Details:
https://bugzilla.mozilla.org/show_bug.cgi?id=1526979
I would like to land a change which should be fairly transparent, but could
go wrong. I'd like to do so next Tuesday around 9am pacific. The landing
process will require a brief tree closure (just a minute or two) followed
by a few pushes then a slightly longer closure to monitor those pushes and
make sure they're looking OK. If so, reopen. If not, rollback and
reopen. I'd like to apply this just for autoland and inbound. If all goes
well, I can apply for other branches later (with less advance preparation).
Please let me know if this is a good plan, or if you'd like me to adjust
anything.
Briefly, the change replaces the mozilla-taskcluster service, which polls
hg repositories and creates decision tasks when changes are discovered;
with a Taskcluster hook that runs immediately when a change is made, based
on a Pulse message. The externally-visible behavior won't change, but it
gets a bunch of advantages including better visibility into decision-task
issues, faster decision task startup, and getting rid of an old and very
confusing service.
This will eventually apply to all
hg.mozilla.org repos, including nss and
comm-central.
Dustin