Yes, continueOnFail is false per default. And yes, the build will simply fail faster instead of timing out (from the Hudson timeout). I've also tweak a bit the server side, and should be able to diagnose the issue when it happens again. So far, I have no clue what makes the whole thing hanging :(
Thanks a lot for your efforts. continueOnFail ist false per default as with the jarsigner? So the build would simply fail faster instead of timing out? Regards, Arthur Once one or two epp are correctly build in the gerrit job, I think it's safe to interrupt it and merge the change, so you can merge it and try a standard build (which actually sign the dmg, whereas the gerrit build doesn't). Of course, it's up to you if you want to wait for the full gerrit build to be finished ;) Seems it could help both of us ;-) Thanks, Markus
Hi Markus, I've made the change on the CBI side and I'm about to submit a review to EPP which add the timeout. If we can try one build with these, it would help me a lot! Any progress or new insights? I've been checking all the builds in the last days: All of them were failing and timed out after about 6 hours. The currently running build https://hudson.eclipse.org/packaging/job/oxygen.epp-tycho-build/298/console was supposed to be the last one for RC3, but after more than 4 hours it seems to be stuck at a similar position. I'm a bit concerned because at some point we need to decide about alternatives.
However, the Apple TSA may be the weak spot causing the hanging. I will add an option to the macOS signing service to use another TSA (codesign --timestamp)
_______________________________________________epp-dev mailing listepp-dev@xxxxxxxxxxxTo change your delivery options, retrieve your password, or unsubscribe from this list, visithttps://dev.eclipse.org/mailman/listinfo/epp-dev
|