Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-03-28
14:20:52 gibi efried: what about https://review.openstack.org/#/c/632904/3/nova/tests/functional/test_servers.py
14:21:06 openstackgerrit Matt Riedemann proposed openstack/nova stable/rocky: Add functional recreate test for bug 1819963 https://review.openstack.org/648409
14:21:07 openstackgerrit Matt Riedemann proposed openstack/nova stable/rocky: Update instance.availability_zone on revertResize https://review.openstack.org/648410
14:21:07 openstack bug 1819963 in OpenStack Compute (nova) rocky "Reverting a resize does not update the instance.availability_zone value to the source az" [Medium,Confirmed] https://launchpad.net/bugs/1819963
14:21:32 openstackgerrit Silvan Kaiser proposed openstack/nova master: Adds systemd detection result caching in Quobyte driver https://review.openstack.org/648093
14:21:36 efried gibi: heh, what about --^
14:22:17 gibi efried: could be. I feel https://review.openstack.org/#/c/632904/3/nova/tests/functional/test_servers.py is more self contained
14:23:06 efried gibi: I was being lazy and looking for one where the recreate was separate from the fix, but it would be easy enough to split out.
14:23:34 efried again with https://review.openstack.org/#/c/648409/1/nova/tests/functional/test_availability_zones.py we wouldn't have to include the refactor part in the demo, just L145-73
14:24:39 gibi efried: OK. you convinced me
14:24:48 kaisers efried: FYI, added the requested tests/nit-corrections to https://review.openstack.org/648093 , pls check if that fits your view
14:25:57 efried kaisers: ack
14:26:27 openstackgerrit Takashi NATSUME proposed openstack/nova stable/stein: doc: Fix openstack CLI command https://review.openstack.org/648412
14:28:03 openstackgerrit Matt Riedemann proposed openstack/nova stable/queens: Add functional recreate test for bug 1819963 https://review.openstack.org/648414
14:28:04 openstackgerrit Matt Riedemann proposed openstack/nova stable/queens: Update instance.availability_zone on revertResize https://review.openstack.org/648415
14:28:04 openstack bug 1819963 in OpenStack Compute (nova) rocky "Reverting a resize does not update the instance.availability_zone value to the source az" [Medium,In progress] https://launchpad.net/bugs/1819963 - Assigned to Matt Riedemann (mriedem)
14:31:04 Zara hi! I have a question; aiui when disk space is overcommitted, disk_available_least shows a negative value (eg: with `openstack hypervisor stats show`). Why is a negative value displayed instead of '0'? It seems like the value could be infinitely low, so I'm wondering how the information would then be used?
14:31:26 openstackgerrit Matt Riedemann proposed openstack/nova stable/queens: Update instance.availability_zone on revertResize https://review.openstack.org/648415
14:32:14 efried Zara: Just spitballing here, but I assume it gives you an indication of how overcommitted you are.
14:32:20 sean-k-mooney Zara: it means that the total amount of worse case diskspace for the current instance on the host exceed the availble diskspace
14:32:36 sean-k-mooney efried: yes it does
14:33:17 sean-k-mooney Zara: i found this really good github gist that explains it https://gist.github.com/JCallicoat/43505cab0535057ca4fb
14:33:20 efried arguably this could be "fixed" to floor out at zero, but that would require a new microversion, so it probably hasn't been considered worth the hassle since you could do that on the client side if you don't care about the "extra information".
14:33:27 sean-k-mooney we porably shoudl have somthing similar in our docs
14:33:44 sean-k-mooney efried: i dont think this is a bug
14:33:49 sean-k-mooney it was by design
14:34:23 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Remove IP proxy methods https://review.openstack.org/643115
14:34:27 efried Right, right, it's a feature, not a bug.
14:34:57 sean-k-mooney efried: that said this was origianly used by the disk filters i nthe scheduler but we do that in placmenet now
14:35:11 sean-k-mooney orperhaps it was the disk wheigher
14:35:45 Zara thanks for the gist (yeah, that says nova-scheduler).
14:36:24 sean-k-mooney hum the weigher actully used free_disk_mb
14:37:02 Zara An official docs note somewhere would be good; I encountered it in the context of someone thinking it was a bug.
14:39:18 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Add functional recreate test for bug 1819963 https://review.openstack.org/648421
14:39:19 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Update instance.availability_zone on revertResize https://review.openstack.org/648422
14:39:19 openstack bug 1819963 in OpenStack Compute (nova) rocky "Reverting a resize does not update the instance.availability_zone value to the source az" [Medium,In progress] https://launchpad.net/bugs/1819963 - Assigned to Matt Riedemann (mriedem)
14:39:36 sean-k-mooney lookin futher we do https://github.com/openstack/nova/blob/9db1c6061d10617b40dbc1d0d64faed79503cb81/nova/scheduler/host_manager.py#L213 so we set free_gb in the context of the host state object to the smalles to free_gb or lesast_gb
14:40:41 sean-k-mooney so yes the disk weighter will prefer host that are less over commited
14:40:43 sean-k-mooney https://github.com/openstack/nova/blob/master/nova/scheduler/weights/disk.py
14:40:58 sean-k-mooney which is why we have the negitive value
14:41:49 openstackgerrit Takashi NATSUME proposed openstack/nova stable/rocky: doc: Fix openstack CLI command https://review.openstack.org/648425
14:43:31 Zara sean-k-mooney: ah, I see, thanks
14:44:57 efried Zara: It would be great if you wanted to find the appropriate spot in the docs and propose a patch to enhance it :)
14:46:06 Zara efried: I'll see what I can do. :)
14:46:14 openstackgerrit Matt Riedemann proposed openstack/nova stable/stein: Add retry_on_deadlock to migration_update DB API https://review.openstack.org/648428
14:56:32 efried Zara: Thanks!
15:23:35 Zara (aw, I found a nice place to add it... but the section is in the ocata docs; I haven't yet found the correlate of those docs for more recent releases (https://docs.openstack.org/ocata/config-reference/compute/schedulers.html )
15:24:50 sean-k-mooney Zara: we generate our config docmuemnt from code now
15:25:21 amodi what was the spam about? :?
15:25:27 Zara ah okay, guess I'll need to add a docstring somewhere
15:25:28 sean-k-mooney Zara: https://github.com/openstack/nova/blob/master/nova/conf/scheduler.py
15:25:54 Zara sean-k-mooney: thanks
15:26:54 sean-k-mooney we do have ohter schduler docs somewhere actully
15:27:50 sean-k-mooney Zara: i think this is what you want https://github.com/openstack/nova/blob/master/doc/source/admin/configuration/schedulers.rst
15:34:14 Zara seank-k-mooney: ahaaa, yes that looks like the same doc
15:34:16 Zara thanks
15:54:41 openstackgerrit Dan Smith proposed openstack/os-traits master: Add COMPUTE_IMAGE_TYPE_* traits https://review.openstack.org/648147
16:02:03 ricolin cdent, jaypipes, zhipeng do you think you guys can help with http://lists.openstack.org/pipermail/openstack-discuss/2019-March/004246.html for Resource Management SIG?
16:02:25 cdent ricolin: there hasn't been any resource mgt sig activity, well, ever
16:02:29 cdent just hope
16:02:59 cdent zhipeng may have some thoughts, but though we wanted to we never really got it off the ground
16:03:40 ricolin cdent, is there any plan for mgt sig in Summit or PTG this time?
16:03:57 cdent not that I'm aware of
16:04:10 cdent but again, zhipeng may have more info
16:19:34 openstackgerrit Zara proposed openstack/nova master: Explain why disk_available_least can be negative https://review.openstack.org/648454
16:21:26 ricolin cdent, thanks, will try to reach out with zhipeng
16:22:27 Zara (my patch may be a bit brief, but I guess it's a start)
16:22:44 openstackgerrit Merged openstack/os-traits master: Add COMPUTE_IMAGE_TYPE_* traits https://review.openstack.org/648147
16:46:47 openstackgerrit Eric Fried proposed openstack/os-traits master: Add contextual comment about image type traits https://review.openstack.org/648466
17:06:29 efried stephenfin: Feel free to carry my +2 forward if kaisers respins
17:06:48 stephenfin efried: I might just do it myself if that suits you?
17:06:53 efried sure
17:06:54 stephenfin mriedem-like
17:08:17 openstackgerrit Stephen Finucane proposed openstack/nova master: Adds systemd detection result caching in Quobyte driver https://review.openstack.org/648093
17:08:47 openstackgerrit Eric Fried proposed openstack/nova master: WIP: Poison sleep() in tests https://review.openstack.org/643763
17:09:05 cdent sean-k-mooney: whatever happened about your lower-constraints fixes? I'm trying to decide if my version still matters https://review.openstack.org/#/c/622972/ . If not I'd like to forget about it.
17:09:26 stephenfin done-zo
17:12:43 sean-k-mooney cdent: i got a little side tracked. it was ment to be the thing i worked on today but that did not happen :(
17:13:31 sean-k-mooney cdent: ill pull down yours and test it locally and see if that still works.
17:14:30 sean-k-mooney cdent: i could be wrong but i assume if i rebased yours it would have teh same rabbitmq issues
17:16:21 cdent sean-k-mooney: I can't remember the issue you're talking about
17:17:18 sean-k-mooney cdent: when using lower constraint in my version a bunch of the unit tests end up using the amqp oslo messageing driver and actully try to connect to rabbitmq
17:18:19 openstackgerrit Merged openstack/os-traits master: Add contextual comment about image type traits https://review.openstack.org/648466
17:19:01 openstackgerrit Balazs Gibizer proposed openstack/nova master: nova-manage: heal port allocations https://review.openstack.org/637955
17:19:02 openstackgerrit Balazs Gibizer proposed openstack/nova master: cache neutron ports in heal allocation https://review.openstack.org/638207
17:20:15 sean-k-mooney cdent: you bumped oslo.messaging to 7.0.0 in your and i didnt bump it in min so i think that is the reason
17:23:13 sean-k-mooney your version seams to be working but you have more version change then i have. just waiting for it too finish locally.
17:23:50 openstackgerrit Zara proposed openstack/nova master: Explain why disk_available_least can be negative https://review.openstack.org/648454
17:24:53 sean-k-mooney Zara: thanks it was only a nit so you didnt have to respin but that does look fien to me
17:25:15 Zara heh, easy enough to edit inline :)
17:42:38 openstackgerrit Ghanshyam Mann proposed openstack/nova master: WIP: Add new default roles in os-services API policies https://review.openstack.org/648480
17:43:32 openstackgerrit Ghanshyam Mann proposed openstack/nova master: WIP:Introduce scope_types in os-services, lock policies https://review.openstack.org/645452
17:43:52 openstackgerrit Ghanshyam Mann proposed openstack/nova master: WIP: Add new default roles in os-services API policies https://review.openstack.org/648480
19:02:02 openstackgerrit Merged openstack/nova master: Temporarily mutate migration object in finish_revert_resize https://review.openstack.org/641137
19:15:53 openstackgerrit Merged openstack/nova master: De-cruftify the finish_resize methods https://review.openstack.org/635349
19:34:55 openstackgerrit Eric Fried proposed openstack/nova master: Delete allocations even if _confirm_resize raises https://review.openstack.org/647566
19:34:55 openstackgerrit Eric Fried proposed openstack/nova master: Error out migration when confirm_resize fails https://review.openstack.org/647546
19:34:56 openstackgerrit Eric Fried proposed openstack/nova master: api-ref: add more details to confirmResize troubleshooting https://review.openstack.org/647741
20:28:58 openstackgerrit Merged openstack/nova master: Explain why disk_available_least can be negative https://review.openstack.org/648454
20:59:14 openstackgerrit melanie witt proposed openstack/nova master: libvirt: set device address tag only if setting disk unit https://review.openstack.org/611974

Earlier   Later