On 3/22/2019 8:50 AM, Brian Rosmaita wrote:
This is mainly addressed to the cinder stable-maint cores.
As Eric noticed at the last Cinder team meeting, while we have had a Stein cycle python-cinderclient release (4.1.0), it was back in October [0]. Because we didn't do another more recent release including the patches that had collected in master, the stable/stein branch was cut at the 4.1.0 tag, and thereby missed some bugfixes. (We don't seem to have missed any major functionality, however.)
I've put together an etherpad of the "missing" patches [1]. It notes the patches I think it's OK to skip and it notes a few dependencies among patches. If any of those look like they should be included, feel free to propose a backport.
I've put up backports to stable/stein with the topic "stein-ketchup" for the rest. There are some that we probably wouldn't normally consider, but this is a bit of an unusual case if we want stable/stein to look the way it would have if we'd cut a release 2 weeks ago. Feel free to -2, I won't take it personally, I just figured it would be faster to make the decision on the reviews rather than try to figure it out up front.
Please review at your earliest convenience: https://review.openstack.org/#/q/topic:stein-ketchup+(status:open+OR+status:...
cheers, brian
[0]http://git.openstack.org/cgit/openstack/python-cinderclient/tag/?h=4.1.0 [1]https://etherpad.openstack.org/p/cinderclient-master-20-march
One the one hand I want to just say, "sorry, that sucks, but you guys should be better about watching your release deadlines" but that only hurts users. I have started going through these backports and I definitely think some of them we can ignore. I really hope asking the stable team to go through and fix this up for people post-freeze doesn't set a precedent. -- Thanks, Matt