stable/grizzly freeze WAS Preparing for 2013.1.1 stable/grizzly
Hi all,
First stable/grizzly release 2013.1.1 is planned for May 9 - early release to pick up critical Grizzly fixes as suggested by Daviey. That puts stable freeze date on May 2, which is already next week!
please consider stable/grizzly frozen now and reserve +2 for critical issues only. There are few open Nova stable reviews but Russell said it's nothing critical, please speak up now if you don't agree. I'll be sending wider call-for-testing announcement shortly. Thanks, Alan
2013/5/3 Alan Pevec <apevec@gmail.com>:
Hi all,
First stable/grizzly release 2013.1.1 is planned for May 9 - early release to pick up critical Grizzly fixes as suggested by Daviey. That puts stable freeze date on May 2, which is already next week!
please consider stable/grizzly frozen now and reserve +2 for critical issues only. There are few open Nova stable reviews but Russell said it's nothing critical, please speak up now if you don't agree.
I got two requests for exceptions on IRC: * Horizon, from Julie: seems to be minor issue but potentially confusing for the first time users, so I'll approve it https://review.openstack.org/#/c/28139/ * Keystone, from Dolph: no patch yet, critical issue "blurring the line between bug fix and feature change" https://bugs.launchpad.net/keystone/+bug/1175838 Cheers, Alan
On 05/03/2013 01:16 PM, Alan Pevec wrote:
2013/5/3 Alan Pevec <apevec@gmail.com>:
Hi all,
First stable/grizzly release 2013.1.1 is planned for May 9 - early release to pick up critical Grizzly fixes as suggested by Daviey. That puts stable freeze date on May 2, which is already next week! please consider stable/grizzly frozen now and reserve +2 for critical issues only. There are few open Nova stable reviews but Russell said it's nothing critical, please speak up now if you don't agree. I got two requests for exceptions on IRC:
* Horizon, from Julie: seems to be minor issue but potentially confusing for the first time users, so I'll approve it https://review.openstack.org/#/c/28139/
* Keystone, from Dolph: no patch yet, critical issue "blurring the line between bug fix and feature change" https://bugs.launchpad.net/keystone/+bug/1175838
I'd hold out for this one. LDAP is broken without it.
Cheers, Alan
2013/5/4 Adam Young <ayoung@redhat.com>:
On 05/03/2013 01:16 PM, Alan Pevec wrote:
2013/5/3 Alan Pevec <apevec@gmail.com>:
Hi all,
First stable/grizzly release 2013.1.1 is planned for May 9 - early release to pick up critical Grizzly fixes as suggested by Daviey. That puts stable freeze date on May 2, which is already next week!
please consider stable/grizzly frozen now and reserve +2 for critical issues only. There are few open Nova stable reviews but Russell said it's nothing critical, please speak up now if you don't agree.
I got two requests for exceptions on IRC:
* Horizon, from Julie: seems to be minor issue but potentially confusing for the first time users, so I'll approve it https://review.openstack.org/#/c/28139/
* Keystone, from Dolph: no patch yet, critical issue "blurring the line between bug fix and feature change" https://bugs.launchpad.net/keystone/+bug/1175838
I'd hold out for this one. LDAP is broken without it.
Above bug was identified as a duplicate of https://bugs.launchpad.net/keystone/+bug/1168726 and Dolph is working on a patch https://review.openstack.org/28197 which hopefully will be ready for backporting soon. BTW regarding freeze period: I noticed 3 Cinder patches were merged post-freeze, without explanation: https://review.openstack.org/28199 https://review.openstack.org/28205 https://review.openstack.org/28206 They all seem like fine backports for stable, but we still need some kind of freeze period before the release. Please let's discuss any freeze exceptions on this list or at least give justification message when approving so it doesn't look it was accidental. Freeze is currently not enforced in Gerrit, but I think I'm now going to put temporary -2 to all open reviews, just in case. Thanks, Alan
On Sun, May 5, 2013 at 5:11 PM, Alan Pevec <apevec@gmail.com> wrote:
2013/5/4 Adam Young <ayoung@redhat.com>:
On 05/03/2013 01:16 PM, Alan Pevec wrote:
2013/5/3 Alan Pevec <apevec@gmail.com>:
Hi all,
First stable/grizzly release 2013.1.1 is planned for May 9 - early release to pick up critical Grizzly fixes as suggested by Daviey. That puts stable freeze date on May 2, which is already next week!
please consider stable/grizzly frozen now and reserve +2 for critical issues only. There are few open Nova stable reviews but Russell said it's nothing critical, please speak up now if you don't agree.
I got two requests for exceptions on IRC:
* Horizon, from Julie: seems to be minor issue but potentially confusing for the first time users, so I'll approve it https://review.openstack.org/#/c/28139/
* Keystone, from Dolph: no patch yet, critical issue "blurring the line between bug fix and feature change" https://bugs.launchpad.net/keystone/+bug/1175838
I'd hold out for this one. LDAP is broken without it.
Above bug was identified as a duplicate of https://bugs.launchpad.net/keystone/+bug/1168726 and Dolph is working on a patch https://review.openstack.org/28197 which hopefully will be ready for backporting soon.
BTW regarding freeze period: I noticed 3 Cinder patches were merged post-freeze, without explanation: https://review.openstack.org/28199 https://review.openstack.org/28205 https://review.openstack.org/28206
They all seem like fine backports for stable, but we still need some kind of freeze period before the release. Please let's discuss any freeze exceptions on this list or at least give justification message when approving so it doesn't look it was accidental. Freeze is currently not enforced in Gerrit, but I think I'm now going to put temporary -2 to all open reviews, just in case.
Thanks, Alan
_______________________________________________ Openstack-stable-maint mailing list Openstack-stable-maint@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-stable-maint
Hey Folks, So it looks like I missed the stable release freeze date on a number of patches. This is completely my fault, for some reason I had the wrong freeze date in mind and thought I had until today. Anyway, there are a three (review links below) fixes that were proposed for Cinder last week that I'd like to ask for folks to consider an exception if possible. I've made sure that said changes don't introduce any huge modifications this time around and are very minimal risk, and they do address real valid customer reported bugs. It would be a shame to have to wait another round to get these in. Thanks, John https://review.openstack.org/#/c/28212/ https://review.openstack.org/#/c/28207/ https://review.openstack.org/#/c/28210/
2013/5/6 Alan Pevec <apevec@gmail.com>:
2013/5/4 Adam Young <ayoung@redhat.com>:
* Keystone, from Dolph: no patch yet, critical issue "blurring the line between bug fix and feature change" https://bugs.launchpad.net/keystone/+bug/1175838 I'd hold out for this one. LDAP is broken without it.
Above bug was identified as a duplicate of https://bugs.launchpad.net/keystone/+bug/1168726 and Dolph is working on a patch https://review.openstack.org/28197 which hopefully will be ready for backporting soon.
Adam, Dolph - I don't see any progress on that review, what shall we do with Keystone 2013.1.1 ? We're supposed to release it in 2 days, do you want to skip this update until the fix for default domain in LDAP is ready? Cheers, Alan
Apologies for the lack of attention on that review; getting that in today and proposed for backport is my highest priority. If we can hold 2013.1.1 for keystone until this is considered for backporting, that would be much appreciated! I want to be clear though, this patch is cutting a feature (multi-domain support for LDAP/AD) from stable/grizzly rather than fixing it; while the feature technically works, it doesn't satisfy the use case it was intended to solve and results in unnecessary post-configuration setup for LDAP/AD deployments that may not be possible in the real world. Trying to backport a real "fix" would require backporting new features that are currently still in the blueprint phase for Havana. -Dolph On Tue, May 7, 2013 at 3:47 AM, Alan Pevec <apevec@gmail.com> wrote:
2013/5/6 Alan Pevec <apevec@gmail.com>:
2013/5/4 Adam Young <ayoung@redhat.com>:
* Keystone, from Dolph: no patch yet, critical issue "blurring the line between bug fix and feature change" https://bugs.launchpad.net/keystone/+bug/1175838 I'd hold out for this one. LDAP is broken without it.
Above bug was identified as a duplicate of https://bugs.launchpad.net/keystone/+bug/1168726 and Dolph is working on a patch https://review.openstack.org/28197 which hopefully will be ready for backporting soon.
Adam, Dolph - I don't see any progress on that review, what shall we do with Keystone 2013.1.1 ? We're supposed to release it in 2 days, do you want to skip this update until the fix for default domain in LDAP is ready?
Cheers, Alan
On 05/07/2013 11:20 AM, Dolph Mathews wrote:
Apologies for the lack of attention on that review; getting that in today and proposed for backport is my highest priority. If we can hold 2013.1.1 for keystone until this is considered for backporting, that would be much appreciated!
I want to be clear though, this patch is cutting a feature (multi-domain support for LDAP/AD) from stable/grizzly rather than fixing it; while the feature technically works, it doesn't satisfy the use case it was intended to solve and results in unnecessary post-configuration setup for LDAP/AD deployments that may not be possible in the real world. Trying to backport a real "fix" would require backporting new features that are currently still in the blueprint phase for Havana.
We are looking at a different patch that will, I think preserve the feature while still fixing the problem. THe developer is doing due dilligence to ensure that it is fixed: https://review.openstack.org/#/c/27364/
-Dolph
On Tue, May 7, 2013 at 3:47 AM, Alan Pevec <apevec@gmail.com <mailto:apevec@gmail.com>> wrote:
2013/5/6 Alan Pevec <apevec@gmail.com <mailto:apevec@gmail.com>>: > 2013/5/4 Adam Young <ayoung@redhat.com <mailto:ayoung@redhat.com>>: >>> * Keystone, from Dolph: no patch yet, critical issue "blurring the >>> line between bug fix and feature change" >>> https://bugs.launchpad.net/keystone/+bug/1175838 >> I'd hold out for this one. LDAP is broken without it. > > Above bug was identified as a duplicate of > https://bugs.launchpad.net/keystone/+bug/1168726 > and Dolph is working on a patch https://review.openstack.org/28197 > which hopefully will be ready for backporting soon.
Adam, Dolph - I don't see any progress on that review, what shall we do with Keystone 2013.1.1 ? We're supposed to release it in 2 days, do you want to skip this update until the fix for default domain in LDAP is ready?
Cheers, Alan
2013/5/8 Adam Young <ayoung@redhat.com>:
We are looking at a different patch that will, I think preserve the feature while still fixing the problem. THe developer is doing due dilligence to ensure that it is fixed:
Hi Adam, that's the older patch linked in LP 1168726 does that mean Dolph will abandon 28197 and 28197 ? How confident are you it can be pulled off today, maybe we'd better postpone this to 2013.1.2 next month (planned release June 6, freeze on May 30) ? Cheers, Alan
spzala's patch is still in flux, but it's also getting a lot of attention. If neither approach merges to master within a few hours, I think it's reasonable to wait until 2013.1.2 -Dolph On Wed, May 8, 2013 at 8:19 AM, Alan Pevec <apevec@gmail.com> wrote:
2013/5/8 Adam Young <ayoung@redhat.com>:
We are looking at a different patch that will, I think preserve the feature while still fixing the problem. THe developer is doing due dilligence to ensure that it is fixed:
Hi Adam,
that's the older patch linked in LP 1168726 does that mean Dolph will abandon 28197 and 28197 ? How confident are you it can be pulled off today, maybe we'd better postpone this to 2013.1.2 next month (planned release June 6, freeze on May 30) ?
Cheers, Alan
2013/5/8 Dolph Mathews <dolph.mathews@gmail.com>:
spzala's patch is still in flux, but it's also getting a lot of attention. If neither approach merges to master within a few hours, I think it's reasonable to wait until 2013.1.2
It's been few hours and I see review 27364 is getting there, but not merged yet so I'd suggest waiting until 2013.1.2 for the backport. Cheers, Alan
participants (4)
-
Adam Young
-
Alan Pevec
-
Dolph Mathews
-
John Griffith