I run Ruby 1.8 indeed. I ended up scheduling a script to check whether the agent hangs and restarts it.
https://gist.github.com/aluetjen/8400615 On Monday, January 13, 2014 8:06:17 AM UTC+1, Toni Schmidbauer wrote: > > At Sun, 12 Jan 2014 08:33:10 -0800 (PST), > Alex wrote: > > * There are two puppet processes. The agent and a child process > > "Puppet: applying configuration" > > * A lck file with the PID of the child process. > > * The lock file is two days old > > * There are no log entries in the agent's log file. > > we've got exactly the same issue here. any chance you are running the > agent with ruby 1.8? > > my assumption is that if the master stops answering requests in a > certain way (in our case we max-ed out httpd processes), there's a bug > in ruby 1.8 where sockets do not get closed. > > https://projects.puppetlabs.com/issues/2089 describes the issue. > > currently we do not have a solution for this (other than upgrading > ruby). as a workaround you could add a cronjob that stops the agent, > kills remaining puppet agent processes and starts the agent again. > > maybe puppetlabs should support rhel software collections because > there's a newer version of ruby included (1.9.3 if i recall > correctly). but you have to change the start/stop script for that. > > regards toni > -- > Don't forget, there is no security | toni at stderr dot at > -- Wulfgar | Toni Schmidbauer > -- 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 [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/puppet-users/8940b140-6dd1-4454-957c-7a81b103c7d2%40googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
