10:07:31 eon root: ==================================================
Sep 7 10:07:31 eon root: ___________ TEST STARTED __________
Sep 7 10:07:38 eon cmclconfd[2061]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory

-StephenD.

[sameer_4]

Hi ,
I already have AUTOSTART_CMCLD to 0 in /etc/rc.config.d/cmcluster.But i still see these errors .Till now i have not got any correct solution for this problem.The only solution is to uninstalll service guard ServiceGuard A.11.15.00

[sameer_4]

yes You are right it is normally during the bootup also and later also ( while nomal running )

During bootup

Aug 4 21:37:49 bnhpqa03 inetd[792]: Configuration complete
Aug 4 21:37:53 bnhpqa03 cmclconfd[1116]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory


During normal operation

Aug 11 20:00:15 bnhpqa03 cmclconfd[3024]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory
Aug 11 20:01:02 bnhpqa03 cmclconfd[3054]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory
Aug 11 20:01:29 bnhpqa03 cmclconfd[3074]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory
Aug 11 20:02:03 bnhpqa03 cmclconfd[3119]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory
Aug 11 20:03:12 bnhpqa03 cmclconfd[3140]: Unable to stat /etc/cmcluster/cmclconfig, No such file or directory
Aug 12 07:18:18 bnhpqa03 xntpd[7804]: synchronized to 10.20.9.24, stratum=3

[Prashant Zanwar_4]

Hi.
Besides changing the values at /etc/rc.config.d, also rename the script in /sbin/init.d & | /sbin/rc2.d. This should take care of startup problem.
Otherwise invoke swremove and select product to remove and remove it.
Hope it helps
rgdz
Prashant

[Sridhar Bhaskarla]

Hi,

If you are not running ServiceGuard, I suggest you uncomment hacl-cfg (both lines) in /etc/inetd.conf and refresh inetd "inetd -c".

That should take care of it.

-Sri

[Sridhar Bhaskarla]

Oh.. where am I.. I meant "comment out" instead of "uncomment". Put "#" marks at the start of those two lines and run "inetd -c".

Having them will make the system to respond to any queries that are coming from other hosts. Hence you are seeing those messages.

-Sri

[Stephen Doud]

Thanks for the content of the syslog.log
I see that several occurences of the error message occur during run-time suggesting that some script/process is run at that time of the day, trying to interrogate the content of the cluster. If you don't need Serviceguard on the server, I suggest that you swremove it, to clear out the product.
This will avoid the system touching any of the Serviceguard files during boot-time, inetd.conf configuration or command execution.
MCOE entitles the server to a license to swinstall it if you reconsider later, but check on how to get a codeword in case you decide to reinstall.

-StephenD.