Puppet users - Feb 2013 - puppet master processing agent report time is very long, how to optimization

There is a puppet cluster, composed by nginx,passenger,puppet. The number of qps
is 1000 currently,
and the processing speed of the whole cluster is pretty good,the average
processing time for the request is 0.25 seconds.
 But it is a little slow from client report, a lot of processing time is tens of
seconds.such
as nginx log:
 
192.168.0.2 puppet  - [07/Feb/2013:15:45:44 +0800] "PUT 
/production/report/123.test.test.cn HTTP/1.1" 200 24 "-"
"-" "-" 95.690
192.168.12.12:8140 200 *95.876*
 
The Normal request:
192.168.0.2 puppet   - [07/Feb/2013:16:03:52 +0800] "GET 
/production/file_metadata/files/123.py HTTP/1.1" 200 324 "-"
"-" "-" 0.014
192.168.12.12:8140 200 0.105
 
puppet version:2.6.11
nginx version:1.0.9
passenger version:3.0.13 and 3.0.17
 
The Server cluster still have not reached the performance bottleneck
 
nginx proxy buffer is large enough
 
passenger conf :
        passenger_root /usr/lib/ruby/gems/1.8/gems/passenger-3.0.17;
        passenger_ruby /usr/bin/ruby;
        passenger_max_pool_size 150;
        passenger_use_global_queue on;
        passenger_max_requests 4000;
        passenger_pool_idle_time 100;
 
how to optimize?

-- 
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to puppet-users+unsubscribe@googlegroups.com.
To post to this group, send email to puppet-users@googlegroups.com.
Visit this group at http://groups.google.com/group/puppet-users?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

On 07/02/13 08:30, serwei2014@gmail.com wrote:
> There is a puppet cluster, composed by nginx,passenger,puppet. The number
of qps is 1000 currently,
> and the processing speed of the whole cluster is pretty good,the average
processing time for the request is 0.25 seconds.
>  But it is a little slow from client report, a lot of processing time is
tens of seconds.such
> as nginx log:
>  
> 192.168.0.2 puppet  - [07/Feb/2013:15:45:44 +0800] "PUT
> /production/report/123.test.test.cn HTTP/1.1" 200 24 "-"
"-" "-" 95.690
> 192.168.12.12:8140 200 _95.876_
Are you using reports for anything?

By default Puppet enables and stores them.  You can disable storage on
your master with:

[master]
reports 
Or you can disable sending them on your agents with:

[agent]
report = false

If you''re using disk stored reports for something useful, then look at
PuppetDB 1.1 instead, or improve your storage.  If you''re using another
report processor (e.g. Foreman), then you need to optimise that.

-- 
Dominic Cleal
Red Hat Engineering

-- 
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to puppet-users+unsubscribe@googlegroups.com.
To post to this group, send email to puppet-users@googlegroups.com.
Visit this group at http://groups.google.com/group/puppet-users?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

The puppet conf:
master:
reports = log,foreman 
 
agent:
 report = true
 
So  The foreman is  the bottleneck?  How to optimise?
 

On Thursday, February 7, 2013 10:59:02 AM UTC, Dominic Cleal wrote:
> On 07/02/13 08:30, serwe...@gmail.com <javascript:> wrote: 
> > There is a puppet cluster, composed by nginx,passenger,puppet. The 
> number of qps is 1000 currently, 
> > and the processing speed of the whole cluster is pretty good,the
average
> processing time for the request is 0.25 seconds. 
> >  But it is a little slow from client report, a lot of processing time
is
> tens of seconds.such 
> > as nginx log: 
> >   
> > 192.168.0.2 puppet  - [07/Feb/2013:15:45:44 +0800] "PUT 
> > /production/report/123.test.test.cn HTTP/1.1" 200 24
"-" "-" "-" 95.690
> > 192.168.12.12:8140 200 _95.876_ 
>
> Are you using reports for anything? 
>
> By default Puppet enables and stores them.  You can disable storage on 
> your master with: 
>
> [master] 
> reports = 
>
> Or you can disable sending them on your agents with: 
>
> [agent] 
> report = false 
>
> If you''re using disk stored reports for something useful, then
look at
> PuppetDB 1.1 instead, or improve your storage.  If you''re using
another
> report processor (e.g. Foreman), then you need to optimise that. 
>
> -- 
> Dominic Cleal 
> Red Hat Engineering 
>
-- 
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to puppet-users+unsubscribe@googlegroups.com.
To post to this group, send email to puppet-users@googlegroups.com.
Visit this group at http://groups.google.com/group/puppet-users?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

On 08/02/13 03:42, serwei2014@gmail.com wrote:
> The puppet conf:
> master:
> reports = log,foreman
>  
> agent:
>  report = true
>  
> So  The foreman is  the bottleneck?  How to optimise?
It''s typically the database.  Make sure you''re using a
"real" database
such as PostgreSQL and not sqlite (~foreman/config/database.yaml).  If
you''re running the database and Foreman on the same system, check
it''s
powerful enough (not one vCPU and a small amount of RAM), check what
storage it''s on etc.

-- 
Dominic Cleal
Red Hat Engineering

-- 
You received this message because you are subscribed to the Google Groups
"Puppet Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to puppet-users+unsubscribe@googlegroups.com.
To post to this group, send email to puppet-users@googlegroups.com.
Visit this group at http://groups.google.com/group/puppet-users?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.