It wasn't a case where a single phone could be isolated. Restarting adb server on the host fixed it for all devices. I talked to ogra about it yesterday to see if he was aware of any recent changes that could impact this. There were some mtp related things including a fix that was due to land in the next image, but he said those were actually expected to reduce conflicts with adb. Due to some of those changes, the override changes we were putting in to prevent mtp could have issues with the session job. We never saw any issues with that, but with the upcoming change, the override would have no effect at all so I took it out yesterday. I was able to reproduce some adb issues at home once yesterday, and I plan to try some more today. The timing of it happening on the same image made me wonder if there was some issue in that image. So if there is some way to reliably reproduce it, that would probably be the best we can hope for to get help debugging. On Dec 6, 2013 6:49 AM, "Evan Dandrea" <[email protected]> wrote:
> I'm generally all for quick fixes, but that sounds like a recipe for > further problems. Can we isolate one of these phones the next time > this happens, create a bug, and ask the phonedations team to have a > look? > > If that sounds sensible, should we add that to the Vanguard project or > Playbook? > > On 5 December 2013 11:39, Para Siva <[email protected]> wrote: > > On 05/12/13 07:27, Paul Larson wrote: > >> > >> The maguro we use for daily smoke testing (maguro-01) is visible, but > >> marked offline in adb even after the adb restart. Rick, can you take a > >> look at this one? At least we have it down to a single device (I > >> hope). So we don't have maguro results. > > > > > > The smoke on maguro is now on maguro-02 and appears progressing. > > > > AFAIK this adb issue in kinnara is the second in three weeks. I am > curious > > if we could run a cron job to restart adb at a quiet time every day > and/or > > killing very old adb processes to minimise this? > > > > Thanks > > > > > > > > > > > > > > -- > > Mailing list: https://launchpad.net/~canonical-ci-engineering > > Post to : [email protected] > > Unsubscribe : https://launchpad.net/~canonical-ci-engineering > > More help : https://help.launchpad.net/ListHelp > > -- > Mailing list: https://launchpad.net/~canonical-ci-engineering > Post to : [email protected] > Unsubscribe : https://launchpad.net/~canonical-ci-engineering > More help : https://help.launchpad.net/ListHelp >
-- Mailing list: https://launchpad.net/~canonical-ci-engineering Post to : [email protected] Unsubscribe : https://launchpad.net/~canonical-ci-engineering More help : https://help.launchpad.net/ListHelp
