Tuesday, 2015-10-06

*** diogogmt has quit IRC00:00
jgriffithpatrickeast: so the trick for me is that our names are special, we can't change them00:00
*** kevincarr1991 has joined #openstack-cinder00:00
patrickeastjgriffith: oh, that does make it harder to upgrade to a different naming scheme then00:02
*** garthb has quit IRC00:04
*** kevincarr1991 has quit IRC00:04
jgriffithyeah, breaks the world :(00:07
aorourke_jgriffith, hey00:16
*** tsekiyama has quit IRC00:17
*** setmason has joined #openstack-cinder00:17
*** pv_ has quit IRC00:18
*** tsekiyama has joined #openstack-cinder00:18
*** takedakn has joined #openstack-cinder00:21
*** ntpttr has joined #openstack-cinder00:21
*** ntpttr has quit IRC00:27
*** takedakn has quit IRC00:28
*** mtanino has quit IRC00:29
*** julim has joined #openstack-cinder00:30
*** tsekiyama has quit IRC00:36
*** julim has quit IRC00:41
*** dims_ has joined #openstack-cinder00:50
*** leeantho has quit IRC00:51
*** lcurtis_ has joined #openstack-cinder00:53
*** _cjones_ has quit IRC00:55
*** dims_ has quit IRC00:55
*** dims_ has joined #openstack-cinder00:58
*** Guest71307 has quit IRC01:02
*** kevincarr1991 has joined #openstack-cinder01:06
*** kevincarr1991 has quit IRC01:11
*** apoorvad_ has joined #openstack-cinder01:11
*** apoorvad has quit IRC01:14
*** gouthamr has quit IRC01:14
*** apoorvad_ has quit IRC01:15
*** akerr is now known as akerr_away01:17
*** haigang has joined #openstack-cinder01:22
*** akerr_away is now known as akerr01:22
*** haigang has quit IRC01:24
*** asselin_ has joined #openstack-cinder01:27
*** markvoelker has quit IRC01:28
*** dims_ has quit IRC01:30
*** asselin_ has quit IRC01:31
*** Lee1092 has joined #openstack-cinder01:32
*** mriedem_away has quit IRC01:35
*** tsekiyama has joined #openstack-cinder01:37
*** fandi has quit IRC01:41
*** tsekiyama has quit IRC01:42
*** Yogi1 has joined #openstack-cinder01:48
*** stevemar_ has quit IRC01:59
*** stevemar_ has joined #openstack-cinder02:00
*** kevincarr1991 has joined #openstack-cinder02:00
*** kevincarr1991 has quit IRC02:05
*** Yogi1 has quit IRC02:05
jungleboyjpatrickeast: Ok, I see that you updated the commit.  Thanks, I will take a look.02:06
patrickeastjungleboyj: thanks!02:09
*** smoriya has joined #openstack-cinder02:12
*** stevemar_ has quit IRC02:41
*** stevemar_ has joined #openstack-cinder02:42
*** stevemar_ has quit IRC02:43
*** stevemar_ has joined #openstack-cinder02:43
*** dave-mccowan has quit IRC02:44
*** jamielennox has quit IRC02:51
*** martyturner has joined #openstack-cinder02:59
*** markvoelker_ has joined #openstack-cinder03:05
*** kevincarr1991 has joined #openstack-cinder03:06
*** fandi has joined #openstack-cinder03:10
*** kevincarr1991 has quit IRC03:10
*** jamielennox has joined #openstack-cinder03:13
*** harlowja_at_home has joined #openstack-cinder03:15
*** haomaiwang has joined #openstack-cinder03:16
*** stevemar_ has quit IRC03:21
*** markvoelker has joined #openstack-cinder03:22
*** stevemar_ has joined #openstack-cinder03:22
*** akerr is now known as akerr_away03:23
*** markvoelker_ has quit IRC03:25
*** markvoelker_ has joined #openstack-cinder03:26
*** markvoelker has quit IRC03:28
*** haomaiwang has quit IRC03:34
*** martyturner has quit IRC03:38
*** haomaiwang has joined #openstack-cinder03:42
*** DericHorn-HP has joined #openstack-cinder03:47
*** links has joined #openstack-cinder03:48
*** lcurtis_ has quit IRC03:50
*** DericHorn-HP has quit IRC03:51
*** pradipta has joined #openstack-cinder03:56
*** DericHorn-HP has joined #openstack-cinder03:57
*** kevincarr1991 has joined #openstack-cinder04:00
*** haomaiwang has quit IRC04:01
*** haomaiwang has joined #openstack-cinder04:01
*** kevincarr1991 has quit IRC04:04
*** harlowja_at_home has quit IRC04:15
*** daneyon has joined #openstack-cinder04:35
*** daneyon_ has joined #openstack-cinder04:35
*** daneyon has quit IRC04:39
*** dims_ has joined #openstack-cinder04:46
*** DericHorn-HP has quit IRC04:54
*** DericHorn-HP has joined #openstack-cinder04:56
*** haomaiwang has quit IRC04:58
openstackgerritwatanabe isao proposed openstack/cinder: Don't merge  https://review.openstack.org/23088005:01
*** haomaiwang has joined #openstack-cinder05:01
*** subscope has joined #openstack-cinder05:01
*** _fortis has quit IRC05:02
*** cbader_ has joined #openstack-cinder05:02
*** MIDENN_ has joined #openstack-cinder05:02
*** morgan has quit IRC05:02
*** cebruns has quit IRC05:03
*** mdenny has quit IRC05:03
*** ByPasS has quit IRC05:03
*** cbader has quit IRC05:03
*** mfisch has quit IRC05:03
*** cebruns has joined #openstack-cinder05:04
*** tristanC has quit IRC05:04
*** tristanC has joined #openstack-cinder05:05
*** mfisch has joined #openstack-cinder05:06
*** mfisch is now known as Guest8345405:06
*** morgan has joined #openstack-cinder05:06
*** ByPasS has joined #openstack-cinder05:08
*** setmason_ has joined #openstack-cinder05:08
*** dsariel_ has quit IRC05:10
*** setmason has quit IRC05:11
*** setmason_ is now known as setmason05:11
*** _fortis has joined #openstack-cinder05:13
*** BharatK has joined #openstack-cinder05:21
*** daneyon_ has quit IRC05:23
*** daneyon has joined #openstack-cinder05:25
*** daneyon_ has joined #openstack-cinder05:28
*** daneyon has quit IRC05:30
*** pradipta has quit IRC05:31
*** setmason has quit IRC05:37
*** setmason has joined #openstack-cinder05:37
*** lpetrut has joined #openstack-cinder05:37
*** bkopilov has joined #openstack-cinder05:38
*** setmason_ has joined #openstack-cinder05:43
*** setmason has quit IRC05:46
*** setmason_ is now known as setmason05:46
*** DericHorn-HP has quit IRC05:47
*** edtubill has joined #openstack-cinder05:51
*** setmason has quit IRC05:53
*** setmason has joined #openstack-cinder05:54
*** kevincarr1991 has joined #openstack-cinder06:00
*** dsariel_ has joined #openstack-cinder06:00
*** haomaiwang has quit IRC06:01
*** haomaiwang has joined #openstack-cinder06:01
openstackgerrityogeshprasad proposed openstack/cinder: Retype support for CloudByte iSCSI cinder driver  https://review.openstack.org/21864506:02
*** kevincarr1991 has quit IRC06:05
*** lpetrut has quit IRC06:07
*** e0ne has joined #openstack-cinder06:22
*** markvoelker_ has quit IRC06:24
*** dsariel_ has quit IRC06:27
*** e0ne has quit IRC06:30
*** haomaiwang has quit IRC06:31
*** uberjay has quit IRC06:33
*** markvoelker has joined #openstack-cinder06:35
*** uberjay has joined #openstack-cinder06:36
*** stevemar_ has quit IRC06:38
*** vlaza has joined #openstack-cinder06:39
*** markvoelker has quit IRC06:40
*** dsariel_ has joined #openstack-cinder06:40
*** nkrinner has joined #openstack-cinder06:41
*** jamielennox is now known as jamielennox|away06:41
*** yrabl has joined #openstack-cinder06:42
*** markvoelker has joined #openstack-cinder06:45
*** markvoelker has quit IRC06:50
*** jwang_ has quit IRC06:54
*** jwang_ has joined #openstack-cinder06:54
*** anshul has joined #openstack-cinder06:56
*** anshul has quit IRC06:57
*** ekarlso has quit IRC06:58
*** bkopilov has quit IRC06:58
*** anshul has joined #openstack-cinder06:58
*** markvoelker has joined #openstack-cinder07:00
*** markvoelker has quit IRC07:04
*** alexschm has joined #openstack-cinder07:05
*** markvoelker has joined #openstack-cinder07:14
*** ollie2 has joined #openstack-cinder07:17
*** markvoelker has quit IRC07:19
*** ekarlso has joined #openstack-cinder07:25
*** rushiagr_away is now known as rushiagr07:27
*** markvoelker has joined #openstack-cinder07:29
*** markvoelker has quit IRC07:33
*** lpetrut has joined #openstack-cinder07:37
*** edtubill has quit IRC07:37
*** lpetrut has quit IRC07:41
*** markvoelker has joined #openstack-cinder07:43
*** Anticimex has quit IRC07:46
*** setmason has quit IRC07:46
*** Anticimex has joined #openstack-cinder07:47
*** markvoelker has quit IRC07:48
*** lpetrut has joined #openstack-cinder07:50
*** xek_ is now known as xek07:53
openstackgerritYusuke Hayashi proposed openstack/cinder: Add Fujitsu ETERNUS DX Volume Driver (again)  https://review.openstack.org/20150007:54
*** markvoelker has joined #openstack-cinder07:58
*** kevincarr1991 has joined #openstack-cinder08:01
*** sgotliv has joined #openstack-cinder08:02
*** markvoelker has quit IRC08:02
*** kevincarr1991 has quit IRC08:06
*** markvoelker has joined #openstack-cinder08:13
*** kmartin has quit IRC08:13
*** rushiagr is now known as rushiagr_away08:13
*** kmartin has joined #openstack-cinder08:13
*** markvoelker has quit IRC08:18
*** smoriya has quit IRC08:24
*** jistr has joined #openstack-cinder08:25
*** bkopilov has joined #openstack-cinder08:27
*** markvoelker has joined #openstack-cinder08:27
*** patrickeast has quit IRC08:30
*** markvoelker has quit IRC08:32
*** patrickeast has joined #openstack-cinder08:32
*** rushiagr_away is now known as rushiagr08:38
*** markvoelker has joined #openstack-cinder08:42
*** markvoelker has quit IRC08:47
*** aix has quit IRC08:48
*** markvoelker has joined #openstack-cinder08:56
*** bkopilov has quit IRC08:58
*** markvoelker has quit IRC09:01
*** bkopilov has joined #openstack-cinder09:04
*** aix has joined #openstack-cinder09:06
*** markvoelker has joined #openstack-cinder09:06
*** markvoelker has quit IRC09:10
*** yrabl has quit IRC09:12
*** markvoelker has joined #openstack-cinder09:13
*** markvoelker has quit IRC09:18
*** yrabl has joined #openstack-cinder09:22
*** markvoelker has joined #openstack-cinder09:22
*** markvoelker has quit IRC09:26
*** markvoelker has joined #openstack-cinder09:32
*** akerr_away has quit IRC09:33
*** markvoelker has quit IRC09:39
*** akerr has joined #openstack-cinder09:41
*** deepakcs has joined #openstack-cinder09:42
*** aix has quit IRC09:45
*** markvoelker has joined #openstack-cinder09:46
*** markvoelker has quit IRC09:50
*** akerr has quit IRC09:52
*** haypo has joined #openstack-cinder09:54
*** akerr has joined #openstack-cinder09:55
*** IanGovett has joined #openstack-cinder09:57
*** aix has joined #openstack-cinder09:57
openstackgerritoliver-leahy-l proposed openstack/cinder: encryption_api_url requires a version  https://review.openstack.org/23003109:59
openstackgerritYusuke Hayashi proposed openstack/cinder: Add Fujitsu ETERNUS DX Volume Driver (again)  https://review.openstack.org/20150010:01
*** markvoelker has joined #openstack-cinder10:01
*** kevincarr1991 has joined #openstack-cinder10:02
*** e0ne has joined #openstack-cinder10:03
*** markvoelker has quit IRC10:05
*** kevincarr1991 has quit IRC10:07
*** akerr has quit IRC10:09
*** markvoelker has joined #openstack-cinder10:11
*** jamielennox|away is now known as jamielennox10:14
*** markvoelker has quit IRC10:16
*** markvoelker has joined #openstack-cinder10:17
*** markvoelker has quit IRC10:22
*** dave-mccowan has joined #openstack-cinder10:24
*** markvoelker has joined #openstack-cinder10:26
*** ociuhandu has quit IRC10:32
*** markvoelker has quit IRC10:34
*** markvoelker has joined #openstack-cinder10:55
*** ociuhandu has joined #openstack-cinder10:57
*** markvoelker has quit IRC11:01
*** ociuhandu has quit IRC11:02
*** ociuhandu_ has joined #openstack-cinder11:02
*** daneyon_ has quit IRC11:03
*** markvoelker has joined #openstack-cinder11:05
*** BharatK has quit IRC11:06
*** akerr has joined #openstack-cinder11:10
*** markvoelker has quit IRC11:10
*** BharatK has joined #openstack-cinder11:17
*** lpetrut has quit IRC11:18
*** deepakcs has quit IRC11:19
*** markvoelker has joined #openstack-cinder11:19
*** markvoelker has quit IRC11:24
*** adrianofr has joined #openstack-cinder11:29
*** kevincarr1991 has joined #openstack-cinder11:30
*** timcl has joined #openstack-cinder11:33
*** markvoelker has joined #openstack-cinder11:34
*** kevincarr1991 has quit IRC11:34
*** markvoelker has quit IRC11:38
*** takedakn has joined #openstack-cinder11:41
openstackgerritoliver-leahy-l proposed openstack/cinder: encryption_api_url requires a version  https://review.openstack.org/23003111:45
*** tpsilva has joined #openstack-cinder11:46
*** ollie2 has left #openstack-cinder11:48
*** markvoelker has joined #openstack-cinder11:48
*** ociuhandu_ has quit IRC11:53
*** markvoelker has quit IRC11:53
*** ociuhandu has joined #openstack-cinder11:54
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Fix Status-Line in HTTP response  https://review.openstack.org/23143811:54
*** bkopilov has quit IRC11:59
*** markvoelker has joined #openstack-cinder12:03
*** akerr has quit IRC12:07
*** markvoelker has quit IRC12:07
*** markvoelker has joined #openstack-cinder12:14
*** gouthamr has joined #openstack-cinder12:18
*** markvoelker has quit IRC12:19
*** rushiagr is now known as rushiagr_away12:21
adrianofrhemnafk: ping12:21
*** geektoor has joined #openstack-cinder12:22
*** haomaiwang has joined #openstack-cinder12:25
*** lpetrut has joined #openstack-cinder12:28
*** markvoelker has joined #openstack-cinder12:29
*** markvoelker has quit IRC12:36
*** crose has joined #openstack-cinder12:37
*** diablo_rojo has joined #openstack-cinder12:37
*** tpsilva has quit IRC12:38
*** ekarlso has quit IRC12:38
*** tpsilva has joined #openstack-cinder12:38
*** takedakn has quit IRC12:39
*** markvoelker has joined #openstack-cinder12:44
*** edmondsw has joined #openstack-cinder12:45
*** akerr has joined #openstack-cinder12:46
*** merooney has joined #openstack-cinder12:48
*** markvoelker has quit IRC12:48
*** merooney has quit IRC12:49
*** jaypipes has joined #openstack-cinder12:50
*** julim has joined #openstack-cinder12:50
*** jungleboyj has quit IRC12:51
*** pckizer has quit IRC12:53
*** pckizer has joined #openstack-cinder12:53
*** timcl has quit IRC12:54
*** ekarlso has joined #openstack-cinder12:56
*** merooney has joined #openstack-cinder12:56
*** porrua has joined #openstack-cinder12:56
*** markvoelker has joined #openstack-cinder12:58
*** sgotliv has quit IRC12:58
*** markvoelker has quit IRC12:59
*** markvoelker has joined #openstack-cinder12:59
*** pckizer has quit IRC12:59
*** pckizer has joined #openstack-cinder12:59
*** markvoelker_ has joined #openstack-cinder13:01
*** haomaiwang has quit IRC13:01
*** haomaiwang has joined #openstack-cinder13:01
*** bswartz has joined #openstack-cinder13:02
openstackgerritDeepak C Shetty (deepakcs) proposed openstack/cinder-specs: Moving assisted-snapshot-improvements spec to mitaka  https://review.openstack.org/23146313:02
*** markvoelker has quit IRC13:03
*** markvoelker has joined #openstack-cinder13:06
*** Yogi1 has joined #openstack-cinder13:08
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Fix Status-Line in HTTP response  https://review.openstack.org/23143813:08
*** markvoelker_ has quit IRC13:10
*** setmason has joined #openstack-cinder13:11
*** nkrinner has quit IRC13:11
*** EinstCrazy has joined #openstack-cinder13:11
*** changbl has quit IRC13:13
*** nkrinner has joined #openstack-cinder13:14
*** EinstCrazy has quit IRC13:14
*** raildo-afk is now known as raildo13:15
*** sgotliv has joined #openstack-cinder13:15
*** bill_az has joined #openstack-cinder13:15
*** ronis has joined #openstack-cinder13:15
*** markvoelker has quit IRC13:17
*** markvoelker has joined #openstack-cinder13:17
*** dave-mccowan has quit IRC13:19
*** Guest92657 has joined #openstack-cinder13:22
*** Guest92657 is now known as grumpycatt13:23
*** thangp has joined #openstack-cinder13:24
*** mriedem has joined #openstack-cinder13:24
*** haomaiwang has quit IRC13:27
thangpgeguileo: ping13:27
*** haomaiwang has joined #openstack-cinder13:27
thangpgeguileo: re - https://review.openstack.org/19502713:28
*** setmason_ has joined #openstack-cinder13:28
thangpgeguileo: I put up a reply... I do not agree that consistencygroup should be an extra field13:28
*** jgregor has joined #openstack-cinder13:30
*** crose has quit IRC13:30
*** kevincarr1991 has joined #openstack-cinder13:31
*** setmason has quit IRC13:31
*** setmason_ is now known as setmason13:31
*** martyturner has joined #openstack-cinder13:32
*** bill_az_ has joined #openstack-cinder13:33
*** dustins has joined #openstack-cinder13:34
*** dave-mccowan has joined #openstack-cinder13:34
*** kevincarr1991 has quit IRC13:35
openstackgerritKendall Nelson proposed openstack/cinder: Hacking check for opt name registration  https://review.openstack.org/22337513:36
geguileothangp: pong13:39
*** vgridnev has joined #openstack-cinder13:39
thangpgeguileo: hey13:40
thangpgeguileo: i put up a reply to 19502713:40
geguileoReading it13:41
thangpgeguileo: we have been using nova as an example of how to do things (good or bad), so it's the same for the patch13:41
thangpgeguileo: if it's not something correct, we can ask dan13:42
*** ronis has quit IRC13:42
*** ronis has joined #openstack-cinder13:43
geguileothangp: I'm looking at the instance nova object13:46
geguileothangp: And it's different13:46
geguileothangp: The way they have their ORM objects and ours are quite different13:46
thangpgeguileo: how so?13:47
*** haomaiwang has quit IRC13:47
thangpgeguileo: maybe another example is https://github.com/openstack/nova/blob/master/nova/objects/floating_ip.py#L4913:49
*** mriedem has quit IRC13:49
thangpgeguileo: and https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/models.py#L91613:49
*** haomaiwa_ has joined #openstack-cinder13:50
dulekgeguileo, thangp: As I've noted they're sometimes skipping relationships in objects and I've found at least one occurrence where they used object properties to make DB calls.13:51
geguileothangp: Ok, and if you look at fixed_ip they don't allow you changing fixed_ip_id13:51
thangpdulek: true13:52
dulekgeguileo, thangp: No obj_extra_fields though. Also I believe they're actually using same sqlalchemy's relationship  to build relationships.13:52
thangpgeguileo: true, and we can add it in13:52
thangpgeguileo: my only point is that you can use fields for properties that are not part of the orm model13:52
*** diogogmt has joined #openstack-cinder13:53
geguileothangp: True13:53
dulekgeguileo: I agree on the non-mutable id. Good catch.13:54
geguileothangp: dulek Mmmm, you guys are probably right on the consistencygroup issue13:55
thangpgeguileo: cool, so we just need to put a catch in the save to not update consistencygroup if its changed, and some other ones you identified elsewhere13:56
geguileothangp: And you are missing the snapshots field as well13:56
geguileoI'll update my review13:57
thangpgeguileo: why snapshot?13:57
geguileoPlural, it's the backref13:57
thangpgeguileo: oh13:58
geguileoI just catched on that now  XD13:58
dulekgeguileo: The more we get into it the worse it becomes. ;)13:58
thangpgeguileo: to put cgsnapshot in snapshot object?13:58
geguileothangp: No, there is a ref in snapshot to cgsnapshot that is called snapshots in cgsnapshot13:59
dulekgeguileo: I guess you've had intuition that there's something wrong. ;)13:59
geguileodulek: Lets call it that   XD13:59
dulekthangp: "backref="cgsnapshots","13:59
thangpgeguileo: dulek: ah ok14:00
dulekthangp, geguileo: There probably are more mistakes like that - I'll take a look.14:00
geguileodulek: Actually backref="snapshots",14:00
*** haomaiwa_ has quit IRC14:01
*** haomaiwang has joined #openstack-cinder14:01
*** breitz has joined #openstack-cinder14:01
dulekgeguileo: Ah, right. I think the one I've mentioned may also be missing.14:02
geguileodulek: Probably, that would be in snapshot object14:02
dulekgeguileo: ConsistencyGroup actually and it's missing there (that's merged).14:03
thangpgeguileo: so we need to add cgsnapshot field in snapshot object, correct?14:04
geguileothangp: I'd have to look at that14:04
*** mriedem has joined #openstack-cinder14:04
thangpgeguileo: and add a list of snapshots in cgsnapshots14:04
geguileothangp: That second part for sure14:04
*** timcl has joined #openstack-cinder14:09
*** diogogmt has quit IRC14:13
mriedemdoes this seem like a long time to be holding a lock in c-vol? http://logs.openstack.org/10/230510/2/gate/gate-grenade-dsvm/e796b39/logs/new/screen-c-vol.txt.gz#_2015-10-06_05_40_15_93614:13
mriedem2015-10-06 05:40:15.936 DEBUG oslo_concurrency.lockutils [req-6b402bc1-3c24-47f8-be65-609bfb98c8fd tempest-TestVolumeBootPatternV2-836916434] Lock "35fbbb48-90a3-45e2-9275-d66c9b8e9b2b-delete_snapshot" released by "cinder.volume.manager._run_flow_locked" :: held 16.103s inner /usr/local/lib/python2.7/dist-packages/oslo_concurrency/lockutils.py:26514:13
*** ntpttr has joined #openstack-cinder14:13
dulekmriedem: Cinder had that lock for whole operation of volume creation.14:15
dulekmriedem: It is blocking snapshot removal while we're creating a volume from it.14:15
mriedemok, i guess that lock is per-volume too14:16
*** sgotliv has quit IRC14:16
*** eharney has joined #openstack-cinder14:17
dulekmriedem: This is kind of a little strange construct. When we're creating a volume from snapshot or source volume, we lock the source resource with <resource-id>-delete_resource.14:18
dulekmriedem: Then we have decorators for that in delete_snapshot and delete_volume methods in manager.14:18
dulekmriedem: So we're protecting the origin resource.14:19
smcginnisall: Liberty RC2 is cut. Nice work.14:19
*** Guest83454 is now known as mfisch14:19
*** mfisch is now known as Guest9449014:20
*** jungleboyj has joined #openstack-cinder14:26
*** dustins has quit IRC14:26
*** yrabl has quit IRC14:27
*** stevemar_ has joined #openstack-cinder14:28
*** setmason has quit IRC14:28
*** yrabl has joined #openstack-cinder14:29
*** bswartz has quit IRC14:29
*** jaypipes has quit IRC14:32
*** markvoelker_ has joined #openstack-cinder14:32
*** kevincarr1991 has joined #openstack-cinder14:33
*** baumann has joined #openstack-cinder14:35
*** markvoelker has quit IRC14:36
*** dustins has joined #openstack-cinder14:37
*** markvoelker_ has quit IRC14:38
kevincarr1991Has anyone had issues attaching volumes using horizon?14:39
*** vgridnev has quit IRC14:41
*** vgridnev has joined #openstack-cinder14:42
*** edtubill has joined #openstack-cinder14:45
*** xyang1 has joined #openstack-cinder14:47
*** daemontool_ has quit IRC14:49
*** asselin_ has joined #openstack-cinder14:49
haypohum, i got some random failures on cinder.tests.unit.test_volume.GetActiveByWindowTestCase.test_snapshot_get_active_by_window on the Python 2.7 check job :-/14:50
*** mc_nair has joined #openstack-cinder14:50
*** vlaza has quit IRC14:51
eharneyhaypo: on patchset 230510?14:52
haypoeharney: https://review.openstack.org/#/c/230510/ yes14:53
*** xyang1 has quit IRC14:53
eharneyhaypo: i've hit this too:  https://bugs.launchpad.net/cinder/+bug/150183814:53
openstackLaunchpad bug 1501838 in Cinder "Tests: lazy load operation of attribute 'snapshot_metadata' cannot proceed" [Undecided,New]14:53
haypoeharney: 8:17 AM: "gate-cinder-python27: SUCCESS",  3:35PM "gate-cinder-python27: FAILURE"14:53
eharneyseems like a bug related to snapshot objects14:53
*** haomaiwang has quit IRC14:54
*** ronis has quit IRC14:54
*** haomaiwang has joined #openstack-cinder14:55
openstackgerritAlex O'Rourke proposed openstack/cinder: 3PAR: Implement v2 replication (managed)  https://review.openstack.org/23120114:55
*** markvoelker has joined #openstack-cinder14:55
*** hemnafk is now known as hemna14:57
*** martyturner has quit IRC14:57
*** diogogmt has joined #openstack-cinder14:58
*** martyturner has joined #openstack-cinder15:00
*** haomaiwang has quit IRC15:01
*** ollie1 has joined #openstack-cinder15:01
*** haomaiwang has joined #openstack-cinder15:01
dulekthangp: ^15:01
Swansonsmcginnis: Notable RC2 fixes?  Other than my driver will actually delete a cg snapshot now, of course.15:05
openstackgerritAlex O'Rourke proposed openstack/cinder: WIP - 3PAR: Implement v2 replication (unmanaged)  https://review.openstack.org/23124515:06
*** david-ly_ is now known as david-lyle15:06
smcginnisSwanson: https://launchpad.net/cinder/liberty/liberty-rc215:06
*** aorourke_ is now known as aorourke15:06
*** diogogmt has quit IRC15:07
smcginnisSwanson: Look for the - blueprints and 5 bugs targeted header.15:07
smcginnisSwanson: Those 5, plus a few more that didn't actually have bugs linked.15:07
Swansonsmcginnis, excellent.  Looks like us driver folk were a problem.15:08
smcginnisSwanson: Just a few driver fixes that were ready and were going to end up getting backported anyway.15:08
*** diogogmt has joined #openstack-cinder15:09
thangpdulek: looks like that test was introduced by https://github.com/openstack/cinder/commit/540bba7f8e049e0eea304ec21ed2da2d0331912515:09
openstackgerritJacob Gregor proposed openstack/cinder: Cleanup/move code in Storwize Driver  https://review.openstack.org/23011715:09
*** martyturner has quit IRC15:09
thangpdulek: https://github.com/openstack/cinder/commit/200a65fa26415b9aa9f3ea7afc7fa2b8f2aee4b515:09
jgriffitheharney: you seen this one:  http://paste.openstack.org/show/475486/15:11
thangpdulek: we just need to return metadata = {}15:12
kevincarr1991Has anyone installed cinder on their controller using Kilo?15:12
eharneyjgriffith: no i haven't... interesting error though15:12
*** dustins has quit IRC15:12
jgriffithkevincarr1991: yes15:13
thangpdulek: it's trying to load snapshot_metadata, but there's none15:13
*** zhangjn has joined #openstack-cinder15:13
dulekthangp: And this Snapshot not bound to session stuff is because of that?15:14
*** zhangjn has quit IRC15:14
kevincarr1991jgriffith: did you simply follow the install guide and do everything on the controller or did you have to make any special modifications15:14
jgriffithkevincarr1991: install guide'ish15:14
* EmilienM 15:14
jgriffithkevincarr1991: I've written scripts based on install guide that I use15:14
*** geektoor_ has joined #openstack-cinder15:15
thangpdulek: usually, the session stuff is related to the context, but probably15:15
kevincarr1991griffith: ahh do you have any tips on what needs to be altered for it to work properly? Right now I have a hard time attaching vm's to instances15:15
*** zhangjn has joined #openstack-cinder15:16
dulekthangp: I'll take a look tomorrow then.15:16
*** yrabl has quit IRC15:16
hemnadiablo_rojo, ping15:16
jgriffithkevincarr1991: that's kinda broad, not sure15:16
thangpdulek: ok15:16
jgriffithkevincarr1991: have you verified you can ping your iqn's from your compute nodes?15:16
*** daneyon has joined #openstack-cinder15:16
jgriffithkevincarr1991: sounds like maybe an issue with network routing?15:16
*** daneyon has quit IRC15:17
kevincarr1991jgriffith: I honestly do not know what iqn is15:17
*** zhangjn has quit IRC15:17
dulekthangp: One more question - https://github.com/openstack/cinder/blob/master/cinder/objects/volume.py#L17915:17
dulekthangp: You're sure there should be .objects?15:17
jgriffithkevincarr1991: what type of device?  iSCSI, FC etc?15:17
diablo_rojohemna: Hello :)15:17
*** zhangjn has joined #openstack-cinder15:17
kevincarr1991jgriffith: I am going to use the default so iSCSI. I plan to use nfs eventually15:18
*** martyturner has joined #openstack-cinder15:18
jgriffithkevincarr1991: let me ask a different way :)15:18
dulekthangp: I'm trying to mimic this stuff in shapshots field in CGSnapshot and got a little doubt in this place.15:18
hemnadiablo_rojo, any idea why the XIV CI has lots of 'worlddump' files from many different dates ?15:18
*** changbl has joined #openstack-cinder15:18
jgriffithkevincarr1991: what did you deploy in the setup that's not working?15:18
*** geektoor has quit IRC15:18
hemnadiablo_rojo, re: http://dal05.objectstorage.softlayer.net/v1/AUTH_58396f85-2c60-47b9-aaf8-e03bc24a1a6f/IBM-XIV-CI/75/223375/12/check/ibm-xiv-iscsi/45423cd/logs/15:18
thangpdulek: yeah... because it a list of objects, instead of a field - VolumeList15:18
jgriffithkevincarr1991: and if it's just attach that isn't working, check your n-cpu logs15:18
hemnadiablo_rojo, looks like files since 9/1 -> 10/615:19
jgriffithkevincarr1991: look for the attach commands15:19
thangpdulek: if we used a object field, then we could just return VolumeList15:19
dulekthangp: Okay, you've convinced me. Well, that was easy, wasn't it? ;)15:19
thangp* VolumeAttachmentList15:19
kevincarr1991jgriffith: I have deployed a basic three node setup. When I installed cinder that is the only thing that is having an issue. I will check nova to see if it has any attach commands issued.15:19
diablo_rojohemna: Yeah. Jay sent out a nunch of angrygrams yesterday about the CI having issues.15:19
thangpdulek: you could do it either way15:19
diablo_rojohemna: *bunch15:20
diablo_rojohemna: We are gonna follow up today on that and make sure it's actively being dealt with.15:21
hemnaok, I was just curious as I was looking into the CI failures on your patch15:21
dulekthangp: Aww, Nova is using list objects directly like VolumeAttachmentList. Which approach is better then?15:21
*** dsariel_ has quit IRC15:21
smcginnisdiablo_rojo, jungleboyj: The IBM CIs aren't looking to happy here: http://ci-watch.tintri.com/project?project=cinder&time=48+hours15:21
diablo_rojohemna: Yeah seems lots of our CI tests are having issues right now unfortuantely. I don't think its specific to my patch.15:22
jungleboyjsmcginnis: I noticed that yesterday and sent a number of angry grams yesterday.  Following up today.15:22
*** martyturner has quit IRC15:22
smcginnisjungleboyj: Cool, thanks.15:22
* smcginnis prep's removal patches15:22
thangpdulek: might be better to use VolumeAttachmentList, but it makes you do more changes15:22
dulekthangp: What kind of changes?15:23
thangpdulek: you have to iterate through the objectlist differently i think15:23
jungleboyjsmcginnis: Sorry.  We are in the process of improving how things are done and have our own asselin lined up, but I am having to push people along to make it happen.15:24
dulekthangp: Ah, right, when *using* objects, not when writing them. Okay, that's fine, I'll stick to Nova's approach.15:24
thangpdulek: sure15:24
asselin_jungleboyj, :)15:24
smcginnisjungleboyj: It doesn't look like it's just you FWIW. Our CI statuses look pretty sad right now.15:24
dulekthangp: Thanks, that was helpful!15:25
smcginnisjungleboyj: Let us know how the asselin_ cloning goes. :)15:25
*** jdurgin1 has joined #openstack-cinder15:25
thangpdulek: np15:25
*** martyturner has joined #openstack-cinder15:25
hemnadiablo_rojo, no worries, we've had loads of CI issues as well.  CI = pain15:25
jungleboyjsmcginnis: I noticed that.  I will keep you up to date on the asselin_ cloning.  His clone will be named Isaac Beckman in Haifa.15:26
asselin_jungleboyj, smcginnis btw, there is a common-ci patch now. my clones' reviews would be appreciated! https://review.openstack.org/#/c/227584/15:26
diablo_rojohemna: Particularly when it gets ignored by the people that should care :)15:26
*** dims_ has quit IRC15:27
jungleboyjasselin_: I will pass those along.15:27
hemnadiablo_rojo, I wouldn't know anything about that15:27
hemnageguileo, ping15:27
smcginnisasselin_: Nice!15:27
geguileohemna: pong15:27
*** daneyon has joined #openstack-cinder15:27
*** zhangjn has quit IRC15:27
hemnageguileo, any interest in a google hangout this morning to chat about the locks stuffs ?15:27
*** zhangjn has joined #openstack-cinder15:28
*** zhangjn has quit IRC15:28
geguileohemna: Any possibility to do it tomorrow?15:28
hemnaI'm game for anytime.15:28
dulekgeguileo, hemna: +1 for geguileo ask.15:28
hemnaI was supposed to set this up last week and got busy15:28
geguileohemna: Thanks15:29
dulekgeguileo, hemna: I would like to be able to participate. Maybe tomorrow 16:00 UTC?15:29
hemnaUTC makes my head hurts....15:29
* hemna is a stupid American15:29
scottdadulek: Isn't that meeting time15:29
dulekhemna: It's like tomorrow in 30 minutes. But scottda's right, so maybe after the meeting?15:29
hemnathat's 9am PST, our cinder meeting time15:30
*** zhangjn has joined #openstack-cinder15:30
geguileohemna: That's what I was going to say  XD15:30
dulekhemna, geguileo: Or before the meeting?15:30
*** alexschm has quit IRC15:30
jungleboyj<russian accent>  Stupid Americans </russian accent>15:30
geguileoI have available the hour before and after the meeting15:30
*** zhangjn has quit IRC15:31
openstackgerritRajini Ram proposed openstack/cinder: Dell Eqlx: Support over subscription in thin provisioning  https://review.openstack.org/23118615:31
*** apoorvad has joined #openstack-cinder15:31
*** apoorvad has quit IRC15:32
*** apoorvad has joined #openstack-cinder15:32
*** dims_ has joined #openstack-cinder15:32
hemnaI can do it at 15:00UTC tomorrow15:33
scottdaI'd like to join as well.15:33
hemnajungleboyj, :)15:34
geguileoOk, so we all can at 15:00 UTC?15:34
hemnaI'm good to go15:34
hemnathat's 8am my time15:34
hemnaI'm usually JUST getting in to the office then15:35
geguileoOh, would you prefer after the meeting then?15:35
dulekI'm fine with that time.15:35
hemnaI'm ok with 17:00UTC as well15:35
hemnawhich is right after the cinder meeting15:35
scottdaI can do 15:00 or 17:0015:35
hemnalets do 17:0015:35
* dulek is looking for a dice with 2 walls15:35
diablo_rojoA dice with two walls...wouldn't that be a coin?15:36
geguileodulek: I usually use a dice and split the decission on odd/even  ;-)15:36
*** vgridnev has quit IRC15:36
dulekdiablo_rojo: Technically there's a little probablity of landing on the edge…15:36
*** apoorvad has quit IRC15:37
diablo_rojodulek: Fair point. I would be truly impressed if that happened.15:37
geguileoYou can always repeat the roll15:37
scottda17:00 is better for me. We could also mention it in the meeting for others who may be interested.15:38
*** krtaylor has quit IRC15:38
scottdaBut getting late in Europe15:38
hemnascottda, +115:38
dulekscottda: I think me and geguileo agreed, so it's fine.15:39
dulek(I'm not sure if DuncanT wouldn't like to attend)15:39
*** zhangjn has joined #openstack-cinder15:40
*** links has quit IRC15:42
*** sgotliv has joined #openstack-cinder15:42
*** dustins has joined #openstack-cinder15:43
*** daneyon_ has joined #openstack-cinder15:43
openstackgerritYuriy Nesenenko proposed openstack/cinder: Implement snapshots-related features for Block Device Driver  https://review.openstack.org/22229215:45
openstackgerritAlex O'Rourke proposed openstack/cinder: 3PAR: Implement v2 replication (managed)  https://review.openstack.org/23120115:46
*** daneyon has quit IRC15:46
*** jistr_ has joined #openstack-cinder15:48
openstackgerritAlex O'Rourke proposed openstack/cinder: WIP - 3PAR: Implement v2 replication (unmanaged)  https://review.openstack.org/23124515:50
*** jistr has quit IRC15:50
*** krtaylor has joined #openstack-cinder15:51
*** aix has quit IRC15:52
*** jistr_ has quit IRC15:53
*** timcl has quit IRC15:55
*** haomaiwang has quit IRC15:56
*** chris_morrell has joined #openstack-cinder15:57
*** mtanino has joined #openstack-cinder15:59
*** garthb has joined #openstack-cinder16:00
jungleboyjsmcginnis: asselin_ It looks like the following may be why our CIs are all hosed:  https://storyboard.openstack.org/#!/story/200037316:06
*** jistr_ has joined #openstack-cinder16:06
smcginnisjungleboyj: Hopefully that's it.16:07
*** e0ne has quit IRC16:08
*** dsariel_ has joined #openstack-cinder16:08
*** dsariel_ has quit IRC16:08
jungleboyjsmcginnis: Will keep you updated.  The moral of the story is that we need to get to the new 3rd Party CI code.16:08
Swansonjungleboyj, any idea when the 3rd party CI stuff will be ready for public consumption?  Not really inclined to be an early adopter on that.16:11
jungleboyjSwanson: That is a question for asselin_16:11
jungleboyjI think it is always a work in progress.16:11
*** gouthamr has quit IRC16:12
Swansonjungleboyj, yeah, but it would be nice if it would get to a set of consistent steps.16:12
scottdaWe support incremental backup in Kilo, but the stable/kilo python-cinderclient doesn't have the 'incremental' option....16:12
scottdaI've filed: https://bugs.launchpad.net/python-cinderclient/+bug/150328716:12
openstackLaunchpad bug 1503287 in python-cinderclient "stable/kilo cinder supports incremental backups, client does not" [Undecided,New]16:12
*** leeantho has joined #openstack-cinder16:12
scottdaHow to I get this backported to stable/kilo cinderclient? Submit a patch on that branch?16:12
jungleboyjSwanson: Yeah, that is part of what I am hoping our CI guy can work with asselin_ on to take care of issues like that.16:14
asselinSwanson, I'd like to get it ready for public consumtopion asap16:14
*** ollie1 has left #openstack-cinder16:14
asselinSwanson, if you'd like to try it out, please do. ping me with any issues16:14
asselinSwanson, https://review.openstack.org/#/c/200330/ https://review.openstack.org/#/c/227584/16:15
*** nkrinner has quit IRC16:16
*** thangp_ has joined #openstack-cinder16:16
*** Guest94490 is now known as mfisch16:17
scottdahemna: jungleboyj You had signed up to help with Nova bugs on this etherpad:16:17
*** geektoor_ has quit IRC16:17
*** mfisch is now known as Guest6262516:18
scottdaI've updated all the Priority:High bugs with some info. I'll be moving to the Medium next, as  I have time.16:18
*** thangp has quit IRC16:18
jungleboyjscottda: Ok, I will put it on our backlog of things to look at.16:18
*** chris_morrell has quit IRC16:19
*** cfriesen has joined #openstack-cinder16:19
*** daneyon_ has left #openstack-cinder16:19
*** links has joined #openstack-cinder16:23
*** jistr_ has quit IRC16:23
Swansonasselin, I'll take a peek16:25
asselinSwanson, thanks16:25
scottdaDisregard question on Backport. Instructions are here: https://wiki.openstack.org/wiki/StableBranch#Proposing_Fixes16:26
*** lpetrut has quit IRC16:26
*** ntpttr has quit IRC16:29
*** tsekiyama has joined #openstack-cinder16:29
*** mudassirlatif has joined #openstack-cinder16:30
*** martyturner has quit IRC16:30
*** rhedlind has joined #openstack-cinder16:32
scottdaPatch is: https://review.openstack.org/#/c/231580/16:33
*** _cjones_ has joined #openstack-cinder16:33
*** apoorvad has joined #openstack-cinder16:35
*** ntpttr has joined #openstack-cinder16:37
*** dims_ has quit IRC16:37
*** vgridnev has joined #openstack-cinder16:38
*** dims_ has joined #openstack-cinder16:41
*** dims_ has quit IRC16:41
*** anshul has quit IRC16:42
*** dims_ has joined #openstack-cinder16:42
*** jaypipes has joined #openstack-cinder16:43
*** dims__ has joined #openstack-cinder16:44
*** dims_ has quit IRC16:45
*** thingee has joined #openstack-cinder16:53
*** timcl has joined #openstack-cinder16:53
cfriesenIf I'm using LVM cinder volumes, where do I specify the "volume_clear" option in the cinder.conf file?  I have it at the top level, but it doesn't seem to take effect.16:54
*** jwcroppe_ has joined #openstack-cinder16:55
cfriesendoes it need to go under a separate subsection?16:55
*** setmason has joined #openstack-cinder16:55
smcginnisscottda: kilo is about to be security fixes only.16:56
*** jwcroppe has quit IRC16:56
jgriffithcfriesen: Make sure it's under the [DEFAULT] section header16:56
*** timcl has quit IRC16:58
*** timcl has joined #openstack-cinder16:59
*** gouthamr has joined #openstack-cinder16:59
guitarzanscottda: TIL clients have stable branches...17:00
scottdasmcginnis: kilo for cinderclient? And what do you think about getting that patch in to support incremental backup in the kilo cinderclient?17:02
*** earlephilhower has joined #openstack-cinder17:03
*** ronis has joined #openstack-cinder17:03
jgriffithscottda: we don't do release version of the client17:03
jgriffithscottda: the whole point is the clients are supposed to *always* work17:04
cfriesenjgriffith: I did,  I put it right near the top.  Running under pdb in cinder/volume/drivers/lvm.py, function _delete_volume(), self.configuration.__dict__ shows 'config_group': 'lvm'.   Does that mean anything?17:04
jgriffithcfriesen: yes it does17:04
jgriffithcfriesen: so you have a [lvm] section in your config17:04
jgriffithcfriesen: you can put it there as well17:04
scottdawell, there's a stable/kilo branch for cinderclient and it's what we ship with kilo version of cinder....17:04
jgriffithcfriesen: but that default should work for you17:04
cfriesenjgriffith: default doesn't work for sure.  will try lvm17:05
scottdaand kilo cinder has incremental backup, but cinderclient does not.17:05
scottdaShould the client support what the server does?17:05
cfriesenjgriffith: I seem to recall some discussion about this on the mailing list...that the defaults didn't propagate down into the various drivers.17:05
jgriffithscottda: yeah, but we've never done backports for client branches that I'm aware of.  That doesn't really work with pypi :)17:05
cfriesenjgriffith: arguably a bug, I think17:05
jgriffithcfriesen: correct17:05
jgriffithcfriesen: but it's odd, because this is a particular case that I've never had problems with in the past17:06
scottdaOK, thanks. I'm never quite sure how the client stable branch thing works.17:06
jgriffithcfriesen: and IIRC it's also how the gate works17:06
jgriffithscottda: yeah... so we started creating the branch just to provide a marker so we knew what was in use when we released17:06
cfriesenjgriffith: well, we have some patches on top of a snapshot of stable/kilo....it's possible we messed something up locally.  I'll see if I can test on pristine stable/kilo.17:06
jgriffithscottda: backporting to client isn't something I've ever come across, not to say it isn't done; just it would be new to me :)17:07
jgriffithscottda: also, based on backport criteria you're getting into the whole feature vs bug debate in your case :)17:07
jgriffithcfriesen: let me know if you do and it doesn't work17:07
*** martyturner has joined #openstack-cinder17:07
scottdahehe..yeah, I get that. Thanks for the info jgriffith17:08
jgriffithcfriesen: I just set it up in devstack on master here and it does still work correctly for me at least17:08
jgriffithscottda: :)  Don't know if that was helpful or not, but hopefully it was17:08
*** porrua has quit IRC17:08
*** porrua has joined #openstack-cinder17:08
hemnaasselin, https://review.openstack.org/23160217:08
asselinhemna, ack17:09
cfriesenjgriffith: putting it under [lvm] works locally, will check with devstack17:10
scottdaNo, that was helpful. It just leaves things in a weird state for something like this, where the feature is in the server, but not in the client, for a particular release (not that there is a client release, but there is that branch).17:10
scottdaA newer client will work, of course, and has the incremental backup option....17:10
scottdaBut we don't want to ship a newer client with older server, since there are other features in the client that are not supported.17:11
jgriffithscottda: yeah, we REALLY screwed the pooch on Kilo and the client if you remember17:11
jgriffithscottda: we never released a client :(17:11
jgriffithscottda: ball was dropped17:11
hemnaasselin, I think that accomplishes what I had set out to do.  it rebind's the PCI device to the hosts's original driver prior to trying to do a passthrough again.17:11
hemnahopefully that might help stabilize things17:11
hemnapatrickeast, ^^17:12
*** mriedem has quit IRC17:13
*** chris_morrell has joined #openstack-cinder17:16
*** e0ne has joined #openstack-cinder17:16
cfriesenjgriffith: I think I might see what's going on.   You alluded to it earlier in a mailing list thread. "Any option that is defined in driver.py needs to be set in the actual enabled-backend stanza IIRC.  This includes iscsi_helper, volume_clear etc."17:17
*** chris_morrell has quit IRC17:17
smcginnisscottda: Sorry, was in a meeting.17:17
smcginnisscottda: What jgriffith said. :)17:17
scottdathanks :)17:18
openstackgerritAlex O'Rourke proposed openstack/cinder: 3PAR: Implement v2 replication (managed)  https://review.openstack.org/23120117:18
*** timcl has quit IRC17:19
*** david-ly_ has joined #openstack-cinder17:21
*** david-lyle has quit IRC17:21
*** timcl has joined #openstack-cinder17:21
*** mriedem has joined #openstack-cinder17:21
*** david-ly_ is now known as david-lyle17:22
*** markstur has quit IRC17:24
*** IlyaG has joined #openstack-cinder17:24
openstackgerritAlex O'Rourke proposed openstack/cinder: WIP - 3PAR: Implement v2 replication (unmanaged)  https://review.openstack.org/23124517:25
*** harlowja has quit IRC17:25
*** raildo is now known as raildo-afk17:26
*** martyturner has quit IRC17:26
*** dims_ has joined #openstack-cinder17:29
asselinhemna, looks promising17:29
hemnathat was basically doing the same thing as the raw host commands I was going to run17:30
*** markstur has joined #openstack-cinder17:30
hemnafor whatever reason virsh wasn't showing that my PCI card was being used by a VM after I attached it17:30
hemnaso I couldn't look for it. :(17:30
hemnaso this seems to do the same thing17:31
asselinhemna, coold17:31
*** dims__ has quit IRC17:32
*** IanGovett has quit IRC17:35
*** harlowja has joined #openstack-cinder17:35
*** timcl has quit IRC17:38
*** timcl has joined #openstack-cinder17:39
*** IanGovett has joined #openstack-cinder17:39
adrianofrhemna: ping17:41
*** raildo-afk is now known as raildo17:43
hemnaadrianofr, yes17:44
adrianofrhemna: Hi. HNAS CI is back and approved my change:  https://review.openstack.org/#/c/228575/17:45
adrianofrhemna: it'd be nice if you take a look again when you have some time17:46
hemnaok coolio.  thanks!17:46
*** erlon has joined #openstack-cinder17:47
adrianofrhemna: thanks :)17:48
*** porrua has quit IRC17:49
*** porrua has joined #openstack-cinder17:51
*** dims_ has quit IRC17:53
*** dave-mccowan has quit IRC17:54
*** rushiagr_away is now known as rushiagr17:56
*** IlyaG has quit IRC17:56
*** IlyaG has joined #openstack-cinder17:57
*** IlyaG has quit IRC17:57
*** IlyaG has joined #openstack-cinder17:57
*** dims_ has joined #openstack-cinder18:01
openstackgerritNate Potter proposed openstack/cinder: Add backup_swift_auth_url to swift backup driver  https://review.openstack.org/22864618:02
*** vgridnev has quit IRC18:05
*** dims_ has quit IRC18:05
*** rushiagr is now known as rushiagr_away18:06
*** dims_ has joined #openstack-cinder18:06
*** dave-mccowan has joined #openstack-cinder18:07
*** vgridnev has joined #openstack-cinder18:07
*** chris_morrell has joined #openstack-cinder18:11
*** links has quit IRC18:13
*** jaypipes has quit IRC18:14
hemnaasselin, thanks for commenting on the CloudByte patch18:18
*** chris_morrell has quit IRC18:18
*** BharatK has quit IRC18:19
openstackgerritEric Harney proposed openstack/cinder: test_backup_swift: Set volume id per test  https://review.openstack.org/23165218:19
openstackgerritNate Potter proposed openstack/cinder: Added more options while uploading volume as image  https://review.openstack.org/23061718:23
openstackgerritEric Harney proposed openstack/cinder: Do not merge: Debugging bug 1501745  https://review.openstack.org/23112218:24
openstackbug 1501745 in Cinder "gate-cinder-python34 failed" [High,Confirmed] https://launchpad.net/bugs/1501745 - Assigned to Eric Harney (eharney)18:24
*** timcl has quit IRC18:26
*** angela-s has joined #openstack-cinder18:26
*** haigang has joined #openstack-cinder18:28
erlondavid-lyle: ping18:30
david-lyleerlon: pong18:30
*** timcl has joined #openstack-cinder18:31
erlondavid-lyle: hi david, remember a day, you and John Griftith where talking about the Horizon option "Boot from image/creates a new volume"'?18:34
openstackLaunchpad bug 1486178 in OpenStack Dashboard (Horizon) "Boot from image (creates a new volume) Doesn't allow specification of volume-type" [High,Fix released] - Assigned to David Lyle (david-lyle)18:34
david-lyleyes, I started working on that18:35
*** pv_ has joined #openstack-cinder18:35
erlondavid-lyle: hmm, it says in the BP that it is resolved but I can't see anything in gerrit18:35
pv_hey guys quick question18:35
pv_is it important to test my driver for backwards compatibility with old openstack releases?18:36
david-lyleerlon: bp?18:36
david-lyleor bug18:36
erlondavid-lyle: bug18:36
erlonpv_: yes, if you driver is in the old release18:36
pv_b/c im trying with a juno devstack and i've placed my driver files in a private juno repo and my stack.sh is erroring out giving me deprecation warnings on some oslo packages18:36
pv_but if its in the new one or aiming to be part of the new one then no?18:37
pv_how would customers use our product if they're running an older version of openstack then?18:37
erlonpv_: i.e., you are changing your driver, if is already in liberty for example, you should make sure that a user upgrading to mitaka won't have problem18:37
david-lyleerlon don't know who marked that committed, looking18:37
jgriffithpv_: no18:38
pv_oh ok, my driver is a new driver and im not sure which release we're aiming for right now but it doesn't exist yet18:38
pv_as part of openstack18:38
jgriffithpv_: the upstream model is that support starts when you introduce/merge it18:38
jgriffithpv_: we don't allow backports of drivers to previous openstack versions18:39
jgriffithpv_: however you are able to have a driver that works and is made available through other channels18:39
jgriffithSoumitra apprantly18:40
erlonpv_: so, we have the same situation here. What I do, I install a version of the system, RH Juno, for example. Place my code in site-packages/cinder/..., then, bring my code up, fixing any import issues I find. Then you can generate a RPM package18:40
*** xyang1 has joined #openstack-cinder18:40
pv_alright thanks so much18:41
pv_i relayed the info onto my superiors18:41
erlonpv_: welcome18:42
erlondavid-lyle: I have looked for something 'fixing' in horizon and nova, but, didn't found anything18:43
erlondavid-lyle: what have you done so far in the fixing?18:43
david-lyleerlon: well I started the horizon side, but I think the issue is bigger18:44
david-lyleI'm not sure how to pass on the volume-type to nova18:44
* david-lyle has to refresh mind18:44
erlondavid-lyle: yes, It is bigger indeed :), it goot me too18:45
erlondavid-lyle: you have to pass it in the API, but nova os-boot does not support that volume-type18:46
erlondavid-lyle: *got18:46
*** baumann has quit IRC18:47
erlondavid-lyle: what I did so far was to change the API in nova to accept a volume-type parameter and pass that to cinder18:47
david-lylewell the BlockDeviceMapping doesn't even seem to have a field for it18:47
david-lylewhich is how we pass on the other information18:48
*** bswartz has joined #openstack-cinder18:49
erlondavid-lyle: yes, I had to add the field18:49
david-lyleerlon: reset state18:51
david-lyleon the bug18:51
david-lyledrive-by status change18:51
*** _cjones_ has quit IRC18:51
*** _cjones_ has joined #openstack-cinder18:52
erlondavid-lyle: ??18:53
david-lyleerlon: some random person appeared to have changed the bug status with no reason18:53
david-lyleset it back to confirmed18:53
david-lyleerlon: do you have a patch proposed to nova then?18:54
david-lyleor do we require a change in workflow18:54
*** setmason has quit IRC18:55
erlondavid-lyle: hmm, Im struggling with my internet to push it18:56
erlondavid-lyle: https://review.openstack.org/23166018:57
erlondavid-lyle: to, call nova directly18:59
david-lyleerlon: I have the UI piece to populate that19:00
erlondavid-lyle: great! do you have the patch proposed?19:01
david-lyleno, because testing proved unsuccessful :-)19:01
david-lyleI'll dig it up and clean it up19:02
jungleboyjHello, this is Nova.19:02
jungleboyjhemna: Get this, my phone rang the other day, I pick it up and the person asks 'Is this OpenStack?'.  I informed them that I work on OpenStack on Cinder.  They just said they were looking for the OpenStack support line.19:03
david-lyleerlon: will have to figure out object version checking too, or just provide users with false hope19:04
david-lyle*I will19:04
erlondavid-lyle: Hmm, I have no idea how that works, I've changed the Object version just to follow what was being done there19:06
Swansonjungleboyj, I presume they meant thingee by that so you should have just coughed up his cell.19:06
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Remove XML API  https://review.openstack.org/23166319:07
jungleboyjSwanson: :-)19:07
david-lyleerlon: wouldn't be the first time we had an ineffective control in the UI19:07
e0nepatchset stats: +11, -6389 lines. I hope, tempest will pass:)19:08
erlondavid-lyle: how can that be? there's no reference to the object version in the API call to nova19:09
david-lyleerlon: I find many things aren't really meant to be used19:10
david-lylewould just be a guessing game19:10
erlondavid-lyle: in nova patchset?19:10
openstackgerritJay Bryant proposed openstack/cinder: Remove extra register_opts() calls in netapp eseries  https://review.openstack.org/23166419:11
david-lyleno, I mean object version, if I call the API how do I know version of the object the nova I'm calling supports?19:11
david-lyleI can only query if the extension is supported19:11
*** setmason has joined #openstack-cinder19:15
scottdadavid-lyle: can the nova api microversion be used to detect the newer version?19:16
david-lylescottda: I don't know that the API is changing, that's part of the problem19:17
erlondavid-lyle: I think it will call the last, it is what is happing with me, in the API request I'm not specifying anything, see the paste above19:17
*** ociuhandu has quit IRC19:18
scottdaIt looks like erlon 's patch is incrementing API to 1.1619:18
david-lyleI think that's just the object version19:19
scottdasorry, you are right.19:20
erlondavid-lyle: does the API have micro versions too?19:20
david-lyleerlon: I believe it should19:20
scottdaerlon: Yes, it does.19:20
david-lylenot sure for extensions19:20
david-lylebut I would hope so19:20
erlonscottda: so, the change should update the API microversion too19:21
scottdaIf it is a change in the Nova API, yes.19:21
scottdaBut the Nova folks can probably give you better guidance than I.19:22
*** e0ne has quit IRC19:23
erlonscottda: hmm, how the API microversions are related to the Object versions?19:23
*** dustins has quit IRC19:24
*** dustins has joined #openstack-cinder19:24
scottdaI don't know that they are. But if you are changing the Nova API to add support for a volume-type, that will require a bump in the microversion.19:24
david-lylenested, undocumented parameter19:25
scottdaBut I'm not sure you are actually changing the Nova api, are you?19:25
david-lylenot sure if it's considered an API change or not19:25
scottdaSorry, I just looked at your review and saw the commit message. But this is only if it's an actual API change.19:26
erlonscottda: well, its an additional parameter. I think is an actual change19:29
cfriesenan additional parameter would certainly require a microversion bump in nova19:29
*** e0ne has joined #openstack-cinder19:32
*** Lee1092 has quit IRC19:33
*** baumann has joined #openstack-cinder19:34
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Remove XML API  https://review.openstack.org/23166319:35
*** IlyaG has quit IRC19:35
*** baumann has quit IRC19:38
*** tpsilva has quit IRC19:40
openstackgerritEric Harney proposed openstack/cinder: test_backup_swift: Set volume id per test  https://review.openstack.org/23165219:45
openstackgerritEric Harney proposed openstack/cinder: test_backup_nfs: Set volume id per test  https://review.openstack.org/23167419:47
*** haigang has quit IRC19:48
openstackgerritEric Harney proposed openstack/cinder: Do not merge: Debugging notification issues in tests  https://review.openstack.org/23112219:49
*** tsekiyama has quit IRC19:51
erlondavid-lyle: talking to nova guys, they pointed out this change: https://review.openstack.org/#/q/status:open+project:openstack/nova+branch:master+topic:bp/enable-volume-types,n,z19:53
erlondavid-lyle: which also implements the volume-type in nova19:53
david-lyleok, cool, but the bp is not approved :(19:54
openstackgerritAlex O'Rourke proposed openstack/cinder: 3PAR: Implement v2 replication (managed)  https://review.openstack.org/23120119:55
*** ronis has quit IRC19:55
*** tsekiyama has joined #openstack-cinder19:57
openstackgerritAlex O'Rourke proposed openstack/cinder: WIP - 3PAR: Implement v2 replication (unmanaged)  https://review.openstack.org/23124519:59
erlondavid-lyle: yep20:00
*** jaypipes has joined #openstack-cinder20:00
*** porrua has quit IRC20:05
*** tsekiyama has quit IRC20:08
*** DericHorn-HP has joined #openstack-cinder20:09
*** tsekiyama has joined #openstack-cinder20:09
*** chris_morrell has joined #openstack-cinder20:10
*** merooney has quit IRC20:10
openstackgerritMichal Dulko proposed openstack/cinder: CGSnapshot Object  https://review.openstack.org/19502720:12
*** bswartz has quit IRC20:14
*** chris_morrell has quit IRC20:14
*** akerr has quit IRC20:18
*** IlyaG has joined #openstack-cinder20:23
*** zhangjn has quit IRC20:24
*** zhangjn_ has joined #openstack-cinder20:24
*** timcl has quit IRC20:29
*** IlyaG has quit IRC20:31
openstackgerritNate Potter proposed openstack/cinder: Added more options while uploading volume as image  https://review.openstack.org/23061720:36
*** jungleboyj has quit IRC20:38
*** mtaylor has joined #openstack-cinder20:39
*** wolsen has quit IRC20:39
*** wolsen has joined #openstack-cinder20:40
*** mordred has quit IRC20:42
*** mtaylor is now known as mordred20:42
*** erlon has quit IRC20:43
*** jgregor has quit IRC20:43
*** adrianofr has quit IRC20:43
*** jamielennox has quit IRC20:43
*** zigo has quit IRC20:43
*** aarefiev has quit IRC20:43
*** amit213 has quit IRC20:43
*** sayali has quit IRC20:43
*** DuncanT has quit IRC20:43
*** krotscheck has quit IRC20:43
*** serverascode has quit IRC20:43
*** vaidy_ril has quit IRC20:43
*** swamireddy has quit IRC20:43
*** mgagne has quit IRC20:43
*** Roamer` has quit IRC20:43
*** erlon has joined #openstack-cinder20:46
*** jgregor has joined #openstack-cinder20:46
*** adrianofr has joined #openstack-cinder20:46
*** jamielennox has joined #openstack-cinder20:46
*** zigo has joined #openstack-cinder20:46
*** aarefiev has joined #openstack-cinder20:46
*** amit213 has joined #openstack-cinder20:46
*** sayali has joined #openstack-cinder20:46
*** DuncanT has joined #openstack-cinder20:46
*** krotscheck has joined #openstack-cinder20:46
*** serverascode has joined #openstack-cinder20:46
*** vaidy_ril has joined #openstack-cinder20:46
*** swamireddy has joined #openstack-cinder20:46
*** mgagne has joined #openstack-cinder20:46
*** Roamer` has joined #openstack-cinder20:46
*** serverascode has quit IRC20:46
*** ntpttr has quit IRC20:47
*** DericHorn-HP has quit IRC20:49
*** DericHorn-HP has joined #openstack-cinder20:51
*** morgan has quit IRC20:52
*** morgan has joined #openstack-cinder20:52
*** serverascode has joined #openstack-cinder20:55
e0nesmcginnis, thingee, jgriffith, DuncanT: hi team! I hate to ask it, but could you please to review https://review.openstack.org/218226? I must to rebase it after almost every spec merge20:58
openstackgerritEric Harney proposed openstack/cinder: Test_backup_swift: Don't leak notifications  https://review.openstack.org/23170720:58
smcginnise0ne: No problem, I'll take a look.20:58
e0nesmcginnis: thanks! I've fixed few typos there in the last patch20:58
openstackgerritAlex O'Rourke proposed openstack/python-cinderclient: Adds v2 replication support  https://review.openstack.org/23170820:58
smcginnise0ne: Funny, I reviewed it (got the green check boxes) but didn't comment on this patch. :)20:59
e0nesmcginnis: FYI, remove xml api patch stats: +11, -6389 :)20:59
smcginnise0ne: Shoot, you're winning. I'm going to have to find more code to remove. :D20:59
e0nesmcginnis: :)21:00
eharneyi'm hoping this patch i just posted ^ will help reduce occurrence of gate failures21:00
*** DericHorn-HP has quit IRC21:01
*** setmason has quit IRC21:01
e0neeharney: great!21:01
eharneykinda have to wait and see, but i think it makes sense21:01
*** raildo is now known as raildo-afk21:09
*** setmason has joined #openstack-cinder21:09
*** edmondsw has quit IRC21:14
*** changbl has quit IRC21:14
*** e0ne has quit IRC21:24
*** gouthamr has quit IRC21:26
*** Yogi1 has quit IRC21:28
*** vgridnev has quit IRC21:29
*** ociuhandu has joined #openstack-cinder21:31
jgriffithhemna: do you know Alex's IRC nick?21:32
jgriffithalso patrickeast around?21:32
kevincarr1991I am unable to attach volumes on my instances. When I try to I get a generic error stating that nova "Failed to attach /dev/sdb" This could mean a network issue correct?21:33
hemnajgriffith, aorourke21:33
*** jgregor has quit IRC21:33
aorourkehemna, yes21:33
jgriffithaorourke: around ?  :)21:33
jgriffithhemna: aorourke I wanted to talk about the replication stuff if you guys have a few minutes?21:34
aorourkeaorourke, yes i have time21:34
jgriffithaorourke: so I'm not sure my comments were making sense in the review21:34
jgriffithaorourke: hemna also in terms of the name/label thing21:34
jgriffithaorourke: hemna why not just modify the specification to have a required param in the dict that is "target_name"21:35
hemnaworks for me21:35
hemnaor as patrickeast mentioned in his abandoned patch, 'name'21:35
hemnaI'm fine either way21:36
jgriffithhemna: haven't seen that21:36
jgriffithaorourke: I'm also not clear on why you need custom conf options for everything in the 3par driver?21:36
jgriffithaorourke: I mean.. if you do, then you do.. that's fine, but seems like there's some overlap21:36
aorourkejgriffith, is there default options for replication_mode, replication_sync_period, etc already?21:37
jgriffithaorourke: yes21:37
jgriffithaorourke: via the extra-specs setting on the volume-type21:37
aorourkejgriffith, ah, I must have missed that. i see now where you recommend using the extra specs21:37
jgriffithaorourke: but.. if you wanted to control that more gloabally via conf that's probably cool too and should be in driver.py21:37
hemnajgriffith, so I believe one of the entries in the conf for us, is how we map the local pool name to the remote pool name21:39
hemnagiving the admin the ability to specify the name of the pool on the target to put the volume during creation21:39
aorourkehemna, that is correct. cpg_map is definitely 3par specific21:39
hemnajust added flexibilty/complexity for us21:39
aorourkebut the others can be consolidated  into driver.py if other plan to use them as well21:39
aorourkejgriffith, I see in your SolidFire implementation you are also using replication_host21:40
*** thangp_ has quit IRC21:41
aorourkewell, "secondary_backend", but using it for the same thing21:41
*** ociuhandu has quit IRC21:42
aorourkeand pure has a "pure_replica_interval_default" which is the same as our "replication_sync_period"21:42
*** ociuhandu has joined #openstack-cinder21:44
openstackgerritEric Harney proposed openstack/cinder: test_backup_swift: Set volume id per test  https://review.openstack.org/23165221:45
*** mriedem has quit IRC21:45
openstackgerritEric Harney proposed openstack/cinder: Test_backup_swift: Don't leak notifications  https://review.openstack.org/23170721:45
*** harlowja has quit IRC21:46
*** IlyaG has joined #openstack-cinder21:46
openstackgerritAlex O'Rourke proposed openstack/python-cinderclient: Adds v2 replication support  https://review.openstack.org/23170821:49
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: Change os-brick to use ostestr  https://review.openstack.org/23173521:51
*** julim has quit IRC21:53
*** stevemar_ has quit IRC21:55
*** stevemar_ has joined #openstack-cinder21:56
*** harlowja has joined #openstack-cinder21:57
*** dave-mccowan has quit IRC21:57
openstackgerritEric Harney proposed openstack/cinder: test_backup_nfs: Set volume id per test  https://review.openstack.org/23167421:57
*** stevemar_ has quit IRC21:58
jgriffithaorourke: sorry... got pulled away22:00
*** dustins has quit IRC22:01
aorourkejgriffith, no worries22:01
Swansonasselin, is the project-config-example git url supposed to be a real url pointing to an example repo or and example url?22:01
jgriffithaorourke: not sure what you're looking at, yes I have the secondary_backend option, but as xyang1 points out that should be in the base driver for all to use22:02
jgriffithaorourke: I don't have the replication period thing that you mention22:02
asselinSwanson, it's supposed to be a real url point to a project-config repo. You can use the default provided (if the docs are correct)22:02
aorourkejgriffith, Pure has the sync period in their driver implementation22:02
jgriffithaorourke: if people need/use that my point was just that it should either be in the extra-specs or in the base driver config I think22:02
asselinSwanson, this is a valid project you can use: https://git.openstack.org/cgit/openstack-infra/project-config-example/22:03
asselinbut it'd be better to fork it first, and use that22:03
jgriffithaorourke: I don't see that merged?22:03
asselinSwanson, idea is that using this url will get you a working system: https://git.openstack.org/openstack-infra/project-config-example22:04
SwansonHmmm...  Getting repo not found on that.22:04
jgriffithaorourke: in other words, I'd ask Danny the same questions :)22:04
aorourkejgriffith, it's not. i was saying some of the config entries we are using for the 3par implementation are also being used in other driver implantations as well. so it would make sense to put them in driver.py for everyone to use22:05
jgriffithaorourke: I am also of the opinion that if you have defaults that you intend to be overridden by type settings, then you should probably just "hide" them internally22:05
jgriffithaorourke: ahh.. yes that :)22:06
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: Change os-brick to use ostestr  https://review.openstack.org/23173522:06
jgriffithI agree with that, barring the point I made about some of them perhaps belonging in the type setting22:06
*** stevemar_ has joined #openstack-cinder22:07
jgriffithaorourke: so my whole goal of stuffing more of the stuff into the type was so we don't have this problem of all sorts of unique conf options for every driver22:07
jgriffithaorourke: I was hoping to just have the absolute bare minimum defined in the core code22:07
jgriffithaorourke: that way it would be easier/flexible for drivers as most have different semantics22:08
aorourkejgriffith, I see. which conf entries are you opposed putting in as defaults?22:08
*** patrickeast has quit IRC22:09
*** adrianofr has quit IRC22:09
*** sweston has quit IRC22:09
*** rex_lee_ has quit IRC22:09
*** boris-42 has quit IRC22:09
openstackgerritNate Potter proposed openstack/cinder: Added more options while uploading volume as image  https://review.openstack.org/23061722:11
jgriffithaorourke: to be clear, I certainly think anything is open for discussion :)22:12
jgriffithaorourke: but, I don't feel that replication_mode, sync_period, target_name or replication_host should be in your driver22:12
jgriffithaorourke: take sync vs async for example22:13
aorourkejgriffith, the idea I was going off of was having the entire driver instance configured to one setting. anything created with 3parfcrep (in our specific case) would use the configured settings22:13
jgriffithaorourke: the agreed upon design was to use capability reporting for a driver to say what it supports22:13
jgriffithaorourke: and then for the choice to be made via the type setting22:14
Swansonasselin, cgit in that url should have been my clue.22:14
jgriffithaorourke: sure, but then there' s one in driver FOO, and BIZ, and BAZ22:14
jgriffithaorourke: and there's no naming scheme22:14
jgriffithaorourke: and we've just duplicated a bunch of conf entries for no real reason22:15
jgriffithsee what I mean?22:15
asselinSwanson, no problem...you're exposing potential issues so that's good22:15
aorourkejgriffith, in the sync vs. async example, how would you change what replication_type reports without having some sort of conf option?22:16
jgriffithaorourke: does the 3par cluster support both?22:16
aorourkejgriffith, yes, sync and async periodic mode22:16
jgriffithaorourke: I'm assumign it does thus the config option :)22:16
jgriffithaorourke: so the 3par driver should report "supported_replication_types = [sync, async]"22:17
jgriffithaorourke: and the admin chooses which are used by volume-type set replication_type=sync22:17
jgriffithor async22:17
jgriffithor one type for each22:17
jgriffithaorourke: it achieves the same thing, but allows it to be dynamic22:18
openstackgerritWalter A. Boring IV (hemna) proposed openstack/os-brick: Fix silent iSCSI login failures  https://review.openstack.org/23174022:18
jgriffithaorourke: so in my case (and others) where the same cluster can do sync or async simultaneously, we offer that flexibility still22:18
*** edtubill has quit IRC22:18
jgriffithaorourke: in the case of those that do not, then the admin just doesn't create two of them22:18
*** apoorvad has quit IRC22:18
jgriffithaorourke: does that make sense?22:18
aorourkejgriffith, couldn't you, in that case, have multiple volume types configured completely differently (one sync the other async) all using the same backend type?22:19
jgriffithaorourke: yes, exactly!22:19
*** jaypipes has quit IRC22:19
jgriffithaorourke: my point is why have it there AND in the config file?22:19
*** IlyaG has quit IRC22:20
*** IlyaG has joined #openstack-cinder22:21
jgriffithaorourke: am I maybe not considering something?22:21
*** cppforlife_ has quit IRC22:22
aorourkejgriffith, hmm ok I believe that makes sense. I will have to explore that a little more in our implementation and get back to you.22:23
jgriffithaorourke: let me see if I can get some updates pushed up tonight/tomorrow that maybe will help22:24
aorourkejgriffith, not related to this directly, but v2 replication in general, i had just filed this an hour or so ago. https://bugs.launchpad.net/cinder/+bug/150343922:25
openstackLaunchpad bug 1503439 in Cinder "Issuing v2 replication commands against a non-replicated volume type puts replication_status into an unrecoverable state" [Undecided,New]22:25
jgriffithaorourke: ahh.. good one :)22:26
jgriffithaorourke: I can fix that up, unless you already had something in mind?22:26
aorourkejgriffith, my thought where simply using a try catch around the volume_rpcapi call. right after it sets it to 'enabling'22:27
aorourkejgriffith, if it fails, set it back to what it was22:27
jgriffithaorourke: yeah, maybe... or the other thing is add an actual mechanism to check the type and make sure it's replicated before even trying22:28
jgriffithaorourke: hmm... actually, there's already a check there22:29
aorourkejgriffith, that is definitely an option. but if it does try because it is of replicated type and fails for a different reason, it can still get stuck in a bad state22:29
*** gouthamr has joined #openstack-cinder22:29
jgriffithaorourke: https://github.com/openstack/cinder/blob/master/cinder/volume/api.py#L161522:30
aorourkejgriffith, that is just to make sure you issue that command against a volume that has a replication_status of disable currently, correct?22:31
aorourkejgriffith, with a non-replicated type volume, replication_status is defaulted to 'disabled'22:32
jgriffithaorourke: if you have a volume reporting replication_status=disabled and it's not replication capabile then the bug is in your driver :)22:32
jgriffithaorourke: we need to fix that :)22:32
jgriffithaorourke: that's not the intent22:32
jgriffithaorourke: the intent was that the default was "None"22:32
aorourkejgriffith, replication_status would be 'disabled' for that volume (because it is not replicated type), but you can still issue enable_replication against it22:33
jgriffithaorourke: no, actually that's not how it was designed22:34
jgriffithaorourke: that status should never be set on a non replication capable volume22:34
jgriffithreplication=disabled implies that it is capable but has been disabled22:35
*** mc_nair has quit IRC22:35
openstackgerritNate Potter proposed openstack/cinder: Added more options while uploading volume as image  https://review.openstack.org/23061722:35
aorourkejgriffith,  so replication_status in the volume is defaulted at None?22:36
jgriffithaorourke: so that's the bug that needs fixing IMO22:36
jgriffithaorourke: no, it's a bug as you point out... just for different reason :)22:36
jgriffithso it is defaulting to "disabled" it shouldn't22:36
jgriffithit should be something else, like "not-supported" or None or Uninitialized or "something"22:37
jgriffithsomething clear22:37
aorourkejgriffith, that would solve the issue22:37
*** gouthamr_ has joined #openstack-cinder22:37
*** sgotliv has quit IRC22:38
*** erlon has quit IRC22:38
jgriffithaorourke: so part of that is from the "old" implementation ;(22:39
aorourkejgriffith, ahh that is unfortunate. is v1 staying in the code?22:40
kevincarr1991jgriffith: we spoke earlier about not being able to attach volumes to instances. I found that nova on my compute node it throwing very generic HTTP 500 errors. Do you have any ideas in where other file locations may be for errors22:40
jgriffithaorourke: afraid so :(22:40
*** IanGovett has quit IRC22:40
*** ociuhandu has quit IRC22:41
*** gouthamr has quit IRC22:41
jgriffithaorourke: unless you can help me convince IBM that it's a bad thing to have  in there :)22:41
*** setmason has quit IRC22:41
jgriffithaorourke: GRRRRRRR   https://github.com/openstack/cinder/blob/master/cinder/volume/flows/api/create_volume.py#L47422:41
jgriffithaorourke: that's kinda stupid IMO22:42
aorourkejgriffith, yeah i agree.....is having it defaulted it 'disabled' crucial to v1 implementation? any idea?22:42
jgriffithadded here:  https://github.com/openstack/cinder/commit/1c8f49bfe9fe3abd713e28922d5551f71228624c22:42
jgriffithaorourke: I'm looking at refactoring it22:43
*** stevemar_ has quit IRC22:43
jgriffithaorourke: one of the problems is that the V1 stuff is such an intertwined mix of things I'm not sure what is going to break yet :(22:43
jgriffithaorourke: of if I'm catching all the little corners it's hiding in22:43
hemnashouldn't it be None for volumes that aren't replication capable ?22:43
*** kevincar_ has joined #openstack-cinder22:43
*** stevemar_ has joined #openstack-cinder22:43
hemnamaybe that's what you guys just said22:44
aorourkehemna, it makes more sense that way22:44
hemnaheh sorry22:44
jgriffithaorourke: I actually filed a bug against this already once I think22:44
aorourkejgriffith, there is still the case where replication_enable can fail for another reason, when the volume is of replicated type, and the replication_status will get stuck as 'enabling'22:44
jgriffithaorourke: and suggested we use something like "unconfigured"22:44
jgriffithaorourke: true dat22:44
jgriffithaorourke: but I'm ok with that in most cases, because:22:44
jgriffith1. You can return a rep status update22:44
jgriffith2. if it failed for a real reason I want it to go to error22:45
openstackgerritNate Potter proposed openstack/cinder: Added more options while uploading volume as image  https://review.openstack.org/23061722:45
*** sweston has joined #openstack-cinder22:46
aorourkejgriffith, I see. so instead of raising an exception you are saying return the model update anyway as an "error" state or something alike22:46
*** stevemar_ has quit IRC22:46
*** kevincarr1991 has quit IRC22:47
jgriffithaorourke: yeah22:47
*** kevincar_ has quit IRC22:47
*** ociuhandu has joined #openstack-cinder22:48
jgriffithaorourke: I already added it: https://github.com/openstack/cinder/blob/master/cinder/volume/manager.py#L309822:49
jgriffithaorourke: but that probably needs some tweaking22:49
jgriffithaorourke: there was some disagreement from folks on what I should to there22:50
jgriffithaorourke: I had been setting the volume to error, which didn't fly :)22:50
jgriffithaorourke: I should probably update the replication_status to error22:51
*** tpeoples has joined #openstack-cinder22:51
*** angela-s has quit IRC22:52
jgriffithaorourke: sighh... that old code really makes this kind of a mess22:52
*** boris-42 has joined #openstack-cinder22:53
aorourkejgriffith, it does...there is no easy way to tell. it should be None or Unconfigured or something22:53
*** ociuhandu has quit IRC22:53
jgriffithaorourke: I'm going to go with unconfigured... sound ok?22:53
jgriffithI guess I better do "not-configured" though, else I'll get a -1 for improper grammar :(22:54
aorourkeyeah, "replication_status" -- "unconfigured" makes sense to me22:54
*** pv_ has quit IRC22:54
*** jwcroppe_ is now known as jwcroppe22:54
*** IanGovett has joined #openstack-cinder22:55
*** setmason has joined #openstack-cinder22:56
*** IlyaG has quit IRC22:58
*** subscope has quit IRC22:59
*** apoorvad has joined #openstack-cinder23:01
*** adrianofr has joined #openstack-cinder23:01
*** mudassirlatif_ has joined #openstack-cinder23:02
jgriffithaorourke: hehe  https://github.com/openstack/cinder/blob/master/cinder/volume/manager.py#L45423:02
*** eharney has quit IRC23:03
jgriffithaorourke: but here's the *good* news https://github.com/openstack/cinder/blob/master/cinder/volume/driver.py#L50223:04
*** mudassirlatif has quit IRC23:05
*** mudassirlatif_ is now known as mudassirlatif23:05
jgriffithaorourke: So, when a driver creates a volume, checks the type and it's replicated and everything is configured and "cool" then the driver should return that status23:05
jgriffithso I think that solves our problem here23:05
jgriffithtesting now, patch coming up in a moment23:06
*** chlong has quit IRC23:06
jgriffithaorourke: thnk I'll update the doc as well to make this more clear23:06
aorourkejgriffith, excellent!23:08
*** MIDENN_ has quit IRC23:09
*** ociuhandu has joined #openstack-cinder23:11
*** rex_lee_ has joined #openstack-cinder23:14
*** ociuhandu has quit IRC23:16
*** IlyaG has joined #openstack-cinder23:17
*** ociuhandu has joined #openstack-cinder23:17
*** gouthamr_ has quit IRC23:19
*** ociuhandu_ has joined #openstack-cinder23:19
openstackgerritJohn Griffith proposed openstack/cinder: Do no use disabled as default replication_status  https://review.openstack.org/23175323:20
*** ociuhandu has quit IRC23:22
*** ociuhandu_ is now known as ociuhandu23:22
*** patrickeast has joined #openstack-cinder23:29
*** uiyice has joined #openstack-cinder23:33
*** hemna is now known as hemnafk23:35
*** tsekiyama has quit IRC23:35
jgriffithd[ ๏ _ ◉ ]b23:36
*** bill_az_ has quit IRC23:37
*** jungleboyj has joined #openstack-cinder23:39
*** uiyice has quit IRC23:40
*** uiyice has joined #openstack-cinder23:40
*** cppforlife_ has joined #openstack-cinder23:40
openstackgerritJohn Griffith proposed openstack/cinder: Fix SolidFire target composition  https://review.openstack.org/23175823:41
openstackgerritJohn Griffith proposed openstack/cinder: Do not use disabled as default replication_status  https://review.openstack.org/23175323:44
*** tsekiyama has joined #openstack-cinder23:44
*** zhangjn_ has quit IRC23:45
*** stevemar_ has joined #openstack-cinder23:50
*** takedakn has joined #openstack-cinder23:50
*** dave-mccowan has joined #openstack-cinder23:51
*** zhangjn has joined #openstack-cinder23:52
*** xyang1 has quit IRC23:52
openstackgerritAlex O'Rourke proposed openstack/cinder: 3PAR: Implement v2 replication (managed)  https://review.openstack.org/23120123:54
*** tsekiyama has quit IRC23:54
*** tsekiyama has joined #openstack-cinder23:56
*** jdurgin1 has quit IRC23:56
openstackgerritAlex O'Rourke proposed openstack/cinder: WIP - 3PAR: Implement v2 replication (unmanaged)  https://review.openstack.org/23124523:57

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!