It seems like you might first want to consider if you really need to poll. Refer to "polling must die", the blog posting from Kohsuke Kawaguchi, for hints on how to configure Jenkins to respond more rapidly to changes, and to generate less load on your git servers as well.
If you still are committed to polling, then you probably should read the log files on the server to confirm that the polling failures are being reported in server log files. If they are reported in server log files, then you can focus your efforts on finding and increasing the limits which cause the errors in the server log files. If the polling failures do not result in entries in server log files, then you may need to investigate the messages from the git polling logs to see if the limit is in the computer hosting the Jenkins master server. If you use slave nodes, and if the polling failures are in the Jenkins server log, then you could add "Polling requires a workspace" as an additional behavior on your jobs. That will move the polling to the slave computer which has the workspace. It doesn't reduce the amount of polling, but it can reduce the amount of polling from a single machine. Mark Waite On Mon, Aug 11, 2014 at 6:10 PM, Grace Lee <g...@proofpoint.com> wrote: > > > *Hi,* > > > > *We often saw git server connection error when doing git polling in > Jenkins. But we didn’t see this error outside of Jenkins.* > > > > *We have 144 Jenkins projects doing SCM polling, while 136 of them doing > git polling.* > > > > *I had ‘Max #of concurrent polling’ configured to be 40 or 60, they didn’t > make much of difference.* > > > > *(I also did side-by-side monitor of the git polling log for same project > on production Jenkins server and test server for about an hour . Only on > the production server, the git connection error was seen. Both the servers > are on same version of Jenkins and plugins. But on test server only this > project was enabled.) * > > > > *Does anyone have any suggestion on how to resolve the git connection > issue on SCM polling?* > > > > *Thanks.* > > -- > You received this message because you are subscribed to the Google Groups > "Jenkins Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to jenkinsci-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/d/optout. > -- Thanks! Mark Waite -- You received this message because you are subscribed to the Google Groups "Jenkins Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to jenkinsci-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
