We have a problem with our windows slaves, where they will have a read timed out error, fail, and the jobs running from them show a connection reset by peer error on the master. We'd like to find out why the servers are having read time outs - or if some other thread has failed in them first. Both ends effectively have a ChannelClosedException.
We are thinking that more logging on the slaves in the slave logs or slave error logs might lead us to what the cause is. We are confident that there is not an actual network connectivity issue between these Windows slaves and the Linux master. We are currently a version or 2 behind the LTS - at 1.480.2. The slaves are installed as Windows services. We have other slaves that run on Linux servers as ssh slaves, which do not have the same problem, and other connections to the same windows slaves are not effected. I've been trying to see if there is some option we can pass along to the slave vm when the master starts it to up the logging level - but I've not found anything documented. Is there some option like -Djava.utils.logging.level=INFO or similar? Have people seen this problem on Windows slaves before and found the solution? A quick search of the past mail has shown someone saw this until uninstalling the "disk usage" plugin - we do not have it installed. -- 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 [email protected]. For more options, visit https://groups.google.com/groups/opt_out.
