[mdt-papyrus.dev] Papyrus HIPP back to normal (Hopefully!)

[LETAVERNIER Camille <Camille.LETAVERNIER@xxxxxx>]

Hi all,

 

As you probably have noticed, we’ve had some trouble with our Hudson instance for the past weeks, which became completely unresponsive and required frequent restarts.

 

It seemed to be related to a deadlock in SCM polling for Hudson 3.2.x. We can’t update to Hudson 3.3.x yet, so I’ve changed the SCM polling timers to avoid having 2 jobs polling Git at the same time (Thus avoid any deadlock).

 

Note that this is just a workaround, so please be careful when configuring new jobs: if you use SCM polling as a job trigger, choose a time that is not already used by another job (Be extra-careful when copy-pasting jobs for specific feature branches)

 

Thanks,
Camille