Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-12-08
22:07:49 melwitt yeah it will make it unavailable https://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.reserved_host_disk_mb
22:08:19 melwitt sorry that was disk. memory is https://docs.openstack.org/nova/latest/configuration/config.html#DEFAULT.reserved_host_memory_mb
22:10:29 mriedem ihrachys: this is the method that sets up the inventory data to send to placement https://github.com/openstack/nova/blob/master/nova/compute/resource_tracker.py#L83
22:11:33 openstackgerrit melanie witt proposed openstack/nova master: Add API and nova-manage tests that use the NoopQuotaDriver https://review.openstack.org/526270
22:11:34 openstackgerrit melanie witt proposed openstack/nova master: Follow up on removing old-style quotas code https://review.openstack.org/524234
22:12:22 ihrachys mriedem, reduced to 512, restarted, and now it seems to have started the instance as expected. I guess the issue was because swap is not considered for the locking
22:12:32 ihrachys thanks mriedem for the pointer to the option!
22:14:24 mriedem yw
22:21:20 openstackgerrit melanie witt proposed openstack/nova master: DNM: Test websocketproxy with TLS https://review.openstack.org/513160
22:34:09 melwitt mriedem: I had to rebase the noop quotas test coverage bc of your quota force bug fix shenanigans https://review.openstack.org/#/c/526270
22:41:18 mriedem done
22:41:33 melwitt thanks \o/
22:49:04 mriedem superdan: were you going to update https://review.openstack.org/#/c/507638/ or save it for monday or are you ok with me updating it?
22:49:25 superdan mriedem: yeah I started locally but likely won't finish until monday
22:49:37 superdan that dependent one is in the gate atm too
22:50:24 mriedem alright
22:50:44 mriedem well i guess i'll just quit and spend quality time with my family
22:51:00 superdan good call
22:52:08 melwitt superdan: were you gonna remove one of the dupe tests in this "unmap compute nodes" one or should I do it? https://review.openstack.org/#/c/524641
22:52:38 superdan melwitt: yes I was
22:52:39 superdan so what you're both saying is, I suck, is that about right?
22:52:56 melwitt okay. no, of course not!
22:55:24 openstackgerrit Dan Smith proposed openstack/nova master: Unmap compute nodes when deleting host mapping https://review.openstack.org/524641
22:55:30 superdan melwitt: ^
22:56:21 melwitt saweet
22:57:12 melwitt mriedem: if you're still around ^
22:59:19 superdan I think you can slam it, he was +2 before right?
23:00:16 melwitt yeah he was. I'm conservative about doing that so thought I'd ask in case he's still there
23:01:17 superdan well, no pressure from me
23:01:23 superdan I'm spent for today anyway
23:01:53 melwitt ++
23:44:46 openstackgerrit Merged openstack/nova stable/pike: Add regression test for rebuild with new image doubling allocations https://review.openstack.org/523213
#openstack-nova - 2017-12-09
01:36:26 openstackgerrit Merged openstack/nova master: xenapi: pass migrate_data to recover_method if live migrate fails https://review.openstack.org/507874
02:06:08 openstackgerrit Merged openstack/nova stable/pike: Fix doubling allocations on rebuild https://review.openstack.org/523214
02:06:14 openstackgerrit Merged openstack/nova master: Refactor the code to check for sufficient hosts https://review.openstack.org/520242
02:41:27 openstackgerrit melanie witt proposed openstack/nova master: Add access_url_base to console_auth_tokens table https://review.openstack.org/334614
02:41:28 openstackgerrit melanie witt proposed openstack/nova master: Add console connection object https://review.openstack.org/320063
02:41:28 openstackgerrit melanie witt proposed openstack/nova master: Add console_auth_token_get() method to DB API https://review.openstack.org/481700
02:41:29 openstackgerrit melanie witt proposed openstack/nova master: Use ConsoleConnection object to generate authorizations https://review.openstack.org/325414
02:41:29 openstackgerrit melanie witt proposed openstack/nova master: Add periodic task to clean expired console tokens https://review.openstack.org/325381
02:41:30 openstackgerrit melanie witt proposed openstack/nova master: Convert websocketproxy to use db for token validation https://review.openstack.org/333990
02:58:35 openstackgerrit Merged openstack/nova master: Return Selection objects from the scheduler driver https://review.openstack.org/495854
02:58:42 openstackgerrit Merged openstack/nova master: Move the to_dict() method to the Selection object https://review.openstack.org/523492
02:58:51 openstackgerrit Merged openstack/nova master: Modify select_destinations() to return objects and alts https://review.openstack.org/510159
02:59:01 openstackgerrit Merged openstack/nova master: Add new style volume attachment support to block_device.py https://review.openstack.org/525787
03:06:16 openstackgerrit Merged openstack/nova master: Implement new attach Cinder flow https://review.openstack.org/330285
03:23:56 openstackgerrit Merged openstack/nova stable/pike: Fix TypeError of _get_project_id when project_id is None https://review.openstack.org/524687
03:24:02 openstackgerrit Merged openstack/nova master: Fix cellsv1 messaging test https://review.openstack.org/525304
03:42:28 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in delete https://review.openstack.org/525231
04:53:10 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in delete (2) https://review.openstack.org/526263
04:53:46 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in delete (3) https://review.openstack.org/526557
04:54:13 openstackgerrit Chen Hanxiao proposed openstack/nova master: InstanceInfo: store more runtime information of instance https://review.openstack.org/526348
05:47:22 openstackgerrit Jiao Pengju proposed openstack/python-novaclient master: Remove deprecated command in nova.rst https://review.openstack.org/526569
06:26:07 openstackgerrit Merged openstack/nova master: placement: update client to set parent provider https://review.openstack.org/385693
06:26:24 openstackgerrit Merged openstack/nova master: [placement] Object changes to support last-modified headers https://review.openstack.org/521639
06:26:34 openstackgerrit Merged openstack/nova master: Scheduler set_inventory_for_provider does nested https://review.openstack.org/520643
06:26:48 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in post https://review.openstack.org/526823
06:30:28 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in delete https://review.openstack.org/525231
06:30:39 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in delete (2) https://review.openstack.org/526263
06:54:36 openstackgerrit Merged openstack/nova master: Reformat _get_all_with_shared https://review.openstack.org/512799
07:12:25 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in delete (3) https://review.openstack.org/526557
07:23:57 openstackgerrit Merged openstack/nova master: Add API and nova-manage tests that use the NoopQuotaDriver https://review.openstack.org/526270
07:24:04 openstackgerrit Merged openstack/nova master: [placement] Fix an error message in API validation https://review.openstack.org/524088
07:24:11 openstackgerrit Merged openstack/nova master: Unmap compute nodes when deleting host mapping https://review.openstack.org/524641
07:24:17 openstackgerrit Merged openstack/nova master: Follow up on removing old-style quotas code https://review.openstack.org/524234
07:50:23 openstackgerrit wanghao proposed openstack/nova master: Fix invalid Image&block device mapping combination https://review.openstack.org/171984
07:55:05 openstackgerrit wanghao proposed openstack/nova master: Fix invalid Image&block device mapping combination https://review.openstack.org/171984
08:09:32 openstackgerrit wanghongxu proposed openstack/nova master: Update cinder info when attach volume that attached https://review.openstack.org/525540
08:23:48 openstackgerrit Takashi NATSUME proposed openstack/nova master: [placement] Add sending global request ID in post https://review.openstack.org/526823
10:52:26 openstackgerrit Chen Hanxiao proposed openstack/nova master: libvirt: guest: introduce blockStats instead of domain.blockStats https://review.openstack.org/526833
18:02:51 openstackgerrit Chris Dent proposed openstack/nova master: Optional separate database for placement API https://review.openstack.org/362766
23:29:20 openstackgerrit Eric Berglund proposed openstack/nova master: WIP: PowerVM Driver: vSCSI https://review.openstack.org/526094
#openstack-nova - 2017-12-10
03:24:27 mnaser (not that I expect an answer for this now): but if someone is familiar with the glance caching code, does it always cache images on the local fs or the image backend?
03:24:51 mnaser asking because we're looking into opening another location in a cell (with its own ceph cluster as it's physically independent)
03:25:36 mnaser nova is smart enough to fallback to simply manually downloading the image if it cant reach the ceph cluster, but it would be nice that it caches it in ceph, and not locally .. allowing the entire cell to get access to the cached image
07:07:42 openstackgerrit Ameed Ashour proposed openstack/nova master: change instance_system_metadata column type https://review.openstack.org/526900
10:03:29 openstackgerrit Hamdy Khader proposed openstack/nova master: Adding NVMEoF for libvirt driver https://review.openstack.org/482640
20:07:30 openstackgerrit Merged openstack/nova master: Fix NoneType error when [service_user] is misconfigured https://review.openstack.org/521947
21:27:13 efried_cya_jan mnaser I know the glance caching code is virt driver-specific. Which virt driver are you using?
21:27:21 mnaser efried_cya_jan: libvirt
21:28:01 efried_cya_jan mnaser If I can point you to a starting point for the relevant source, can you take it from there?
21:28:15 mnaser efried_cya_jan: absolutely
21:28:19 efried_cya_jan stand by
21:28:34 mnaser efried_cya_jan: i ended up in the imagebackend code + spawn + image.cache stuff, but wasnt sure if it was caching inside ceph
21:29:47 efried_cya_jan mnaser Did you find libvirt's override of ComputeDriver.manage_image_cache ?
21:30:16 mnaser no, i didn't know about that part actually, i was just following the code which cached the image on first download
21:30:17 efried_cya_jan mnaser Which appears to end up in nova.virt.libvirt.imagecache.ImageCacheManager.update
21:31:10 mnaser efried_cya_jan: looks like that code is all about caching files locally
21:31:32 mnaser bummer, guess that'll have to be worked on
21:31:40 efried_cya_jan mnaser Does that answer your question?
21:31:53 mnaser efried_cya_jan: it does! thanks, i didn't know about that component
21:32:06 efried_cya_jan I'm not sure what good it would do to cache images "remotely". What benefit would you gain over leaving them in glance?
21:33:10 mnaser efried_cya_jan: our use cache is we're adding a new cell (in an entirely different facility). we use ceph for all storage, but the ceph cluster will not span across both cells. this means that we cant take advantage of new VMs doing a COW of the glance image (because glance ceph storage cannot be accessed by the second cell)
21:33:26 mnaser so it would have to pull down the image and import to ceph (which takes time and bandwidth)
21:33:47 efried_cya_jan Hum, is that a common use case?
21:33:59 mnaser the alternative is to somehow write something which mirrors images to both clusters (adding multiple locations) and the image backend code will loop until it finds the location of the image stored locally
21:34:24 mnaser i'm assuming cells is a common use case, but cells with 2 different ceph clusters to each one might be something that hasn't really been explored a lot yet
21:34:24 efried_cya_jan You need GPFS aka Spectrum Scale.
21:35:13 mnaser like to stick to open source :>
21:35:53 mnaser i think the best alternative is to get images mirrored across both ceph rbd clusters

Earlier   Later