On Tue, 2013-01-29 at 13:26 +0200, Gary Kotton wrote:
On 01/29/2013 01:02 PM, Mark McLoughlin wrote:
Hi Dave,
On Tue, 2013-01-29 at 10:31 +0000, Daviey Walker wrote:
Hi Mark,
Hope you are well.
You've probably noticed that I accepted a few merges to stable/folsom. I spotted on a comment for one that stable/folsom was in freeze until this Thursday. I didn't realise this, which is why you will have seen a few land today.
I apologise for this, and whilst i'm confident they were safe and the regression potential is low; I am willing to help solve any issues this causes (Including backing them out). Thanks for the heads-up. Partly my bad, I should have been keeping a closer watch out.
My understanding is that the release date is the 31st. I do not recall any update about the stable branch being in a frozen state. At last nights Quantums meeting I mentioned that I was planning to backport some patches and there were no objections.
On Jan 22: http://lists.openstack.org/pipermail/openstack-stable-maint/2013-January/000... 2012.2.3 is scheduled for January 31st. I'm planning on calling whatever we have on January 25th a RC and getting aggressive about punting anything risky until the next release. and: http://eavesdrop.openstack.org/meetings/project/2013/project.2013-01-22-21.0... 2012.2.3 scheduled for Jan 31, RC Jan 24 I -2ed some Nova changes on Friday: https://review.openstack.org/20485 https://review.openstack.org/20337
The process isn't perfect, but I'm reluctant to get into the complication of doing the equivalent of milestone-proposed for stable releases. We'll probably end up doing that eventually, though.
I do not think that we need to be extreme about it, but a sync a few days before to understand that we are all on the same page would suffice.
That's what we did for 2012.2.1: http://lists.openstack.org/pipermail/openstack-stable-maint/2012-November/00... and ended up with a regression in Nova: http://lists.openstack.org/pipermail/openstack-stable-maint/2012-December/00... A week really is a fairly minimal amount of time to lock things down before a release.
The three changes merged since we froze were:
https://review.openstack.org/19264 nova: Fix to include error message in instance faults
https://review.openstack.org/20644 quantum: LinuxBridge: update status according to admin_state_up
This is not a MUST have for this release. It is very nice to have an improves the usability. What is the risk - very low. It would be nice to have https://review.openstack.org/#/c/20677/ in conjunction with this one. Please note that the patch https://review.openstack.org/#/c/20676/ depends on the aforementioned patch.
Please add your justifications to the review - i.e. how user-visible the issue, how low risk the fix is, how well tested it is etc. Thanks, Mark.