Monday, 2015-01-05

*** openstack has joined #openstack-cinder15:06
*** lpetrut has quit IRC15:10
*** EduardBiceriMate has quit IRC15:12
*** mdenny has joined #openstack-cinder15:13
*** med_ has quit IRC15:13
*** eharney has joined #openstack-cinder15:13
*** eharney has quit IRC15:13
*** eharney has joined #openstack-cinder15:13
*** med_ has joined #openstack-cinder15:13
*** tristanC_ has joined #openstack-cinder15:13
*** med_ has quit IRC15:14
*** med_ has joined #openstack-cinder15:14
*** tristanC has quit IRC15:14
*** afazekas has quit IRC15:21
*** cl__ has joined #openstack-cinder15:26
*** tsekiyama has joined #openstack-cinder15:27
*** scottda_ has joined #openstack-cinder15:27
*** jungleboyj has joined #openstack-cinder15:28
cl__Hi, anyone knows whether creating a boot volume from a raw image will result in a qcow2 format volume?15:29
*** nellysmitt has joined #openstack-cinder15:30
*** fischerw has joined #openstack-cinder15:30
*** sgotliv has quit IRC15:30
*** mtanino has joined #openstack-cinder15:30
*** scottda_ has quit IRC15:32
*** lpetrut has joined #openstack-cinder15:33
*** mtanino_ has joined #openstack-cinder15:33
*** lpabon has joined #openstack-cinder15:34
*** mtanino has quit IRC15:35
*** nshaikh has quit IRC15:35
*** nlevinki has quit IRC15:36
*** alecv has quit IRC15:41
*** tristanC_ is now known as tristanC15:42
*** mtanino_ has quit IRC15:43
*** sgotliv has joined #openstack-cinder15:45
*** EduardBiceriMate has joined #openstack-cinder15:46
*** Masahiro has joined #openstack-cinder15:47
*** Yogi11 has joined #openstack-cinder15:49
*** bswartz has quit IRC15:51
*** tshefi has quit IRC15:51
*** Yogi1 has quit IRC15:51
*** Masahiro has quit IRC15:52
*** teju has joined #openstack-cinder15:52
*** ryanpetrello_ has joined #openstack-cinder15:54
*** mwichmann has joined #openstack-cinder15:55
*** mudassirlatif has joined #openstack-cinder15:56
*** ryanpetrello has quit IRC15:57
*** ryanpetrello_ is now known as ryanpetrello15:57
*** beekneemech is now known as bnemec16:01
*** mudassirlatif has quit IRC16:03
*** mudassirlatif has joined #openstack-cinder16:04
*** KurtMartin has joined #openstack-cinder16:04
*** openstackgerrit has quit IRC16:06
*** openstackgerrit has joined #openstack-cinder16:07
*** mudassirlatif has quit IRC16:12
*** david-ly_ is now known as david-lyle16:13
*** teju has quit IRC16:15
*** Yogi11 has quit IRC16:17
openstackgerritMitsuhiro Tanino proposed openstack/python-cinderclient: Add command to show pool information for backends
*** mtanino has joined #openstack-cinder16:20
*** Yogi11 has joined #openstack-cinder16:20
*** david-lyle has quit IRC16:22
*** dannywilson has joined #openstack-cinder16:22
*** jwcroppe has joined #openstack-cinder16:23
*** mudassirlatif has joined #openstack-cinder16:25
*** _cjones_ has joined #openstack-cinder16:25
*** nellysmitt has quit IRC16:30
*** lpabon has quit IRC16:30
*** nellysmitt has joined #openstack-cinder16:31
*** rhe00_ has quit IRC16:31
*** jdurgin has joined #openstack-cinder16:32
*** junhongl has quit IRC16:34
*** junhongl has joined #openstack-cinder16:35
*** Longgeek has quit IRC16:35
*** Longgeek has joined #openstack-cinder16:35
*** nellysmitt has quit IRC16:35
*** lcurtis has joined #openstack-cinder16:38
*** rushiagr_away is now known as rushiagr16:40
*** rushiagr is now known as rushiagr_away16:42
*** Mandell has joined #openstack-cinder16:44
*** hemna has joined #openstack-cinder16:44
*** thingee has quit IRC16:47
*** nellysmitt has joined #openstack-cinder16:47
*** gary-smith has joined #openstack-cinder16:48
hemnaand so the break is over.  back to work!16:48
*** openstackgerrit has quit IRC16:51
*** openstackgerrit has joined #openstack-cinder16:51
*** nuritv has joined #openstack-cinder16:52
*** Mandell_ has joined #openstack-cinder16:52
*** Mandell has quit IRC16:55
*** mudassirlatif has quit IRC16:57
*** openstackgerrit has quit IRC17:04
*** openstackgerrit has joined #openstack-cinder17:04
*** cnesa has quit IRC17:05
*** dulek has quit IRC17:06
*** EmilienM is now known as EmilienM|afk17:07
jungleboyjhemna: That is the word on the street.17:11
*** pwehrle has quit IRC17:14
*** Yogi11 has quit IRC17:15
*** kfox1111 has joined #openstack-cinder17:19
*** KurtMartin has quit IRC17:22
kfox1111how do you git update a revew when somene else has posted an updated patch set?17:23
jungleboyjYou need to download the review first with the latest patchset.17:23
*** mikedillion has quit IRC17:24
jungleboyjgit review -d <review number>17:24
*** KurtMartin has joined #openstack-cinder17:24
jungleboyjkfox1111: ^^17:24
kfox1111k. thx.17:24
*** KurtMartin has quit IRC17:24
*** KurtMartin has joined #openstack-cinder17:25
*** jistr has quit IRC17:25
kfox1111is there a way in the website to show the diff if things between patch sets? I'm interested in only what changed.17:27
*** kmartin has quit IRC17:27
*** jpich has quit IRC17:28
jungleboyjYes.  On the left hand side select the previous patch instead of the base.17:28
*** rwsu has joined #openstack-cinder17:28
*** vedams-seeni_ has joined #openstack-cinder17:30
vedams-seeni_Hi everyone17:30
vedams-seeni_anyone did benchmarking I/O ?17:31
vedams-seeni_help me please17:31
kfox1111jungleboyj: the review -d keeps updating the branch that depends on it, not the branch that is for that feature. is that expected? is there a way to tell it which branch to update?17:31
jungleboyjkfox1111: Wait, what are you trying to do?17:32
jungleboyjYou want to pull down the latest patch, make changes and upload a new patch.  Right?17:32
kfox1111jungleboyj: I see a dropdown saying Reference version: base, and if I switch it to any of the other patches, I see a lot of changes that weren't made I think.17:32
*** jdurgin has quit IRC17:33
kfox1111jungleboyj: I have two reviews.... 143553 - posix-backup-driver and 139737 - chunked-backup-base-class17:33
kfox1111143553 depends on 139737.17:33
jungleboyjCan you provide the links?17:33
kfox1111I want to work on 139737.17:34
kfox1111while I was out on vacation, Tom helpfully committed some changes. I need to pull what he did, so I can continue to work.17:35
*** Masahiro has joined #openstack-cinder17:36
jungleboyjkfox1111: So, the easiest thing is just to go to a fresh checkout and then use the git command in the 'Download' section of the review.17:37
jungleboyjThat will bring you up to the latest patch is out there.17:37
jungleboyjOh, I see why I had confused you.  I mean git-review -d , not git -d .17:38
jungleboyjkfox1111: Sorry about that.17:39
jungleboyjgit-review -d does basically the same thing as using the command from gerrit.17:39
*** Masahiro has quit IRC17:41
vedams-seeni_anyone did benchmarking i/o?17:42
tbarronkfox1111: i have some pending changes in  Our internal gerrit wasn't available over the holidays so I haven't got them pushed up yet.17:42
tbarronI'll probably just email them to you if you are planning to push up changes.17:43
vedams-seeni_hi tbarron17:43
vedams-seeni_hi junleboyj17:43
tbarronkfox1111: mostly just fixes for some good remarks from e0ne17:43
jungleboyjvedams-seeni_: Hi.  No, I have not done benchmarking.17:44
vedams-seeni_ok, thanks for your reply17:44
tbarronkfox1111: and also one substantive  addition, namely an overridable method for getting "extra_metadata"17:44
*** leeantho has joined #openstack-cinder17:44
tbarrondefault method returns None.  But we'll need it for class extensions.17:45
tbarronvedams-seeni_: nor I ...17:45
kfox1111tbarron: ok. I'll wait until you post then. thanks again for the help. :)17:45
kfox1111cool. yeah. that'll be nice.17:45
kfox1111thx. tbarron, nah, I was just going to see where things were at and address a couple of the minor comments that were made. but if you already have changes to push, its probably easier for you to do the couple of tweaks.17:46
tbarronkfox1111: either way, if you want to push first, that's fine.17:47
kfox1111I havent made any changes yet. just trying to figure out how pulling works in gerrit.17:47
kfox1111doesn't seem like multiple people working on the same feature is a common thing with gerrit.17:48
tbarronkfox1111: yeah, always doing commit --amend with multiple updaters is not the best :-)17:49
tbarronI'll get my stuff up there today and then won't likely be in the way going forwards - can send you remarks/suggestions on the side or in the review itself.17:50
kfox1111ok. sounds like a plan. thanks. :)17:50
*** tries has joined #openstack-cinder17:51
*** tries has joined #openstack-cinder17:51
kfox1111have you gotten a chance to look at unit test for the posix driver at all, or should I start working on that?17:51
*** mudassirlatif has joined #openstack-cinder17:52
tbarronkfox1111: I haven't done anything yet.  I think we should do UTs for posix in a more contemporary fashion, using mock(), than what we see for Swift.17:55
*** nkrinner has quit IRC17:55
tbarronJust test the actual methods in the posix extension.17:55
tbarronin isolation from the workings of chunking, I/O, etc.17:56
*** thingee has joined #openstack-cinder17:56
kfox1111yeah. I tried copying swifts unit tests for the previous posix driver, and it was rejected for using old code. :/17:56
tbarronSo it will mostly be testing that os.join() yields the expected results :-)17:57
kfox1111unformtunatly, I'm not very familior with the new way, so it was REALLY slowing me down. :/17:57
tbarronI can work on some unit tests for posix if you want.17:58
kfox1111that would be awesome.17:58
*** jordanP has quit IRC18:00
tbarronkk - look for a swift/chunking update later today, posix tomorrow if I can keep the time clear.  In the mean time, maybe think about e0ne's suggestion that you use context handlers for _write_all_to_callback_on_flush(), etc.18:01
tbarronIn the changeset I'll push up, I convert that function to a context handler by adding __enter__ and __exit__ methods.18:01
tbarron(You may decide to keep that change or not)  but I don't change the caller to do a with ...18:02
tbarronThe logical direction of e0ne's suggestion is to dp that as well.  Doing something like 'with put_object_writer as writer: ...'18:03
tbarronand 'with get_object_reader as reder: ....'18:03
tbarronBut the current get_object_reader is doing StringIO.StringIO(), which isn't a context handler.18:04
tbarronSo you may want to push back on the suggestion, reasonable as it is, on the grounds that these methods are meant to be overriden, and we don't want to impose the requirement on all implementations that they be context handlers.18:05
tbarronI dunno what I really think though.18:05
*** Longgeek has quit IRC18:06
kfox1111k. sounds like a plan. :)18:06
*** sgotliv has quit IRC18:07
*** ryanpetrello has quit IRC18:10
*** ryanpetrello has joined #openstack-cinder18:10
*** sweston has quit IRC18:12
*** cppforlife has quit IRC18:12
*** mgagne has quit IRC18:12
*** vedams-seeni_ has quit IRC18:13
*** krtaylor has quit IRC18:13
*** spligak has quit IRC18:14
*** jbernard has quit IRC18:14
*** jordandh has quit IRC18:14
*** jbernard has joined #openstack-cinder18:16
*** spligak has joined #openstack-cinder18:16
*** krtaylor has joined #openstack-cinder18:17
*** sweston has joined #openstack-cinder18:17
*** jordandh has joined #openstack-cinder18:17
*** mgagne has joined #openstack-cinder18:17
*** cppforlife has joined #openstack-cinder18:17
*** mgagne is now known as Guest5831918:17
thingeeanyone in core take a look at
thingeewould like to get this through soon18:19
*** patrickeast has joined #openstack-cinder18:20
thingeejungleboyj: question with
thingeeis icehouse backports for security only?
jungleboyjthingee: I am not aware of that restriction being set yet.18:21
thingeeok, was going off that releases page18:22
thingeefine with me otherwise18:22
*** nikesh_vedams has joined #openstack-cinder18:22
jungleboyjthingee: Ok.  Yeah, I haven't heard that we are only taking security fixes there.  Given that it is a minor change I am not worried about letting it in.18:24
thingeejungleboyj: sounds good, thanks18:24
*** rlucio has joined #openstack-cinder18:25
*** annashen has joined #openstack-cinder18:27
flip214hemna: is rebased, please remove your -1. thanks.18:27
*** EmilienM|afk is now known as EmilienM18:28
hemnaflip214, ok looking18:28
*** tries has quit IRC18:29
nikesh_vedamshi i am running driver_certs/ on my cinder driver for dothill array but 2-3 tests cases are failing and these failed cases are passing when running separately18:29
*** btran has joined #openstack-cinder18:29
*** e0ne has joined #openstack-cinder18:30
*** e0ne is now known as e0ne_18:30
*** 64MABL776 has joined #openstack-cinder18:30
flip214hemna: thank you.18:31
flip214No code changes, apart from some whitespace changes that flake8 wanted.18:32
nikesh_vedamsdo any one have idea of setting up a devstack setup for HPMSA18:35
hemnayah looks like it has a dep on jgriffith's target change.   I'm reviewing that one first.18:36
hemnamaybe I can +A both of them :P18:36
flip214hemna: you get a medaillon of code-review, +4 health, for that idea ;)18:38
*** junhongl has quit IRC18:38
nikesh_vedamstempest test cases are passing individually but 2-3 test cases are failing on running driver_certs/ so driver is failing18:39
*** e0ne_ has quit IRC18:40
*** tries has joined #openstack-cinder18:41
*** junhongl has joined #openstack-cinder18:42
*** ebalduf has joined #openstack-cinder18:43
*** jdurgin has joined #openstack-cinder18:43
*** nellysmitt has quit IRC18:43
rushilnikesh_vedams: Set concurrency to 1 in ~/devstack/driver_certs/ in line number 97. Or you can keep trying till it passes.18:45
thingeejgriffith: happy day
*** e0ne has joined #openstack-cinder18:47
thingeejgriffith: now to get through gate18:47
thingeeDuncanT: you around today?18:48
*** leeantho has quit IRC18:49
jgriffiththingee: W00T!!!!!!!!18:50
jgriffiththingee: Happy New year!!!18:50
e0nejungleboyj: hi! are you around?18:51
thingeejgriffith: thanks for your work on it18:55
jgriffiththingee: you bet, thanks for reviewing it and supporting the idea :)  Now the "real" work will start18:58
thingeeso the cinder object stuff looks pretty cool. have some comments on it...looks like it's breaking some unit tests still though19:01
thingeemostly because there are a lot of fields that are optional that's it's raising keyerrors on.19:01
e0nejungleboyj: i would like to ask you to take a look on once more. your -2 was too quick there19:01
*** EduardBiceriMate has quit IRC19:03
jungleboyje0ne: Looking.19:06
jungleboyje0ne: What am I missing?19:06
jungleboyje0ne: Oh, I see your comment now.  Hold on.19:07
e0nejungleboyj: it's not syncing with oslo-incubator..19:07
openstackgerritNavneet Singh proposed openstack/cinder: Manage/unmanage impl for NetApp ONTAP iscsi driver
*** dustins_ has joined #openstack-cinder19:10
openstackgerritNavneet Singh proposed openstack/cinder: Implement Cinder Manage/Unmanage APIs in NFS driver.
*** eharney has quit IRC19:10
*** kfox1111 has quit IRC19:11
*** junhongl has quit IRC19:11
jungleboyje0ne: Ah, jeez, I don't know what I was thinking when I looked at that.  Thanks e0ne19:11
e0nejungleboyj: no problem.19:12
jungleboyje0ne: Fixed my friend.19:12
*** dustins has quit IRC19:12
e0nejungleboyj: thanks a lot for it! and thanks for reviewing my patch.19:13
openstackgerritNavneet Singh proposed openstack/cinder: NetApp eseries implementation for manage/unmanage
*** junhongl has joined #openstack-cinder19:14
*** Mandell_ has quit IRC19:14
*** tristanC has quit IRC19:20
*** tristanC has joined #openstack-cinder19:20
*** nellysmitt has joined #openstack-cinder19:23
*** nellysmitt has quit IRC19:24
*** Masahiro has joined #openstack-cinder19:25
*** ebalduf has quit IRC19:25
*** Masahiro has quit IRC19:29
*** nellysmitt has joined #openstack-cinder19:30
jungleboyje0ne: No problem.19:30
jungleboyjthingee: Given your earlier question.  What is your thought here:
*** sgotliv has joined #openstack-cinder19:31
*** btran has quit IRC19:32
*** nellysmitt has quit IRC19:34
*** 64MABL776 has quit IRC19:34
*** markstur has joined #openstack-cinder19:53
openstackgerritCorey Bryant proposed openstack/cinder: Pre python 3.2 compatibility for
*** Yogi1 has joined #openstack-cinder20:03
*** Mandell has joined #openstack-cinder20:06
*** thangp has joined #openstack-cinder20:07
*** xyang has joined #openstack-cinder20:11
*** eharney has joined #openstack-cinder20:14
*** xyang has quit IRC20:17
*** bswartz has joined #openstack-cinder20:24
*** nellysmitt has joined #openstack-cinder20:31
jgriffiththingee: jungleboyj rushiagr_away FYI:
*** sfineberg has joined #openstack-cinder20:32
openstackgerritTomoki Sekiyama proposed openstack/cinder: Failover to alternative iSCSI portals on login failure
openstackgerritTomoki Sekiyama proposed openstack/cinder: Enhance iSCSI multipath support
*** nellysmitt has quit IRC20:35
jungleboyjjgriffith: I am confused.20:36
*** leeantho has joined #openstack-cinder20:37
jungleboyjjgriffith: It looks like sean suggested doing it and then we caught the fact we didn't need to do that.20:37
jungleboyjjgriffith: What am I missing?20:37
jgriffithjungleboyj: huh?20:38
*** kvidvans has quit IRC20:38
smcginnisjgriffith, jungleboyj: Yeah, looks like my mistake. Didn't pay attention that it was a backport.20:38
jungleboyjjgriffith: You noted that it was bad juju to mix using LI and not using it.  Ahhhh.20:38
*** kvidvans has joined #openstack-cinder20:38
jgriffithjungleboyj: :)20:38
jungleboyjjgriffith: Ok.  And Sean made the opposite mistake earlier.20:39
jgriffithjungleboyj: smcginnis no worries I was just pointing it out so we catch them going forward20:39
jgriffiththere are sure to be MANY more20:39
jungleboyjjgriffith: Ok, that is a good point.  I will make sure to keep that in mind.20:39
smcginnisLet's just switch everything back. :D20:39
jungleboyjsmcginnis: :-p20:39
jungleboyjYou are clsoe enough I can come up and give you my thoughts on that.  Careful boy.  ;-)20:40
smcginnisjungleboyj: Come visit any time!20:40
*** xyang has joined #openstack-cinder20:40
jungleboyjsmcginnis: Was up there just a month ago to see Manheim Steamroller at the Orpheum.20:40
smcginnisjungleboyj: You probably make it downtown more than I do. Worked in that area for years. Now I rarely go there.20:41
jungleboyj:-)  Yeah, we are up there once a quarter or so.  Hells Kitchen was awesome.20:41
smcginnisjungleboyj: Half the time that I actually do go downtown we end up going there. Good place.20:42
jungleboyjsmcginnis: Our first trip there last time.  Usually do Dakota instead.  I think we will do Hells Kitchen again though.  My boys liked it.20:43
smcginnisjungleboyj: Yeah, cool vibe my kids like too.20:43
*** annashen has quit IRC20:44
jgriffithsmcginnis: +1 :)20:45
*** annashen has joined #openstack-cinder20:46
thingeejungleboyj: I was going to get clarification on the stable branch stuff.20:52
thingeejungleboyj: I would think that we would want to be selective on the older branches20:53
jungleboyjthingee: Agreed.20:54
thingeejgriffith: thanks for the catch20:54
*** Yogi11 has joined #openstack-cinder20:54
*** Yogi1 has quit IRC20:56
*** IanGovett has quit IRC20:57
*** Mandell has quit IRC20:59
*** rwsu has quit IRC21:00
*** alexpilotti has quit IRC21:00
*** rwsu has joined #openstack-cinder21:04
*** Mandell has joined #openstack-cinder21:04
*** btran has joined #openstack-cinder21:08
*** IanGovett has joined #openstack-cinder21:11
*** Masahiro has joined #openstack-cinder21:14
openstackgerritCorey Bryant proposed openstack/cinder: Use assertRaisesRegexp() in
jgriffiththingee: checking out objects :)21:16
*** Masahiro has quit IRC21:18
*** e0ne has quit IRC21:18
thingeejgriffith: nice21:19
thingeejgriffith: like I was saying earlier, these failures
jgriffiththingee: hehe21:23
*** nellysmitt has joined #openstack-cinder21:23
thingeejgriffith: I'm guess from maybe things not being properly mocked?21:23
jgriffiththingee: few minor missed from the ref obj21:23
jdandreaDuncanT: yt?21:23
jgriffiththingee: not sure, probably... or they're actually just missing in the new object21:24
jdandreaWe have another MIA vol attachment in the wild (even though OS claims its attached).21:24
jgriffiththingee: hmm... they seem to be in the obj21:24
thingeejgriffith: I thought that too. But if you take the first failure for example, is defined .,....yeah21:24
openstackgerritMitsuhiro Tanino proposed openstack/python-cinderclient: Add complementary explanation for <host> format of cinder migrate
*** Yogi1 has joined #openstack-cinder21:25
jgriffiththingee: oh... hehe21:25
jgriffiththingee: yeah, so this is going to need a good bit of test work I think maybe21:25
jgriffiththingee: one of the things I was thinking about the other day....21:25
jgriffiththingee: we're really bad in our unit tests about sprinkling fake objects all over21:26
jgriffiththingee: and making them "just what we need"21:26
jgriffiththingee: would be good to have a class with "fake objects" we just call and use everywhere21:26
thingeejgriffith: this change will be good all around :)21:26
jgriffiththingee: and have those be full blown fakes21:26
jgriffiththingee: haha... indeed21:26
*** Yogi11 has quit IRC21:26
*** fifieldt_ has quit IRC21:27
jgriffithmaybe I'll try and work some of that up here this week... something like utils.get_fake_volume/snapshot/backup etc21:27
jdandreaDuncanT: I tried instigating a scsi rescan on the VM per our last chat. It doesn't change what's in /proc/partitions.21:27
thingeejgriffith: don't worry about it, you already got the rest of the target work. :)21:28
jgriffiththingee: I'm ADD member!21:28
*** fifieldt_ has joined #openstack-cinder21:28
jgriffiththingee: can't work on one thing at a time... that's Crazy talk!21:28
*** annashen_ has joined #openstack-cinder21:29
*** annashen has quit IRC21:29
*** jamielennox|away is now known as jamielennox21:30
*** kmartin has joined #openstack-cinder21:30
thingeemorganfainberg: ping21:31
openstackgerritMitsuhiro Tanino proposed openstack/python-cinderclient: Add command to show pool information for backends
morganfainbergthingee, pong back in 20mins21:31
thingeemorganfainberg: k thanks21:31
jdandreaDuncanT: If I detach and reattach the volume, it apears in the VM! Hmm.21:33
*** KurtMartin has quit IRC21:33
*** jamielennox is now known as jamielennox|away21:35
*** nellysmitt has quit IRC21:35
*** jungleboyj has quit IRC21:37
*** akerr has quit IRC21:38
jgriffiththingee: hey... it merged!!21:41
jgriffiththingee: ok, now to keep an eye on the gate21:41
*** gugl2 has quit IRC21:47
*** dustins_ has quit IRC21:48
*** gugl has joined #openstack-cinder21:49
*** sseverson has joined #openstack-cinder21:50
morganfainbergthingee, ping - back21:50
thingeemorganfainberg: hi, since v1 of cinder will be gone next release, I would like to have the volume service type in keystone to point to the root endpoint for discovery of available versions. I think I remember reading some code in keystone that already supports this, but wanted some direction.21:52
*** lpetrut has quit IRC21:52
*** thangp has quit IRC21:53
thingeecurrently I have volume and volumev2 that is being created in gate to support both endpoints.21:53
thingeevolume and volumev2 service type&21:53
*** annashen_ has quit IRC21:55
*** bswartz has quit IRC21:57
*** chlong has quit IRC21:59
morganfainbergthingee, jamielennox|away can give more info22:03
morganfainbergbut yes, it should22:03
thingeeok, I'll wait for jamielennox|away22:03
thingeemorganfainberg: thank you22:03
morganfainbergthingee, i'll circle back when i'm off the phone, battling with travel :P22:03
*** Yogi1 has quit IRC22:04
*** annashen has joined #openstack-cinder22:08
kauferthingee: If possible, could you review my comments to your -1 here:
*** kfox1111 has joined #openstack-cinder22:15
thingeekaufer: my mistake, thanks for pointing that out22:17
kauferthingee: No problem22:17
*** KurtMartin has joined #openstack-cinder22:17
*** tries is now known as tries__22:20
*** kmartin has quit IRC22:21
*** jungleboyj has joined #openstack-cinder22:29
*** mwichmann has quit IRC22:40
*** xyang has quit IRC22:42
*** ebalduf has joined #openstack-cinder22:44
*** kaufer has quit IRC22:46
jgriffiththingee: so I'm keeping an eye on status but honestly things are so bad it's hard to really "see" anything :(22:47
jgriffithgood thing is it mostly seems to be startup failures more than anything which are easy enough to spot22:47
*** btran has quit IRC22:48
*** jamielennox|away is now known as jamielennox22:51
*** chlong has joined #openstack-cinder22:57
*** IanGovett has quit IRC22:58
*** Masahiro has joined #openstack-cinder23:02
*** jcru has quit IRC23:03
*** Masahiro has quit IRC23:07
*** gugl has quit IRC23:13
*** chlong_ has joined #openstack-cinder23:14
morganfainbergjamielennox, ping re: discoverability and the catalog ^ thingee had a question23:16
openstackgerritWalter A. Boring IV (hemna) proposed openstack/cinder: Add volume multi attach support
*** btran has joined #openstack-cinder23:18
jamielennoxthingee: so the problem we've had in keystone is that there is software out there expecting that the 'identity' service type in the catalog returns the url with a /v2.0 suffix23:18
jamielennoxremoving that is just going to break stuff23:18
jamielennoxso we've hacked around it in the client so that if you ask for a version that > 2 and you see the suffix /v2.0 then just chop it off before we do discovery23:19
*** gugl has joined #openstack-cinder23:19
jamielennoxthe intention being that in some future release where we've removed the assumption of /v2.0 the switch to the base url will be transparent (the chop just won't find anything and everything will progress as normal)23:20
jamielennoxthere is support for handling all this within the session object, when making a request you specify version=(2, 0)  or whatever you require and that triggers the client to do discovery to determine the URL for the version23:21
jamielennoxyou will need to add the hack to chop off the cinder suffix - from memory this is hard within cinder because the project_id is a part of the url23:22
jamielennox is where we do the hack for keystone23:23
jamielennoxwe can either make the cinder hack in keystoneclient, or make a public function for add_discover_hack so that it can be controlled from cinderclient23:24
*** hemna has quit IRC23:24
thingeejamielennox: with v1 going away in cinder, I'm using this as an opportunity in Kilo and forward to remove the version suffix from the catalog. So I'm not sure we need the hack23:24
thingeejamielennox: Unless I'm missing something still23:25
jamielennoxok, so we skip that bit and worry about it if there is compatibility problems - you're lucky we never managed to drop the API23:25
thingeeI might be an unpopular person next release, regardless of the warnings23:26
thingeebut even nova is default to the newer version23:26
*** Mandell has quit IRC23:26
thingeethis has been there since G though23:26
jamielennoxfrom memory even in version 2 the catalog provides the project_id right?23:26
thingeejamielennox: yeah23:26
jamielennoxthats a pain23:27
jamielennoxso what's missing in the session to support your case is adding the project_id to a url23:27
*** km has joined #openstack-cinder23:27
*** km has left #openstack-cinder23:28
jamielennoxi've assumed that the URLs you get back from GET / can be used directly23:28
jamielennoxhowever that was expecting nova v3 to happen where they remove that23:28
thingeejamielennox: / GET in can be used directly without a project_id23:30
thingeein cinder*23:30
jamielennoxthingee: right - i mean though that it will return http://host:port/v223:30
jamielennoxwhich is not the full url for making cinder calls23:30
thingeeI see23:31
thingeejamielennox: so what would you recommend for me to move forward in K?23:32
openstackgerritMudassir Latif proposed openstack/cinder: add replication and PG setup methods on PureISCSIDriver
jamielennoxthingee: we should be able to hack it in
jamielennoxjust prepend the project_id to the path that is given23:34
jamielennoxthen it should be just a case of adding a version=(2, 0) to the requests coming from the v2 client23:35
*** scottda_ has joined #openstack-cinder23:35
*** nellysmitt has joined #openstack-cinder23:36
jamielennoxgetting the project_id from the plugin is kind of annoying, but i have a review up to allow getting it directly
thingeejamielennox: taking a step back, so my question actually was to start with, how does the discovery work with keystone?23:37
jamielennoxthingee: sure, the managers make a request for /users from service_type='identity', version=(3, 0), interface='public'23:39
jamielennoxit consults the service catalog and gets the URL for identity service, public interface23:39
*** chlong has quit IRC23:39
*** chlong_ has quit IRC23:39
jamielennoxit then queries the URL it retrieves, gets the list of available versions on the server and finds the url matching version=(3, 0)23:40
jamielennoxit then prepends /users to make the full URL23:40
*** nellysmitt has quit IRC23:41
thingeejamielennox: got it23:42
*** jungleboyj has quit IRC23:44
*** chlong_ has joined #openstack-cinder23:46
*** chlong has joined #openstack-cinder23:46
*** chlong_ has quit IRC23:46
openstackgerritMudassir Latif proposed openstack/cinder: add replication and PG setup methods on PureISCSIDriver
mudassirlatifHi, anybody familiar with replication … can you please review volume replication WIP for the Pure ISCSI driver ( ). Thanks23:56
*** lcurtis has quit IRC23:56
*** ho has joined #openstack-cinder23:59

Generated by 2.14.0 by Marius Gedminas - find it at!