Bingo! I set the timeout to 15s on server and agent both (some of our network links are pretty crappy) and now everyone's connecting as before. Thanks for the tip!
On Wednesday, October 1, 2014 6:24:07 PM UTC-4, Josh Cooper wrote: > > > > On Wed, Oct 1, 2014 at 10:47 AM, Bret Wortman <br...@thewortmans.org > <javascript:>> wrote: > >> I guess not. I upgraded the server to match but the problem persists. >> >> >> On Wednesday, October 1, 2014 1:38:16 PM UTC-4, Bret Wortman wrote: >>> >>> We're running through Passenger and Apache. Puppetdb back-end. Puppet >>> 3.6.2-1 on F20 on the server, Puppet 3.7.1-1 C6.5 on the client. Hmmmm. >>> Could the newer client be the problem? >>> >>> On Wednesday, October 1, 2014 1:31:00 PM UTC-4, Henrik Lindberg wrote: >>>> >>>> On 2014-01-10 19:21, Bret Wortman wrote: >>>> > This is happening on some of my clients. It'll happen about 1 out of >>>> 20 >>>> > runs, but it's really getting to be a problem. There's nothing >>>> > indicative in /var/log/messages or any other server-side log that I >>>> can >>>> > find, and this is all the output I'm seeing on the client: >>>> > >>>> > # puppet agent - t >>>> > Info: Retrieving pluginfacts >>>> > Info: Retrieving plugin >>>> > Info: Loading facts >>>> > Error: Could not retrieve catalog from remote server: end of file >>>> reached >>>> > Warning: Not using cache on failed catalog >>>> > Error: Could not retrieve catalog; skipping run >>>> > # >>>> > >>>> > Any guidance on how to get more details out of Puppet about what's >>>> going >>>> > wrong here? It's persistent, but again, doesn't happen 100% of the >>>> time. >>>> > And it's only happening on certain agents, so I'm pretty sure it's >>>> > related to a module somewhere that's doing something wrong, but for >>>> the >>>> > life of me, I can't suss out which module is causing the problem. >>>> > >>>> > >>>> > Bret Wortman >>>> > >>>> >>>> Are you using Webrick ? If so, there are versions of Puppet 3x that >>>> have >>>> issues with concurrent operation (more than one agent at the same time) >>>> that can cause all kinds of weird behavior - typically resulting in >>>> truncated streams. >>>> >>>> Try latest 3.6, or 3.7. The best fix is to stop using Webrick in favor >>>> of Passenger. Webrick is really not for production use. >>>> >>>> - henrik >>>> >>>> -- >>>> >>>> Visit my Blog "Puppet on the Edge" >>>> http://puppet-on-the-edge.blogspot.se/ >>>> >>>> -- >> 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...@googlegroups.com <javascript:>. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/puppet-users/f159d59a-aa57-4648-8e34-cb39686469fa%40googlegroups.com >> >> <https://groups.google.com/d/msgid/puppet-users/f159d59a-aa57-4648-8e34-cb39686469fa%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> >> For more options, visit https://groups.google.com/d/optout. >> > > You may be running into https://tickets.puppetlabs.com/browse/PUP-3238. > Is the passenger KeepAliveTimeout shorter than the the puppet agent's > timeout? If so, the server may close an idle connection, e.g. while the > agent is running facter. > > See also > https://docs.puppetlabs.com/puppet/3.7/reference/release_notes.html#performance-improvements > > Josh > > -- > Josh Cooper > Developer, Puppet Labs > -- 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/751f4b93-383a-44d1-9ee5-e6539cc62559%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
