Monday, 2019-05-13

*** boxiang has quit IRC00:48
*** altlogbot_1 has quit IRC00:50
*** altlogbot_1 has joined #openstack-cinder00:53
*** altlogbot_1 has quit IRC00:54
*** altlogbot_3 has joined #openstack-cinder00:55
*** lixiaoy1 has joined #openstack-cinder01:04
*** baojg has quit IRC01:08
openstackgerritrenminmin proposed openstack/cinder master: Fix rbd mirror retype bug  https://review.opendev.org/65798501:23
*** baojg has joined #openstack-cinder01:29
*** ircuser-1 has quit IRC01:32
*** brinzhang has joined #openstack-cinder01:49
*** v12aml has quit IRC02:10
*** whoami-rajat has joined #openstack-cinder03:05
*** psachin has joined #openstack-cinder03:19
*** udesale has joined #openstack-cinder03:56
*** david-lyle is now known as dklyle03:58
*** dklyle has quit IRC04:03
*** itlinux has joined #openstack-cinder04:10
*** Luzi has joined #openstack-cinder04:42
*** dpawlik has joined #openstack-cinder04:44
*** e0ne has joined #openstack-cinder05:16
*** boxiang has joined #openstack-cinder05:18
*** lpetrut has joined #openstack-cinder05:39
*** udesale has quit IRC05:45
*** udesale has joined #openstack-cinder05:45
*** udesale has quit IRC05:49
*** udesale has joined #openstack-cinder05:49
*** lpetrut has quit IRC05:55
*** lpetrut has joined #openstack-cinder06:06
*** sapd1_x has joined #openstack-cinder06:11
*** pcaruana has joined #openstack-cinder06:16
*** boxiang has quit IRC06:17
*** boxiang has joined #openstack-cinder06:17
*** lpetrut has quit IRC06:22
*** baojg has quit IRC06:28
*** e0ne has quit IRC06:29
openstackgerritMerged openstack/cinder master: move 3par exception  https://review.opendev.org/65818906:38
*** udesale has quit IRC06:47
*** udesale has joined #openstack-cinder06:47
*** vishalmanchanda has joined #openstack-cinder06:57
*** rcernin has quit IRC07:06
*** tesseract has joined #openstack-cinder07:17
*** itlinux has quit IRC07:35
*** helenafm has joined #openstack-cinder07:40
*** sahid has joined #openstack-cinder07:50
*** tkajinam has quit IRC08:31
*** e0ne has joined #openstack-cinder08:32
*** lixiaoy1 has quit IRC08:54
*** sapd1_x has quit IRC08:57
*** sapd1_x has joined #openstack-cinder09:24
*** wxy-xiyuan has joined #openstack-cinder09:39
*** dpawlik has quit IRC09:41
*** dpawlik has joined #openstack-cinder09:41
*** kambiz has quit IRC09:44
*** lpetrut has joined #openstack-cinder10:12
*** udesale has quit IRC11:17
*** m75abrams has joined #openstack-cinder11:20
openstackgerritrenminmin proposed openstack/cinder master: Fix rbd mirror retype bug  https://review.opendev.org/65798511:39
*** dviroel has joined #openstack-cinder11:39
openstackgerritSimon O'Donovan proposed openstack/cinder master: PowerMax driver - Rapid TDEV Deallocation  https://review.opendev.org/65879311:42
*** boxiang has quit IRC11:51
*** boxiang has joined #openstack-cinder11:51
*** boxiang has quit IRC11:54
*** boxiang has joined #openstack-cinder11:57
*** carloss has joined #openstack-cinder11:58
*** rosmaita has joined #openstack-cinder12:04
*** lennyb has quit IRC12:12
*** baojg has joined #openstack-cinder12:28
*** psachin has quit IRC12:46
*** baojg has quit IRC12:49
*** pcaruana|afk| has joined #openstack-cinder12:55
*** pcaruana has quit IRC12:57
*** rchurch has joined #openstack-cinder12:58
*** dirk has quit IRC12:58
openstackgerritMerged openstack/cinder stable/stein: Make sure we install cinder requirements during the correct tox phase  https://review.opendev.org/65829712:58
*** wxy-xiyuan has quit IRC12:59
*** jbernard_ has joined #openstack-cinder12:59
*** rchurch_ has quit IRC12:59
*** xyang_ has joined #openstack-cinder12:59
*** carloss has quit IRC12:59
*** jdillaman has joined #openstack-cinder13:00
*** jbernard has quit IRC13:00
*** xyang has quit IRC13:00
*** xyang_ is now known as xyang13:00
*** dirk has joined #openstack-cinder13:00
*** carloss has joined #openstack-cinder13:02
openstackgerritBrin Zhang proposed openstack/cinder master: Handle collections.abc deprecations  https://review.opendev.org/65880813:03
*** brinzhang has quit IRC13:09
*** kambiz has joined #openstack-cinder13:09
*** Luzi has quit IRC13:10
openstackgerritMerged openstack/cinder master: Add python37 job  https://review.opendev.org/65561213:10
*** udesale has joined #openstack-cinder13:10
*** enriquetaso has joined #openstack-cinder13:12
*** wxy-xiyuan has joined #openstack-cinder13:14
*** vishalmanchanda has quit IRC13:16
*** udesale has quit IRC13:18
*** lseki has joined #openstack-cinder13:18
*** udesale has joined #openstack-cinder13:18
*** mchlumsky has joined #openstack-cinder13:23
*** dpawlik has quit IRC13:35
*** gmann has joined #openstack-cinder14:06
*** gmann is now known as gmann_pto14:07
*** gnufied has joined #openstack-cinder14:10
*** baojg has joined #openstack-cinder14:11
*** jbernard_ is now known as jbernard14:13
openstackgerritSimon O'Donovan proposed openstack/cinder master: PowerMax driver - Rapid TDEV Deallocation  https://review.opendev.org/65883014:23
*** mriedem has joined #openstack-cinder14:25
*** baojg has quit IRC14:32
*** abishop has joined #openstack-cinder14:50
*** pcaruana|afk| has quit IRC14:55
*** __hemna has quit IRC14:57
*** ircuser-1 has joined #openstack-cinder15:15
*** itlinux has joined #openstack-cinder15:37
*** pcaruana|afk| has joined #openstack-cinder15:44
*** mvkr has joined #openstack-cinder15:46
*** __hemna has joined #openstack-cinder15:57
*** __hemna has quit IRC15:58
*** hemna has joined #openstack-cinder15:59
hemnamorning16:00
*** helenafm has quit IRC16:03
mriedemso i thought i knew how this worked in cinder but i might have forgotten. if i boot from volume i've got an in-use volume. then if i'm going to migrate the server, nova will create an "empty" volume attachment for the dest host - this is a volume attachment without a connector, and the volume status changes to "attaching",16:08
mriedemnow when nova updates the dest host attachment with the connector, it should make the volume as in-use again right? i'm seeing a failure like this instead: "Invalid input received: Invalid volume: Volume  b2aba195-6570-40c4-82bb-46d3557fceeb status must be available or  downloading to reserve, but the current status is attaching."16:08
mriedemmaybe the issue is i haven't deleted the source host attachment yet?16:09
smcginnismriedem: Looking at blame in the code, doesn't appear we've ever allowed that with the new attach APIs.16:12
mriedemah yeah i think that's the problem. looking at how nova does a migration, we create the empty dest host attachment first which reserves the volume, then we delete the source host attachment, and once we migrate to the dest host, we update the dest host attachment with the connector and finally complete it to mark the volume as in-use16:12
smcginnisWe sepcifically only check for available or downloading and also allow in-use only if multiattach is set.16:12
mriedemyeah...shazbot, my cross-cell stuff has hit a snag for volume-backed servers then...16:13
mriedemi'm trying to sniff out if a dest host is going to work in another cell by creating an export there first (updating the attachment using the dest host connector), but i haven't deleted the existing source host attachment for the volume, which is different from how normal resize works16:14
smcginnisYeah, that could be tricky.16:14
smcginnisWe could add "attaching", but I think we would need to enforce that the update is for the same host that started the attachment or something to prevent a race of someone else coming in and attaching it.16:15
mriedemi'd have to (1) create the dest host empty attachment, (2) delete the source host attachment, (3) update the dest host attachment with the dest host connector, and failing that, re-connect the source host attachment16:15
smcginnisWhich I think defeats the purpose of having that "reserved" state.16:15
smcginnismriedem: That sounds reasonable.16:15
mriedemreasonable, but pretty sucky16:16
mriedemi.e. lots of things to go wrong there16:16
mriedembut without creating the export on the dest host we don't really have a good pre-check that that host will work16:17
mriedemright?16:17
smcginnisI think you are right.16:18
smcginnisWe would have to add a new API call to validate that beforehand.16:18
*** tesseract has quit IRC16:20
*** whoami-rajat has quit IRC16:31
*** whoami-rajat has joined #openstack-cinder16:36
*** sahid has quit IRC16:37
*** e0ne has quit IRC16:42
*** tejdeep has joined #openstack-cinder16:51
*** tejdeep has quit IRC16:53
*** imacdonn has joined #openstack-cinder16:57
*** udesale has quit IRC16:59
*** tejdeep has joined #openstack-cinder17:07
*** mchlumsky has quit IRC17:23
openstackgerritRajat Dhasmana proposed openstack/cinder master: Bump bandit version  https://review.opendev.org/65887317:26
*** lpetrut has quit IRC17:27
*** mchlumsky has joined #openstack-cinder17:29
*** jroll has quit IRC17:34
*** jroll has joined #openstack-cinder17:35
*** itlinux has quit IRC17:50
*** itlinux has joined #openstack-cinder17:53
*** sapd1_x has quit IRC17:56
*** dklyle has joined #openstack-cinder17:59
openstackgerritJohn Cates proposed openstack/cinder master: Change _unmap_vdisk_from_host to unmap all connected hosts if connector is unspecified  https://review.opendev.org/65888418:07
*** dklyle has quit IRC18:08
*** enriquetaso has quit IRC18:09
*** enriquetaso has joined #openstack-cinder18:11
*** e0ne has joined #openstack-cinder18:17
*** tejdeep has quit IRC18:17
*** dklyle has joined #openstack-cinder18:23
*** dklyle has quit IRC18:34
*** tejdeep has joined #openstack-cinder18:39
*** e0ne has quit IRC18:54
*** dpawlik has joined #openstack-cinder19:03
*** m75abrams has quit IRC19:15
*** gmann_pto has quit IRC19:33
*** e0ne has joined #openstack-cinder20:14
openstackgerritDirk Mueller proposed openstack/cinder stable/rocky: Make sure we install cinder requirements during the correct tox phase  https://review.opendev.org/65891320:21
*** e0ne has quit IRC20:26
hemnasimondodsley: https://review.opendev.org/#/c/658096/20:30
*** itlinux has quit IRC20:32
openstackgerritDirk Mueller proposed openstack/cinder master: Use testenv deps instead of repeating the dependencies  https://review.opendev.org/65891420:33
openstackgerritDirk Mueller proposed openstack/cinder stable/queens: Make sure we install cinder requirements during the correct tox phase  https://review.opendev.org/65891520:34
*** enriquetaso has quit IRC20:35
openstackgerritDirk Mueller proposed openstack/cinder stable/pike: Make sure we install cinder requirements during the correct tox phase  https://review.opendev.org/65891620:35
*** dpawlik has quit IRC20:35
dirksmcginnis: looks like there were more missing deps in tox.ini (see https://review.opendev.org/658914)20:35
smcginnisdirk: Was just looking at that. I don't think so, I think this just adds more to some targets that don't need them.20:38
dirksmcginnis: reason for the constraint arg being injected (incorrectly imho) intot he deps option is the lower-constraints stuff20:39
dirkI was just trying to find a setenv or similar trick for this20:39
dirksmcginnis: any idea why -U is missing in the install command? that looks pretty broken20:40
smcginnisdirk: Why does that affect l-c?20:40
smcginnis-U is likely another oversight.20:40
*** pcaruana|afk| has quit IRC20:46
dirksmcginnis: is  your comment a -1? e.g. the releasenotes target is already installing all cinder deps (instead of just pulling in reno..)20:47
smcginnisdirk: I think that could get cleaned up, and api-ref and bandit could probably be left alone if we just move the constraints arg into the install_command like most other projects have it.20:48
smcginnisThat seems "more correct".20:48
dirksmcginnis: imho reno and linters should not be part of test-requirements (test-requirements should imho ~ match what is needed for unit/functional testing)20:48
smcginnisI'm assuming they were probably added to test-requirements because they are not needed to actually run the code.20:49
dirkright, but they're also not needed for tests20:51
*** openstackgerrit has quit IRC21:09
*** dklyle has joined #openstack-cinder21:13
_alastor_jungleboyj: Another day, another weird set of error from bare-metal tempest21:19
_alastor_jungleboyj: May 13 14:14:15 tlx3cp cinder-volume[29059]: ERROR oslo_messaging.rpc.server ImageNotAuthorized: Not authorized for image e1e6b8f0-416c-4263-8a81-cbd1f84efba4.21:19
_alastor_jungleboyj: I haven't changed the devstack setup, just the box where stack.sh was run21:20
_alastor_jungleboyj: It's like devstack decided this time not to give Cinder any Glance permissions at all21:21
*** dklyle has quit IRC21:24
*** mchlumsky has quit IRC21:38
*** whoami-rajat has quit IRC21:41
*** openstackgerrit has joined #openstack-cinder21:46
openstackgerritDirk Mueller proposed openstack/cinder master: Use testenv deps instead of repeating the dependencies  https://review.opendev.org/65891421:46
*** mriedem is now known as mriedem_afk21:53
*** lixiaoy1 has joined #openstack-cinder22:48
*** tkajinam has joined #openstack-cinder22:56
*** rcernin has joined #openstack-cinder23:04
*** rcernin has quit IRC23:07
*** rcernin has joined #openstack-cinder23:07
*** carloss has quit IRC23:17
*** dklyle has joined #openstack-cinder23:18
*** mriedem_afk has quit IRC23:31
*** hemna has quit IRC23:59

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