We have masters and slaves set up in VPC at my job. The only thing to make sure of is that the master / slave communication port is open in your security groups; this should be set to a fixed value via the "TCP port for JNLP slave agents" field in http://yourjenkins/configureSecurity (even though the field description mentions JNLP, this port is used for SSH connection as well).
Are your slaves running in the same VPC as your master (this is the case for me) ? If not, there might be some additional network access issues. On Thursday, May 22, 2014 5:05:25 AM UTC-7, Guillaume Boucherie wrote: > > After more research and test I find that the problem is link to AWS VPC. > I'm did a simple test, running two instance in AWS (one for master and one > for slave) and create a job that simply echo 1000 times "Lorem ipsum dolor > sit amet, consectetur adipiscing elit. Maecenas et.". > When I did this on AWS VPC I can reproduce the problem after 2-3 job > execution, on classic EC2 its never failed. > > Does anyone have already trying to setup a Jenkins master/slave in AWS VPC > ? > > Thanks > > Guillaume Boucherie > > > Le lundi 5 mai 2014 11:32:23 UTC+2, Guillaume Boucherie a écrit : >> >> Hi, >> >> I'm facing very unstable ssh connection between master/slaves on my >> Jenkins installation. >> Both master and slaves are AWS instance that run on AWS Linux system. >> After some time the ssh connection stop, current job that run are freeze >> and failed after 15min. >> I attached the stacktrace from jenkins.log >> >> Does anybody face the same problem ? >> Maybe its related to ssh configuration ? >> I'm running Jenkins 1.554.1 and salve version used is 2.36. >> >> Thanks >> Guillaume Boucherie >> > -- 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.
