[Pacemaker] ocf_log

[Dejan Muhamedagic]

Hi,

On Fri, Apr 17, 2009 at 11:28:38AM +0000, Steve Kwee wrote:
> Andrew Beekhof <beekhof at ...> writes:                                             
> 
> > 
> > On Fri, Apr 17, 2009 at 01:04, Steve Kwee <skwee at ...> wrote:
> > > Hi,
> > >
> > > Can someone point me to the proper way to debug OCF RAs ?
> > > Looking  at the source of the existing RAs I found ocf_log and ha_log
> > > and ha_debug.
> > > I followed the source to .ocf-shellfuncs  and /etc/ha.d/shellfuncs
> > > but I did not understand where to configure the logging.
> > > I played around with ha.cf but did not get any of the messages in the
> > > drbd and LVM RAs.
> > 
> > Right, its controlled by the logging settings in ha.cf

It looks like it's openais.

> > Try adding:
> > 
> > logfacility	daemon
> > debug       1
> > 
> > this should ensure the RA logs show up in the same place as the rest
> > of the cluster logging.
> > or, simply log all your test messages as errors so that they are 
> > always printed.
> > 
> > ocf_log err "Some message"
> > 
> > >
> > > Background: I am using linux ha since a week now and at least I have a 
> > > simple
> > > drbd-cluster running. I use the current 1.0.3 pacemaker binaries from 
> > > openSUSE
> > > ha-clustering on openSUSE-11.1
> > > As it seems these binaries support only openAIS and not heartbeat, 
> > > so I go with
> > > that and it seems to work fine.
> > > As basic-configuration for openAIS and ha.cf I used the one from 
>     the wiki at
> > > www.clusterlabs.org
> > >
> > > Thanks
> > >
> > >
> > >
> > > _______________________________________________
> > > Pacemaker mailing list
> > > Pacemaker at ...
> > > http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> > >
> > 
> Thank you for the quick answer.
> 
> In my /etc/ha.cf I now have:
> debug		1
> use_logd	false
> logfacility    	daemon
> 
> I get all crit, error, warning and info messags in the syslog.
> 
> But I have still some questions.
> 
> 1. Is http://www.linux-ha.org/wiki/Ha.cf#debug_-_set_debug_level
> still valid concerning the changing of the debuglevel ?
> I says:
> ..Additionally, the debug level of the system can be dynamically changed by
> sending the heartbeat process SIGUSR1 and SIGUSR2 signals. SIGUSR1 raises the
> debug level, and SIGUSR2 lowers it. A sample debug directive is shown below.
> 
> Which process should the signal be sent to ?
> Converning the logging of RA-Agents the lrmd ?

Right. lrmd will capture and log all output from resource agents.

> If not how do I change the debuglevel ?
> I guess there is a better option than "rcopenais restart"
> 
> 2. Is there a relation between ha.cf/debug and the loglevels 
> and if where can i find it ?

With openais you should take a look at openais.conf (there's a
man page). In the service pacemaker stanza you can set use_logd
to yes and then set the facility in logd.cf.

Thanks,

Dejan


> 3. I was not able to get any message of type debug
> I can force my own test messages to become visibile by using "err"
> 
> How can I get the debug-level messages i.e. from the drbd RA ?
> Messages of level debug go to /etc/ha.d/shellfuncs:ha_debug
> 
> What is the appropriate logging solution and how can I configure it in
> a openAIS / pacemacker setup ?
> 
> I read a lot of hints on using the logger daemon ha_logd in combination with
> ha_logger. Is this still valid and how do I configure it ?
> ha.cf: use_logd	true
> seems not to work
> 
> Sorry for so many questions.
> I can post a summary on logging when things sort out
> 
> 
> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker