Re: [webkit-dev] Commit Queue Outage

2011-04-15 Thread Eric Seidel
CQ backlog is cleared. Recent CQ changes of note: 1. The CQ now runs with --exit-after-N-failures=10 instead of 1. 2. The CQ now knows how to upload layout-test-results.zip files when tests fail during a commit run. 3. The CQ can now land when the tree is red with up to 9 failures. (Keeps a lis

Re: [webkit-dev] [webkit-changes] [83955] trunk

2011-04-15 Thread Adam Roben
On Apr 15, 2011, at 1:19 PM, Oliver Hunt wrote: > Don't roll out patches that don't break any core builders, without making > some attempt to diagnose the problem or providing information that will make > it possible to fix the problem. The GTK bots are core builders. -Adam __

Re: [webkit-dev] [webkit-changes] [83955] trunk

2011-04-15 Thread Oliver Hunt
Indeed, i missed the gtk release one, but the point stands that just rolling out a patch (which should not roll out the changelogs) without providing any information about the crash doesn't help anyone. --Oliver On Apr 15, 2011, at 10:30 AM, Adam Roben wrote: > On Apr 15, 2011, at 1:19 PM, Oli

Re: [webkit-dev] Commit Queue Outage

2011-04-15 Thread Adam Barth
While we're discussing commit-queue infrastructure changes, I should mention that we're thinking about moving some or all of the commit-queue nodes over to EC2. Moving to EC2 means can use faster machines and we can scale up or down the number of machines easily. Currently, we run the commit-queue

Re: [webkit-dev] [webkit-changes] [83955] trunk

2011-04-15 Thread Adam Barth
As everyone knows, I'm a proponent of rolling things out more frequently, but that change looks huge. I think we should have a bias against rolling out large changes because they're much harder and disruptive to roll in and out than small changes. (Of course, we should also have a bias against ac

Re: [webkit-dev] Commit Queue Outage

2011-04-15 Thread Evan Martin
I think what we ended up doing for Chrome is virtualize the non-Mac builders but still use hardware for the Mac buildbots. It means our Mac builders sometimes lag behind the other platforms (because it is more costly in terms of both resources and people to add additional Mac builders) but we at l

Re: [webkit-dev] Commit Queue Outage

2011-04-15 Thread Ryosuke Niwa
On Fri, Apr 15, 2011 at 10:50 AM, Adam Barth wrote: > > That means we'd likely need to switch the commit-queue over to using one of > the Linux ports. The consequences of switching ports are somewhat difficult > to foresee, which is why we're being cautious. While this might result in less test

Re: [webkit-dev] Commit Queue Outage

2011-04-15 Thread Eric Seidel
The model we're working towards would involve running tests on some sub-set of platforms (certainly mac!) before landing. This would involve making the EWSes run tests and making the commit-queue dependent on their output. This is all in a very experimental stage at the moment. -eric On Fri, Ap