On 2013-11-18 7:17 AM, Ed Morley wrote:
On 16/11/2013 15:17, smaug wrote:
the recent OOM cases have been really annoying. They have slowed down
development, even for those who
haven't been dealing with the actual issue(s).
Could we handle this kind of cases differently. Perhaps clone the bad
state of m-i to
some other repository we're tracking using tbpl, backout stuff from m-i
to the state where we can
run it, re-open it and do the fixes in the clone.
Unfortunately as Nick mentioned - this wasn't possible, otherwise we
would just have performed a backout similar to those performed several
times a day when something breaks the tree in a more 'normal' way.
The closure was due to a seemingly chronic issue, that had only been
highlighted by recent landings (and no one particular landing, since the
one backout that was performed, still didn't make the failures disappear
entirely). Even if we had just reverted the last week's worth of changes
it would not fix the root cause - which was that any single patch could
potentially tip us off the edge of being OOM again.
But we still reopened without the root cause being fixed, didn't we?
What am I missing?
Cheers,
Ehsan
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
