On Thu, Nov 29, 2012 at 3:35 PM, GordonJB <[email protected]> wrote: > Hi all, > > I'm currently getting a lot of update failures showing up with no logs in > the dashboard. This happens for about half of our 28 nodes, about once an > hour. When re-run half an hour later, everything seems fine. > > Through a combination of server restarts and our Puppet master locking up > entirely yesterday, almost all of our nodes are now trying to do their half > hourly update at the same time. Could this be the reason half of them fail? > If so, how can I avoid this?
Are you using any webserver in front of Puppet (Apache/Passenger for example)? If not, you're probably running into Webrick's (Puppet's default HTTP server) single-threaded limitation, whereby it can only service a request from a single node at a time. Depending on your client's update schedule (by default it's once every 30 minutes) and number of clients you may end up with multiple clients trying to request a catalog run at the same time. You can either set up Apache/Passenger to sit in front of Puppet for you, thereby enabling more clients to be handled concurrently. Or you can have your clients check in less frequently and schedule them (via cron, for example) so as to avoid multiple clients checking in with the master concurrently (this obviously does not scale well at all). If you want to go the Passenger route, there's a guide at http://docs.puppetlabs.com/guides/passenger.html. Hope this helps, Matt. -- You received this message because you are subscribed to the Google Groups "Puppet Users" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/puppet-users?hl=en.
