Wednesday, 2015-02-04

*** dkingshott has quit IRC00:04
*** vilobhmm has quit IRC00:05
*** vilobhmm has joined #openstack-cinder00:05
*** eharney has quit IRC00:06
*** annashen has joined #openstack-cinder00:07
*** markvoelker has quit IRC00:07
*** xyang has quit IRC00:08
*** rmesta has quit IRC00:09
*** xyang has joined #openstack-cinder00:09
*** alexpilotti has quit IRC00:10
*** bswartz has joined #openstack-cinder00:10
thingeehemna, xyang: I really don't think we can do brocade in k-3.00:13
thingeekmartin: ^00:14
hemnathingee, do they have a code patch up yet, or just the cinder-specs ?00:15
thingeeboth, but it's 5K.00:15
xyanghemna: they have one patch up, wip00:15
thingeeI want that tested well by folks00:15
hemnaugh 5k00:15
xyangthingee: there is a 3000 line cinder.conf.sample got submitted:)00:15
thingeebrocade folks aren't going to be able to tell me anything if it works.00:15
thingeealso here's k-3 currently
hemnawell besides the fact that the brocade folks are active in here or the ML00:16
*** jasondotstar has joined #openstack-cinder00:16
thingeejnrao thought my name was Mark for starters00:16
hemnagoodh 23 BPs for K300:16
xyangthingee: you got a new name?:)00:17
thingeeI think we're oversubscribed, and k-2 has been an eye opener for me00:17
thingeeI've had to punt way too much00:17
*** vilobhmm has quit IRC00:17
hemnayah we are oversubscribed at this point00:17
hemnalooking for the brocade cinder reviews00:18
thingeethis is why at the midcycle meet up I was mentioning we need to only look at reviews with bps that are approved...even ones that don't require a spec. Also we need to stop accept bps at some point00:18
xyanghemna: he was pinging me today.00:18
thingeeof course bugs too00:18
hemnaso the 5k lines patch is 3k lines of cinder.conf.sample00:18
hemnawhich goes away00:18
hemnaso it's really 2k lines of code to review00:19
xyang800 tests00:19
xyang800 line test00:19
thingeeit doesn't matter. we still all these other patches from vendors. They were submit way earlier.00:20
*** vilobhmm has joined #openstack-cinder00:20
thingeeas far as I'm concerned those get priority...and we don't have time for the brocade patches at this time00:20
*** vilobhmm has quit IRC00:20
*** vilobhmm has joined #openstack-cinder00:21
hemnabrocade has 2 cinder-specs for the same feature00:21
hemna   and
openstackgerritAnish Bhatt proposed openstack/cinder-specs: Add open-iscsi transport support to brick
anishhemna: done ^00:22
hemnaI can't -2 cinder-specs or I would right now.00:23
anishI hope you're talking about brocade :)00:24
xyanghemna: thingee fixed it:)00:25
hemnaanish, I am00:25
*** oomichi has joined #openstack-cinder00:26
*** jnrao has quit IRC00:26
anishI guess they thought a new approach needed a new spec ?00:26
hemnaIf they were actively involved here working with us on getting their stuff in, then I'd say give them an exception.00:26
hemnaI'd like to see their stuff land, but I haven't heard from them in here at all.00:27
hemnaanish, so I'd mark yours as WIP00:27
openstackgerritAnish Bhatt proposed openstack/cinder-specs: [WIP] Add open-iscsi transport support to brick
anishdone and done00:29
anishit'd be nice if the WIP tag automagically set workflow to -100:29
*** jcru has joined #openstack-cinder00:32
*** primechu_ has quit IRC00:33
openstackgerritabhiram moturi proposed openstack/cinder: Enhance unit tests for zfssa drivers
*** jcru has quit IRC00:34
*** theanalyst has quit IRC00:36
*** theanalyst has joined #openstack-cinder00:40
*** jasondotstar has quit IRC00:41
tsekiyamahemna: ok, i have added 'enforce_multipath' arg to specify it should raise or failover to single path when multipathd is not running.
tsekiyamahemna: The default in cinder is failover to single path.00:42
hemnatsekiyama, ok cool00:42
*** markvoelker has joined #openstack-cinder00:52
*** annashen has quit IRC00:53
*** david-lyle is now known as david-lyle_afk00:57
thingeejungleboyj: I hope you're not being serious about this ... this is coming off as a backdoor of what we all already agreed on.00:59
*** markvoelker has quit IRC01:00
*** primechuck has joined #openstack-cinder01:02
*** dannywilson has quit IRC01:02
*** nellysmitt has joined #openstack-cinder01:11
*** nellysmitt has quit IRC01:16
*** tsekiyam_ has joined #openstack-cinder01:16
*** Mandell has quit IRC01:16
*** markvoelker has joined #openstack-cinder01:19
*** mtanino has quit IRC01:19
*** tsekiyama has quit IRC01:20
*** tsekiyam_ has quit IRC01:21
*** jasondotstar has joined #openstack-cinder01:24
*** hemna is now known as hemnafk01:28
*** _cjones_ has quit IRC01:29
*** MentalRay has joined #openstack-cinder01:29
*** _cjones_ has joined #openstack-cinder01:29
*** jamielennox is now known as jamielennox|away01:33
*** _cjones_ has quit IRC01:34
*** Lee1092 has joined #openstack-cinder01:37
*** mriedem has quit IRC01:38
openstackgerritVilobh Meshram proposed openstack/cinder: Introducing micro_states for create workflow
*** theanalyst has quit IRC01:48
*** jasondotstar has quit IRC01:50
*** leeantho has quit IRC01:50
*** theanalyst has joined #openstack-cinder01:51
*** scottda_ has joined #openstack-cinder01:52
*** scottda_ has quit IRC02:09
*** kaisers has joined #openstack-cinder02:11
*** patrickeast has quit IRC02:11
*** rwsu is now known as rwsu-afk02:11
*** mberlin1 has joined #openstack-cinder02:11
*** kaisers1 has quit IRC02:12
*** mberlin has quit IRC02:13
openstackgerritwanghao proposed openstack/cinder: Implement function of import/export snapshots
*** haomaiwang has joined #openstack-cinder02:20
*** tsekiyama has joined #openstack-cinder02:27
*** vilobhmm has quit IRC02:29
*** tsekiyama has quit IRC02:31
*** diegows has quit IRC02:33
*** Yogi1 has joined #openstack-cinder02:39
*** rushiagr_away is now known as rushiagr02:48
*** theanalyst has quit IRC02:53
*** Yogi1 has quit IRC02:56
*** theanalyst has joined #openstack-cinder02:56
*** tellesnobrega_ has joined #openstack-cinder03:03
*** jcru has joined #openstack-cinder03:06
*** nellysmitt has joined #openstack-cinder03:12
jungleboyjthingee: Not trying to backdoor.  Just surprised we are still allowing huge updates to drivers but won't take something like that.  I am going to have them split it up to make it as small as possible and carry the protocol internally then.03:13
*** hemna_ has joined #openstack-cinder03:14
*** nellysmitt has quit IRC03:16
rhe00_thingee: would it make sense to talk about Cinder presentations in Vancouver at the meeting tomorrow? (submission deadline is February 9th). I remember Kurt bringing it up the idea of cross company presentations at the meetup last week but I don't think we came to a conclusion.03:17
*** hemna_ is now known as hemna03:17
*** rushiagr is now known as rushiagr_away03:19
thingeejungleboyj: I agree IMO driver array updates are questionable03:23
thingeexyang: ^03:23
thingeethe whole reason to stop drivers after k-1 was to allow time for other things.03:24
xyangthingee, jungleboyj: so those blueprints were approved a while back.  I am sorry that we gave everyone a headache:(03:24
xyangthingee, jungleboyj: Cinder has been allowing driver update with multiple changes in it.  we planned based on that assumption03:25
xyangthingee, jungleboyj: also our internal review process takes 2 months03:25
xyangthingee, jundleboyj: so we are working on one big patch while waiting for it to be approved.03:26
xyangthingee, jungleboyj: we'll definitely have to change this in L.  no more big driver update like this03:26
thingeereally I think we just have to learn as we go. New protocols in drivers is a new driver imo. it's exactly why people introduce a new volume driver class..because of another protocol03:26
thingeexyang: +103:27
thingeeI think we learned from this, and we just have to be explicit about what k-1 means.03:27
thingeeeven more explicit*03:27
*** harlowja is now known as harlowja_away03:28
xyangthingee: agreed.03:29
*** annashen has joined #openstack-cinder03:29
thingeebut I'm really curious to get thoughts in tomorrow's meeting about what to do with all these proposed patches that are new features in drivers. we can't do them all in k-3. We need to target what was submitted earlier and ready and just do those for now03:31
thingeecore stuff of course gets priority in this milestone.03:31
xyangthingee: do we have to cut things off so early? I mean we have allowed driver features in the 3rd milestone in the past. anyway, we can discuss tomorrow03:33
thingeethere's 22 things targeted. that doesn't cover everything in status: open project: openstack/cinder03:33
thingeexyang: we only did maybe 14 things for k-203:34
thingeexyang: if I accept anymore things in, we won't be productive and nothing will get merged because people will just be focused on separate things03:35
xyangthingee: we merged about 28 blueprints in juno, mostly in j-303:35
hemnathingee, agreed about the adding of a protocol.  It's a new driver.  In particular the IBM driver I -1'd was an FC driver that was being patched to support iSCSI03:36
hemnaand the FC driver didn't extend from the base FibreChannelDriver base class03:36
hemnaalso a bad deal03:36
xyangthingee: I remember 0 bp merged in J-103:37
hemnaI agree that we need to have a cutoff.03:37
xyangthingee: so we are in good shape:)03:37
hemnathingee,  I thought you mentioned in the ML post that the cutoff was Feb. 15th for specs/blueprints03:37
thingeecut off is the 15th, I'm just saying we should be saying at some point we have enough and we can't accept anymore.03:38
hemnathingee, +103:38
thingeeand take a look at what's proposed, what's a priority at this time03:38
anishthingee: there are  expected outputs for fake_excecute, hence the redefinition.03:38
hemnaNova's cutoff is tomorrow AFAIK03:38
anishI meant there are multiple expected outputs03:38
thingeeyou can't exactly figure that out with most of them being driver specific things03:38
hemnaanything after the 5th needs to have a core sponsor03:38
anishthingee: okay if I use mock in place instead ?03:39
thingeeanish: that's what the return_value is for03:39
thingeeanish: see my examples03:39
anishjust to make sure I'm not misunderstanding you, I'm talking about keeping fake_execute out of setUp(self) and replacing stubs.Set with mock03:41
anishor are you saying I can have multiple (and different) return values even if it is setup in setUp(self) ?03:42
jungleboyjthingee: I am going to ask the Edwin to shrink the patch down to just the iSCSI change and submit patches to fix the other stuff separately.  Not going to push that issue.03:45
*** rushiagr_away is now known as rushiagr03:45
hemnajungleboyj, are you talking about ?03:48
jungleboyjhemna: Yes.03:49
hemnaso that patch has other issues that I commented on03:49
hemnaI think it's a really bad idea to mix FC and iSCSI in the same driver03:50
hemnait's what caused me to have to create a decorator for FCZM zoning03:50
hemnawhich has to get added to all FC drivers to do zoning.03:50
hemnabesides the fact that the existing code is an FC driver that doesn't extend from FibreChannelDriver03:51
xyanghemna: I remember that. you initially was going to fix it differently but have to redo it completely because of the storwize driver03:51
*** Mandell has joined #openstack-cinder03:51
hemnaxyang, yup03:52
jungleboyjhemna:  Ok, the problem is that they wrote that based on the Storwize driver so they inherited that badness.03:52
hemnayah, so lets not make it worse IMO03:52
thingeeanish: hang on, I'll give you another example of what I think you're trying to do03:53
jungleboyjhemna: Yes dear.  ;-)03:53
*** MentalRay has quit IRC03:54
jungleboyjhemna: I will shoot them a note and work with them to make that change.03:55
hemnaok sounds good03:56
*** mitz_ has quit IRC03:56
hemnaI think jdg will eventually create an FC target object.  We'll need to look into refactoring FC drivers at that point03:56
*** mitz has joined #openstack-cinder03:56
hemnaand hopefully centralize the FCZM usage at that point to 1 place.03:56
thingeeanish: alright another comment posted
jungleboyjhemna: On a totally unrelated note ... I was amazed today that my dad was able to load the driver for an HP printer and get it working by himself today.  That was nearly a miracle!03:57
thingeethat would allow you to change the value of a method in each test case.03:57
thingeewith very little code.03:57
hemnajungleboyj, nice!   just don't tell him about the ink :P03:57
openstackgerritJay Bryant proposed openstack/cinder-specs: Fix the volume replication spec
jungleboyjhemna: I bought him the first few rounds.  NOt as bad as some of ink.03:58
anishthingee: aah, much clearer. I did not know you could overload the return value like that03:58
thingeeanish: whoops that last line in setUp() should be...04:00
lifelessthingee: .stop)04:01
lifelessthingee: not .stop())04:01
lifelessthingee: you want to call stop later, not immediately04:01
thingeelifeless: thanks :)04:01
anishlifeless: thingee thanks !04:02
* anish secretly misses C04:02
thingeeafter college, I was so happy to get back into scripting04:02
anishonce you learn the 40 or so functions, there is nothing new :P04:03
anishnot even that many I think04:03
anishI do kernel dev, so this is a nice break for me04:03
thingeeanish: nice, I just started in that space last year04:04
anishare you enjoying the complete lack of everything python is known for04:05
thingeewhen I had to deal with xml in c, I was so pissed04:06
hemnathingee, you had me at "when I had to deal with xml..."04:06
thingeeactually that was in libvirt...yeah kernel is fine actually. Took more time to wrap my head around things.04:07
anishI'm having the same problem with stuff like mock04:07
anishit takes a bit of reading, but then it's easey04:08
thingeeI have some good mentors though.04:08
*** Yogi1 has joined #openstack-cinder04:08
anishthough, nothing is as terrible as parsing iscsi text response packets in C04:08
anishso there's that04:08
lifelessyacc FTW04:09
anishin control utilities, yes04:12
anishsometimes you do need some parsing in the kernel04:12
*** annashen has quit IRC04:12
* anish is pushing to change control utilities to python04:12
thingeeanish: so after those fix ups, I'm good with the target driver patch04:13
anishok, i'll get it done tonight04:13
anisheverything but the mock tests is done04:13
thingeeanish: cool, I'll +2 it then and hopefully jdg will have a chance to look at it soon04:14
anishI wanted to check if "except exception.CinderException as TargetException" is okay04:14
*** annashen has joined #openstack-cinder04:15
thingeeanish: should be...everything else that is explicitly raised in your code inerherits from it04:15
thingeeat least in that code path04:16
anishaite !04:16
openstackgerritZhiyuan Cai proposed openstack/cinder-specs: Change-since filter for resource retrieving
anishaargh. I keep looking at scst code with is 145778, while mine is 14775804:17
anishodd coincidence04:17
thingeeoh truth is my company and them were racing to be the target subsystem in the kernel04:18
thingeeI think that was around 201204:18
anishthingee: mix of decorator & context manager is okay ? (in different functions)04:18
thingeeanish: so truth is with that, I could send you a youtube of what openstack prefers. I think ameade has an opinion on it, but I don't.04:19
anishthingee: are you working for lio or something ?04:19
thingeeanish: :)04:19
thingeeyeah my company was formerly rising tide systems04:19
thingeenow datera04:19
anishaah, we're trying to hook into lio here as well04:20
anishyou know I can see your office from my window04:21
thingeeso is scst04:21
thingeeI thought you were in NY?04:21
anishnope, right here in mountain view04:22
thingeeha alright04:22
thingeeI hate that building by the way04:22
thingeeugly thing next to this trader joes and walmart parking lot04:22
anishI'm surprised they aren't tearing that down04:23
anishfrom what I hear, walmart refused to move04:23
anishany opinions on atlantis computing ?04:24
*** Yogi1 has quit IRC04:24
anishthingee: out of curiosity, do send me that video04:25
anishI'm assuming it's the video by sdague04:26
*** julim has quit IRC04:27
anishthe first 2 min of this are "see who's reviewed your code so far" for me04:27
*** _cjones_ has joined #openstack-cinder04:30
thingeeanish: atlantis computing...not sure sorry.04:30
thingeeseries d though in 201304:30
anishthey're in your building04:31
thingeeI barely come in :)04:31
*** annashen has quit IRC04:31
thingeethey had me relocate from LA, but I'm mostly traveling anyways.04:32
*** _cjones_ has quit IRC04:32
thingeeI'll probably move back to LA after my lease though. I haven't really been able to get comfortable here =/04:32
thingeein sunnyvale though04:32
*** _cjones_ has joined #openstack-cinder04:33
thingeewhere my backyard is tech wasteland of a big abandon hp building, symatec, synopsys, linkedin, etc.04:33
thingeeoh and murphy st...where six people thought they were unique in opening an irish pub there.04:34
anishthat is not an irish pub04:34
anishthey have a neon dance floor04:34
hemnaok dinner time.....04:35
anishtheir sister establishment in MV has a neon floor04:35
anishthingee: though dish dash, can't beat taht04:35
*** _cjones_ has quit IRC04:37
thingeeeveryone talks about dish dash. still haven't been there04:38
anishyou really should04:38
anishit is the one good place in sunnyvale04:38
rushiagrjordanP: I recall netapp has 4 drivers, and they test all of them at once04:39
rushiagrwoops, that was wayy back up in my scrollback :004:40
openstackgerrityatin proposed openstack/cinder: Added more options while uploading volume to glance
*** BharatK has joined #openstack-cinder04:45
thingeea 2PB ceph cluster
thingeeapparently expansion causes down time04:51
openstackgerritChao Zheng Li proposed openstack/cinder: Add support and Unittest of consistency group for XIV/DS8K driver proxy
*** Ilja has joined #openstack-cinder04:57
*** changbl has joined #openstack-cinder04:57
*** avishay has joined #openstack-cinder04:58
rushiagrthingee: why didn't they just add two nodes at a time? Or Ceph will have perf degradation then too?04:59
thingeebeats me, not my problem anymore05:00
rushiagrthingee: heh05:00
openstackgerritChao Zheng Li proposed openstack/cinder: Add support and Unittest of consistency group for XIV/DS8K cinder driver proxy layer
*** oomichi has quit IRC05:04
*** mwichmann has quit IRC05:08
*** annashen has joined #openstack-cinder05:09
*** david-lyle_afk has quit IRC05:11
*** david-lyle_afk has joined #openstack-cinder05:11
avishaywe should have a contest to see who can replace the xiv/ds8k driver with the fewest lines of code05:12
*** nellysmitt has joined #openstack-cinder05:13
openstackgerritChao Zheng Li proposed openstack/cinder: Add support and Unittest of consistency group for XIV/DS8K cinder driver proxy
rushiagravishay: I'll also nominate vmware's and emc's drivers05:14
avishayrushiagr: look at the driver :)05:15
rushiagravishay: oh my05:16
rushiagravishay: this is in a different league altogether :)05:17
avishayrushiagr: :)05:17
*** nellysmitt has quit IRC05:17
*** harlowja_away has quit IRC05:19
*** test2222 has joined #openstack-cinder05:23
*** test2222 has quit IRC05:27
thingeeavishay: can you review ?05:28
avishaythingee: i did...05:28
thingeeah right!05:28
avishaythingee: :)05:29
avishaythingee: you probably confused me with the other avishay :)05:30
*** sgotliv has joined #openstack-cinder05:30
openstackgerritChao Zheng Li proposed openstack/cinder: Add support and Unittest of consistency group for XIV/DS8K cinder driver proxy layer
*** pradipta_away is now known as pradipta05:35
*** madhuri_ has joined #openstack-cinder05:38
*** BharatK has quit IRC05:39
*** mtreinish has quit IRC05:40
madhuri_Hi, I am getting this error while creating volumes.
madhuri_I can't figure out the issue. Can anyone please help?05:41
thingeemadhuri_: that means the cinder-volume service is not able to talk to the backend you're trying to create volumes onto05:41
thingeemadhuri_: can you restart the cinder-volume service and paste the output when it first starts up?05:42
madhuri_Is there any issue while the volume-group?05:42
thingeemadhuri_: dunno, need the start up information to know what's going on05:43
madhuri_thingee: Here
*** lpetrut has joined #openstack-cinder05:48
thingeemadhuri_: according to the error, something is bad with the rootwrap config file05:48
madhuri_thingee: Yes, after correcting it. It is working fine05:49
*** Mandell has quit IRC05:49
madhuri_But one question here.05:49
thingeemadhuri_: great!05:49
thingeesecond question you mean :)05:49
madhuri_I followed the openstack installation guide. But I din't find any rootwrap configuration.05:50
madhuri_thingee: Yes :)05:50
madhuri_So how this file was wrong.05:50
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
thingeediving into the code, it means the code that would parse the config wasn't able to. I believe that /etc/cinder/rootwrap.conf comes from installing cinder with whatever package you used by whatever distro05:51
thingeeif that file never got created, I dunno...bad package maybe? =/05:52
madhuri_thingee: Ok. Thank you :)05:52
thingeemadhuri_: you're welcome05:52
openstackgerritTina Tang proposed openstack/cinder: Pool-aware scheduler support in EMC VNX Cinder driver
*** madhuri_ has quit IRC05:53
*** nshaikh has joined #openstack-cinder05:55
*** hemna has quit IRC05:55
*** BharatK has joined #openstack-cinder05:55
*** mtreinish has joined #openstack-cinder05:56
*** coolsvap_ is now known as coolsvap05:57
*** Mandell has joined #openstack-cinder05:57
*** jcru has quit IRC05:58
*** annashen has quit IRC05:59
*** annashen has joined #openstack-cinder06:01
*** xyang has quit IRC06:02
openstackgerritChao Zheng Li proposed openstack/cinder: Add support and Unittest of consistency group for XIV/DS8K cinder driver proxy
*** abehl has joined #openstack-cinder06:06
*** markvoelker has quit IRC06:08
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Imported Translations from Transifex
*** nlevinki has joined #openstack-cinder06:13
*** nlevinki has quit IRC06:14
*** nlevinki has joined #openstack-cinder06:15
*** annashen has quit IRC06:17
*** sgotliv has quit IRC06:22
*** eharney has joined #openstack-cinder06:29
*** tellesnobrega_ has quit IRC06:32
*** tellesnobrega__ has joined #openstack-cinder06:32
*** primechuck has quit IRC06:38
*** markvoelker has joined #openstack-cinder06:39
*** sgotliv has joined #openstack-cinder06:40
*** markvoelker has quit IRC06:43
*** nshaikh has quit IRC06:50
*** nshaikh has joined #openstack-cinder06:51
*** sgotliv has quit IRC06:52
*** Mandell has quit IRC07:01
*** Mandell has joined #openstack-cinder07:04
*** chlong has joined #openstack-cinder07:07
*** TobiasE has joined #openstack-cinder07:07
*** BharatK has quit IRC07:08
*** BharatK has joined #openstack-cinder07:09
*** harlowja_at_home has joined #openstack-cinder07:10
*** nellysmitt has joined #openstack-cinder07:13
*** nellysmitt has quit IRC07:18
*** chlong has quit IRC07:28
*** eharney has quit IRC07:31
*** nkrinner has joined #openstack-cinder07:32
openstackgerritChao Zheng Li proposed openstack/cinder: Add support and Unittest of consistency group for XIV/DS8K cinder driver proxy Follow the pep8 code style to refactor code
*** Miouge has joined #openstack-cinder07:35
*** markvoelker has joined #openstack-cinder07:39
*** lpetrut has quit IRC07:42
*** nellysmitt has joined #openstack-cinder07:42
*** markvoelker has quit IRC07:44
*** ronenkat has joined #openstack-cinder07:47
*** ronenkat_ has joined #openstack-cinder07:47
*** ronenkat has quit IRC07:51
*** e0ne has joined #openstack-cinder07:52
*** dulek has joined #openstack-cinder07:57
*** harlowja_at_home has quit IRC08:02
avishaye0ne: hey! want to put this one through?
*** ronis has joined #openstack-cinder08:02
e0neavishay: hi! i'll take a look on it08:03
*** Miouge_ has joined #openstack-cinder08:09
*** Miouge has quit IRC08:11
*** Miouge_ is now known as Miouge08:11
e0neavishay: unfortunately, i'll -1 it. i'm totally disagree with raising exception in sqlalchemy/api in this way08:11
*** lpetrut has joined #openstack-cinder08:13
*** tshefi has joined #openstack-cinder08:14
avishaye0ne: yes i agree with you08:22
e0nethanks. i hope, we'll get it in K-208:25
avishaye0ne: he is responsive, should get merged soon anyway08:26
e0neavishay: thanks for review i'll fix it08:27
avishaye0ne: i'm sorry but i don't agree that this should be implemented at all08:27
avishaye0ne: i think 99% of the time the user cares only about name, status, and size08:28
*** eharney has joined #openstack-cinder08:28
avishaye0ne: for 1% that he wants to see snapshots, he can do another call instead of slowing down all other operations08:28
*** jistr has joined #openstack-cinder08:29
e0neavishay: not sure about statistics. our custemers ask it to make aotomation more clear08:29
*** sgotliv has joined #openstack-cinder08:29
e0neavishay: i'm going to do it only for API08:29
e0nefor 'volume get' action08:30
avishaye0ne: i still disagree, but if others over-ride my decision, that's ok08:30
e0neavishay: ok. lets see how community will vote on it.08:33
e0neavishay: anyway, thanks for review08:33
avishaye0ne: sure!08:36
avishaye0ne: it can be a flag, but not sure we want flags like that on volume list/get08:37
* e0ne bb in 2hours08:38
*** markvoelker has joined #openstack-cinder08:40
*** Mandell has quit IRC08:40
e0neavishay: something like 'cinder show --detailed vol' could be an option08:41
avishaye0ne: agreed08:41
*** markvoelker has quit IRC08:45
openstackgerritJeegn Chen proposed openstack/cinder: EMC VNX Cinder Driver Update
*** cinderella has joined #openstack-cinder08:50
cinderellagood morning to everyone08:50
cinderellaam having odd problems with cinder, can anybody help me?08:50
cinderellaI'm getting a "failed to copy image to volume: permission denied"08:50
cinderellawhenever I try to create a volume with an image attached08:51
*** jordanP has joined #openstack-cinder08:53
*** PaulCuzner has joined #openstack-cinder08:53
*** jpich has joined #openstack-cinder08:55
cinderellaseems like I solved, I have a new problem now09:01
*** PaulCuzner has quit IRC09:03
*** karimb has joined #openstack-cinder09:04
openstackgerritgit-harry proposed openstack/cinder: Remove use of len for testing truth
*** PaulCuzner has joined #openstack-cinder09:09
*** takedakn has joined #openstack-cinder09:12
*** e0ne has quit IRC09:12
openstackgerritChao Zheng Li proposed openstack/cinder: Add consistency group support for XIV/DS8K cinder driver proxy
*** jordanP has quit IRC09:18
cinderellaguys, now cinder list shows nothing, while in horizon I have 2 volumes created. What is happening there?09:19
rushiagrcinderella: hi09:20
rushiagrcinderella: what is 'creating a volume with image attached'?09:22
*** Miouge has quit IRC09:25
*** afazekas has joined #openstack-cinder09:25
*** jordanP has joined #openstack-cinder09:25
*** jordanP has quit IRC09:25
*** mdbooth has quit IRC09:28
*** takedakn has quit IRC09:29
openstackgerritwanghao proposed openstack/cinder: query volume detail support volume_glance_metadata
*** Miouge has joined #openstack-cinder09:32
*** PaulCuzner has left #openstack-cinder09:32
*** PaulCuzner has joined #openstack-cinder09:33
*** mdbooth has joined #openstack-cinder09:33
openstackgerritlvdongbing proposed openstack/python-cinderclient: Support cinder list filter by tenant
eharneyrushiagr: please see comments/question on
rushiagreharney: looking09:38
*** boris-42 has joined #openstack-cinder09:39
*** markvoelker has joined #openstack-cinder09:41
rushiagreharney: the two defs were unbound before (not belonging to any class). Can we not move them directly to the file, instead of making them class methods in and then making them unbound again in
rushiagri.e., just move those two unbound functions from to make them unbound methods in remotefs.py09:43
eharneyrushiagr: it would work, but i'm not sure why it would be better to have them not belonging to a class, when they are decorators only used for methods related to that class09:44
rushiagreharney: then I don't see a point in them being unbound again in file..09:45
eharneyrushiagr: they're in the RemoteFSSnapDriver class?09:46
eharneyso they aren't unbound09:46
rushiagreharney: oh, sorry, glusterfs.py09:46
eharneyrushiagr: the point there is so you don't have to repeat a ridiculously long decorator name on each function09:47
eharneywe do this fairly often i think...09:47
*** markvoelker has quit IRC09:47
*** avishay has quit IRC09:50
cinderellarushiagr: brb09:50
rushiagreharney: hmm09:50
rushiagrCan we not use those methods directly in glusterfs class as they're already part of the parent class? I'm not sure how staticmethods work in inherited class09:51
rushiagrlet me find that out09:52
*** alecv has joined #openstack-cinder09:52
erloneharney: ping09:52
eharneyrushiagr: i was just wondering the same thing actually, not sure09:52
eharneyerlon: hi09:52
erloneharney: good morning09:52
eharneygood morning09:53
erloneharney: forget to send the comment, weird, it left the review -1 but not the inline comment09:53
erloneharney: fixed that09:53
*** enterprisedc has quit IRC09:54
eharneyerlon: ok, looking now, thanks09:54
openstackgerritAnish Bhatt proposed openstack/cinder: Add support for chiscsi iscsi helper
*** e0ne has joined #openstack-cinder09:57
*** avishay has joined #openstack-cinder09:58
eharneyerlon: replied there for posterity... but basically, the system from the spec is implemented there, it just uses different (pre-existing) names for the options rather than the ones shown in the example in the spec09:58
*** jordanP has joined #openstack-cinder09:58
eharneyjust to avoid introducing more options than needed09:58
rushiagreharney: A staticmethod can simply be accessed in inherited class using a self.<staticmethodname>. I'm just not sure how the decorator syntax will be. Maybe @self.<staticmethodasdecorator>?10:00
openstackgerritTina Tang proposed openstack/cinder: Pool-aware scheduler support in EMC VNX Cinder driver
eharneyrushiagr: yeah i think i'll try that... seems like a nice way to go10:00
rushiagror maybe that 'self' is silly here and remove10:00
rushiagreharney: okay10:00
eharneyi think it'd be some super() thing since self is for instances and it's a staticmethod10:00
eharneyi dunno, i'll go learn python i guess :)10:01
eharneyself doesn't work for decorators because there is no "self" var...10:02
erloneharney: hmm, Im not familiar with the option parsing, so, when you access a variable like 'self.configuration.nas_ip', cinder will get it from a section?10:02
eharneyerlon: no, the sections are just sections from how you configure multi-backend10:02
rushiagreharney: I just did this:
eharneyerlon: if not using multi-backend they are just in default10:03
rushiagreharney: yeah, how this will work with decorators is unknown to me10:03
eharneyrushiagr: you have to bind it to a class namespace instead of an instance (which is what i was doing in the patch)10:03
*** e0ne is now known as e0ne_10:04
erloneharney: there's no section for each share, like in the specs [nfs1] nas_ip = x.x.x.x \n nas_share_path = /yyy10:04
erloneharney: the way described in the  spec gives that impression10:05
eharneyerlon: the sections are just whatever you define in enabled_backends10:05
eharneyerlon: it will work that way10:05
eharneyerlon: when using this type of config:
erloneharney: hmm, ok10:05
erlongot it10:05
erloneharney: [nfs1] is a backend name10:05
rushiagreharney: what about making them classmethods?10:05
eharneyrushiagr: i think that would just result in the same thing in this case, since they don't need class information (which is why they are static)10:07
rushiagreharney: hmm..that's right10:08
openstackgerritAnish Bhatt proposed openstack/cinder: Add support for chiscsi iscsi helper
eharneyrushiagr: so i dunno, i'm not really seeing a better way to do it10:11
*** e0ne_ is now known as e0ne10:12
openstackgerritPh. Marek proposed openstack/cinder: DRBD: Use correct function object after DBus disconnect.
rushiagreharney: so other remotefs drivers won't be able to use the two staticmethods you created10:19
eharneyrushiagr: sure they will...why not?10:20
eharneyrushiagr: the NFS snaps patch which depends on  this one does...10:20
rushiagreharney: I meant the remotefssnapdriver10:21
openstackgerritYuriy Nesenenko proposed openstack/cinder: Group test files in the right package in cinder.tests
eharneyrushiagr: remotefssnapdriver is an interface that actual drivers inherit from...10:22
cinderellarushiagr: I'm back10:22
cinderellayou there?10:22
eharneyrushiagr: if your driver is just a remotefsdriver w/o snaps then it doesn't need these10:22
rushiagreharney: I think making those decorators unbound functions is cleaner, but I think the current way works well too, as remotefssnapdriver is just the base (sort of abstract) class10:22
rushiagreharney: yes, got that10:23
eharneyrushiagr: i disagree that having methods floating outside of classes is likely to be cleaner i guess10:23
rushiagreharney: I'm actually not aware of pros vs cons of these two approaches10:26
rushiagrremoved my -110:26
eharneyit's mostly just about code organization in this case10:26
cinderellarushiagr: I'm quite in a mess with volumes. Seems like my cinder is in an inconsistent state10:26
eharneyrushiagr: ok thanks... i'll look around and see if there are some useful ideas in this area in general for how to handle this in Cinder10:27
rushiagreharney: I'll just comment on the patch saying that there is an alternate approach, to get opinions from others as well10:28
rushiagrcinderella: hey10:28
rushiagrcinderella: so what's the issue?10:28
cinderellaI simply can't attach volumes (that are correctly created) to instances.10:28
rushiagrcinderella: can you tell me more about your environment?10:29
rushiagrcinderella: also, can you see any exceptions/errors in the logs?10:29
cinderellafine rushiagr10:30
cinderellaI have my openstack installed in one single machine10:30
cinderellawith all of the components (nova, cinder, glance, etc.)10:30
cinderellaI followed this guide:
cinderellaeverything works fine, and now I was fighting with cinder10:31
cinderellaI can create volumes and no problems10:31
cinderellabut I can't attach them to machine (actually, I'm trying to attach volumes to a CentOS VMs)10:31
cinderellamachines = instances10:32
cinderellaand now, when I try to delete volumes, they go on "error_deleting" and can't remove them10:32
openstackgerritMichal Dulko proposed openstack/cinder-specs: Integrate persistence support of TaskFlow
cinderellaalso, when I try to run "cinder list", nothing shows: but I can see in horizon some volumes I created10:32
cinderellait's all clear, rushiagr? need more info?10:32
rushiagrcinderella: wow, that's weird10:32
*** Ilja has quit IRC10:32
rushiagrhow did you attach those volumes?10:33
cinderellarushiagr: via horizon. status "attaching" hangs for a while, then nothing happens10:33
cinderellawhere can I paste my logs?10:33
cinderella(sorry for my english in advance :) )10:34
rushiagrcinderella: you can use paste.openstack.org10:34
rushiagrcinderella: no issues10:34
rushiagrcinderella: can you refresh and see if volumes are present or not?10:34
rushiagrrefresh = refresh that page in horizon10:34
cinderellarushiagr: yes: they are there and are stuck in "error_deleting"10:35
cinderellarushiagr: cinder list shows absolutely nothing10:35
rushiagrcinderella: oh, they're maybe coz you tried to delete them10:35
rushiagrcinderella: maybe you're logged into horizon as a different user/tenant10:35
rushiagrcinderella: check that too10:35
cinderellait's fine10:36
cinderellaI had also this issue once10:36
cinderellaand solved via mysql workaroudn10:36
cinderellamy biggest problems is the attachment10:36
rushiagrcinderella: you were trying attachment via horizon or CLI?10:37
*** navid__ has joined #openstack-cinder10:39
cinderellarushiagr: horizon10:39
*** nshaikh has quit IRC10:40
DuncanTIf horizon shows volumes but cinder list doesn't, you /must/ be using different creds or have a serious setup problem - they both use the same rest interface10:40
cinderellarushiagr: I resolved the volume "error_deleting" problem,10:41
cinderellaDuncanT: solved it10:42
cinderellanow I'd like to understand why I can't attach volumes10:42
DuncanTWhat was your solution?10:42
cinderellaDuncanT: we went into the mysql DB10:43
*** markvoelker has joined #openstack-cinder10:43
DuncanTThat doesn't solve anything, and may have caused more problems10:43
DuncanTYou should never need to go into the database10:43
cinderellaDuncanT: I know, but cinder delete wasn't working, cinder force-delete wasn't working... not much choices10:44
DuncanTIf you delete the volumes out of the cinder db and they haven't been deleted on the backend, you now have no (easy) way of cleaning up10:44
DuncanTYou need to work through /why/ cinder delete wasn't' working - it is probably a similar reason as to why attach doesn't work10:45
openstackgerritYuriy Nesenenko proposed openstack/cinder: Group test files in the right package in cinder.tests
cinderellaDuncanT: the weird thing is that deleting and creating volumes was working like a charm and then - quite suddenly - deleting stopped working10:45
cinderellawhile attaching never worked10:45
*** jasondotstar has joined #openstack-cinder10:46
DuncanTCan you run create then delete, and post of the cinder-volume log for both operations, please?10:46
cinderellabtw you were probably right - now I created a volume via horizon and cinder-list is showing nothing10:46
*** alexpilotti has joined #openstack-cinder10:47
cinderellalog incoming10:47
*** markvoelker has quit IRC10:48
DuncanTAre you certain you are using the same user in cinder and the cli? Because it really shouldn't be possible for volumes to show in one and not the other10:48
DuncanT2015-02-04 11:45:04.150 22912 WARNING cinder.volume.manager [-] Unable to update stats, driver is uninitialized10:48
DuncanTSo are your volumes ever showing as 'available' or do they get stuck in 'creating'?10:49
*** MentalRay has joined #openstack-cinder10:50
cinderellaDuncanT: they are available10:50
DuncanTDo you only have one cinder-volume service?10:50
cinderellaDuncanT: yes, just one10:50
DuncanTThen you have something weird going on, because with the driver uninitialised, there should be no way for a volume to end up available, it should never get past the scheduler10:51
*** haomaiwang has quit IRC10:51
*** rushiagr has quit IRC10:52
cinderellaDuncanT: actually you were right, I had a user issue :)10:52
cinderellaI sourced the right creds and now I see correctly the volumes10:52
DuncanTOk, that is one issue solved.10:53
cinderella(sorry for the dumb issue)10:53
DuncanTCan you restart cinder-volume and post the logs from the restart? Hopefully that will tell us why the driver can't initialise10:53
cinderellaDuncanT: ok, they create and delete with no problems?10:53
cinderellano ?, they create and delete with no problems.10:53
cinderellaok DuncanT, gonna do that10:53
cinderellalogs incoming10:55
openstackgerritMichal Dulko proposed openstack/cinder-specs: Integrate TaskFlow's persistence support into volume
openstackgerritMichal Dulko proposed openstack/cinder-specs: Integrate TaskFlow's persistence into scheduler
*** Longgeek has joined #openstack-cinder10:55
openstackgerritMichal Dulko proposed openstack/cinder-specs: Integrate TaskFlow's persistence support into API
*** enterprisedc has joined #openstack-cinder10:55
DuncanTHopefully there are some more logs after that?10:56
cinderellaDuncanT: no - I didn't do any operation after the restart of the service10:56
DuncanTThere should still be something from driver init, and the stats periodic task every 30 sexconds10:57
*** MentalRay has quit IRC10:57
cinderellaDuncanT: I'm "catting" the log continuosly - no new lines, just "2015-02-04 11:57:42.563 26296 INFO cinder.volume.manager [-] Updating volume status" each minute10:58
cinderellaDuncanT: maybe the driver warning was caused by the inconsistent situation?10:59
DuncanTDo you have debug and verbose turned on in cider.conf? If not, please turn them on and restart10:59
cinderellaI try to create a volume and paste logs10:59
cinderellaDuncanT: I have verbose, but not debug11:00
DuncanTI generally recommend turninng on debug when you have any issues, it makes tracing things far easier11:00
cinderellaDuncanT: new logs with debug = True ->
cinderellathis is cinder-volume.log11:02
*** rushiagr has joined #openstack-cinder11:02
cinderellaDuncanT: any weird stuff?11:02
*** navid__ is now known as nshaikh11:03
DuncanTNo, that looks fine11:03
cinderellaDuncanT: great! can you help me with attaching?11:05
DuncanTSure, run create then attach and post the cinder-volume log again11:05
cinderellaDuncanT: want me to do that via CLI or horizon?11:06
DuncanTEither, it is all the same to cinder11:06
cinderellaDuncanT: ok, created VM and THEN started attaching11:08
cinderellanow, horizon is hanging with "attaching"11:08
cinderellaI'm waiting... still "attaching"11:09
cinderellaDuncanT: what logs should I paste?11:10
cinderellajust cinder-volume?11:10
DuncanTStart with cinder-vol11:10
cinderella(because I noticed in nova logs I have an error)11:10
cinderellaDuncanT: this is cinder-volume.log ->
cinderellaDuncanT: here you are nova logs where I found an error11:12
cinderellaDuncanT: this is the error I found in nova-compute.log ->
cinderella(tried to attach to /dev/htc, if this can help)11:13
DuncanTOk, so nova can't find the export11:13
cinderellaDuncanT: what should I do?11:14
DuncanTGive me a sec, I'm just trying to remember how target side iscsi commands work to get some more debug11:15
DuncanTIt has been a while since I've do anything target-side, I've forgotten far too much11:15
cinderellait's fine DuncanT11:16
cinderellaif it can help, the VMs is a CentOS-6.6-x86_64-minimal.iso11:17
cinderellamaybe I should do something "Into the VM" so that the attaching may happen?11:17
DuncanTNo, this problem is before the VM11:18
DuncanTNova is trying to find the target on the cinder node and failing11:19
DuncanTOn the node running nova-volume, can you ping ?11:20
cinderellaDuncanT: it's all in one server, all the components are running on the same machine11:20
cinderella(this is our first time with openstack - waiting budget for more machines)11:21
DuncanTOk. Can you check that the server address is please?11:21
*** e0ne is now known as e0ne_11:21
cinderellaDuncanT: yes, I'm pretty sure it is (SSH-ing it right now)11:21
DuncanTI'm guessing at a network problem of some sort, since the cinder-vol log looks good11:21
cinderellaDuncanT: what do you think about this workaround? seems similar to mine:
cinderellaand yes, I have tgt and iscsitarget running together11:23
DuncanTAh, then yes, that could well be your problem11:23
cinderellaDuncanT: what do you suggest to do?11:23
cinderellauninstall one of the two?11:23
DuncanTTry uninstalling iscsitarget and rebooting11:24
cinderellayou mean hard reboot my server? fine, gonna do that11:24
DuncanTYeah, you'll need a hard reboot after uninstalling11:24
DuncanTIf this works, please file a bug, we should at least detect the situation and print a suitable warning11:25
cinderellaok DuncanT - doing it right now11:25
cinderellaDuncanT: I never "filed a bug", maybe I'll need help in that :)11:26
DuncanTSure, go to and use the 'report a bug' button in the top right11:26
*** avishay has quit IRC11:27
cinderellaDuncanT: rebooted, now I'll try to attach11:29
cinderellaiscsitarget uninstalled11:29
cinderellaDuncanT: "attaching"...11:30
DuncanTYou uninstalled before rebooting, right?11:31
cinderellaDuncanT: naturally11:31
DuncanTGood. Just wanted to be sure :-)11:31
*** e0ne_ has quit IRC11:31
cinderellaDuncanT: have you got experience in using a NAS with cinder? because maybe this is our next step, when we get cinder working :)11:32
DuncanTI've not really played with the filesystem based cinder drivers much, other than a bit of security work a while ago. Sorry.11:33
*** Longgeek has quit IRC11:33
cinderellaDuncanT: no problems11:33
DuncanTDoes you attach work now?11:34
cinderellaDuncanT: boot is quite long, just a sec :)11:34
DuncanTNo rush11:34
cinderellaDuncanT: I'm attaching via horizon11:36
cinderella"attaching" is hanging11:36
cinderellahope it isn't taking too long...11:37
cinderellaDuncanT: error :(11:38
DuncanTAnything in the nova log?11:38
cinderellayep, gonna post11:38
DuncanTHmmm, same11:41
DuncanTCan you run 'sudo iscsiadm -m session -P3' please?11:41
cinderellaiSCSI Transport Class version 2.0-870 version 2.0-871 iscsiadm: No active sessions.11:41
*** pschaef has joined #openstack-cinder11:42
DuncanTsudo ietadm --op show --tid=111:43
*** tellesnobrega__ has quit IRC11:43
cinderella"iscsitarget" not installed11:43
cinderellayou can install it via apt-get etc etc11:43
cinderellasorry - "ietadm" not installed11:44
*** markvoelker has joined #openstack-cinder11:44
cinderella(not pasting it since it is italian)11:44
DuncanTOk, that's odd, since that is the command cinder was supposed to run. Can you post up the cinder-vol log again please?11:45
cinderellaDuncanT: here you are:
*** BharatK has quit IRC11:48
*** markvoelker has quit IRC11:49
DuncanTSo 12:35:53.611 is the initialise connection11:49
DuncanT12:37:55.820 there's a terminate connection11:50
DuncanTI can't see where the iscsi target is being set up anymore :-(11:51
*** Ilja has joined #openstack-cinder11:54
DuncanTSorry, I'm out of ideas on that one. I suspect it is something simple and I'm missing it11:55
*** diegows has joined #openstack-cinder11:57
cinderellaDuncanT: no problems, thanks anyway :)12:00
cinderellaI completely don't know where to investigate now12:00
*** cinderella is now known as cinderella_away12:00
DuncanTFinding out whether there is actually an iscsi target exported would be the first thing12:01
cinderella_awayDuncanT: you there later?12:01
DuncanTI'm completely forgotten how though12:01
*** david-lyle_afk is now known as david-lyle12:01
DuncanTI'll be here for a while, yes. Hopefully one of the US crowd will be on soon and they'll know more than me12:01
DuncanTI honestly think I'm going senile or something12:01
cinderella_awayahahah DuncanT don't worry12:03
cinderella_awayI'm 25 and can't figure this, so :D12:03
*** cinderella_away has quit IRC12:03
DuncanTI *should* be able to, my brain is just not working12:03
*** MIDENN_ has joined #openstack-cinder12:13
*** hemna_ has joined #openstack-cinder12:13
*** guitarza1 has joined #openstack-cinder12:13
*** alexpilotti has quit IRC12:13
*** Anticime1 has joined #openstack-cinder12:15
*** rodrigod` has joined #openstack-cinder12:18
*** jasondotstar has quit IRC12:20
*** afazekas has quit IRC12:20
*** mdenny has quit IRC12:20
*** Anticimex has quit IRC12:20
*** hemnafk has quit IRC12:20
*** rodrigods has quit IRC12:20
*** svasheka has quit IRC12:20
*** sbadia has quit IRC12:20
*** guitarzan has quit IRC12:20
*** rodrigod` is now known as rodrigods12:22
*** rodrigods has quit IRC12:22
*** rodrigods has joined #openstack-cinder12:22
*** svasheka has joined #openstack-cinder12:24
*** sbadia has joined #openstack-cinder12:25
*** afazekas has joined #openstack-cinder12:27
*** EmilienM|afk is now known as EmilienM12:27
*** david-lyle is now known as david-lyle_afk12:36
*** david-lyle_afk is now known as david-lyle12:36
*** Mandell has joined #openstack-cinder12:41
*** david-lyle is now known as david-lyle_afk12:44
*** david-lyle_afk is now known as david-lyle12:45
*** markvoelker has joined #openstack-cinder12:45
*** Mandell has quit IRC12:46
*** takedakn has joined #openstack-cinder12:47
*** markvoelker has quit IRC12:49
*** IanGovett has joined #openstack-cinder12:52
*** e0ne has joined #openstack-cinder12:53
*** aix has joined #openstack-cinder13:01
*** e0ne is now known as e0ne_13:04
*** e0ne_ is now known as e0ne13:05
*** bswartz has quit IRC13:05
*** Miouge has quit IRC13:09
*** pradipta has quit IRC13:13
openstackgerritYuriy Nesenenko proposed openstack/cinder: Group test files in the right package in cinder.tests
*** akerr has joined #openstack-cinder13:17
*** markvoelker has joined #openstack-cinder13:20
*** tshefi has quit IRC13:22
*** xyang has joined #openstack-cinder13:22
*** tshefi has joined #openstack-cinder13:22
*** e0ne is now known as e0ne_13:29
*** karimb has quit IRC13:29
*** nshaikh has quit IRC13:34
*** nshaikh has joined #openstack-cinder13:35
*** e0ne_ is now known as e0ne13:38
*** nshaikh has quit IRC13:41
*** karimb has joined #openstack-cinder13:43
*** bswartz has joined #openstack-cinder13:45
*** rushiagr is now known as rushiagr_away13:45
*** bswartz has quit IRC14:00
*** ronenkat has joined #openstack-cinder14:02
*** bswartz has joined #openstack-cinder14:03
*** ronenkat_ has quit IRC14:05
*** mwichmann has joined #openstack-cinder14:05
*** dulek has quit IRC14:06
*** takedakn has quit IRC14:07
*** dustins has joined #openstack-cinder14:07
*** dulek has joined #openstack-cinder14:07
openstackgerritNavneet Singh proposed openstack/cinder: Implement Cinder Manage/Unmanage APIs in NFS driver
*** kaufer has joined #openstack-cinder14:12
openstackgerritxing-yang proposed openstack/cinder: Support over subscription in thin provisioning
*** xyang has quit IRC14:16
openstackgerritMichal Dulko proposed openstack/cinder: WIP: Add persistence to scheduler workflow
openstackgerritMichal Dulko proposed openstack/cinder: Common changes for TaskFlow's persistence support
openstackgerritMichal Dulko proposed openstack/cinder: Serialize data passed to scheduler workflow
*** rushiagr_away is now known as rushiagr14:18
*** julim has joined #openstack-cinder14:19
*** Yogi1 has joined #openstack-cinder14:20
*** julim has quit IRC14:21
*** jcru has joined #openstack-cinder14:21
*** julim has joined #openstack-cinder14:24
*** erlon has quit IRC14:26
*** Mandell has joined #openstack-cinder14:28
*** jecarey has quit IRC14:28
*** chenleji has joined #openstack-cinder14:29
*** jungleboyj has quit IRC14:30
*** Longgeek has joined #openstack-cinder14:33
*** Longgeek has quit IRC14:35
openstackgerritnikeshmahalka proposed openstack/cinder: Add iSCSI SCST Target support to cinder
*** brut has joined #openstack-cinder14:38
*** Mandell has quit IRC14:41
nikesh_vedams_thingee jdb : now in new patch no db dependency and also not inheriting volume driver class in target driver class(,looking forward for further reviews in k-314:41
*** e0ne is now known as e0ne_14:44
*** mriedem has joined #openstack-cinder14:45
*** mdbooth has quit IRC14:45
*** eharney has quit IRC14:45
*** dulek has quit IRC14:46
*** dulek has joined #openstack-cinder14:47
*** e0ne_ is now known as e0ne14:47
*** sgotliv has quit IRC14:47
*** mdbooth has joined #openstack-cinder14:50
*** EduardBiceriMate has joined #openstack-cinder14:52
*** rushil has joined #openstack-cinder14:54
*** brut has quit IRC14:55
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
*** hemna has joined #openstack-cinder14:57
*** stefan_amann has joined #openstack-cinder14:57
*** Yogi11 has joined #openstack-cinder14:58
*** dustins_ has joined #openstack-cinder15:00
*** Yogi1 has quit IRC15:01
*** dustins has quit IRC15:03
*** rushil has quit IRC15:04
*** thangp has joined #openstack-cinder15:04
*** primechuck has joined #openstack-cinder15:07
*** BharatK has joined #openstack-cinder15:09
*** ronenkat_ has joined #openstack-cinder15:10
*** ronenkat has quit IRC15:14
*** EmilienM has quit IRC15:14
*** julim has quit IRC15:15
*** EmilienM has joined #openstack-cinder15:16
*** tbarron has joined #openstack-cinder15:17
*** julim has joined #openstack-cinder15:18
*** jungleboyj has joined #openstack-cinder15:21
*** mtanino has joined #openstack-cinder15:22
openstackgerritStefan Amann proposed openstack/cinder: Adjust Cinder to support FCP on System z systems
*** IanGovett has quit IRC15:23
*** IanGovett has joined #openstack-cinder15:24
*** MentalRay has joined #openstack-cinder15:24
*** Yogi11 has quit IRC15:26
*** ronenkat_ is now known as ronenkat15:26
*** rmesta has joined #openstack-cinder15:27
*** Yogi1 has joined #openstack-cinder15:27
*** jecarey has joined #openstack-cinder15:34
*** nlevinki has quit IRC15:34
*** sgotliv has joined #openstack-cinder15:35
*** EmilienM is now known as EmilienM|afk15:35
*** e0ne is now known as e0ne_15:37
*** BharatK has quit IRC15:39
*** newbism has joined #openstack-cinder15:42
*** diegows has quit IRC15:44
rushiagrCan somebody get this out of the way? (Fix up help message for reset-state call). Too many +1's15:45
*** e0ne_ is now known as e0ne15:47
*** lpetrut has quit IRC15:47
*** BharatK has joined #openstack-cinder15:51
openstackgerritMichal Dulko proposed openstack/cinder: WIP: Add persistence to scheduler workflow
*** xyang has joined #openstack-cinder15:55
*** erlon has joined #openstack-cinder15:56
thingeerushiagr: I'm assuming it hasn't been pushed forward yet because of the k-2 priorities15:56
flip214 /join #openstack-meeting15:57
*** dannywilson has joined #openstack-cinder15:57
rushiagrthingee: yeah, quite possible15:57
flip214meeting now?16:00
*** tshefi has quit IRC16:00
rushiagrmeeting time?16:02
flip214thingee: ^^16:02
*** patrickeast has joined #openstack-cinder16:04
*** ronis has quit IRC16:04
*** avishay has joined #openstack-cinder16:04
*** diemt has joined #openstack-cinder16:04
*** tsekiyama has joined #openstack-cinder16:04
*** diegows has joined #openstack-cinder16:04
*** abehl has quit IRC16:05
*** sgotliv has quit IRC16:06
*** dannywilson has quit IRC16:13
*** dannywilson has joined #openstack-cinder16:13
*** ebalduf has joined #openstack-cinder16:16
*** abehl has joined #openstack-cinder16:16
*** dustins_ has quit IRC16:17
*** EmilienM|afk is now known as EmilienM16:18
*** dustins has joined #openstack-cinder16:19
*** lpetrut has joined #openstack-cinder16:22
*** IanGovett has quit IRC16:24
*** coolsvap is now known as coolsvap_16:28
*** Anticime1 is now known as Anticimex16:29
xyangwinston-d: hi, I updated the over subscription patch. addressed your comments. can you take a look again?  thanks!16:29
*** dannywilson has quit IRC16:30
*** Mandell has joined #openstack-cinder16:30
winston-dxyang: sure16:31
*** IanGovett has joined #openstack-cinder16:33
xyangwinston-d: I'd like it to make k-2 if possible.  thanks!16:34
*** IanGovett1 has joined #openstack-cinder16:35
*** MentalRay has quit IRC16:36
*** abehl has quit IRC16:36
*** IanGovett has quit IRC16:37
openstackgerritShay Halsband proposed openstack/cinder: XtreamIO version 4.0 support
*** rwsu-afk is now known as rwsu16:40
*** lan_ has quit IRC16:40
*** diegows has quit IRC16:41
*** IanGovett has joined #openstack-cinder16:41
*** lan has joined #openstack-cinder16:42
*** primechuck has quit IRC16:43
*** IanGovett1 has quit IRC16:44
*** MentalRay has joined #openstack-cinder16:48
*** e0ne is now known as e0ne_16:54
flip214thingee: just for clarification, please: If I understand #meeting correctly, changing the DRBD driver to support the DRBD transport to Nova nodes would classify it as "new driver", and it's too late for that?16:56
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
flip214I'm waiting for jgriffiths feedback since, oh, about 3 weeks on that ...16:56
*** avishay has quit IRC16:57
*** diegows has joined #openstack-cinder16:57
*** annegent_ has joined #openstack-cinder16:58
*** ronenkat has quit IRC16:59
flip214hmmm, all we can do, is sit and wait ...17:00
*** patrickeast has quit IRC17:00
smcginniskmartin: We have a FC driver and would potentially be interested in a joint preso.17:00
rhe00_kmartin: the same for us. I also have other ideas for presentation topics.17:01
xyangkmartin: I am interested in the FC session too17:01
*** dustins has quit IRC17:01
*** rajinir_r has joined #openstack-cinder17:01
rhe00_could we set up a meeting in a meeting channel so we don't have to fill up this channel with the discussion?17:02
nikesh_vedams_is 14 feb deadline confirmed?17:02
smcginnisEven if we don't all have a lot to say, having a large group from different companies up on stage could help show that FC really is an option in cinder.17:02
smcginnisrhe00_: Just start an etherpad?17:02
kmartinsmcginnis, rhe00_ xyang i think it ok to talk in here, it's cinder related17:02
*** vilobhmm has joined #openstack-cinder17:03
kmartinI have started on an abstract for the Past, present and future of Fibre Channel in OpenStack17:03
*** atan8 has joined #openstack-cinder17:04
*** primechuck has joined #openstack-cinder17:04
smcginnisNotes and brainstorm here?
*** vilobhmm_ has joined #openstack-cinder17:04
*** nkrinner has quit IRC17:05
*** Yogi1 has quit IRC17:05
*** chenleji has quit IRC17:05
vilobhmm_are we meeting today can’t see any activity in openstack-meeting room17:06
rhe00_vilobhmm: meeting started an hour ago, already over17:07
*** vilobhmm has quit IRC17:07
*** vilobhmm_ is now known as vilobhmm17:07
*** Adriano_ has joined #openstack-cinder17:07
nikesh_vedams_thingee:in meeting you mentioned 15 feb deadline,is it regarding merging of code  in kilo?17:09
*** Yogi1 has joined #openstack-cinder17:09
duleknikesh_vedams_: Merging specs and getting BPs approved17:10
*** amann has joined #openstack-cinder17:10
dulekGuys, anyone to share some info for the mid-cycle about bug 1408763?17:11
hemnathingee, sorry I missed the meeting this morning.   I was presenting to the FC standards body at the same time.17:11
*** Adriano__ has joined #openstack-cinder17:11
*** pschaef has quit IRC17:12
hemnathingee, also FWIW, I posted on the ML last night to nova and cinder about brick and our plans.17:12
vilobhmmthingee, sorry me too missed the meeting…would like to know about the timeline for the code cutoff for kilo release17:13
nikesh_vedams_dulek: ok so for already approved bps which are still under code review and targeted to k-3,deadline would be expected 19th march17:14
duleknikesh_vedams_: That's true.17:14
nikesh_vedams_sounds great,looking forward for in k-317:15
dulekAnyone knows Anyone knows if jgriffith will be around today?17:16
*** amann has quit IRC17:18
*** stefan_amann has quit IRC17:18
*** alecv has quit IRC17:20
*** atan8 has quit IRC17:22
vilobhmmthingee : thanks17:22
thingeeflip214: how does drbd work currently?17:23
*** leeantho has joined #openstack-cinder17:25
*** jistr has quit IRC17:25
nikesh_vedams_thingee: thanks17:25
*** gtt116__ has joined #openstack-cinder17:25
nikesh_vedams_dulex: march 10 for code freeze :)17:26
*** ronis has joined #openstack-cinder17:26
*** annashen has joined #openstack-cinder17:27
duleknikesh_vedams_: I knew it was in march. ;)17:27
nikesh_vedams_ok sorry17:28
*** bswartz has quit IRC17:28
*** gtt116_ has quit IRC17:28
*** MentalRay has quit IRC17:29
*** bswartz has joined #openstack-cinder17:30
*** e0ne_ is now known as e0ne17:30
*** jpich has quit IRC17:31
thingeexyang: I'm guessing this posted now for your legal reasons :)17:31
*** BharatK has quit IRC17:33
thingeekmartin, hemna: is it possible for tempest to be useful today with brocade setup to a fc driver?17:33
kmartinthingee, is should have a CI IMO17:34
thingeekmartin: but if a fc driver is setup with brocade, would that indirectly test it still with tempest?17:35
thingeeI imagine it would, but confirming17:35
xyangthingee: I'll talk to them about breaking the manage/unmanage out.  This is not a new driver though17:36
xyangthingee: they are going to use this updated driver to continue CI testing17:37
hemnathingee, so yes, if you have CI setup to use FCZM, then tempest indirectly tests the FCZM17:37
thingeexyang: ibm is introducing another protocol as well17:37
thingeewe just discussed this17:37
xyangthingee: it has a FC and an iscsi driver17:37
kmartinthingee, we prezone the fabric for our CI currently17:37
thingeekeyword, protocol17:37
xyangthingee: that is not the transport protocol we are talking about.17:37
thingeethis protocol vX and vX17:37
thingeeit's too late17:37
xyangthingee: that is their internal protocol17:37
hemnathingee, but yes, it's possible.17:38
thingeedoesn't matter. keyword protocol. I can't make it anymore clear17:38
thingeethis is ridiculous. You didn't have a problem when I said no to IBM, but you have a problem when I say no to xtreme io17:39
xyangthingee: ok, I'll talk to them and understand more about the change.  I'll ask them to fix the CI issue in the existing driver first17:39
xyangthingee: not ridiculous17:39
xyangthingee: I think that is the XML parsing protocol, but because it is confusing, I'll find out17:40
xyangthingee: I'll ask them to fix the CI first17:40
thingeexyang: I wouldn't argue about this honestly. I have half the mind to say no your other 5K plus patches for driver updates. You're already getting enough review resources as it is.17:41
thingeeI don't care if they are mostly doc strings17:41
xyangthingee: ok17:41
kmartinthingee, Is this a bad time to have my two driver BPs added to K3? and
* jungleboyj pictures kmartin spraying gasoline on a roaring fire17:44
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
thingeeI'm not concerned about drivers introducing new features. My mind is just blown about this because we just talked about this and no one raised objections.17:45
*** p01s0n has joined #openstack-cinder17:45
kmartinthingee, I quickly added these at the meetup, it was not clear that they were required...the 1st already has code posted(sorry) the second is started17:45
winston-dxyang: i'm afraid you need another version of overprovision. i have two more nits for you to handle.17:46
*** dulek has quit IRC17:47
*** dulek has joined #openstack-cinder17:47
thingeekmartin: done17:49
e0neAbel Lopez: are you around?17:49
thingeekmartin: were you able to see the proposed for kilo in series goal?17:50
thingeekmartin: I might be able to use that for people to get my attention17:50
kmartinthingee, yes17:50
kmartinis that new?17:51
thingeekmartin: ok they're targeted17:51
*** _cjones_ has joined #openstack-cinder17:51
thingeekmartin: I don't think so...just not sure anymore what I'm able to do versus other people in launch pad...was just confirming17:51
kmartinthingee, thanks...were you notified when I set that17:52
p01s0ni am facing some issue with attaching volume to instance with emc vnx as backend ,its showing this error "Unable to fetch connection information from backend: Bad or unexpected response from the storage volume backend API: Failed to add volume-0c4be426-39e3-4321-aa54-14e30ec7c7d5 into SG"17:52
xyangwinston-d: sure, I'll change that.17:52
*** annegent_ has quit IRC17:53
winston-dxyang: thx, i'll review that first thing tomorrow. need to get some sleep now.17:53
*** winston-d is now known as winston-d_zZZ17:53
xyangp01s0n: hi, please send me logs and I'll ask the VNX team to take a look.  thanks17:53
xyangwinston-d: sure.  good night.  Thanks for help!17:54
*** coolsvap_ is now known as coolsvap17:57
rhe00_patrickeast: ping17:58
p01s0nxyang: cinder-volume log , configuration :
p01s0nxyang: can you have a look17:58
xyangp01s0n: It's not clear to me what is the problem. what version is this?18:00
thingeeHow to contribute a driver has been updated
p01s0nxyang : icehouse18:01
thingeebswartz: is anyone at netapp able to pick this up...looks  like it's failing one unit test and I'm sure Navneet is gone for the day
bswartzthingee: I'll round someone up18:02
bswartzthingee: you can also shout about this stuff in #openstack-netapp -- most of our team hangs out there when they're online18:03
thingeeotherwise no worries, I can bump it to k-3, but it's close to merge18:03
thingeegood to know, thanks18:04
xyangp01s0n: ok, I'll ask someone to take a look and get back to you.  Did you send me an email or is this a new problem?18:04
p01s0nxyang : i havent sent any mail,i will send a detailed mail18:04
xyangp01s0n: sure. you know my email, right18:05
p01s0nxyang :yes18:05
*** vilobhmm has quit IRC18:05
xyangp01s0n: ok18:05
*** aix has quit IRC18:05
*** karimb has quit IRC18:06
thingeekmartin: I get emails about all blueprint activity. It's not useful18:06
thingeee0ne: is there going to be a problem with this just doing a raise in a couple of places instead of giving an actually exception to raise?
thingeee0ne: line 3302 and 333418:08
e0nethingee: not sure that i've got your point. in just re-raise catched exception18:08
rhe00_thingee: what's the process to update would be nice if it referenced Kilo instead of Havana and Icehouse.18:08
*** rushil has joined #openstack-cinder18:09
*** dustins has joined #openstack-cinder18:09
thingeerhe00_: that's the doc directory in the cinder tree18:09
e0nethingee: do you think that we need to raise more specific excepitons?18:09
flip214thingee: the currently merged DRBD driver uses jgriffiths iscsi target driver.18:09
thingeerhe00_: so same as code18:10
jungleboyjthingee: I have one blueprint that we have code up for but the BP hasn't been approved yet.
rhe00_thingee: ok, got it.18:10
jungleboyjthingee: Would appreciate feedback on that one.18:10
thingeee0ne: my point is it's just doing 'raise' instead 'raise SomeException'18:10
flip214using the drbd protocol up to the nova nodes should be a) much faster b) more resilient against network interruptions.18:10
*** ebalduf has quit IRC18:10
thingeee0ne: I'm asking if that's going to be a problem18:10
thingeeflip214: might be a problem then yes.18:10
*** harlowja has joined #openstack-cinder18:11
e0nethingee: what kind of problem are you talking about?18:11
*** jasondotstar has joined #openstack-cinder18:11
thingeeflip214: are you going to be able to provide a ci for iscsi and your own connector?18:11
thingeeby march 19th?18:11
thingeealso is the nova part going to land?18:11
*** dkingshott has joined #openstack-cinder18:11
flip214thingee: we're planning to. sadly the available hardware isn't good enough, and the replacement has not yet arrived.18:11
flip214it should be here in the next 2 weeks, though ...18:12
flip214and I hope that I can transfer the CI that I had up to now over...18:12
thingeeflip214: a little advice, I would worry about making sure what you already got in tree stays in tree.18:12
thingeenot some new connector. I know you got pressure from your company about it, but I would convey that to them18:13
*** jasondotstar has quit IRC18:14
thingeejungleboyj: is the nova side going to happen in K?18:15
thingeeah looks like some stuff has been merged on that side18:15
thingeeand you're doing it in small patches for different features, nice18:15
jungleboyjIt is approved and under review.18:16
jungleboyjMet with these guys in Paris.  They seem to get the community.  I think.18:16
thingeejungleboyj: so I do have an outstanding question on this18:17
thingeeit's in the comments18:17
e0nethingee: it's in a API, maybe raising InvalidParameterValue will be better18:18
thingeee0ne: I agree18:18
e0neok, i'll add it to a review18:18
jungleboyjhemna: ^^18:18
thingeehemna, jungleboyj: this might just be a problem with brick though. but I really don't want a bunch of if system this, do this all over the place18:19
*** dannywilson has joined #openstack-cinder18:20
*** coolsvap is now known as coolsvap_18:20
jungleboyjTalking about
thingeeanish: comments posted for,publish18:21
jungleboyjhemna: thingee has questions for you in the spec review.18:21
thingeehemna: this touches brick and is introducing a problem imo.18:22
hemnayah the conditionals are a bit of a PITA18:22
e0nethingee, anish: i've added comments to i hope, next patch will be a final18:22
hemnaI'm not sure how else he'd do it at the moment18:22
jungleboyjhemna: I had the same thought and same reaction.18:23
hemnaif this is for volume discovery18:23
thingeeanish, bswartz, xyang: I want to do a cut of k-2 by feb 5th 2:00 UTC18:23
*** patrickeast has joined #openstack-cinder18:24
hemnahrmm yah I dunno18:24
anishthingee: so on the first comment, I actually mention this in the reply18:24
anishI _absolutely_ could not figure out how to mix configuration.safe_get & self.fake_volumes_dir18:25
anishany pointers ?18:26
jungleboyjhemna: thingee So, given this discussion, can we target the BP and finish working the details through the review?18:26
jungleboyjI guess Spec review.18:26
thingeejungleboyj: IMO figure this is the point of the spec, no?18:26
thingeeyou're more than welcome to work on the code, which has already why not discuss this now?18:27
jungleboyjthingee: Agreed.18:28
* e0ne need to go offline. bb in ~2hours18:28
jungleboyjOh wait, this is approved and targetted.18:28
jungleboyjI swore it wasn't when I looked yesterday.18:29
openstackgerritPatrick East proposed openstack/cinder: Make PureISCSIDriver iSCSI port discovery more flexible
thingeeanish: not sure if that helps18:29
thingeeanish: as an example18:29
thingeejungleboyj: no you're right I just targeted it, but I forgot about this issue in the spec18:30
thingeejungleboyj: which is why I want it discussed18:30
hemnayah looks like it's targeted for k3 already18:30
jungleboyjthingee: Gotcha.18:30
anishthingee: looking, thanks18:31
jungleboyjhemna: How else can we do this without checking out the platform?18:32
hemnaI don't know really18:32
hemnathe paths are very different for that platform vs x8618:32
*** p01s0n has quit IRC18:32
rhe00_patrickeast: any chance you can take a moment and point me to your nodepool yaml. Would be super to have a working real life example while trying to figure this out.18:32
jungleboyjThe bigger question is how big an issue is this going to be?  Are we really going to have a bunch of different platforms doing 'if' in brick?18:33
thingeejungleboyj: I don't know. that's the problem18:34
thingeebut it's definitely making the main connector code in brick more complicated18:34
thingeejungleboyj, hemna:
thingeeif hemna doesn't care, then ok with me.18:37
thingeeI don't have an answer for this.18:37
hemnaugh yah that code is a pita18:37
hemnathe alternative is to not even check for the system18:37
hemnaso that loop is all about creating a possible list of entries that might show up18:38
*** rushiagr is now known as rushiagr_away18:38
hemnathe search looks until it finds the first one.18:39
hemnaso we could just remove the ifs and create s390x possible entries as well.18:39
hemnanot idea either18:39
thingeecan I get some cores on these quick ones
thingeedriver removals ^18:39
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
*** e0ne is now known as e0ne_18:41
*** jordanP has quit IRC18:41
*** MentalRay has joined #openstack-cinder18:41
thingeehemna: ok well I already have a +1 from you on the spec.18:42
*** p01s0n has joined #openstack-cinder18:42
thingeehemna: let me know if you're fine with it as is and we can continue....but honestly I don't know how many more of these we'll see in brick18:42
hemnaso, if we start to see more of these, then a refactor will be needed IMHO18:43
hemnato make it more manageable18:43
*** alop has joined #openstack-cinder18:43
p01s0nxyang: i think the issue is with vnx itself,each time i need to map LUN to storage group after that attach to instance is working18:43
thingeehemna: not disagreeing with you, but if we look on the target side, it was a pita for some time. Thank goodness jdg refactored things.18:44
thingeehemna: what we found was things just didn't work unless you were tgt or lio.18:44
thingeeiet, iser didn't work, etc18:44
alope0ne_: Thanks for your feedback,18:44
*** e0ne_ is now known as e0ne18:45
hemnaso the other side is18:45
hemnawhere is the CI to test this ?18:45
thingeealop: I'm fine with it after that...assuming you're the person working on db purge18:45
thingeehemna: haha18:45
*** e0ne is now known as e0ne_18:45
alopthingee: I'm that happy bald guy who keeps talking to you like he knows you :)18:45
thingeehemna, jungleboyj: you know I was going to bring that up, but honestly, I think a cert test now is we did with target drivers. CI requirements need to be updated in L IMO18:46
thingeealop: gotta be more specific :P18:46
*** e0ne_ is now known as e0ne18:46
*** e0ne is now known as e0ne_18:47
hemnayah sounds good18:47
*** diemt has quit IRC18:47
thingeejungleboyj: is that fair?18:47
*** e0ne_ is now known as e0ne18:48
*** e0ne has quit IRC18:48
patrickeastrhe00_:  sry for the delay, just got in and had a scrum meeting… here is the nodepool.yaml i am currently running
rhe00_patrickeast: thanks!18:51
patrickeastrhe00_: let me know if you have questions about the fields, happy to explain what its doing18:51
*** afazekas has quit IRC18:51
tbarronthingee: i'm gonna look at and see if I can figure why that unit test is failing18:53
tbarronthingee: off to a meeting in 7 min, but just wanted to let you know we're not ignoring the issue18:53
*** annegent_ has joined #openstack-cinder18:54
rhe00_patrickeast: this will help fill in some of the missing pieces in my understanding of nodepool configuration. thanks again!18:54
thingeetbarron: thanks18:54
tbarronthingee: and appreciate the opportunity to correct it before K-2 cut18:54
alopthingee: looks like Ivan just went offline, did I finally get the exception piece right?18:58
*** annegent_ has quit IRC18:59
*** Yogi11 has joined #openstack-cinder18:59
*** annegent_ has joined #openstack-cinder19:00
*** bswartz has quit IRC19:01
*** dustins_ has joined #openstack-cinder19:01
*** Yogi1 has quit IRC19:01
*** bswartz has joined #openstack-cinder19:03
*** dustins has quit IRC19:03
nikesh_vedams_jdg: there?19:04
*** rushil has quit IRC19:04
*** vilobhmm has joined #openstack-cinder19:05
*** ks_ne has quit IRC19:06
jungleboyjthingee: So, the answer is that we are ok with the if's for now.  Need to readdress if we have a lot of platforms show up.19:07
jungleboyjthingee: Need to get a cert test result.  Which I think is fair.19:07
*** bill_az_ has joined #openstack-cinder19:07
xyangp01s0n: so you get it resolved? anyway, if you still have issues, shoot me an email with logs and we'll take a look19:09
xyangthingee: ok, I'm updating the over subscription patch now.  will see if winston-d can take a look before the cut off time.  need another core too.  Is jdg around?19:10
*** ebalduf has joined #openstack-cinder19:11
*** sgotliv has joined #openstack-cinder19:15
*** ebalduf has quit IRC19:15
*** Yogi11 has quit IRC19:20
anishthingee: ok, last stubs is gone too19:21
*** e0ne has joined #openstack-cinder19:24
anishregarding your other comment, that test is just checking the format is correct19:25
anishif I understand your comment correctly, you want me to mock the read here, is that correct ?19:26
nikesh_vedams_going to bed :) will wait for futher reviews ""19:29
nikesh_vedams_also tomorrow try to upload logs too19:30
*** asmith_ has joined #openstack-cinder19:37
*** annashen has quit IRC19:37
Swansonthingee: do we need a blueprint to add to our unit tests?  We have an additional module we want to add to our "suite" of tests.19:37
*** annashen has joined #openstack-cinder19:38
*** diegows has quit IRC19:41
thingeejungleboyj: yup!19:41
thingeexyang: I don't know where jdg is =/... but yeah if you can round up another core on it. I'm kind of confused with it atm.19:42
thingeexyang: I'm assuming there is a spec update waiting for me on it?19:42
jungleboyjthingee: Works for me.  I will shoot a note to those guys about how to proceed.19:42
thingeeanish: if possible yes.19:43
thingeejungleboyj: ok, I'll push the spec through. thanks19:43
anishthingee: well it's just a test of grep then19:43
anishcan i just remove it ?19:43
jungleboyjthingee: Thank you and hemna for talking this through.19:44
thingeeSwanson: yes's just for tracking purposes. let me know the url and I'll target it19:44
xyangthingee: I'm talking about the over subscription patch.  are we talking about the same thing?19:44
thingeexyang: yes19:44
thingeexyang: at the midcycle meetup, we agreed on spec updates for any changes with the code19:45
xyangthingee: oh, that.  yes, but I can't update the spec when the code is still changing:)19:45
thingeexyang: at this point with winston-d_zZZ's comments, I'm sure the spec is outdated. I once understood the proposal, but have been lost since the midcycle meetup19:45
xyangthingee: the design keeps changing19:45
xyangthingee: I can't update the spec until the code is merged19:46
thingeexyang: haha ok understandable. I'm just lost at this point19:46
thingeexyang: so it makes reviewing things hard19:46
e0nealop: thanks for contibution19:46
xyangthingee: yes, winston-d made some new suggestions, so I changed code accordingly19:46
xyangthingee: we are getting close.19:47
xyangthingee: I'll need to log off now to submit code now.  will be back19:47
thingeexyang: so I would recommend, if there are suggestions to changes, don't change the code first. change the spec first, get agreement, and then change the code.19:47
thingeeyou can change the code if you want in parallel to make sure the suggestions would actually work, but if we don't understand what we're reviewing then we can't review it19:48
*** xyang has quit IRC19:48
*** xyang has joined #openstack-cinder19:48
openstackgerritxing-yang proposed openstack/cinder: Support over subscription in thin provisioning
e0nealop:, line #3300: not sure that it work correct:(19:49
aloplet me test19:49
*** harlowja is now known as harlowja_away19:50
alope0ne: I'm just going to drop the interpolation then19:50
*** dustins_ has quit IRC19:51
*** ebalduf has joined #openstack-cinder19:52
*** DaveChen has quit IRC19:52
e0nealop: something like shoul works19:52
*** dustins has joined #openstack-cinder19:52
*** bswartz has quit IRC19:53
alope0ne: ah, even better19:53
*** EmilienM is now known as EmilienM|afk19:53
vilobhmme0ne : ping19:55
*** xyang has quit IRC19:55
e0nevilobhmm: hi19:55
*** xyang has joined #openstack-cinder19:55
vilobhmmnot sure why the tempest tests are failing
vilobhmmI can’t see anything specific to the changes I have made here19:55
xyangthingee: just submitted another patch to address winston-d's comments. I'll update the spec too.19:56
e0nevilobhmm: is it work on your env when you do it manually?19:57
*** dkingshott has quit IRC19:58
vilobhmmyes i dont see any of these19:58
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
*** dulek has quit IRC19:58
*** annegent_ has quit IRC19:58
*** diegows has joined #openstack-cinder19:58
e0nevilobhmm: may be it'a an infra issue. try recheck19:59
*** ronenkat has joined #openstack-cinder20:00
anishthingee: can I just drop than one test case then ? it's testing open &  a re.seach ? everything else is done20:01
patrickeastvilobhmm: in the volume service logs there are errors
*** annashen has quit IRC20:02
vilobhmmpatrickeast : i see but when i run run_test in my virtual enn they dont seem to be highlighted..thanks for catching this one20:03
*** annegent_ has joined #openstack-cinder20:04
e0nevilobhmm: run_tests runs only unit-tests20:04
*** Ilja has quit IRC20:06
*** dkingshott has joined #openstack-cinder20:06
openstackgerritAbel Lopez proposed openstack/cinder: Purge deleted rows
thingeeSwanson: done...I think if I remember correctly, the code is in review?20:10
*** rhe00_ has quit IRC20:13
*** xyang has quit IRC20:14
*** xyang1 has quit IRC20:15
*** newbism has quit IRC20:15
Swansonthingee: I think smcginnis put something in review.  This is separate from that.  I need to get my ducks in a row on this before I push this up.  Specifically we want more test time on it.20:17
*** Yogi11 has joined #openstack-cinder20:19
jungleboyjDo we still have documentation on how to just run the certification test anywhere?20:19
*** dustins has quit IRC20:20
*** bswartz has joined #openstack-cinder20:20
*** dustins has joined #openstack-cinder20:23
*** Mandell has quit IRC20:24
openstackgerritAnish Bhatt proposed openstack/cinder: Add support for chiscsi iscsi helper
openstackgerritScott DAngelo proposed openstack/python-cinderclient: reset-state should warn that it is DB only
thingeejungleboyj: no20:25
*** rushil has joined #openstack-cinder20:28
jungleboyjthingee: Ok.  Wanted to provide the info back to the z Series guys.  How do you run it?20:30
*** PaulCuzner has quit IRC20:31
thingeejbernard: we still aiming for k-3 with this ?20:33
*** harlowja_away is now known as harlowja20:33
thingeejungleboyj: setup devstack, run
thingeejungleboyj: have you really never done this before with any of your drivers?20:34
*** Mandell has joined #openstack-cinder20:34
jungleboyjthingee: I really have not.  The driver developers have been doing that.20:34
jungleboyjI get stuck doing a lot more management stuff here than you guys realize.20:35
thingeejungleboyj: well, it's a good thing you have internal folks working on all these ci's that can help, right?20:35
* jungleboyj is afraid I am now being looked at as the bastard child.20:35
thingeejungleboyj: instead of running tempest directly, you would run that script20:35
jungleboyjthingee: Yeah.  Ok.  Thanks.20:36
*** lcurtis has joined #openstack-cinder20:36
jungleboyjSorry for the dumb quesiton.20:36
scottdaxyang: Let me know if you have time to talk about force_detach API20:36
thingeejungleboyj: it's not dumb, I just assumed you were it on all those cert tests. My mistake20:36
* hemna is currently waiting on a tempest run20:36
jungleboyjthingee: No, I don't have access to all the different backends.20:36
thingeealright I'll bb, lunch time20:36
jungleboyjHave a good lunch.20:36
thingeejust got my burning man tickets, so I'm in a good mood right now.20:37
hemnanap time for me.20:37
* hemna is home w/ the flu20:37
jungleboyjhemna: Bummer!20:37
jungleboyjhemna: Hope you feel better.20:37
*** rhe00_ has joined #openstack-cinder20:38
*** p01s0n has quit IRC20:39
markstur_thingee, representing cinder at burning man20:40
jungleboyjHe he, burning block heads.20:40
*** annashen has joined #openstack-cinder20:43
kauferjungleboyj: Process question for you, if you have a minute?20:44
jungleboyjkaufer: Shoot.20:44
kauferjungleboyj: Spec has been approved for kilo and is under review -- the API working group sorting guidelines just merged ( so I need to update the code to reflect the new guidelines20:44
akerranyone else's ci system getting caught by being down off and on all day today?20:44
patrickeastakerr: yep, mine is20:45
kauferjungleboyj: My question is, do I need to update the spec to reflect the new sorting syntax and get that re-approved?  The spec is no longer aligned with the sorting guidelines.20:45
akerrpatrickeast: whew!  not us :)20:46
jungleboyjkaufer: How timely.20:46
patrickeastakerr: i asked in -infra and they mentioned that the gate ci’s actually cache the images on the nodes which is why gate doesn’t seem to be broken right now, i’m looking into how to make that work on my ci too20:47
akerrpatrickeast: we do some cacheing on ours as well, but I guess not enough20:47
jungleboyjkaufer: So you are only the second person to kind-of hit this.20:47
openstackgerritDan Smith proposed openstack/cinder: Add support for chiscsi iscsi helper
jungleboyjkaufer: So, if the spec is no longer accurate, what I think makes the most sense is to push a new patch for your code to the community to meet the new guidelines.  Then create a patch for your spec that updates it.20:48
jungleboyjkaufer: Push that up for review and reference it from your review that has the latest pagination code included.20:49
jungleboyjThat way we know you are getting the spec updated.20:49
jungleboyjLets see how that works.  Then we will need to formalize a process around how to handle this in the future.20:50
kauferjungleboyj: So I can do both in parallel?20:50
jungleboyjthingee: ^^^20:50
jungleboyjI would say so given that this is a change that is coming from the API working group's changes.20:50
jungleboyjJust need to make sure they both get merged about the same time.20:50
*** kvidvans has joined #openstack-cinder20:51
kauferjungleboyj: Ok, I'll start with the spec update (should be simple) ... I just wanted to make sure that it was still acceptable to make spec changes20:51
jungleboyjYes, if your spec merged you will need to do it as a new patch though.20:53
jungleboyjThere is no freeze, that I can imagine on making specs accurate.  We will need to talk about that in the future though.20:54
kauferjungleboyj: So a patch to the existing file in specs/kilo, right?20:54
jungleboyjNeed to drop.  Back in a bit.20:54
*** upenrc has joined #openstack-cinder20:55
upenrcHi, I am unable to either attach a new cinder volume to a nova instance OR BOOT from Image(Create a new volume).... However, creating an empty volume on my V7000 backend works.20:55
upenrcHere's more details abt the issue:
*** nellysmitt has quit IRC20:56
smcginnisjungleboyj: ^^20:56
upenrcpl. check20:57
*** ronenkat has quit IRC20:57
vilobhmmcan we run the tempest test in our dev envoirnment so as to be on the safer side before jenkins complains about tempest errors harlowja pointed me to …looks quite old is there something new that can be tried20:58
vilobhmmalthough this is a excellent writeup just that it was written in 201320:58
*** jungleboyj has quit IRC20:59
upenrcvilobhmm - responding to me?21:02
vilobhmmupenrc : sorry didnt got you21:02
upenrcsorry i tht u were responding to me21:03
upenrcHere's more details abt the issue:
*** jnrao has joined #openstack-cinder21:05
*** kvidvans has quit IRC21:06
*** kvidvans has joined #openstack-cinder21:06
*** kvidvans has quit IRC21:06
*** kvidvans has joined #openstack-cinder21:06
thingeekaufer: yes, spec update would be great.21:07
*** kvidvans has quit IRC21:07
thingeekaufer: did things get approved in openstack-specs and api-wg?21:07
thingeekaufer: juggling a bunch of things atm sorry21:07
*** kvidvans has joined #openstack-cinder21:07
vilobhmmupenrc : what do you see in your cinder-api logs ? I havent tried with lvm but played around with other storage i can see this when nova-api can’t talk to cinder-api21:07
kauferthingee: Approved in the API-WG for the server sorting guidelines, the CLI openstack-spec is still under review21:07
*** annegent_ has quit IRC21:10
upenrcvilobhmm i only have cinder volume log posted in the linkk21:10
upenrcpl. be specific21:10
*** EduardBiceriMate is now known as EduardBiceri_afk21:11
upenrcoh checking api.log21:11
vilobhmmyup 2015-02-04 14:12:20.962 21245 ERROR cinder.volume.driver [req-a9780d29-e87f-46af-9f28-7f27a55c7b17 2c37f6bd55b8402bae2ab04402562e4f 4cf6d19846284f4bb5cb3731395de51b - - -] Unable to fetch connection information from backend: create_host: No initiators or wwpns supplied.21:12
upenrcvilobhmm - here: 402bae2ab04402562e4f 4cf6d19846284f4bb5cb3731395de51b - - -] Connected to AMQP server on 2015-02-02 13:00:49.523 3786 ERROR cinder.api.middleware.fault [req-a2ab13c4-2b92-40df-95e3-41118f5dc8a2 2c37f6bd55b8402bae2ab04402562e4f 4cf6d19846284f4bb5cb3731395de51b - - -] Caught error: Volume driver not ready.21:12
upenrc^^error above21:12
vilobhmmhave you made proper conf changes to your nova.conf, cinder.conf to use LVM ?21:12
vilobhmmsome config is missing21:13
upenrccinder.conf - yes - empty volume creation works21:13
upenrchavent touched nova.conf21:13
upenrcwhat properties to edit in nova.conf21:14
*** Yogi11 has quit IRC21:14
*** dustins has quit IRC21:14
upenrc##### VOLUMES ##### # iscsi target user-land tool to use iscsi_helper=ietadm volume_api_class=nova.volume.cinder.API # Region name of this node (string value) os_region_name=RegionOne21:15
*** karimb has joined #openstack-cinder21:19
*** jcru has quit IRC21:19
upenrc^ error21:20
*** Adriano_ has quit IRC21:22
*** Adriano__ has quit IRC21:22
openstackgerritVilobh Meshram proposed openstack/cinder: Introducing micro_states for create workflow
*** foobloo2 has joined #openstack-cinder21:24
*** Yogi1 has joined #openstack-cinder21:24
foobloo2is there a blueprint for cinder change block tracking?21:24
*** dustins has joined #openstack-cinder21:24
*** dustins has quit IRC21:25
*** Yogi1 has quit IRC21:25
*** markstur_ has quit IRC21:26
openstackgerritVilobh Meshram proposed openstack/cinder: Introducing micro_states for create workflow
harlowjavilobhmm u might also have luck in #openstack-infra about the tempest question21:28
thingeefoobloo2: what?21:29
*** markstur has joined #openstack-cinder21:29
*** dustins has joined #openstack-cinder21:31
foobloo2thingee, a certain storage company is saying something to the effect that there is some upstream effort re change block tracking. Not sure if they are implying cinder of libvirt itself.21:34
foobloo2or libvirt**21:35
*** jungleboyj has joined #openstack-cinder21:35
*** ronis has quit IRC21:41
*** sgotliv has quit IRC21:41
foobloo2they're saying something about leveraging the cinder backup api to track changed blocks to swift21:41
foobloo2err track changed blocks and push them to swift21:44
*** julim has quit IRC21:45
*** EmilienM|afk is now known as EmilienM21:45
guitarza1it's been hand waved about, but is there a serious proposal somewhere?21:45
foobloo2that's what I am trying to figure out...there is vague reference to it in raksha, but that's all i've been able to find21:46
thingeetbarron, DuncanT ^21:46
thingeefoobloo2: those are the people you'll want to talk to21:46
foobloo2thingee, thanks21:46
*** foobloo2 is now known as foobloo21:47
openstackgerritAnish Bhatt proposed openstack/cinder: Add support for chiscsi iscsi helper
anishthingee: ^ final21:54
*** thangp has quit IRC21:57
openstackgerritNagendra Rao Jaladanki proposed openstack/cinder-specs: Spec for Brocade Zone Driver Friendly Zone names
openstackgerritSteven Kaufer proposed openstack/cinder-specs: Cinder Pagination Sorting Enhancements
*** upenrc has quit IRC21:59
*** enterprisedc has quit IRC22:00
jnraoHi all, we posted any update to BP:
jnraoCan you please take a look at it?22:00
*** karimb has quit IRC22:02
*** annegent_ has joined #openstack-cinder22:04
*** enterprisedc has joined #openstack-cinder22:04
*** annegent_ has quit IRC22:05
kauferjungleboyj: Thx for the guidance on updating the spec, it's up:
jungleboyjkaufer: You are welcome!22:08
*** bswartz has quit IRC22:13
*** dustins has quit IRC22:15
*** akerr has quit IRC22:17
*** jecarey has quit IRC22:20
*** lcurtis has quit IRC22:20
*** Lee1092 has quit IRC22:20
*** kaufer has quit IRC22:20
openstackgerritTom Barron proposed openstack/cinder: Implement Cinder Manage/Unmanage APIs in NFS driver
tbarronthingee: ^^^^^^ fixed that unit test22:21
tbarronthingee: but let's see what jenkins says22:22
thingeetbarron: thanks22:22
thingeeanish: thanks22:22
*** dulek has joined #openstack-cinder22:24
*** jcru has joined #openstack-cinder22:25
*** tellesnobrega_ has joined #openstack-cinder22:27
e0nethingee: just fyi: i tested dulek's patches for persistance workflow. it work with some limitation. it resume flow on a new host works too22:27
*** jcru has joined #openstack-cinder22:28
*** ebalduf has quit IRC22:28
*** dulek has quit IRC22:29
e0nethingee, dulek: i tested scheduler and manager flows22:29
*** tbarron has quit IRC22:36
*** mriedem has quit IRC22:41
*** vilobhmm has quit IRC22:44
thingeee0ne: excellent!22:44
thingeeexcited to check things out22:44
thingeeI gotta do an interview in a few, but I'll bb22:44
*** vilobhmm has joined #openstack-cinder22:45
*** MentalRay has quit IRC22:45
e0nethingee: i'll publish comments to dulek, there some broken code:(22:45
e0nethingee: and will send review-request for volume flow with some notes tomorrow before your morning22:46
e0ne*manager flow22:46
*** lpetrut has quit IRC22:53
*** nellysmitt has joined #openstack-cinder22:57
* e0ne tried to work with not latest patchset from gerrit :(22:57
*** lpetrut has joined #openstack-cinder22:58
*** lpetrut has quit IRC22:59
*** jaypipes has quit IRC23:00
*** casusbelli has joined #openstack-cinder23:01
*** nellysmitt has quit IRC23:01
*** Mandell has quit IRC23:06
vilobhmme0ne, thingee : can you please review this
vilobhmmpassed jenkins23:07
vilobhmmxyang : ^^23:08
*** Mandell has joined #openstack-cinder23:15
*** VManiac16 has quit IRC23:19
*** annashen has quit IRC23:22
*** e0ne has quit IRC23:24
*** primechuck has quit IRC23:26
*** primechuck has joined #openstack-cinder23:27
*** primechuck has quit IRC23:27
*** ebalduf has joined #openstack-cinder23:29
*** ebalduf has quit IRC23:34
*** dulek has joined #openstack-cinder23:39
*** rmesta has quit IRC23:40
vilobhmmthingee : ping23:46
*** bswartz has joined #openstack-cinder23:47
anishsmcginnis: re : can you confirm setup.cfg ?23:52
*** annegent_ has joined #openstack-cinder23:55
*** markvoelker has quit IRC23:56
*** lan has quit IRC23:57
smcginnisanish: You should probably check with Captian Oslo. :)23:57
smcginnisjungleboyj: ^^23:57
*** lan has joined #openstack-cinder23:57
jungleboyjDid someone way Captain Oslo !?!23:58
anishI was gonna say jay of the jungle23:58
anishdropped all openstack/common file re : , but I am not sure if setup.cfg is supposed to be modified23:58
*** lcurtis has joined #openstack-cinder23:59

Generated by 2.14.0 by Marius Gedminas - find it at!