Puppet caught TERM; calling stop - error

[Will S. G.]

I'm running puppet on CentOS 6.3 x86_64:

• facter.i386           1:1.6.12-2.el6    @puppet
• puppet.noarch         2.7.19-1.el6      @puppet
• ruby-shadow.x86_64    1.4.1-13.el6      @puppet

Puppetmaster:

• facter.x86_64         1:1.6.12-2.el6    @puppet
• hiera.noarch          1.0.0-2.el6       @puppet
• puppet.noarch         2.7.19-1.el6      @puppet
• puppet-dashboard.noarch 1.2.11-1.el6      @puppet
• puppet-server.noarch  2.7.19-1.el6      @puppet
• ruby-mysql.x86_64     2.8.2-1.el6       @puppet
• ruby-shadow.x86_64    1.4.1-13.el6      @puppet

I keep getting seeing the following error on my puppet agent:

Wed Oct 03 23:25:43 -0700 2012 Puppet (notice): Starting Puppet client version 2.7.19
Wed Oct 03 23:26:11 -0700 2012 Puppet (notice): Caught TERM; calling stop
Wed Oct 03 23:26:16 -0700 2012 Puppet (notice): Reopening log files
Wed Oct 03 23:26:55 -0700 2012 Puppet (err): Could not retrieve catalog from remote server: Error 400 on SERVER: could not obtain a database connection within 5 seconds.  The max pool size is currently 5; consider increasing it.

I can't seem to figure out what's wrong. What's even more weird is that only some of the agents are displaying this error.

Any suggestions?

[Will S. G.]

One other thing I noticed, /etc/sysconfig/puppet was rewritten with default values on the hosts that are experiencing this issue. I killed the proc, then wrote the values that I use on the working agents, but I continue to get the same error message. Is there anyone else that has seen this? I've been at it for a while, I cannot figure this out.

[Will S. G.]

This may be related to the latest kernel: 2.6.32-279.9.1.el6.x86_64. My other CentOS 6.3 hosts with 2.6.32-279.5.2.el6.x86_64 kernel do not experience this issue. Anyone else have this problem?

[Garrett Honeycutt]

On 10/4/12 8:32 AM, Will S. G. wrote:
> Wed Oct 03 23:26:55 -0700 2012 Puppet (err): Could not retrieve catalog
> from remote server: Error 400 on SERVER: could not obtain a database
> connection within 5 seconds. The max pool size is currently 5; consider
> increasing it.

The puppet master is attempting to connect to a database to get
information needed to compile the catalog and is timing out. Try setting
playing around with setting dbconnections = some_positive_integer in
your puppet.conf. If you are still having issues, post your puppet.conf
in your reply.

Regards,
-g

--
Garrett Honeycutt

206.414.8658
http://puppetlabs.com

[Will S. G.]

Hi Garrett,

I ended up using mod_passenger with Apache for *:8140 to address the database connection and load issue. I no longer see that particular error. However, there is still that pesky Caught TERM; calling stop error, which seems to be related to the recent kernel upgrade. Any thoughts?

[Jo Rhett]

On Oct 10, 2012, at 11:37 AM, Will S. G. wrote:

However, there is still that pesky Caught TERM; calling stop error, which seems to be related to the recent kernel upgrade. Any thoughts?

Look for resources which would notify => Service['puppet']. Run puppet with --debug and you'll see what resource triggered it.

(or whatever service is restarting, if its not puppet)

-- 

Jo Rhett

Net Consonance : net philanthropy to improve open source and internet projects.