Thursday, 2018-07-26

openstackgerritTakashi NATSUME proposed openstack/nova master: Making consistent used of GiB and MiB in API ref  https://review.openstack.org/55998500:19
openstackgerritTakashi NATSUME proposed openstack/nova master: Fix 500 error while passing 4-byte unicode data  https://review.openstack.org/40751400:20
*** cdent has joined #openstack-placement00:58
openstackgerritChris Dent proposed openstack/nova master: [placement] Extract base functional test case from test_direct  https://review.openstack.org/58577701:01
openstackgerritChris Dent proposed openstack/nova master: [placement] Use base test in placement functional tests  https://review.openstack.org/58577801:01
*** cdent has quit IRC01:08
openstackgerritMerged openstack/nova master: Remove support for /os-floating-ips-bulk REST API  https://review.openstack.org/58291201:11
openstackgerritMerged openstack/nova master: Remove support for /os-floating-ip-dns REST API  https://review.openstack.org/58294301:14
openstackgerritMerged openstack/nova master: Remove "DEPRECATED" tag from Obsolete APIs  https://review.openstack.org/58309201:14
openstackgerritMerged openstack/nova stable/pike: cleanup mapping/reqspec after archive instance  https://review.openstack.org/58077501:14
openstackgerritMerged openstack/nova master: doc: remove rocky-specific nova-scheduler min placement version  https://review.openstack.org/58552001:14
*** edmondsw has joined #openstack-placement01:17
*** edmondsw has quit IRC01:21
openstackgerritArtom Lifshitz proposed openstack/nova master: DNM? Extra logs for volume detach device tags cleanup  https://review.openstack.org/58403202:16
openstackgerritXiaohan Zhang proposed openstack/nova master: compute node local_gb_used include swap disks The ComputeNode.local_gb_used value is set in the ResourceTracker._update_usage() method:  https://review.openstack.org/58592802:18
*** lei-zh has joined #openstack-placement02:29
*** openstack has joined #openstack-placement02:51
*** ChanServ sets mode: +o openstack02:51
*** edmondsw has joined #openstack-placement03:05
*** edmondsw has quit IRC03:09
*** edleafe has quit IRC03:14
*** edleafe has joined #openstack-placement03:15
*** mriedem_afk has quit IRC03:20
*** lei-zh has quit IRC03:28
*** edmondsw has joined #openstack-placement04:53
*** edmondsw has quit IRC04:58
*** e0ne has joined #openstack-placement05:06
*** lei-zh has joined #openstack-placement05:30
*** e0ne has quit IRC06:00
openstackgerrithuanhongda proposed openstack/nova master: hypervisor-stats shows wrong disk usages with shared storage  https://review.openstack.org/14987806:22
*** edmondsw has joined #openstack-placement06:41
*** edmondsw has quit IRC06:47
*** ttsiouts has joined #openstack-placement07:10
*** tssurya has joined #openstack-placement07:12
openstackgerritXiaohan Zhang proposed openstack/nova master: compute node local_gb_used include swap disks  https://review.openstack.org/58592807:22
openstackgerritZhenyu Zheng proposed openstack/nova master: Increase sleep time in ServerGroupTestV21 to avoid potential failure  https://review.openstack.org/58597807:31
*** ttsiouts has quit IRC07:33
openstackgerritMerged openstack/nova master: Online data migration for queued_for_delete flag  https://review.openstack.org/58450407:52
openstackgerritMerged openstack/nova master: Annotate flows and handle PortBindingDeletionFailed in ComputeManager  https://review.openstack.org/55137107:53
openstackgerritMerged openstack/nova master: Remove stevedore extensions server_create method  https://review.openstack.org/58388207:53
*** avolkov has joined #openstack-placement07:56
*** ttsiouts has joined #openstack-placement08:10
*** avolkov has quit IRC08:11
*** e0ne has joined #openstack-placement08:17
*** edmondsw has joined #openstack-placement08:30
*** tetsuro has quit IRC08:31
*** edmondsw has quit IRC08:34
*** cdent has joined #openstack-placement08:54
*** ttsiouts has quit IRC09:29
*** ttsiouts has joined #openstack-placement09:31
*** cdent has quit IRC09:58
*** cdent has joined #openstack-placement09:58
*** ttsiouts has quit IRC10:10
*** lei-zh has quit IRC10:11
openstackgerritMerged openstack/nova master: ironic: Report resources as reserved when needed  https://review.openstack.org/51792110:11
*** edmondsw has joined #openstack-placement10:18
openstackgerritTakashi NATSUME proposed openstack/nova master: Adds view builders for keypairs controller  https://review.openstack.org/34728910:18
*** edmondsw has quit IRC10:23
openstackgerritKashyap Chamarthy proposed openstack/nova master: libvirt: Remove usage of migrateToURI{2} APIs  https://review.openstack.org/56725810:42
*** edmondsw has joined #openstack-placement10:49
*** takashin_ has joined #openstack-placement10:57
*** takashin has quit IRC10:58
openstackgerritChris Dent proposed openstack/nova master: [placement] Retry allocation writes server side  https://review.openstack.org/58604811:05
cdentefried, gibi : you may find that ^ interesting, especially the fact that I've made rp conflict exception different from consumer. It probably needs rather close review as it is a bit weird and hard to test.11:06
openstackgerritTakashi NATSUME proposed openstack/nova master: [placement] api-ref: add traits parameter  https://review.openstack.org/57804811:07
cdentefried, gibi : I was inspired to escalate my attention to that because allocation rp generation conflicts happen a lot with clustered virt drivers11:15
*** ttsiouts has joined #openstack-placement11:15
openstackgerritAndrey Volkov proposed openstack/osc-placement master: Get resource provider by uuid or name  https://review.openstack.org/52779111:27
openstackgerritAndrey Volkov proposed openstack/osc-placement master: Provide a useful message in the case of 500-error  https://review.openstack.org/58605612:00
openstackgerritAndrey Volkov proposed openstack/osc-placement master: Provide a useful message in the case of 500-error  https://review.openstack.org/58605612:01
openstackgerritBalazs Gibizer proposed openstack/nova master: Use placement 1.28 in scheduler report client  https://review.openstack.org/58366712:23
openstackgerritOpenStack Release Bot proposed openstack/os-traits stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/58610112:35
openstackgerritOpenStack Release Bot proposed openstack/os-traits stable/rocky: Update UPPER_CONSTRAINTS_FILE for stable/rocky  https://review.openstack.org/58610212:35
openstackgerritOpenStack Release Bot proposed openstack/os-traits master: Update reno for stable/rocky  https://review.openstack.org/58610312:35
openstackgerritOpenStack Release Bot proposed openstack/osc-placement stable/rocky: Update .gitreview for stable/rocky  https://review.openstack.org/58611312:36
*** openstackgerrit has quit IRC12:36
*** mriedem has joined #openstack-placement12:36
efriedpatchbot down?13:15
mriedemwe should add the patchbot to the nova channel btw13:20
efriedsorry, maybe I don't mean patchbot. Whichever one tells us about patches being proposed. I just rebased and pushed the reshaper series and it didn't show up. Must be recent, though, cause there's messages from the patchbot 45 mins ago.13:21
cdentefried: "openstackgerrit left the chat room. (Killed (Sigyn (Spam is off topic on freenode.)))"13:22
efriedwhoops13:23
efriedThere was a lot of spam last night; wonder if they lowered the threshold.13:23
cdentthe spam is still flowing today, so they are probably being defensive13:24
cdentefried: anything in particular to which I should draw my attention today13:24
efriedlooks like they're aware in -infra. I'll just sit back.13:24
*** mriedem is now known as mriedem_afk13:25
efriedcdent: Sounds like we're going to punt reshaper.13:25
cdentmriedem: you too might find https://review.openstack.org/586048 interesting, bug fix. It turns out the bug is more serious than I though (especially for vmware)13:25
efriedcdent: But if you want to look at the latest updates (if you haven't already).13:25
cdentefried: punt all the way to stein? no FFE?13:25
cdentis there discussion somewhere?13:25
efriedcdent: That's what mriedem_afk thinks. in -nova. I think we have a meeting soon, though, so we should discuss there.13:26
efriedyeah, half an hour. I'll put it on the agenda.13:26
cdenthmm. I'm not sure I really mind it being punted in any sense other than "but but we're so close"13:28
cdentit's not a feature that I'm personally super wed to13:28
cdentand I guess there's no FFU tool on the immediate horizon13:28
cdentefried: I think we should try to merge the server side, even if we don't merge the client side13:43
cdentotherwise we're causing ourselves some roadblocks for extraction (if we can continue to consider that a thing)13:43
efriedcdent: how so?13:45
*** ttsiouts has quit IRC13:47
cdentefried: code checked out of /nova that we want checked into /placement13:48
efriedcdent: Sorry, I'm still not seeing why having a pending API change would affect that.13:49
cdentefried: it's an ordering thing13:49
*** ttsiouts has joined #openstack-placement13:50
cdentif we have pending changes to nova/ap/openstack/placement and at the start of stein we go ahead and pull the trigger and move all that code to placement we'll have extra merege conflicts to get those pending changes in13:50
cdentwhich people won't like13:51
cdentso we'll say "let's wait until all the pending stuff merges before we do the cut over"13:51
cdentwhich will eventually mean: forget it, let's wait until T13:51
cdentat which point we'll never do it13:51
efriedmm. Well, I would hope we would merge the reshaper series very early, like as soon as stein is branched.13:51
cdentso would I, but it would be even easier if we just merged the api now and simply didn't use it13:52
efriedIs that, taking the risk that we'll find something broken and have to burn another microversion to fix it, the lesser evil?13:52
cdentand "hope" is frequently for naught13:52
cdentI don't really care about burning microversions13:53
*** mriedem_afk is now known as mriedem13:53
cdentand ideally when placement is its own thing we can throw off some of the nova-imposed overhead for needing a spec when everyone agrees13:53
mriedemefried: -nova has the one that shows things proposed,13:56
mriedemi would also like the one that's in the meeting channel where you dump a gerrit link and it reports the title13:56
efriedmriedem: it's gerritbot that does the proposed things, and it's in both of these channels. patchbot is in neither, afaik.14:03
*** avolkov has joined #openstack-placement14:03
*** diga has joined #openstack-placement14:16
*** tssurya has quit IRC14:23
*** avolkov has quit IRC14:32
*** takashin_ has left #openstack-placement14:50
*** takashin has joined #openstack-placement14:50
*** avolkov has joined #openstack-placement14:58
*** takashin has left #openstack-placement15:03
cdentmriedem, efried : I'm doing a simple repeat of the analysis I did last year at http://lists.openstack.org/pipermail/openstack-dev/2017-January/110953.html15:16
cdenttoday instead of 5 requests per iteration (when no instances) we have ten15:16
mriedem:(15:16
cdentit's effectively for the same reason: we call _udpate at the start and end of _update_available _resource15:16
mriedemand this time we actually care about aggregates b/c of shared storage providers,15:17
* cdent nods15:17
mriedemalthough that's not really supported yet15:17
mriedemand i'm guessing traits is now in there too15:17
cdentyes15:17
cdentbut the redundancies are: the two _update calls and the fact that the thing that _udpate calls requests inventory twice15:18
cdentso there are 4 calls to inventory, and then 2 of rps, traits, aggregates and 1 of allocations15:18
cdentwait that's 11.15:18
* cdent looks at data15:19
cdentyeah, 1115:19
* cdent sighs15:19
cdentanyway, I'll write it up and we can think about it15:21
*** ttsiouts has quit IRC15:24
*** avolkov has quit IRC15:31
*** avolkov has joined #openstack-placement15:33
mriedemcdent: for shared storage buckets https://review.openstack.org/#/c/149878/15:55
cdenthmmm, interesting, will look at that more closely asap15:56
mriedemit's 3 years old :)15:56
mriedembut coincides with my desire long-term to change the compute hypervisor-stats API to just be a proxy to placement15:56
cdentone would hope that could be done15:57
*** edmondsw has quit IRC15:57
*** edmondsw has joined #openstack-placement15:57
*** edmondsw has quit IRC15:59
*** e0ne has quit IRC16:02
cdentmriedem, efried : https://anticdent.org/novas-use-of-placement.html16:15
efriedack16:15
efriedcdent: s/claim/claims/16:17
cdentnoted16:17
efried"sort of noise" but probably the main reason it's not easy to simplify16:17
cdentthat claim is correct, the actor is processes16:18
efriedcdent: ah, see it now, thx16:19
cdentthe other interaction are the conductor and nova-manage commands16:19
cdentas you'll see as you get further in "this discussion" is _very_ focussed16:19
efriedcdent: Could be read as, "The nova-scheduler processes(v) [request candidates](n)" or "The [nova-scheduler processes](n) request(v) candidates"16:20
efriedthat was msy confusion.16:20
* cdent nods16:20
efriedcdent: Having a hard time parsing, "Can get_provider_tree_and_ensure_root tries to manage inventory twice be rationalized for simple cases?"16:22
cdentthat's a "typo"16:22
cdentcan the way16:22
efriedah16:26
cdentefried: does it otherwise make sense?16:31
*** diga has quit IRC16:37
*** openstackgerrit has joined #openstack-placement16:49
openstackgerritChris Dent proposed openstack/nova master: [placement] Retry allocation writes server side  https://review.openstack.org/58604816:49
openstackgerritChris Dent proposed openstack/nova master: [placement] Use base test in placement functional tests  https://review.openstack.org/58577816:54
*** e0ne has joined #openstack-placement17:20
*** e0ne has quit IRC17:37
*** edmondsw has joined #openstack-placement17:48
openstackgerritDan Smith proposed openstack/nova master: Add additional functional tests for NUMA networks  https://review.openstack.org/58538517:49
openstackgerritDan Smith proposed openstack/nova master: Add additional functional tests for NUMA networks  https://review.openstack.org/58538517:55
openstackgerritMerged openstack/nova master: [placement] Extract base functional test case from test_direct  https://review.openstack.org/58577718:19
*** e0ne has joined #openstack-placement18:22
openstackgerritkarim proposed openstack/nova master: Updated AggregateImagePropertiesIsolation filter illustration  https://review.openstack.org/58631718:29
efriedcdent: Yeah, and we should totally figure out how to trim it down.18:30
openstackgerritMatt Riedemann proposed openstack/nova master: Fix accumulated nits from port binding for live migration series  https://review.openstack.org/58399418:57
mriedemnutrislim19:08
mriedemhttps://www.youtube.com/watch?v=BORkjEKVVSc19:08
openstackgerritDan Smith proposed openstack/nova master: Assorted cleanups from numa-aware-vswitches series  https://review.openstack.org/58265119:15
openstackgerritDan Smith proposed openstack/nova master: Add additional functional tests for NUMA networks  https://review.openstack.org/58538519:15
openstackgerritMerged openstack/nova master: Port binding based on events during live migration  https://review.openstack.org/43487019:35
openstackgerritMerged openstack/nova master: conductor: use port binding extended API in during live migrate  https://review.openstack.org/52253719:45
*** cdent has quit IRC19:59
*** e0ne has quit IRC20:13
*** e0ne has joined #openstack-placement20:59
*** e0ne has quit IRC21:00
*** edmondsw has quit IRC21:43
*** edmondsw has joined #openstack-placement21:45
*** edmondsw has quit IRC21:49
*** mriedem is now known as mriedem_afk22:14
*** avolkov has quit IRC22:38
*** mriedem_afk is now known as mriedem22:47
*** knikolla has quit IRC23:01
openstackgerritMerged openstack/nova master: Stop using HostAPI.service_delete  https://review.openstack.org/58268023:50
*** takashin has joined #openstack-placement23:53

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!