Tuesday, 2018-10-23

*** lixiaoy1 has joined #openstack-cinder00:10
*** rcernin has joined #openstack-cinder00:11
*** rcernin_ has quit IRC00:12
*** rcernin has quit IRC00:13
*** rcernin has joined #openstack-cinder00:14
*** chhagarw has joined #openstack-cinder00:47
*** gouthamr has joined #openstack-cinder00:47
*** brinzhang has joined #openstack-cinder00:50
daikk115enriquetaso, Absolutely yes. I will.00:50
*** chhagarw has quit IRC00:51
*** daikk115 has quit IRC00:53
*** LiangFang has joined #openstack-cinder01:15
*** imacdonn has quit IRC01:23
*** imacdonn has joined #openstack-cinder01:23
*** markvoelker has joined #openstack-cinder01:25
openstackgerritYikun Jiang proposed openstack/cinder master: Forbidden to revert volume to a different size snapshot  https://review.openstack.org/61149101:29
openstackgerritMerged openstack/cinder master: Remove WindowsDriver mapping  https://review.openstack.org/59532601:37
openstackgerritMerged openstack/cinder master: Fix debug message for _copy_volume_data  https://review.openstack.org/61246701:37
*** Dinesh_Bhor has joined #openstack-cinder01:42
*** mhen has quit IRC01:44
*** mhen has joined #openstack-cinder01:47
*** TuanDA has joined #openstack-cinder01:49
*** idlemind has joined #openstack-cinder01:53
*** Dinesh_Bhor has quit IRC01:59
openstackgerritYikun Jiang proposed openstack/cinder master: Microversion 3.56 - Support volume re-image  https://review.openstack.org/60634602:12
*** Dinesh_Bhor has joined #openstack-cinder02:35
*** TuanDA has quit IRC02:36
*** TuanDA has joined #openstack-cinder02:36
*** psachin has joined #openstack-cinder02:56
*** munimeha1 has quit IRC03:01
openstackgerrithzliaoyuehua proposed openstack/cinder master: Fix TypeError when retrieving pools filtered by volume-type  https://review.openstack.org/60406703:17
*** Dinesh_Bhor has quit IRC03:37
*** LiangFang has quit IRC03:49
*** chhagarw has joined #openstack-cinder03:58
*** udesale has joined #openstack-cinder03:58
*** faizy98 has joined #openstack-cinder04:14
*** faizy_ has quit IRC04:16
*** faizy98 has quit IRC04:34
*** Dinesh_Bhor has joined #openstack-cinder04:42
*** jiaopengju has quit IRC05:20
*** jiaopengju has joined #openstack-cinder05:23
*** LiangFang has joined #openstack-cinder05:24
*** Luzi has joined #openstack-cinder05:44
*** spsurya has joined #openstack-cinder05:48
*** faizy98 has joined #openstack-cinder05:57
openstackgerritFrancois Deppierraz proposed openstack/cinder master: cinder-volume: Stop masking IOError different than ENOSPC  https://review.openstack.org/61239306:03
*** hamdyk has joined #openstack-cinder06:06
*** francois has joined #openstack-cinder06:21
openstackgerritFrancois Deppierraz proposed openstack/cinder master: cinder-volume: Stop masking IOError different than ENOSPC  https://review.openstack.org/61239306:23
openstackgerritFrancois Deppierraz proposed openstack/cinder master: cinder-volume: Stop masking IOError different than ENOSPC  https://review.openstack.org/61239306:32
*** Dinesh_Bhor has quit IRC06:52
*** pcaruana has joined #openstack-cinder06:56
*** rcernin has quit IRC07:06
*** Dinesh_Bhor has joined #openstack-cinder07:16
openstackgerritYong Huang proposed openstack/cinder master: [Unity] Storage-assisted migration support  https://review.openstack.org/61262107:48
openstackgerritMerged openstack/cinder master: Fix unexpected behavior in _clone_image_volume  https://review.openstack.org/60623508:05
openstackgerritMerged openstack/cinder master: Fix wrong uuid recognized when create group  https://review.openstack.org/60566608:05
openstackgerritFrancois Deppierraz proposed openstack/cinder master: cinder-volume: Stop masking IOError different than ENOSPC  https://review.openstack.org/61239308:08
*** lixiaoy1 has quit IRC08:12
*** lixiaoy1 has joined #openstack-cinder08:13
*** pvradu has joined #openstack-cinder08:16
*** e0ne has joined #openstack-cinder08:23
*** mvkr has quit IRC08:26
chhagarwjgriffith: want some clarification on the comment https://bugs.launchpad.net/cinder/+bug/179367608:29
openstackLaunchpad bug 1793676 in Cinder "get_volume_stats is called twice during volume driver initialization" [Undecided,New]08:29
chhagarwjgriffith: For doing the replication check, do we need to make get_volume_stats, I was thinking if we can have a new API, get_replication_status which can be used to get the replication capabilities instead of getting the complete statistics.08:31
*** gouthamr has quit IRC08:32
openstackgerritFrancois Deppierraz proposed openstack/cinder master: cinder-volume: Stop masking IOError different than ENOSPC  https://review.openstack.org/61239308:40
*** mvkr has joined #openstack-cinder08:51
*** lixiaoy1 has quit IRC08:53
*** Dinesh_Bhor has quit IRC08:54
*** gouthamr has joined #openstack-cinder09:11
mnaserjungleboyj: smcginnis thank you for getting this through so quickly :)09:15
*** mvkr has quit IRC09:36
*** mvkr has joined #openstack-cinder09:50
*** udesale has quit IRC09:51
*** luizbag has joined #openstack-cinder09:52
*** udesale has joined #openstack-cinder09:52
*** sri_ has joined #openstack-cinder09:54
*** pvradu has quit IRC10:00
*** pvradu has joined #openstack-cinder10:00
sri_Hello, folks!, Does Ceph monitor IPs are still hardcoded in the database ? did anybody tried replacing ceph monitors(IP) recently ?10:03
*** TuanDA has quit IRC10:04
*** pvradu_ has joined #openstack-cinder10:05
*** Dinesh_Bhor has joined #openstack-cinder10:07
*** pvradu has quit IRC10:09
e0nesri_: hi. we store it in the config only10:14
*** Dinesh_Bhor has quit IRC10:15
sri_e0ne, thanks for conforming that, so if we add a new mon's or remove mon's all we need to do is update the cinder.conf file right ?10:16
e0nesri_: yes10:16
*** brinzh has joined #openstack-cinder10:17
sri_e0ne, cool, how does running instances pickup the new chages ?10:18
e0nesri_: you have to restart cinder-volume process10:19
*** brinzhang has quit IRC10:20
*** sdinescu has joined #openstack-cinder10:21
*** ianychoi has quit IRC10:22
sri_e0ne, hmmmm, I think that will not update the Qemu xml!, and also do we need perform any one of this operations to get the chages migrate, live-migration, evacuate, host-evacuate, host-evacuate-live, host-servers-migrate and shelve/unshelve will also cause the instance to read the stale monitor IP from the database10:22
*** ianychoi has joined #openstack-cinder10:25
sri_e0ne, i might be wrong, I get that information form this thread https://bugzilla.redhat.com/show_bug.cgi?id=1414124#c110:28
openstackbugzilla.redhat.com bug 1414124 in openstack-nova "Ceph Monitor hardcoded IPs in Nova database" [High,Assigned] - Assigned to lyarwood10:28
e0nesri_: oh.. I don't know how it works in nova10:29
*** LiangFang has quit IRC10:30
sri_e0ne, ack, anyway thanks for your help :)10:32
e0neyou're welcome10:32
*** erlon has joined #openstack-cinder10:58
*** ganso has joined #openstack-cinder11:03
*** yikun has quit IRC11:09
*** udesale has quit IRC11:14
*** pck has joined #openstack-cinder11:28
*** eharney has joined #openstack-cinder11:45
*** markvoelker has quit IRC11:45
*** mmethot has joined #openstack-cinder11:54
*** abishop has joined #openstack-cinder12:00
*** dave-mccowan has joined #openstack-cinder12:07
openstackgerritEric Harney proposed openstack/python-cinderclient master: Fix doc build error  https://review.openstack.org/61170812:22
*** abishop has quit IRC12:24
*** abishop has joined #openstack-cinder12:30
*** abishop has quit IRC12:31
*** udesale has joined #openstack-cinder12:31
*** markvoelker has joined #openstack-cinder12:36
*** markvoelker has quit IRC12:37
*** brinzh has quit IRC12:40
*** lixiaoy1 has joined #openstack-cinder12:48
*** dustins has joined #openstack-cinder12:51
*** dustins is now known as dschoenb|worksho12:52
*** dschoenb|worksho is now known as dustins12:52
*** abishop has joined #openstack-cinder12:58
*** lixiaoy1 has quit IRC13:05
*** lixiaoy1 has joined #openstack-cinder13:06
*** psachin has quit IRC13:10
*** e0ne has quit IRC13:19
*** mriedem has joined #openstack-cinder13:34
*** lixiaoy1 has quit IRC13:38
openstackgerritMerged openstack/cinder master: Fix multiattach set to false after retype  https://review.openstack.org/60565013:46
openstackgerritMerged openstack/cinder stable/pike: Fix image volume cache max size and max count limits  https://review.openstack.org/61247013:57
*** edmondsw has joined #openstack-cinder14:00
*** Luzi has quit IRC14:06
openstackgerritMerged openstack/os-brick stable/queens: The validation of iscsi session should be case insensitive  https://review.openstack.org/61121314:17
*** daikk115 has joined #openstack-cinder14:19
openstackgerritMerged openstack/cinder stable/queens: Handling type's multiattach extra specs in API task flow  https://review.openstack.org/59319114:25
openstackgerritMerged openstack/cinder stable/pike: Storwize: self assign the SCSI lun id for volume attaching  https://review.openstack.org/60912314:25
*** noonedeadpunk has joined #openstack-cinder14:29
*** e0ne has joined #openstack-cinder14:29
noonedeadpunkHi everyone. After Q -> R upgrade, cinder-volume started failing with this error http://paste.openstack.org/show/732829/ Previously this problem was non-critical, and didn't influence on cinder14:30
noonedeadpunkWe have some manually created volumes inside this ceph pool, and we were going to migrate these images to cinder (via cinder manage) within some time14:31
noonedeadpunkI've found some common bug https://bugs.launchpad.net/cinder/+bug/1698786 which has been fixed by e0ne some time ago14:32
openstackLaunchpad bug 1698786 in Cinder "cinder-volume fails on start when rbd pool contains partially deleted images" [Undecided,Fix released] - Assigned to Ivan Kolodyazhny (e0ne)14:32
jgriffithchhagarw: still around?14:38
jgriffithchhagarw: I'll add comments in the bug report; so I like the idea of scaling that down14:39
jgriffithchhagarw: but it's not as simple as just a property or config setting14:39
jgriffithchhagarw: in most cases the actual device should be queried14:40
jgriffithchhagarw: I never had a problem with stats because for the devices I worked on it wasn't *expensive*14:40
jgriffithbut I realize it might be for some14:40
jgriffithgiven it's a periodic call I'm not sure how much it gets you to cut out just 1 of the calls at init though14:41
openstackgerritCorey Bryant proposed openstack/cinder master: PY3: Ensure rados.Object.read/write use byte data  https://review.openstack.org/61249014:42
*** hamdyk has quit IRC14:43
openstackgerritLiang Fang proposed openstack/cinder master: WIP: solution 2 of convey image size to glance  https://review.openstack.org/60999414:46
whoami-rajateharney:  Hi, can you please respond to my query on https://review.openstack.org/#/c/587610/ . Thanks.15:30
*** ianychoi_ has joined #openstack-cinder15:36
eharneywhoami-rajat: responded there... i like it more than i did the last time i looked at it15:38
*** mvkr has quit IRC15:39
*** e0ne has quit IRC15:39
*** ianychoi has quit IRC15:40
mriedemhello cinder people. looking at the backup API reference, it says only available (not in-use/reserved/etc) volumes can be backed up https://developer.openstack.org/api-ref/block-storage/v3/#backups-backups16:02
mriedembut i'm not really finding in the code where that status check happens...16:02
mriedemoh nvm i found it16:03
mriedemelif volume['status'] not in ["available", "in-use"]:16:03
mriedemoh look there, you can backup attached volumes...16:03
smcginnismriedem: There's a force option in the request args - https://developer.openstack.org/api-ref/block-storage/v3/?expanded=create-a-backup-detail#create-a-backup16:03
mriedemyeah the api ref is misleading when it says "Backup and restore operations can only be carried out on volumes that are in an unattached and available state."16:04
mriedemshould have a caveat about the force flag16:04
mriedemi'll push a change16:04
*** udesale has quit IRC16:06
mriedemstill trying to hash this out https://review.openstack.org/#/c/530214/16:11
mriedemif you backup a volume-backed server, do you get a volume snapshot or a volume backup16:11
mriedemi would think the latter16:12
smcginnisYes, that later.16:12
smcginnisIt will take a snapshot to be able to backup from, but it will be a backup.16:12
mriedemi mean, as a user i would expect a volume backup16:12
smcginnisAnd that snapshot will be deleted after the backup is complete.16:12
mriedemthe createImage api in nova creates a volume snapshot of each volume attached to the server, and then records metadata about those snapshots in the image16:13
smcginnisIt doesn't do anything else with the snaps?16:13
mriedemyou get an image snapshot and N volume snapshots where N is the number of volumes attached to the server, including the root volume,16:15
mriedemmetadata about those volume snapshots is recorded in the image meta,16:15
mriedemand if you create a new server from that image, nova-compute will actually create a volume-backed server from the volume snapshot and attach new volumes for any of the data volume snapshots16:16
mriedemotherwise referred to as image-defined BDMs16:16
mriedemand that's apparently a pretty common thing in EC216:16
mriedembut it's not used often in openstack i don't think - it's pretty confusing16:16
smcginnisIt concerns me there's nothing protecting those externally referenced snapshots, but short of making them full backups (and requiring all that to be configured) or uploading them as glance images, I guess that's OK.16:16
mriedemanywho, we don't really even want the createBackup API in nova anymore,16:18
mriedemit could all be done externally, it's just orchestration16:18
mriedemcreate a snapshot with some metadata for rotations and then manage rotations yourself16:18
mriedemsame could be done with volume-backed servers16:18
smcginnisDoes nova do scheduling and retention for backups?16:19
mriedemoh, like a cron16:20
smcginnisYeah, that meaning of scheduling.16:20
smcginnisCinder does not. We just provide the mechanism to get a backup with the expectation that you will use Karbor or your own scripting to actually orchestrate anything beyond that.16:21
mriedemnova is the same16:21
mriedemyou create backups with a type (daily/weekly) and max number16:21
mriedemso the next time you create a backup, we check to see if any older backups should be deleted16:21
mriedembut that could all be done externally16:21
smcginnisWhere does that daily/weekly come into play?16:22
smcginnisOh wait, I see.16:22
smcginnisSo nova does do a little more orchestration beyond just creating a backup.16:22
smcginnisWe don't do that.16:22
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Fixes bug, when cinder-volume fails  on broken RBD image  https://review.openstack.org/61274116:23
smcginnisIt's entirely up to something external to manage the lifetime of the backups and anything else that needs to happen to them.16:23
mriedemright, and that's kind of what we want people to be doing as well16:23
mriedemsince we already have the createImage API which handles volume-backed servers16:23
smcginnisMakes sense.16:24
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Handle rbd.OSError on broken RBD image  https://review.openstack.org/61274116:25
*** pvradu_ has quit IRC16:37
openstackgerritRajat Dhasmana proposed openstack/cinder master: Fix and Optimize : retype tests  https://review.openstack.org/61274916:41
openstackgerritMatt Riedemann proposed openstack/cinder master: api-ref: cleanup status conditions for backup/restore APIs  https://review.openstack.org/61275516:56
*** pvradu has joined #openstack-cinder17:04
*** daikk115 has quit IRC17:06
*** daikk115 has joined #openstack-cinder17:07
*** e0ne has joined #openstack-cinder17:08
*** pvradu has quit IRC17:09
*** daikk115 has quit IRC17:12
jungleboyjsmcginnis:  How many times did we talk about backup scheduling at the PTG?17:15
*** lbragstad is now known as lbragstad_f00d17:17
e0nejungleboyj, smcginnis: what did I missed about backup scheduling?17:18
e0nejungleboyj, smcginnis: we'd got, a least, 2 discussions about this AFAIR17:18
jungleboyje0ne:  I think it was from the User feedback where there were a number of requests to be able to schedule backups.  The answer we kept coming back to is that that isn't really Cinder's job but for one of the external services or user automation.17:40
*** tpsilva has joined #openstack-cinder17:40
e0nejungleboyj: oh.. got it. it's not about backup via scheduler feature17:40
jungleboyje0ne:  No.  It was about people want to set their systems to do automatic backups on a scheduler.17:41
*** pvradu has joined #openstack-cinder17:42
e0neI understood it now. usually, we recommend to use something like mistral for such automation17:42
jungleboyje0ne:  ++17:42
*** e0ne has quit IRC17:45
*** lbragstad_f00d is now known as lbragstad17:45
openstackgerritDai Dang Van proposed openstack/cinder master: Update backup flow for CephBackupDriver  https://review.openstack.org/61064517:48
openstackgerritDai Dang Van proposed openstack/cinder master: Update backup flow for CephBackupDriver  https://review.openstack.org/61064517:49
*** pvradu has quit IRC17:53
*** gary_perkins has quit IRC18:00
*** e0ne has joined #openstack-cinder18:31
*** e0ne has quit IRC18:31
*** irclogbot_1 has joined #openstack-cinder18:35
*** gary_perkins has joined #openstack-cinder18:51
*** chhagarw has quit IRC18:57
openstackgerritErlon R. Cruz proposed openstack/cinder master: SF: Handle qos values on extend volume  https://review.openstack.org/60695919:11
*** e0ne has joined #openstack-cinder19:15
*** pcaruana has quit IRC19:23
*** david-lyle has joined #openstack-cinder19:27
*** dklyle has quit IRC19:28
*** lbragstad has quit IRC19:50
*** lbragstad has joined #openstack-cinder19:53
*** david-lyle is now known as dklyle19:53
*** openstackgerrit has quit IRC20:06
*** luizbag has quit IRC20:06
*** mvkr has joined #openstack-cinder20:12
*** openstackgerrit has joined #openstack-cinder20:40
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Handle rbd.OSError on broken RBD image  https://review.openstack.org/61274120:40
openstackgerritDmitriy Rabotjagov (noonedeadpunk) proposed openstack/cinder master: Handle rbd.OSError on broken RBD image  https://review.openstack.org/61274120:58
*** mchlumsky has quit IRC21:01
*** erlon has quit IRC21:03
*** e0ne has quit IRC21:04
*** dustins has quit IRC21:19
*** dklyle has quit IRC21:19
*** dklyle has joined #openstack-cinder21:20
openstackgerritMerged openstack/cinder master: Drop drv_cfg from rootwrap filters  https://review.openstack.org/61164621:33
openstackgerritMerged openstack/cinder master: api-ref: cleanup status conditions for backup/restore APIs  https://review.openstack.org/61275521:33
*** spsurya has quit IRC21:38
*** rtjure has quit IRC21:39
*** dustins has joined #openstack-cinder21:41
*** eharney has quit IRC22:13
*** abishop has quit IRC22:13
*** dustins has quit IRC22:15
*** rcernin has joined #openstack-cinder22:24
*** mriedem has quit IRC22:26
*** threestrands has joined #openstack-cinder23:02
*** tpsilva has quit IRC23:11
*** ganso has quit IRC23:17
*** idlemind has quit IRC23:38
*** lixiaoy1 has joined #openstack-cinder23:40

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