Re: [Puppet Users] PuppetDB low catalog-duplication rate Puppet DB 4.3.0

2017-06-28 Thread Christopher Wood
I had a broadly similar issue in that I had a low catalog duplication rate and 
I had to change some puppet manifests around to fix that.

Back in 2015 I was doing this to get mcollective plugin sources for the file 
resource:

source => regsubst(keys($plugins), '^', 'puppet:///modules/mco/plugins/')

But obviously keys() returns things in any old order and every catalog was 
different. The solution was to sort it:

source => regsubst(sort(keys($plugins)), '^', 'puppet:///modules/mco/plugins/')

In your place I would grab some different catalogs for the same host, 
pretty-format them and diff to see what's different. That will show you what's 
changing between runs. The easy way for this is to do a bunch of agent runs and 
then use curl in between them on the puppetdb host.

curl http//localhost:8080/pdb/query/v4/catalogs/host.domain.com | python -m 
json.tool >/tmp/cat1

On Wed, Jun 28, 2017 at 11:11:17AM -0700, Mike Sharpton wrote:
>Hey all,
>I am hoping there is someone else in the same boat as I am.  We are
>running Puppet 4.2.2, along with PuppetDB 4.3.0.  I am seeing low
>duplication rate which I think is contributing to our queuing problems in
>PuppetDB.  The queue will fluctuate from 0-100 queued, to up to 2000.  We
>have around 4500 nodes, and we are using 8 threads on our PuppetDB server.
> I am seeing that the low duplication rate is caused by hashes not
>matching and a full insert running which is expensive on the DB instead of
>just updating the time stamp.  I don't know why these would not be
>matching, and may need help as far as how to find something like this.  I
>see items in PuppetDB3 for this, but not 4.  I see that using timestamp
>and other items which change each time will cause the catalog to never be
>the same, but I would think we would have 0% duplication if this was the
>case.  I am also seeing that things are improved in 4.4.0 as far as
>performance and a missing index is corrected that may speed things.  I am
>wondering what others have done/seen with this and whether upgrading to
>4.4.0 would do me good.  I am thinking it would as many things appear to
>fixed around the issues I am seeing.  Thanks in advance,
>Mike
> 
>--
>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 [1]puppet-users+unsubscr...@googlegroups.com.
>To view this discussion on the web visit
>
> [2]https://groups.google.com/d/msgid/puppet-users/bde7abd4-fccb-420b-b3d8-d4c674ca5705%40googlegroups.com.
>For more options, visit [3]https://groups.google.com/d/optout.
> 
> References
> 
>Visible links
>1. mailto:puppet-users+unsubscr...@googlegroups.com
>2. 
> https://groups.google.com/d/msgid/puppet-users/bde7abd4-fccb-420b-b3d8-d4c674ca5705%40googlegroups.com?utm_medium=email&utm_source=footer
>3. https://groups.google.com/d/optout

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/puppet-users/20170628192618.5kwip7wdxr62hajo%40iniquitous.heresiarch.ca.
For more options, visit https://groups.google.com/d/optout.


[Puppet Users] PuppetDB low catalog-duplication rate Puppet DB 4.3.0

2017-06-28 Thread Mike Sharpton
Hey all,

I am hoping there is someone else in the same boat as I am.  We are running 
Puppet 4.2.2, along with PuppetDB 4.3.0.  I am seeing low duplication rate 
which I think is contributing to our queuing problems in PuppetDB.  The 
queue will fluctuate from 0-100 queued, to up to 2000.  We have around 4500 
nodes, and we are using 8 threads on our PuppetDB server.  I am seeing that 
the low duplication rate is caused by hashes not matching and a full insert 
running which is expensive on the DB instead of just updating the time 
stamp.  I don't know why these would not be matching, and may need help as 
far as how to find something like this.  I see items in PuppetDB3 for this, 
but not 4.  I see that using timestamp and other items which change each 
time will cause the catalog to never be the same, but I would think we 
would have 0% duplication if this was the case.  I am also seeing that 
things are improved in 4.4.0 as far as performance and a missing index is 
corrected that may speed things.  I am wondering what others have done/seen 
with this and whether upgrading to 4.4.0 would do me good.  I am thinking 
it would as many things appear to fixed around the issues I am seeing. 
 Thanks in advance,

Mike

-- 
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+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/puppet-users/bde7abd4-fccb-420b-b3d8-d4c674ca5705%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.