[Puppet Users] Re: Monitoring the puppetmaster
Pete Emerson pemer...@gmail.com wrote: We've got over 150 hosts hitting the one puppetmaster, and based on what I've seen via searching it seems like we're hitting into scalability issues with Webrick, and the recommendation is to switch to Mongrel or Passenger. Looks to me like Passenger is where the focus is, so I'm working on migrating to 0.25 and Passenger, with multiple master nodes for redundancy and scalability. FWIW, we were also seeing our 0.24 puppetmaster stop responding from time to time, requiring a restart. Since upgrading to 0.25.1rc1, reliability has improved a lot, and the load decreased too, while still using the puppetmaster with the included webrick. Matthias --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Puppet Users group. To post to this group, send email to puppet-users@googlegroups.com To unsubscribe from this group, send email to puppet-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/puppet-users?hl=en -~--~~~~--~~--~--~---
[Puppet Users] Re: Monitoring the puppetmaster
We had this issue while we were using webrick and ended up with the
following in cron running every 15 minutes:
if [ `/usr/bin/puppetlast |grep -v puppetlast |sort -n -k 4 |head -n
1|awk '{print \$4}'` -ge 15 ]; then /etc/init.d/puppetmaster restart ;fi
- Jeff
On 09/22/2009 02:38 PM, Pete Emerson wrote:
I'm using 0.24 with Webrick (in the process of migrating to 0.25 / passenger).
Occasionally, the puppetmasterd becomes unavailable, and we see error
messages along the lines of:
Could not call puppetmaster.getconfig: #Errno::ECONNRESET: Connection
reset by peer
I believe the puppetmasterd does not completely die, so it is still in
the process list.
I'm wondering what a good way to monitor this would be.
I see that I can telnet into port 8140, is there something simple I
can send that would give me an indication that everything is okay or
not?
Any suggestions on monitoring this would be appreciated.
Pete
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
Puppet Users group.
To post to this group, send email to puppet-users@googlegroups.com
To unsubscribe from this group, send email to
puppet-users+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~--~~~~--~~--~--~---
[Puppet Users] Re: Monitoring the puppetmaster
Sounds promising. Did puppetlast appear after 0.24.7-4? I don't see it
on my box.
[r...@admin]# rpm -qa | grep ^puppet
puppet-server-0.24.7-4.el5
puppet-0.24.7-4.el5
[r...@admin]# rpm -ql puppet puppet-server | grep puppetlast
[r...@admin]# find /usr -name puppetlast
[r...@admin]#
Pete
On Tue, Sep 22, 2009 at 12:45 PM, Jeff Adams jeff.ad...@kw.com wrote:
We had this issue while we were using webrick and ended up with the
following in cron running every 15 minutes:
if [ `/usr/bin/puppetlast |grep -v puppetlast |sort -n -k 4 |head -n
1|awk '{print \$4}'` -ge 15 ]; then /etc/init.d/puppetmaster restart ;fi
- Jeff
On 09/22/2009 02:38 PM, Pete Emerson wrote:
I'm using 0.24 with Webrick (in the process of migrating to 0.25 /
passenger).
Occasionally, the puppetmasterd becomes unavailable, and we see error
messages along the lines of:
Could not call puppetmaster.getconfig: #Errno::ECONNRESET: Connection
reset by peer
I believe the puppetmasterd does not completely die, so it is still in
the process list.
I'm wondering what a good way to monitor this would be.
I see that I can telnet into port 8140, is there something simple I
can send that would give me an indication that everything is okay or
not?
Any suggestions on monitoring this would be appreciated.
Pete
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
Puppet Users group.
To post to this group, send email to puppet-users@googlegroups.com
To unsubscribe from this group, send email to
puppet-users+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~--~~~~--~~--~--~---
[Puppet Users] Re: Monitoring the puppetmaster
Ah. I grabbed puppetlast from
http://reductivelabs.com/trac/puppet/attachment/ticket/1188/puppetlast
It doesn't work with 0.24.7-4:
# ./puppetlast
puppetlast
error: undefined method `version' for #Puppet::Node::Facts:0x2c0b8958
However, it's pretty clear (even to my untrained ruby eyes) what this
script does, so it looks like what I can do is simply check the last
modified time of the last yaml file like this:
stat `ls -tr /var/lib/puppet/yaml/facts/*.yaml | tail -n1`
and then check to make sure that's within reasonable limits.
Pete
On Tue, Sep 22, 2009 at 12:55 PM, Pete Emerson pemer...@gmail.com wrote:
Sounds promising. Did puppetlast appear after 0.24.7-4? I don't see it
on my box.
[r...@admin]# rpm -qa | grep ^puppet
puppet-server-0.24.7-4.el5
puppet-0.24.7-4.el5
[r...@admin]# rpm -ql puppet puppet-server | grep puppetlast
[r...@admin]# find /usr -name puppetlast
[r...@admin]#
Pete
On Tue, Sep 22, 2009 at 12:45 PM, Jeff Adams jeff.ad...@kw.com wrote:
We had this issue while we were using webrick and ended up with the
following in cron running every 15 minutes:
if [ `/usr/bin/puppetlast |grep -v puppetlast |sort -n -k 4 |head -n
1|awk '{print \$4}'` -ge 15 ]; then /etc/init.d/puppetmaster restart ;fi
- Jeff
On 09/22/2009 02:38 PM, Pete Emerson wrote:
I'm using 0.24 with Webrick (in the process of migrating to 0.25 /
passenger).
Occasionally, the puppetmasterd becomes unavailable, and we see error
messages along the lines of:
Could not call puppetmaster.getconfig: #Errno::ECONNRESET: Connection
reset by peer
I believe the puppetmasterd does not completely die, so it is still in
the process list.
I'm wondering what a good way to monitor this would be.
I see that I can telnet into port 8140, is there something simple I
can send that would give me an indication that everything is okay or
not?
Any suggestions on monitoring this would be appreciated.
Pete
--~--~-~--~~~---~--~~
You received this message because you are subscribed to the Google Groups
Puppet Users group.
To post to this group, send email to puppet-users@googlegroups.com
To unsubscribe from this group, send email to
puppet-users+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/puppet-users?hl=en
-~--~~~~--~~--~--~---
[Puppet Users] Re: Monitoring the puppetmaster
On Tue, Sep 22, 2009 at 1:38 PM, Pete Emerson pemer...@gmail.com wrote: I'm using 0.24 with Webrick (in the process of migrating to 0.25 / passenger). Occasionally, the puppetmasterd becomes unavailable, and we see error messages along the lines of: Could not call puppetmaster.getconfig: #Errno::ECONNRESET: Connection reset by peer I believe the puppetmasterd does not completely die, so it is still in the process list. I'm wondering what a good way to monitor this would be. I see that I can telnet into port 8140, is there something simple I can send that would give me an indication that everything is okay or not? Any suggestions on monitoring this would be appreciated. Pete Strangely enough, we had a similar situation here. Whenever Puppet (the client) would reload its own configs, it would die and not start again. So we wrote a nice little nagios script that monitored whether puppet was running on each machine. From that, we could restart or send an email or anything we want really. Cheers, Clint --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Puppet Users group. To post to this group, send email to puppet-users@googlegroups.com To unsubscribe from this group, send email to puppet-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/puppet-users?hl=en -~--~~~~--~~--~--~---
[Puppet Users] Re: Monitoring the puppetmaster
We've got over 150 hosts hitting the one puppetmaster, and based on what I've seen via searching it seems like we're hitting into scalability issues with Webrick, and the recommendation is to switch to Mongrel or Passenger. Looks to me like Passenger is where the focus is, so I'm working on migrating to 0.25 and Passenger, with multiple master nodes for redundancy and scalability. Pete On Tue, Sep 22, 2009 at 12:41 PM, Clint Savage her...@gmail.com wrote: On Tue, Sep 22, 2009 at 1:38 PM, Pete Emerson pemer...@gmail.com wrote: I'm using 0.24 with Webrick (in the process of migrating to 0.25 / passenger). Occasionally, the puppetmasterd becomes unavailable, and we see error messages along the lines of: Could not call puppetmaster.getconfig: #Errno::ECONNRESET: Connection reset by peer I believe the puppetmasterd does not completely die, so it is still in the process list. I'm wondering what a good way to monitor this would be. I see that I can telnet into port 8140, is there something simple I can send that would give me an indication that everything is okay or not? Any suggestions on monitoring this would be appreciated. Pete Strangely enough, we had a similar situation here. Whenever Puppet (the client) would reload its own configs, it would die and not start again. So we wrote a nice little nagios script that monitored whether puppet was running on each machine. From that, we could restart or send an email or anything we want really. Cheers, Clint --~--~-~--~~~---~--~~ You received this message because you are subscribed to the Google Groups Puppet Users group. To post to this group, send email to puppet-users@googlegroups.com To unsubscribe from this group, send email to puppet-users+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/puppet-users?hl=en -~--~~~~--~~--~--~---
