dirk | hwoarang: what's the question about the wheel job? | 06:36 |
---|---|---|
*** sdague has joined #openstack-requirements | 12:18 | |
*** sdague has quit IRC | 14:20 | |
*** zenpwner has joined #openstack-requirements | 14:43 | |
*** zenpwner has quit IRC | 14:44 | |
fried_rice | Hey kids - what are the chances of getting a pypowervm bump backported to pike? | 16:53 |
*** fried_rice is now known as efried | 16:53 | |
efried | dirk prometheanfire ^ | 16:53 |
efried | smcginnis tonyb ^ | 16:55 |
dirk | efried: min version bump or changing uc to a new version? | 18:06 |
efried | dirk Both | 18:06 |
dirk | efried: by default dependencies don't change for projects following stable policy. What is the reason for the new version requirement? | 18:07 |
efried | dirk Found a bug that'll cause an OVS-based VM to lose network connectivity if the host reboots. | 18:08 |
efried | dirk To fix it in the driver, we effectively have to monkey-patch the lib, which is... not ideal. | 18:08 |
efried | But doable. | 18:08 |
efried | dirk https://review.openstack.org/#/c/519030/2/nova_powervm/virt/powervm/vif.py@616 if you're interested. | 18:09 |
dirk | efried: hmm... In the case of of pypowervm I personally wouldn't mind a version update as it affects only those that are affected. | 18:12 |
efried | dirk Okay, cool. So once we cut the new version, what's the protocol? Do we wait for generate-constraints to bump u-c in master, then propose the g-r bump in master, then... backport something? Or propose a new something in pike? | 18:14 |
*** lxnch_ has joined #openstack-requirements | 18:45 | |
*** lxnch has quit IRC | 18:48 | |
*** rakhmerov has quit IRC | 21:53 | |
*** rakhmerov has joined #openstack-requirements | 21:58 | |
prometheanfire | efried: I would say no to the min version bump | 22:58 |
prometheanfire | but uc bump makes sense | 22:58 |
prometheanfire | we can mask that bad version though | 22:59 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!