Heartbeat dont start automatically on boot-up

rodrigoroma · 03-22-2011, 02:38 PM

Hi, I have configured heartbeat+DRBD on a SLES10 SP3 cluster and that's working fine except when both nodes are down and I turn on only one node. When this happens, the heartbeat don't starts automatically (the service status says it's unused, so the services, IP address and partition don't start automatically too).
I can turn on the heartbeat manually (service heartbeat start) and get the services, ip address and partition initialized too, but I'd like to make it starts automatically because it's possible to have one node broken and the person that will turn on the other node can forget to manually starts the service.

-When the heartbeat is unused on the first node and I turn on the second node, the heartbeat on the first node starts automatically and everything is fine.
-The command "chkconfig heartbeat on" was executed.

Thanks!

alunduil · 03-22-2011, 03:56 PM

What is the output of chkconfig --list | grep -i heartbeat?

If there is no output you may need to do a checkconfig --add heartbeat before chkconfig heartbeat on.

Regards,

Alunduil

rodrigoroma · 03-23-2011, 07:14 AM

The output is:

Code:

heartbeat 0:off 1:off 2:off 3:on 4:off 5:on 6:off

So I think it's ok... What else can I do?

alunduil · 03-23-2011, 07:29 AM

It looks like it should start but just to be certain what's your defualt runlevel and is it possible to reboot as a test?

Regards,

Alunduil

rodrigoroma · 03-23-2011, 10:22 AM

The default runlevel is 3 and I have already rebooted a lot of times.

alunduil · 03-23-2011, 12:08 PM

Is there anything in the log showing why it's failing? Also, does it start normally when you issue /etc/init.d/heartbeat restart?

Regards,

Alunduil

rodrigoroma · 03-29-2011, 12:47 PM

After turning on the server, there's nothing new in the log.

Code:

cnab-1:/var/log # service heartbeat start
Starting High-Availability services2011/03/29_12:55:43 INFO:  Resource is stopped
heartbeat: udpport setting must precede media statementsheartbeat[4430]: 2011/03/29_12:55:43 WARN: Core dumps could be lost if multiple dumps occur.
heartbeat[4430]: 2011/03/29_12:55:43 WARN: Consider setting non-default value in /proc/sys/kernel/core_pattern (or equivalent) for maximum supportability
heartbeat[4430]: 2011/03/29_12:55:43 WARN: Consider setting /proc/sys/kernel/core_uses_pid (or equivalent) to 1 for maximum supportability
heartbeat[4430]: 2011/03/29_12:55:43 info: Version 2 support: false
heartbeat[4430]: 2011/03/29_12:55:43 WARN: Logging daemon is disabled --enabling logging daemon is recommended
heartbeat[4430]: 2011/03/29_12:55:43 info: **************************
heartbeat[4430]: 2011/03/29_12:55:43 info: Configuration validated. Starting heartbeat 2.1.4
                                                                      done

Remember that the heartbeat in the primary server starts automatically when I turn on the second one (and vice-versa).