| casusbelli | jgriffith: Hi! I'm still banging at sos-ci. Just wanted to ask what machine and what images are u using? | 00:00 |
|---|---|---|
| jgriffith | casusbelli: I'm running a super-micro box, with Juno deployed on it | 00:00 |
| mtreinish | which seems like a big statement | 00:00 |
| jgriffith | mtreinish: yeah, that's why I haven't done it already | 00:00 |
| casusbelli | jgriffith: I've serious performance issues, wheel venv buildup takes 1.5 hours... | 00:00 |
| jgriffith | mtreinish: FWIW, do multi-node and these problems "disappear" | 00:00 |
| jgriffith | mtreinish: http://goo.gl/JQ9ots | 00:01 |
| jgriffith | mtreinish: so that's "an" issue | 00:01 |
| jgriffith | mtreinish: the problem is I stopped what I was going to propose because it screws up existing/expected behavior | 00:01 |
| casusbelli | jgriffith: are you using ubuntu cloud images on the internal devstack? | 00:02 |
| jgriffith | mtreinish: need to put something together to keep track of the attachment and put it back in place for the user | 00:02 |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/os-brick: Fix comments style according to Hacking rules https://review.openstack.org/157468 | 00:02 |
| jgriffith | casusbelli: yep, ubuntu Trusty | 00:02 |
| casusbelli | jgriffith: plain or modified? | 00:02 |
| jgriffith | plain | 00:03 |
| jgriffith | casusbelli: are you running into an issue? | 00:03 |
| mtreinish | casusbelli: you're running second level guests with ubuntu? | 00:03 |
| *** esker has quit IRC | 00:03 | |
| casusbelli | jgriffith: yeah | 00:03 |
| jgriffith | mtreinish: +1 | 00:03 |
| mtreinish | s/casusbelli/jgriffith | 00:03 |
| casusbelli | jgriffith: x2 | 00:03 |
| jgriffith | casusbelli: what's the issue? | 00:03 |
| casusbelli | jgriffith: i switched from centos to ubuntu images in order to reduce required changes on the existing sos-ci code | 00:04 |
| casusbelli | jgriffith: Issue is primarily performance | 00:04 |
| mtreinish | jgriffith: hmm, ok yeah tracking that would probaly require storing it somewhere. like a new db column or something (without knowing the internals of the rescue operation at all :) ) | 00:04 |
| *** emagana has joined #openstack-cinder | 00:04 | |
| jgriffith | casusbelli: there's no changes... it's just python and ansible | 00:04 |
| casusbelli | jgriffith: there's some ansible stuff that expects user 'ubuntu', etc. | 00:04 |
| casusbelli | jgriffith: but thath's not the issue | 00:05 |
| jgriffith | mtreinish: yeah... and honestly I thought maybe rather than doing that in Nova (which is *hard* to get in to code) I could create a new cinder API | 00:05 |
| jgriffith | temporary-detach or something cheesie | 00:05 |
| jgriffith | like that | 00:05 |
| *** emagana has quit IRC | 00:05 | |
| casusbelli | jgriffith: big point is that 2nd level devstack build takes hours, esp. wheel based venv compilation of numpy, etc. | 00:05 |
| jgriffith | so have a special call in Cinder... maybe even rescue | 00:05 |
| jgriffith | that tells Cinder, detach this... but keep track of it | 00:05 |
| jgriffith | and then an unrescue, that says "put it back the way it was" | 00:06 |
| mtreinish | jgriffith: sure that could work, but we'd need to still teach nova how to do that right? | 00:06 |
| mtreinish | so you'll have to get a nova patch in anyway | 00:06 |
| jgriffith | mtreinish: yeah, but on the nova side it would become a simple method add to nova/volume | 00:06 |
| *** annashen_ has joined #openstack-cinder | 00:06 | |
| jgriffith | mtreinish: but yes, nova patch regardless... but have you seen how difficult it can be to make significant changes to Nova lately :) | 00:07 |
| casusbelli | jgriffith: Are you using any special settings for the 1st level devstack VMs? I tried to improve by host-mode'ing the nova vcpus but that did not make much of a difference... | 00:07 |
| jgriffith | casusbelli: so that has nothing to do with sos, Ubuntu or Cent | 00:07 |
| jgriffith | casusbelli: sounds like youre network if bad | 00:07 |
| mtreinish | jgriffith: sure, I'm just wondering if we should bully some nova core's like dansmith into helping push this through quickly | 00:07 |
| casusbelli | jgriffith: it's on the local machine, devstack in devstack | 00:07 |
| jgriffith | mtreinish: oh... I like that | 00:07 |
| jgriffith | mtreinish: but that means I have to go back to working on it again LOL | 00:08 |
| jgriffith | casusbelli: does your CPU support nesed virt? | 00:08 |
| jgriffith | casusbelli: and what I'm saying is this is purely an issue with network time for you to deploy devstack | 00:09 |
| casusbelli | jgriffith: Not sure, i7 3770 8 core i think | 00:09 |
| jgriffith | casusbelli: so the other thing is check out your network on the instance | 00:09 |
| jgriffith | sounds to me like it's painfully slow | 00:09 |
| mtreinish | jgriffith: heh fair point. Tomorrow I can corner dansmith and we can try to get the rescue task force started :) | 00:09 |
| jgriffith | but if the build is "hours" then you have issues with your VCPU as well.. and perhaps memory | 00:10 |
| jgriffith | haha... I like that "rescue task force" :) | 00:10 |
| jgriffith | mtreinish: in all seriousness I'm more than happy to work on it | 00:10 |
| casusbelli | jgriffith: But e.g. numpy compilation from source takes devstack almost 1 hour, that's while he's compiling the source on 1 to 2 cores, from what i saw | 00:10 |
| openstackgerrit | Anish Bhatt proposed openstack/cinder: Add partial unit tests for IET driver https://review.openstack.org/158925 | 00:10 |
| anish | mtanino: ^ | 00:10 |
| jgriffith | mtreinish: I'll be offline most of tomorrow but back at it tomorrow night and Thurs | 00:10 |
| mtanino | anish: hi | 00:11 |
| jgriffith | mtreinish: meanwhile, we seem to have something "bigger" happening in the tempest tests right now | 00:11 |
| jgriffith | so I'll did into that tonight | 00:11 |
| mtreinish | casusbelli: yeah that sounds like a virt config or other hardware setup issue. It takes me 2-4min to compile it on my desktop | 00:11 |
| jgriffith | I've been hung up for about 8 minutes now.. something died | 00:11 |
| mtreinish | jgriffith: ok, cool. Thanks | 00:11 |
| *** annashen_ has quit IRC | 00:11 | |
| jgriffith | mtreinish: thank you | 00:11 |
| *** ho__ has joined #openstack-cinder | 00:13 | |
| *** emagana has joined #openstack-cinder | 00:13 | |
| jgriffith | casusbelli: see what I'm saying? | 00:13 |
| mtanino | anish: nice :) | 00:14 |
| casusbelli | jgriffith: hmm, ok. If you're goin' with vanilla ubuntu trusty cloud images, too and the host is not that different it has to be something like that | 00:14 |
| jgriffith | casusbelli: well.. what i'm saying is that the instance has little to do with what you described | 00:15 |
| openstackgerrit | Anish Bhatt proposed openstack/cinder: Simplify cxt test_create_export() unit test https://review.openstack.org/158926 | 00:15 |
| jgriffith | casusbelli: devstack for the most part is installing a bunch of pip libs and git repos over the network | 00:15 |
| casusbelli | jgriffith: Gotta read in more detail about that nested virtualization & co | 00:15 |
| jgriffith | casusbelli: log on to that instance and do a speed test | 00:15 |
| casusbelli | jgriffith: yep, i know. Have been looking over devstacks should for quite some time now. :-D | 00:15 |
| jgriffith | casusbelli: also... run stack.sh and in another window top | 00:15 |
| jgriffith | casusbelli: check this guy out: http://goo.gl/cCR8i | 00:16 |
| jgriffith | casusbelli: also, you never answered, what flavor is your instance? | 00:16 |
| jgriffith | casusbelli: vcpu cound, mem and disk? | 00:16 |
| casusbelli | jgriffith: sry, didn't see that. standard m3 flavor, i think | 00:17 |
| casusbelli | jgriffith: but the vm never comes near max RAM or multiple busy vcpus | 00:17 |
| casusbelli | jgriffith: it justs dances on 1 or 2 cores and takes hours with the compilation, eventually either timing out or failing to load the sources for one of the wheeled components | 00:18 |
| anish | mtanino: I think you need to manually take back https://bugs.launchpad.net/cinder/+bug/1409918 | 00:18 |
| openstack | Launchpad bug 1409918 in Cinder "iet target driver not implemented in new iscsi target driver model" [High,In progress] - Assigned to Anish Bhatt (anish7) | 00:18 |
| casusbelli | jgriffith: but i'll check the virt config and thanks for the speed-cli link | 00:19 |
| mtanino | anish: Ah, if it does not take back after pushing next patch, I will change it. thanks | 00:22 |
| casusbelli | jgriffith: Hmm, nested virtualization is definitely something i've been missing out on and that i'm taking a closer look at now | 00:25 |
| xyang1 | patrickeast: Hi | 00:25 |
| *** david-lyle is now known as david-lyle_afk | 00:25 | |
| patrickeast | xyang1: hey | 00:25 |
| xyang1 | patrickeast: I mixed things up:) | 00:25 |
| patrickeast | xyang1: ahh ok, scared me at first, thought i might have totatlly misunderstood how this was supposed to work :D | 00:26 |
| xyang1 | patrickeast: Since used is from ['total'], i used it as denominator | 00:26 |
| xyang1 | patrickeast: So it should be 10/100 < 1 | 00:27 |
| casusbelli | jgriffith: ok, the cpu does support this but i've to check if the OS does the ootb | 00:27 |
| xyang1 | patrickeast: Provisioned/total < 1 | 00:28 |
| patrickeast | xyang1: i think the confusing part is the [‘tota’] that comes back from the array is actually our physically used total on the array | 00:28 |
| patrickeast | total* | 00:28 |
| xyang1 | patrickeast: Ya, that is confusing:) | 00:28 |
| patrickeast | xyang1: ok, cool, so i think we have are doing the right thing, i just confirmed with dannywilson (he sits next to me) and we will go ahead with this algorithm but expand the comments a bit to explain what the fields are | 00:31 |
| patrickeast | xyang1: thanks for all the reviews and support with these changes! | 00:32 |
| xyang1 | patrickeast: No problem. Sorry for the false alarm | 00:32 |
| *** Apoorva_ has quit IRC | 00:34 | |
| *** Apoorva has joined #openstack-cinder | 00:35 | |
| openstackgerrit | Anish Bhatt proposed openstack/cinder: Add partial unit tests for IET driver https://review.openstack.org/158925 | 00:37 |
| casusbelli | jgriffith: Ok, thanks for the hints, i'm digging into the state of nested virt... | 00:37 |
| openstackgerrit | Anish Bhatt proposed openstack/cinder: Simplify cxt test_create_export() unit test https://review.openstack.org/158926 | 00:40 |
| openstackgerrit | Anish Bhatt proposed openstack/cinder: Add partial unit tests for IET driver https://review.openstack.org/158925 | 00:41 |
| anish | apologies for the flood | 00:41 |
| *** timcl has joined #openstack-cinder | 00:43 | |
| jgriffith | casusbelli: I'd recommend just running some tests onyour instance manually first | 00:45 |
| jgriffith | casusbelli: and I don't know what m.7 translates too at all so I can't offer much there | 00:45 |
| *** timcl has quit IRC | 00:45 | |
| *** ebalduf has joined #openstack-cinder | 00:46 | |
| *** Ilja has quit IRC | 00:46 | |
| casusbelli | jgriffith: m.3 is 2vcpus, 4GB RAM and 40GB HD | 00:48 |
| jgriffith | casusbelli: ok...little light | 00:49 |
| jgriffith | but should be ok | 00:49 |
| casusbelli | jgriffith: and yep, i'll do the manual checks. I helps already that i narrowed it down in this area, previously i did not know if maybe specially prepared images, etc. where right | 00:49 |
| casusbelli | jgriffith: ok :) | 00:49 |
| *** ebalduf has quit IRC | 00:50 | |
| *** _cjones_ has quit IRC | 00:51 | |
| *** emagana has quit IRC | 00:53 | |
| *** emagana has joined #openstack-cinder | 00:54 | |
| casusbelli | jgriffith: sry for my rudimentary english, language quality seriously declines after beeing awake and at the desk for almost 20h... | 00:54 |
| *** emagana has quit IRC | 00:54 | |
| *** emagana has joined #openstack-cinder | 00:54 | |
| casusbelli | jgriffith: bye | 00:55 |
| *** casusbelli has quit IRC | 00:55 | |
| *** dannywilson has quit IRC | 00:56 | |
| *** _cjones_ has joined #openstack-cinder | 00:57 | |
| *** Yogi1 has quit IRC | 01:00 | |
| patrickeast | so uh, maybe dumb question, are the tests in test_db_api.py using an actual database? is it being cleared out between tests?/should i worry about leaving it dirty? | 01:03 |
| jgriffith | patrickeast: it runs against sqlite | 01:05 |
| jgriffith | patrickeast: or at least used to | 01:05 |
| jgriffith | patrickeast: and it should have cleanup methods | 01:05 |
| jgriffith | wait... | 01:05 |
| jgriffith | patrickeast: yeah | 01:05 |
| *** boichev has joined #openstack-cinder | 01:05 | |
| patrickeast | jgriffith: ok cool, i don’t see any of the cleanup methods called explicitly by the test cases, so i’m assuming that happens automagically somewhere… that a safe assumption? | 01:07 |
| *** theanalyst has quit IRC | 01:07 | |
| jgriffith | patrickeast: :) | 01:07 |
| *** annashen has quit IRC | 01:08 | |
| patrickeast | maybe i’ll just add my tests and not look into it too much | 01:08 |
| patrickeast | the ol’ toss in in and run away approach | 01:08 |
| *** boichev2 has quit IRC | 01:08 | |
| *** SergeyLukjanov has quit IRC | 01:08 | |
| *** timcl has joined #openstack-cinder | 01:08 | |
| *** arif-ali has quit IRC | 01:08 | |
| patrickeast | it in* | 01:08 |
| jgriffith | patrickeast: so there used to be a pretty obvious call to fake-db or something along those lines | 01:08 |
| jgriffith | patrickeast: you *should* be safe :) | 01:08 |
| jgriffith | patrickeast: but I make no promises or guarantees | 01:08 |
| patrickeast | jgriffith: alright, i’ll play around a little and see if anything breaks, thanks for the info | 01:09 |
| *** openstack has joined #openstack-cinder | 01:12 | |
| *** SergeyLukjanov has joined #openstack-cinder | 01:14 | |
| *** tsekiyam_ has joined #openstack-cinder | 01:16 | |
| *** _cjones_ has quit IRC | 01:16 | |
| *** kbyrne has quit IRC | 01:16 | |
| jgriffith | patrickeast: that's what i'd recommend :) | 01:16 |
| jgriffith | patrickeast: just kidding | 01:16 |
| jgriffith | kinda | 01:16 |
| jgriffith | not really | 01:16 |
| jgriffith | sorta | 01:16 |
| * jgriffith runs away now | 01:16 | |
| *** delattec has quit IRC | 01:16 | |
| *** sgotliv has quit IRC | 01:16 | |
| *** mdbooth has quit IRC | 01:16 | |
| *** bnemec has quit IRC | 01:16 | |
| *** cburgess has quit IRC | 01:16 | |
| *** lifeless has quit IRC | 01:16 | |
| *** wolsen has quit IRC | 01:16 | |
| *** flip214 has quit IRC | 01:16 | |
| *** tristanC has quit IRC | 01:16 | |
| *** morganfainberg has quit IRC | 01:16 | |
| *** git-harry has quit IRC | 01:16 | |
| *** ondergetekende_ has quit IRC | 01:16 | |
| *** timcl has quit IRC | 01:16 | |
| *** Yogi1 has joined #openstack-cinder | 01:16 | |
| *** thingee has joined #openstack-cinder | 01:17 | |
| *** Longgeek has joined #openstack-cinder | 01:18 | |
| *** Longgeek has quit IRC | 01:18 | |
| *** tsekiyama has quit IRC | 01:19 | |
| *** mtanino has quit IRC | 01:19 | |
| *** arif-ali has joined #openstack-cinder | 01:20 | |
| *** mdbooth has joined #openstack-cinder | 01:20 | |
| *** bnemec has joined #openstack-cinder | 01:20 | |
| *** cburgess has joined #openstack-cinder | 01:20 | |
| *** lifeless has joined #openstack-cinder | 01:20 | |
| *** wolsen has joined #openstack-cinder | 01:20 | |
| *** flip214 has joined #openstack-cinder | 01:20 | |
| *** tristanC has joined #openstack-cinder | 01:20 | |
| *** morganfainberg has joined #openstack-cinder | 01:20 | |
| *** git-harry has joined #openstack-cinder | 01:20 | |
| *** ondergetekende_ has joined #openstack-cinder | 01:20 | |
| *** lifeless has quit IRC | 01:20 | |
| *** tsekiyam_ has quit IRC | 01:20 | |
| *** lifeless has joined #openstack-cinder | 01:21 | |
| *** Manshoon has joined #openstack-cinder | 01:21 | |
| *** sgotliv has joined #openstack-cinder | 01:25 | |
| *** Longgeek has joined #openstack-cinder | 01:25 | |
| *** Manshoon has quit IRC | 01:26 | |
| *** Lee1092 has joined #openstack-cinder | 01:27 | |
| *** sigmavirus24 is now known as sigmavirus24_awa | 01:31 | |
| *** Mandell has quit IRC | 01:35 | |
| *** bswartz has quit IRC | 01:35 | |
| *** jwang_ has quit IRC | 01:35 | |
| *** ozamiatin has quit IRC | 01:35 | |
| *** ekarlso has quit IRC | 01:35 | |
| *** jdurgin has quit IRC | 01:35 | |
| *** dansmith has quit IRC | 01:35 | |
| *** sseverson has quit IRC | 01:35 | |
| *** jgriffith has quit IRC | 01:35 | |
| *** anteaya has quit IRC | 01:35 | |
| *** nibalizer has quit IRC | 01:35 | |
| *** tries has quit IRC | 01:35 | |
| *** leeantho has quit IRC | 01:35 | |
| *** rcallawa has quit IRC | 01:35 | |
| *** greghaynes has quit IRC | 01:35 | |
| *** coolsvap_ has quit IRC | 01:35 | |
| *** j_king_ has quit IRC | 01:35 | |
| *** svasheka has quit IRC | 01:35 | |
| *** dosaboy has quit IRC | 01:35 | |
| *** baffle has quit IRC | 01:35 | |
| *** frakt has quit IRC | 01:35 | |
| *** s0nea has quit IRC | 01:35 | |
| *** naotokl has quit IRC | 01:35 | |
| *** kbyrne has joined #openstack-cinder | 01:39 | |
| *** Mandell has joined #openstack-cinder | 01:39 | |
| *** ozamiatin has joined #openstack-cinder | 01:39 | |
| *** rcallawa has joined #openstack-cinder | 01:39 | |
| *** bswartz has joined #openstack-cinder | 01:39 | |
| *** jwang_ has joined #openstack-cinder | 01:39 | |
| *** ekarlso has joined #openstack-cinder | 01:39 | |
| *** greghaynes has joined #openstack-cinder | 01:39 | |
| *** jdurgin has joined #openstack-cinder | 01:39 | |
| *** dansmith has joined #openstack-cinder | 01:39 | |
| *** sseverson has joined #openstack-cinder | 01:39 | |
| *** coolsvap_ has joined #openstack-cinder | 01:39 | |
| *** j_king_ has joined #openstack-cinder | 01:39 | |
| *** jgriffith has joined #openstack-cinder | 01:39 | |
| *** anteaya has joined #openstack-cinder | 01:39 | |
| *** svasheka has joined #openstack-cinder | 01:39 | |
| *** dosaboy has joined #openstack-cinder | 01:39 | |
| *** nibalizer has joined #openstack-cinder | 01:39 | |
| *** tries has joined #openstack-cinder | 01:39 | |
| *** baffle has joined #openstack-cinder | 01:39 | |
| *** frakt has joined #openstack-cinder | 01:39 | |
| *** s0nea has joined #openstack-cinder | 01:39 | |
| *** naotokl has joined #openstack-cinder | 01:39 | |
| *** timcl has joined #openstack-cinder | 01:39 | |
| *** Mandell has quit IRC | 01:40 | |
| *** bswartz has quit IRC | 01:40 | |
| *** jwang_ has quit IRC | 01:40 | |
| *** ozamiatin has quit IRC | 01:40 | |
| *** ekarlso has quit IRC | 01:40 | |
| *** jdurgin has quit IRC | 01:40 | |
| *** dansmith has quit IRC | 01:40 | |
| *** sseverson has quit IRC | 01:40 | |
| *** jgriffith has quit IRC | 01:40 | |
| *** anteaya has quit IRC | 01:40 | |
| *** nibalizer has quit IRC | 01:40 | |
| *** tries has quit IRC | 01:40 | |
| *** kbyrne has quit IRC | 01:40 | |
| *** rcallawa has quit IRC | 01:40 | |
| *** greghaynes has quit IRC | 01:40 | |
| *** coolsvap_ has quit IRC | 01:40 | |
| *** j_king_ has quit IRC | 01:40 | |
| *** svasheka has quit IRC | 01:40 | |
| *** dosaboy has quit IRC | 01:40 | |
| *** baffle has quit IRC | 01:40 | |
| *** frakt has quit IRC | 01:40 | |
| *** s0nea has quit IRC | 01:40 | |
| *** naotokl has quit IRC | 01:40 | |
| *** Yogi1 has quit IRC | 01:40 | |
| *** amoturi_ has quit IRC | 01:40 | |
| *** fischerw has quit IRC | 01:40 | |
| *** markstur has quit IRC | 01:40 | |
| *** david-lyle_afk has quit IRC | 01:40 | |
| *** delattec has joined #openstack-cinder | 01:40 | |
| *** timcl has quit IRC | 01:41 | |
| *** cdelatte has joined #openstack-cinder | 01:41 | |
| *** kbyrne has joined #openstack-cinder | 01:41 | |
| *** Mandell has joined #openstack-cinder | 01:41 | |
| *** ozamiatin has joined #openstack-cinder | 01:41 | |
| *** rcallawa has joined #openstack-cinder | 01:41 | |
| *** bswartz has joined #openstack-cinder | 01:41 | |
| *** jwang_ has joined #openstack-cinder | 01:41 | |
| *** ekarlso has joined #openstack-cinder | 01:41 | |
| *** greghaynes has joined #openstack-cinder | 01:41 | |
| *** jdurgin has joined #openstack-cinder | 01:41 | |
| *** dansmith has joined #openstack-cinder | 01:41 | |
| *** sseverson has joined #openstack-cinder | 01:41 | |
| *** coolsvap_ has joined #openstack-cinder | 01:41 | |
| *** j_king_ has joined #openstack-cinder | 01:41 | |
| *** jgriffith has joined #openstack-cinder | 01:41 | |
| *** anteaya has joined #openstack-cinder | 01:41 | |
| *** svasheka has joined #openstack-cinder | 01:41 | |
| *** dosaboy has joined #openstack-cinder | 01:41 | |
| *** nibalizer has joined #openstack-cinder | 01:41 | |
| *** tries has joined #openstack-cinder | 01:41 | |
| *** baffle has joined #openstack-cinder | 01:41 | |
| *** frakt has joined #openstack-cinder | 01:41 | |
| *** s0nea has joined #openstack-cinder | 01:41 | |
| *** naotokl has joined #openstack-cinder | 01:41 | |
| *** delattec has quit IRC | 01:42 | |
| *** rushil_ has quit IRC | 01:43 | |
| *** primechuck has joined #openstack-cinder | 01:44 | |
| *** ebalduf has joined #openstack-cinder | 01:47 | |
| *** Longgeek has quit IRC | 01:49 | |
| *** cdelatte has quit IRC | 01:51 | |
| *** Mandell has quit IRC | 01:51 | |
| *** ebalduf has quit IRC | 01:51 | |
| *** Mandell has joined #openstack-cinder | 01:52 | |
| *** timcl has joined #openstack-cinder | 01:52 | |
| *** timcl has quit IRC | 01:55 | |
| *** scottda has joined #openstack-cinder | 01:57 | |
| mfisch | thingee: thanks for the response today about cinderv2 | 01:57 |
| *** jamielennox is now known as jamielennox|away | 01:58 | |
| *** Yogi1 has joined #openstack-cinder | 01:59 | |
| *** fischerw has joined #openstack-cinder | 01:59 | |
| *** markstur has joined #openstack-cinder | 01:59 | |
| *** david-lyle_afk has joined #openstack-cinder | 01:59 | |
| *** scottda has quit IRC | 02:01 | |
| *** IanGovett has quit IRC | 02:02 | |
| *** thingee has quit IRC | 02:04 | |
| *** jamielennox|away is now known as jamielennox | 02:06 | |
| *** harlowja has quit IRC | 02:07 | |
| *** harlowja_ has joined #openstack-cinder | 02:07 | |
| *** Yogi1 has quit IRC | 02:10 | |
| *** fischerw has quit IRC | 02:10 | |
| *** markstur has quit IRC | 02:10 | |
| *** david-lyle_afk has quit IRC | 02:10 | |
| *** esker has joined #openstack-cinder | 02:10 | |
| *** annashen has joined #openstack-cinder | 02:10 | |
| *** Yogi1 has joined #openstack-cinder | 02:16 | |
| *** fischerw has joined #openstack-cinder | 02:16 | |
| *** markstur has joined #openstack-cinder | 02:16 | |
| *** david-lyle_afk has joined #openstack-cinder | 02:16 | |
| *** hemna is now known as hemnafk | 02:16 | |
| *** annashen has quit IRC | 02:16 | |
| *** jamielennox is now known as jamielennox|away | 02:17 | |
| *** theanalyst has joined #openstack-cinder | 02:17 | |
| *** Longgeek has joined #openstack-cinder | 02:18 | |
| *** emagana has quit IRC | 02:19 | |
| *** emagana has joined #openstack-cinder | 02:20 | |
| *** timcl has joined #openstack-cinder | 02:23 | |
| *** cdelatte has joined #openstack-cinder | 02:23 | |
| *** timcl has quit IRC | 02:24 | |
| *** emagana has quit IRC | 02:24 | |
| *** jamielennox|away is now known as jamielennox | 02:27 | |
| *** changbl has joined #openstack-cinder | 02:27 | |
| *** vilobhmm has quit IRC | 02:28 | |
| *** Apoorva_ has joined #openstack-cinder | 02:31 | |
| *** cdelatte has quit IRC | 02:32 | |
| *** timcl has joined #openstack-cinder | 02:34 | |
| *** Apoorva has quit IRC | 02:34 | |
| *** Apoorva_ has quit IRC | 02:36 | |
| *** jungleboyj has quit IRC | 02:36 | |
| *** jamielennox is now known as jamielennox|away | 02:38 | |
| *** kaisers2 has joined #openstack-cinder | 02:45 | |
| *** kaisers3 has joined #openstack-cinder | 02:45 | |
| *** kaisers has quit IRC | 02:46 | |
| *** kaisers1 has quit IRC | 02:47 | |
| *** jamielennox|away is now known as jamielennox | 02:47 | |
| *** ebalduf has joined #openstack-cinder | 02:47 | |
| *** timcl has quit IRC | 02:49 | |
| *** nikesh_vedams has quit IRC | 02:51 | |
| *** Mandell has quit IRC | 02:51 | |
| *** ebalduf has quit IRC | 02:52 | |
| *** nikesh_vedams has joined #openstack-cinder | 02:54 | |
| *** Yogi1 has quit IRC | 02:55 | |
| *** markvoelker has quit IRC | 02:56 | |
| *** markvoelker has joined #openstack-cinder | 02:57 | |
| *** markvoelker has quit IRC | 03:01 | |
| *** bkopilov has quit IRC | 03:09 | |
| *** bkopilov has joined #openstack-cinder | 03:09 | |
| *** devlaps has joined #openstack-cinder | 03:14 | |
| *** akerr is now known as akerr_away | 03:17 | |
| *** jamielennox is now known as jamielennox|away | 03:22 | |
| *** Manshoon has joined #openstack-cinder | 03:22 | |
| *** Manshoon has quit IRC | 03:27 | |
| *** annashen has joined #openstack-cinder | 03:30 | |
| *** jamielennox|away is now known as jamielennox | 03:33 | |
| *** annashen has quit IRC | 03:37 | |
| *** devlaps has quit IRC | 03:40 | |
| openstackgerrit | xing-yang proposed openstack/python-cinderclient: Add support to incremental backups in cinder https://review.openstack.org/112101 | 03:47 |
| *** ebalduf has joined #openstack-cinder | 03:48 | |
| openstackgerrit | rajiv proposed openstack/python-cinderclient: cinderclient accepts arguments after metadata without -- separator https://review.openstack.org/156499 | 03:48 |
| *** primechuck has quit IRC | 03:48 | |
| *** mtanino has joined #openstack-cinder | 03:51 | |
| openstackgerrit | Patrick East proposed openstack/cinder: Add CHAP support to PureISCSIDriver https://review.openstack.org/137175 | 03:51 |
| openstackgerrit | Patrick East proposed openstack/cinder: Add DB table for driver specific data https://review.openstack.org/151837 | 03:51 |
| *** Longgeek has quit IRC | 03:52 | |
| *** ebalduf has quit IRC | 03:52 | |
| *** mriedem has quit IRC | 03:56 | |
| *** dannywilson has joined #openstack-cinder | 03:57 | |
| *** fischerw has quit IRC | 03:58 | |
| *** dannywilson has quit IRC | 04:01 | |
| *** asselin_ has joined #openstack-cinder | 04:02 | |
| *** asselin__ has joined #openstack-cinder | 04:04 | |
| *** xyang has joined #openstack-cinder | 04:04 | |
| *** markvoelker has joined #openstack-cinder | 04:06 | |
| *** harlowja_ is now known as harlowja_away | 04:06 | |
| *** asselin_ has quit IRC | 04:07 | |
| *** ho has joined #openstack-cinder | 04:08 | |
| *** ho__ has quit IRC | 04:08 | |
| *** pdeore_ has joined #openstack-cinder | 04:13 | |
| *** coolsvap_ is now known as coolsvap | 04:16 | |
| *** btran has quit IRC | 04:19 | |
| *** annegent_ has joined #openstack-cinder | 04:20 | |
| *** ho__ has joined #openstack-cinder | 04:28 | |
| *** ho has quit IRC | 04:28 | |
| *** Apoorva has joined #openstack-cinder | 04:30 | |
| *** Mandell has joined #openstack-cinder | 04:40 | |
| *** deepakcs has joined #openstack-cinder | 04:41 | |
| *** annashen has joined #openstack-cinder | 04:41 | |
| *** patrickeast has quit IRC | 04:41 | |
| *** BharatK has joined #openstack-cinder | 04:42 | |
| *** rcallawa has quit IRC | 04:43 | |
| *** mtanino has quit IRC | 04:46 | |
| *** Longgeek has joined #openstack-cinder | 04:46 | |
| *** ebalduf has joined #openstack-cinder | 04:49 | |
| *** krtaylor has quit IRC | 04:51 | |
| *** ebalduf has quit IRC | 04:53 | |
| *** tbarron1 has joined #openstack-cinder | 04:57 | |
| *** lpetrut has joined #openstack-cinder | 04:57 | |
| *** tbarron has quit IRC | 05:01 | |
| *** annashen has quit IRC | 05:03 | |
| *** krtaylor has joined #openstack-cinder | 05:03 | |
| *** zhithuang has joined #openstack-cinder | 05:03 | |
| *** krtaylor has quit IRC | 05:04 | |
| openstackgerrit | xing-yang proposed openstack/cinder: Create Consistency Group from CG Snapshot API https://review.openstack.org/145952 | 05:06 |
| *** xyang has quit IRC | 05:06 | |
| *** Longgeek has quit IRC | 05:07 | |
| *** Longgeek has joined #openstack-cinder | 05:07 | |
| *** krtaylor has joined #openstack-cinder | 05:09 | |
| *** PaulCuzner has quit IRC | 05:19 | |
| *** pradipta has joined #openstack-cinder | 05:20 | |
| *** Manshoon has joined #openstack-cinder | 05:23 | |
| *** Manshoon has quit IRC | 05:28 | |
| *** PaulCuzner has joined #openstack-cinder | 05:28 | |
| *** vilobhmm has joined #openstack-cinder | 05:38 | |
| *** abehl has joined #openstack-cinder | 05:43 | |
| *** abehl has quit IRC | 05:43 | |
| *** asselin__ has quit IRC | 05:43 | |
| *** rmstar has quit IRC | 05:46 | |
| *** annegent_ has quit IRC | 05:47 | |
| *** Apoorva has quit IRC | 05:48 | |
| *** annegent_ has joined #openstack-cinder | 05:48 | |
| *** Longgeek has quit IRC | 05:50 | |
| *** ebalduf has joined #openstack-cinder | 05:50 | |
| *** Longgeek has joined #openstack-cinder | 05:52 | |
| *** lpetrut has quit IRC | 05:53 | |
| openstackgerrit | John Griffith proposed openstack/cinder: Revert "Remove retry_deactivation directive on lvremove" https://review.openstack.org/159003 | 05:53 |
| *** annegent_ has quit IRC | 05:53 | |
| *** ebalduf has quit IRC | 05:54 | |
| *** xyang has joined #openstack-cinder | 05:59 | |
| openstackgerrit | Wang Peter proposed openstack/cinder: CG Modification Support in EMC VNX Cinder Driver https://review.openstack.org/154784 | 06:03 |
| openstackgerrit | Wang Peter proposed openstack/cinder: CG Modification Support in EMC VNX Cinder Driver https://review.openstack.org/154784 | 06:05 |
| *** BharatK has quit IRC | 06:05 | |
| *** zhithuang has quit IRC | 06:07 | |
| *** jungleboyj has joined #openstack-cinder | 06:13 | |
| *** Longgeek has quit IRC | 06:19 | |
| *** rwsu is now known as rwsu-afk | 06:25 | |
| *** nshaikh has joined #openstack-cinder | 06:26 | |
| *** xyang1 has quit IRC | 06:27 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/cinder: Imported Translations from Transifex https://review.openstack.org/152401 | 06:31 |
| *** annashen has joined #openstack-cinder | 06:31 | |
| *** annashen has quit IRC | 06:42 | |
| *** rmstar has joined #openstack-cinder | 06:42 | |
| *** anshul has joined #openstack-cinder | 06:45 | |
| *** ebalduf has joined #openstack-cinder | 06:51 | |
| *** ebalduf has quit IRC | 06:55 | |
| *** nkrinner has joined #openstack-cinder | 06:56 | |
| *** rmstar has quit IRC | 06:57 | |
| *** dannywilson has joined #openstack-cinder | 06:57 | |
| *** lpetrut has joined #openstack-cinder | 06:57 | |
| *** dannywilson has quit IRC | 07:00 | |
| *** dannywilson has joined #openstack-cinder | 07:01 | |
| *** zhithuang has joined #openstack-cinder | 07:04 | |
| *** rmstar has joined #openstack-cinder | 07:06 | |
| *** zhithuang has quit IRC | 07:09 | |
| *** zhithuang has joined #openstack-cinder | 07:10 | |
| *** zhithuan1 has joined #openstack-cinder | 07:12 | |
| *** zhithuan1 is now known as winston-d_ | 07:12 | |
| *** zhithuang has quit IRC | 07:15 | |
| *** maike has joined #openstack-cinder | 07:18 | |
| *** TobiasE has joined #openstack-cinder | 07:20 | |
| *** Ilja has joined #openstack-cinder | 07:23 | |
| *** vilobhmm has quit IRC | 07:26 | |
| openstackgerrit | Danny Wilson proposed openstack/cinder: Over subscription for Pure Storage iSCSI driver. https://review.openstack.org/157681 | 07:29 |
| *** Miouge has joined #openstack-cinder | 07:30 | |
| *** markvoelker has quit IRC | 07:32 | |
| *** markvoelker has joined #openstack-cinder | 07:33 | |
| *** markvoelker has quit IRC | 07:37 | |
| *** BharatK has joined #openstack-cinder | 07:39 | |
| *** ekarlso has quit IRC | 07:48 | |
| *** nkrinner has quit IRC | 07:50 | |
| *** ebalduf has joined #openstack-cinder | 07:51 | |
| *** nkrinner has joined #openstack-cinder | 07:52 | |
| *** winston-d_ has quit IRC | 07:56 | |
| *** ebalduf has quit IRC | 07:56 | |
| *** e0ne has joined #openstack-cinder | 07:57 | |
| *** winston-d_ has joined #openstack-cinder | 08:00 | |
| *** pwehrle has joined #openstack-cinder | 08:00 | |
| *** ronis has joined #openstack-cinder | 08:01 | |
| *** TobiasE1 has joined #openstack-cinder | 08:04 | |
| *** haomaiw__ has quit IRC | 08:04 | |
| *** TobiasE has quit IRC | 08:04 | |
| *** zhipeng has joined #openstack-cinder | 08:04 | |
| *** dulek has joined #openstack-cinder | 08:07 | |
| *** dannywilson has quit IRC | 08:07 | |
| *** e0ne has quit IRC | 08:08 | |
| *** zhipeng has quit IRC | 08:09 | |
| *** anurag1 has joined #openstack-cinder | 08:09 | |
| *** dalgaaf has joined #openstack-cinder | 08:23 | |
| *** zhipeng has joined #openstack-cinder | 08:23 | |
| *** Ilja has quit IRC | 08:24 | |
| *** afazekas_ has joined #openstack-cinder | 08:30 | |
| *** markvoelker has joined #openstack-cinder | 08:35 | |
| *** winston-d_ has quit IRC | 08:38 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements https://review.openstack.org/157642 | 08:40 |
| *** markvoelker has quit IRC | 08:40 | |
| *** nellysmitt has joined #openstack-cinder | 08:40 | |
| *** ho__ has quit IRC | 08:41 | |
| *** karimb has joined #openstack-cinder | 08:43 | |
| *** Ilja has joined #openstack-cinder | 08:45 | |
| *** e0ne has joined #openstack-cinder | 08:52 | |
| *** ekarlso has joined #openstack-cinder | 08:52 | |
| *** ebalduf has joined #openstack-cinder | 08:52 | |
| *** Mandell has quit IRC | 08:53 | |
| *** jistr has joined #openstack-cinder | 08:53 | |
| *** ebalduf has quit IRC | 08:57 | |
| *** maike has quit IRC | 09:04 | |
| *** maike has joined #openstack-cinder | 09:04 | |
| *** anurag2 has joined #openstack-cinder | 09:05 | |
| *** anurag1 has quit IRC | 09:06 | |
| *** nkrinner has quit IRC | 09:08 | |
| *** Ilja has quit IRC | 09:08 | |
| *** nkrinner has joined #openstack-cinder | 09:09 | |
| *** maike has quit IRC | 09:09 | |
| *** alecv has joined #openstack-cinder | 09:10 | |
| *** lpetrut has quit IRC | 09:12 | |
| *** markvoelker has joined #openstack-cinder | 09:16 | |
| *** nkrinner_ has joined #openstack-cinder | 09:19 | |
| *** markvoelker has quit IRC | 09:21 | |
| *** nkrinner has quit IRC | 09:22 | |
| *** nkrinner_ has quit IRC | 09:22 | |
| *** nkrinner has joined #openstack-cinder | 09:22 | |
| *** nkrinner has quit IRC | 09:22 | |
| *** nkrinner has joined #openstack-cinder | 09:22 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Delay requests in scheduler to wait for metrics https://review.openstack.org/156219 | 09:30 |
| *** anurag has joined #openstack-cinder | 09:31 | |
| *** anurag2 has quit IRC | 09:32 | |
| *** e0ne has quit IRC | 09:33 | |
| *** emaisne has joined #openstack-cinder | 09:34 | |
| *** e0ne has joined #openstack-cinder | 09:35 | |
| *** Miouge has quit IRC | 09:35 | |
| *** e0ne is now known as e0ne_ | 09:35 | |
| *** yuriy_n17 has quit IRC | 09:38 | |
| *** yuriy_n17 has joined #openstack-cinder | 09:39 | |
| *** emaisne_ has joined #openstack-cinder | 09:39 | |
| *** emaisne has quit IRC | 09:39 | |
| *** emaisne__ has joined #openstack-cinder | 09:40 | |
| *** Miouge has joined #openstack-cinder | 09:42 | |
| *** nkrinner_ has joined #openstack-cinder | 09:43 | |
| *** emaisne_ has quit IRC | 09:43 | |
| *** e0ne_ is now known as e0ne | 09:44 | |
| *** nkrinner has quit IRC | 09:44 | |
| *** Ilja has joined #openstack-cinder | 09:44 | |
| *** jordanP has joined #openstack-cinder | 09:44 | |
| *** anurag has left #openstack-cinder | 09:46 | |
| openstackgerrit | Anton Arefiev proposed openstack/cinder: Import only modules: H302 https://review.openstack.org/145780 | 09:50 |
| *** anuragpalsule has joined #openstack-cinder | 09:50 | |
| *** e0ne has quit IRC | 09:52 | |
| *** ebalduf has joined #openstack-cinder | 09:53 | |
| nikesh_vedams | hi | 09:54 |
| *** TobiasE has joined #openstack-cinder | 09:57 | |
| *** ebalduf has quit IRC | 09:58 | |
| *** TobiasE1 has quit IRC | 09:59 | |
| *** nikesh_vedams has quit IRC | 10:02 | |
| *** anshul has quit IRC | 10:05 | |
| *** emaisne__ has quit IRC | 10:06 | |
| *** anshul has joined #openstack-cinder | 10:06 | |
| *** markvoelker has joined #openstack-cinder | 10:17 | |
| *** Ilja has quit IRC | 10:19 | |
| *** Ilja has joined #openstack-cinder | 10:20 | |
| *** Miouge has quit IRC | 10:20 | |
| *** markvoelker has quit IRC | 10:22 | |
| *** Miouge has joined #openstack-cinder | 10:27 | |
| *** Miouge has quit IRC | 10:28 | |
| openstackgerrit | Anton Arefiev proposed openstack/cinder: Add entry create and cast tasks to manage workflow https://review.openstack.org/139071 | 10:34 |
| openstackgerrit | Anton Arefiev proposed openstack/cinder: Fix lvm manage existing volume https://review.openstack.org/156939 | 10:34 |
| openstackgerrit | Gorka Eguileor proposed openstack/cinder: Use iscsi_helper instead of target_helper on logs https://review.openstack.org/159072 | 10:40 |
| *** erlon has joined #openstack-cinder | 10:45 | |
| *** casusbelli has joined #openstack-cinder | 10:47 | |
| *** ebalduf has joined #openstack-cinder | 10:54 | |
| *** TobiasE has quit IRC | 10:58 | |
| *** ebalduf has quit IRC | 10:58 | |
| openstackgerrit | Gorka Eguileor proposed openstack/cinder: Use iscsi_helper instead of target_helper on logs https://review.openstack.org/159072 | 11:00 |
| *** aix has joined #openstack-cinder | 11:02 | |
| *** lpetrut has joined #openstack-cinder | 11:06 | |
| *** dannywilson has joined #openstack-cinder | 11:08 | |
| *** rushiagr_away is now known as rushiagr | 11:12 | |
| *** dannywilson has quit IRC | 11:12 | |
| *** pradipta has quit IRC | 11:13 | |
| *** asselin has quit IRC | 11:16 | |
| *** mdenny has quit IRC | 11:16 | |
| *** kmartin has quit IRC | 11:16 | |
| openstackgerrit | Anusha rayani proposed openstack/cinder: Filter cinder volumes by name https://review.openstack.org/159081 | 11:17 |
| *** markvoelker has joined #openstack-cinder | 11:18 | |
| *** markvoelker has quit IRC | 11:23 | |
| *** zhipeng has quit IRC | 11:25 | |
| *** Manshoon has joined #openstack-cinder | 11:26 | |
| *** Miouge has joined #openstack-cinder | 11:28 | |
| *** Manshoon has quit IRC | 11:30 | |
| *** timcl has joined #openstack-cinder | 11:32 | |
| *** karimb has quit IRC | 11:33 | |
| *** timcl has quit IRC | 11:36 | |
| *** EmilienM is now known as EmilienM|afk | 11:39 | |
| *** mdbooth has quit IRC | 11:40 | |
| *** mdbooth has joined #openstack-cinder | 11:41 | |
| *** nkrinner_ is now known as nkrinner | 11:46 | |
| *** karimb has joined #openstack-cinder | 11:50 | |
| *** Ilja has quit IRC | 11:51 | |
| *** ebalduf has joined #openstack-cinder | 11:54 | |
| *** timcl has joined #openstack-cinder | 11:56 | |
| *** rcallawa has joined #openstack-cinder | 11:57 | |
| *** IanGovett has joined #openstack-cinder | 11:57 | |
| *** ebalduf has quit IRC | 11:59 | |
| *** Miouge has quit IRC | 12:02 | |
| *** nkrinner has quit IRC | 12:04 | |
| *** Miouge has joined #openstack-cinder | 12:04 | |
| *** ganso_ has joined #openstack-cinder | 12:08 | |
| openstackgerrit | Michal Dulko proposed openstack/cinder: Delay requests in scheduler to wait for metrics https://review.openstack.org/156219 | 12:09 |
| dulek | DuncanT: Hi! | 12:10 |
| DuncanT | dulek: Hi | 12:10 |
| dulek | DuncanT: Can you explain your latest comment on 156219? I'm starting to think I'm missing something. | 12:11 |
| DuncanT | Sure, one sec | 12:11 |
| *** aix has quit IRC | 12:11 | |
| dulek | I don't get how we can call self.driver_api.schedule_create_volume every second. | 12:11 |
| *** e0ne has joined #openstack-cinder | 12:11 | |
| DuncanT | By just calling it... use ether sleep or (better) looping call - it will stop raising an exception if it finds a host | 12:12 |
| *** nkrinner has joined #openstack-cinder | 12:13 | |
| dulek | DuncanT: And I should stop after 60 seconds. Hm, this seem crude. | 12:13 |
| dulek | Also - it will most probably place the volume on the first host that sent metrics. Scheduler should not schedule when it doesn't know whole situation. | 12:14 |
| DuncanT | Well, 1 second is arbitary, but waiting 60 should usually be unnecessary most of the time with winston's other patch in place | 12:14 |
| *** timcl has quit IRC | 12:16 | |
| DuncanT | IMO it is better to start scheduling once you have a suitable host... the odd volume going to a suitable but not necessarily optimal place is better than random creates stalling for 60 seconds whenever you start a new scheduler (which happens in an active/active scheduler environment at random times) | 12:16 |
| dulek | That's true, but calling scheduler_driver till it returns without exception seem like a wrong solution. | 12:18 |
| dulek | In the latest patchset I've modified scheduler.manager to count down the timeout. | 12:18 |
| *** jaypipes has joined #openstack-cinder | 12:18 | |
| dulek | The delay I mean | 12:18 |
| *** markvoelker has joined #openstack-cinder | 12:19 | |
| dulek | So requests will wait 60 seconds _at most_. | 12:19 |
| *** Lee1092 has quit IRC | 12:20 | |
| dulek | We're probably unable to estimate the time needed to collect the metrics even after Winston's patch? | 12:20 |
| DuncanT | Yeah, the clean fix is to modify oslo.messaging so that we get the stats updates but not the schedule request until either 60 seconds has passed or else we have heard from everybody in the service table of the db, but that requires some pretty big changes by the looks of things | 12:23 |
| DuncanT | Until then, it is going to end up being something ugly, we just need to minimise the ugliness | 12:24 |
| *** markvoelker has quit IRC | 12:24 | |
| *** deepakcs has quit IRC | 12:25 | |
| dulek | Agreed, thing is ugliness metric isn't same for everyone. ;) | 12:25 |
| dulek | So we have two approaches - asking every second till no exception comes or delay passes or to wait 60 or less seconds. | 12:25 |
| dulek | Maybe this is good topic for quick discussion at the tomorrow's meeting?\ | 12:26 |
| *** danieru has quit IRC | 12:26 | |
| *** danieru has joined #openstack-cinder | 12:27 | |
| *** lpetrut has quit IRC | 12:32 | |
| dulek | DuncanT: I've added it on the tomorrow's agenda. | 12:34 |
| dulek | DuncanT: Got to go now, thanks for help! | 12:34 |
| DuncanT | Erm, the meeting is today | 12:34 |
| *** timcl has joined #openstack-cinder | 12:35 | |
| DuncanT | dulek: Functional uglyness is worse than code uglyness in general | 12:36 |
| *** dulek has quit IRC | 12:38 | |
| *** timcl1 has joined #openstack-cinder | 12:39 | |
| *** timcl has quit IRC | 12:39 | |
| *** aix has joined #openstack-cinder | 12:43 | |
| *** lpetrut has joined #openstack-cinder | 12:46 | |
| *** sandywalsh has quit IRC | 12:49 | |
| *** lpetrut has quit IRC | 12:50 | |
| *** lpetrut has joined #openstack-cinder | 12:53 | |
| *** ebalduf has joined #openstack-cinder | 12:55 | |
| *** annegent_ has joined #openstack-cinder | 12:58 | |
| *** ebalduf has quit IRC | 13:00 | |
| *** EmilienM|afk is now known as EmilienM | 13:03 | |
| *** annegent_ has quit IRC | 13:03 | |
| *** alecv has quit IRC | 13:12 | |
| *** timcl1 has quit IRC | 13:12 | |
| *** markvoelker has joined #openstack-cinder | 13:20 | |
| aarefiev | hi, could somebody from cores look at patch https://review.openstack.org/#/c/145780/. Everyday rebasing is a little tired:), thanks. | 13:22 |
| *** cdelatte has joined #openstack-cinder | 13:23 | |
| *** tbarron has joined #openstack-cinder | 13:23 | |
| *** akerr_away is now known as akerr | 13:24 | |
| *** Miouge has quit IRC | 13:25 | |
| *** markvoelker has quit IRC | 13:25 | |
| *** rcallawa has quit IRC | 13:27 | |
| *** akerr has quit IRC | 13:28 | |
| *** sandywalsh has joined #openstack-cinder | 13:28 | |
| *** dulek has joined #openstack-cinder | 13:29 | |
| *** Miouge has joined #openstack-cinder | 13:32 | |
| openstackgerrit | Wang Peter proposed openstack/cinder: CG Modification Support in EMC VNX Cinder Driver https://review.openstack.org/154784 | 13:35 |
| *** bswartz has quit IRC | 13:44 | |
| *** mtanino has joined #openstack-cinder | 13:45 | |
| *** nellysmitt has quit IRC | 13:48 | |
| *** nellysmitt has joined #openstack-cinder | 13:49 | |
| *** nellysmitt has quit IRC | 13:53 | |
| *** ebalduf has joined #openstack-cinder | 13:56 | |
| *** bill_az has joined #openstack-cinder | 13:57 | |
| *** scottda has joined #openstack-cinder | 13:59 | |
| *** eharney has joined #openstack-cinder | 14:00 | |
| *** ebalduf has quit IRC | 14:01 | |
| *** mtanino has quit IRC | 14:03 | |
| *** eharney has quit IRC | 14:04 | |
| *** Miouge has quit IRC | 14:07 | |
| *** tbarron has quit IRC | 14:07 | |
| *** eharney has joined #openstack-cinder | 14:08 | |
| *** lpetrut has quit IRC | 14:08 | |
| *** anuragpalsule has quit IRC | 14:10 | |
| *** timcl has joined #openstack-cinder | 14:10 | |
| *** bswartz has joined #openstack-cinder | 14:10 | |
| *** kaufer has joined #openstack-cinder | 14:13 | |
| *** alecv has joined #openstack-cinder | 14:13 | |
| *** dulek has quit IRC | 14:15 | |
| *** kallebe has joined #openstack-cinder | 14:17 | |
| *** annegent_ has joined #openstack-cinder | 14:17 | |
| *** TobiasE has joined #openstack-cinder | 14:17 | |
| *** markvoelker has joined #openstack-cinder | 14:21 | |
| *** Yogi1 has joined #openstack-cinder | 14:22 | |
| *** lpetrut has joined #openstack-cinder | 14:22 | |
| *** Miouge has joined #openstack-cinder | 14:22 | |
| *** nshaikh has quit IRC | 14:24 | |
| kallebe | Hello. Does anyone here have experience using SAN generic driver? I tried to find some tutorial to configure cinder in order to use it but did not find anything | 14:25 |
| *** markvoelker has quit IRC | 14:26 | |
| *** Miouge has quit IRC | 14:29 | |
| *** pdeore_ has quit IRC | 14:29 | |
| *** jungleboyj has quit IRC | 14:31 | |
| *** kvidvans has joined #openstack-cinder | 14:34 | |
| *** timcl has quit IRC | 14:34 | |
| *** nlevinki has joined #openstack-cinder | 14:35 | |
| *** Mandell has joined #openstack-cinder | 14:35 | |
| *** Miouge has joined #openstack-cinder | 14:41 | |
| *** BharatK has quit IRC | 14:41 | |
| *** annegent_ has quit IRC | 14:42 | |
| *** david-lyle_afk is now known as david-lyle | 14:48 | |
| openstackgerrit | nikeshmahalka proposed openstack/cinder: Add CHAP persistence to SCST target helper https://review.openstack.org/157356 | 14:49 |
| *** esker has quit IRC | 14:50 | |
| *** Ilja has joined #openstack-cinder | 14:51 | |
| *** nlevinki has quit IRC | 14:52 | |
| *** markvoelker has joined #openstack-cinder | 14:52 | |
| *** markvoelker has quit IRC | 14:53 | |
| *** Manshoon has joined #openstack-cinder | 14:54 | |
| *** Mandell has quit IRC | 14:56 | |
| *** tsekiyama has joined #openstack-cinder | 14:56 | |
| *** ebalduf has joined #openstack-cinder | 14:57 | |
| *** deepakcs has joined #openstack-cinder | 14:59 | |
| *** Mandell has joined #openstack-cinder | 15:00 | |
| *** ebalduf has quit IRC | 15:01 | |
| *** Yogi1 has quit IRC | 15:02 | |
| *** alexpilotti has joined #openstack-cinder | 15:02 | |
| *** BharatK has joined #openstack-cinder | 15:03 | |
| *** xyang1 has joined #openstack-cinder | 15:05 | |
| *** timcl has joined #openstack-cinder | 15:06 | |
| *** jungleboyj has joined #openstack-cinder | 15:07 | |
| DuncanT | aarefiev: It is another of the cleanup type patches that doesn't actually fix any bugs or implement any new features, and will cause a rebase hit for just about every open review due to the number of files it touches. I'm really not seeing the big win of merging it | 15:07 |
| *** Yogi1 has joined #openstack-cinder | 15:07 | |
| *** nkrinner has quit IRC | 15:08 | |
| *** openstackgerrit has quit IRC | 15:08 | |
| *** alexpilotti has quit IRC | 15:08 | |
| *** openstackgerrit has joined #openstack-cinder | 15:08 | |
| DuncanT | aarefiev: Most of the code looks less readable to me after your patch, not more | 15:08 |
| DuncanT | aarefiev: I'm not saying you're definitely wrong to want this change, I'm saying convince me as to the benefit | 15:09 |
| *** rushil has joined #openstack-cinder | 15:10 | |
| jungleboyj | xyang: Morning. | 15:10 |
| *** amoturi has quit IRC | 15:11 | |
| *** amoturi has joined #openstack-cinder | 15:12 | |
| *** thangp has joined #openstack-cinder | 15:12 | |
| *** BharatK has quit IRC | 15:13 | |
| *** primechuck has joined #openstack-cinder | 15:14 | |
| xyang | jungleboyj: morning | 15:15 |
| jungleboyj | xyang: Hey, quick question, are there any plans for a vplex cinder driver? | 15:16 |
| xyang | jungleboyj: you are the 2nd person asking me that today:). no plan | 15:16 |
| xyang | jungleboyj: you can use ViPR driver for that | 15:17 |
| jungleboyj | xyang: Interesting. Who was the other person. | 15:17 |
| smcginnis | DuncanT: in aarefiev's defence, I think this does need to happen at some time. Not sure if there really is a good time for it. | 15:17 |
| xyang | jungleboyj: haiming | 15:17 |
| smcginnis | DuncanT: We do have code reviews that get -1's because of not following this. | 15:17 |
| smcginnis | DuncanT: This would clean it up and enforce it so it's explicit. | 15:17 |
| smcginnis | DuncanT: Just my 2 cents (or pence?) so I'll be quiet now. :) | 15:18 |
| *** Mandell has quit IRC | 15:18 | |
| *** tbarron has joined #openstack-cinder | 15:20 | |
| openstackgerrit | nikeshmahalka proposed openstack/cinder: Use Unique SCST Group names in SCST cinder helper driver. https://review.openstack.org/159156 | 15:23 |
| *** mriedem has joined #openstack-cinder | 15:24 | |
| *** sigmavirus24_awa is now known as sigmavirus24 | 15:24 | |
| *** dustins has joined #openstack-cinder | 15:25 | |
| *** BharatK has joined #openstack-cinder | 15:25 | |
| *** annegent_ has joined #openstack-cinder | 15:28 | |
| *** mriedem has quit IRC | 15:29 | |
| *** kallebe has quit IRC | 15:30 | |
| *** esker has joined #openstack-cinder | 15:35 | |
| *** mriedem has joined #openstack-cinder | 15:37 | |
| *** hemnafk is now known as hemna | 15:38 | |
| *** mtanino has joined #openstack-cinder | 15:38 | |
| *** BharatK has quit IRC | 15:38 | |
| jungleboyj | xyang: *Sigh* Ok, same person asking me. | 15:38 |
| DuncanT | smcginnis: Yeah, I know it gets -1s at review time sometimes... should we just tell people to not do that instead? Is the code really better if you only import modules? Looking at some of the code after that patch is applied, it looks less readable to me, not more, but I'm quite willing to hear arguements | 15:38 |
| jungleboyj | xyang: Thanks for the info. | 15:39 |
| smcginnis | DuncanT: I think we can tell people not to do that, but it still would end up happening. | 15:39 |
| smcginnis | DuncanT: This would enable the pep8 validation to make sure they get that feedback right away. | 15:40 |
| *** Manshoon has quit IRC | 15:40 | |
| DuncanT | smcginnis: But the new version of the code looks worse to me, not better | 15:40 |
| smcginnis | DuncanT: I somewhat agree with that statement. | 15:41 |
| DuncanT | smcginnis: Having an arbitrary rule that makes the code less readable seems really, really dumb | 15:41 |
| *** Manshoon has joined #openstack-cinder | 15:41 | |
| smcginnis | DuncanT: Personally I don't have a problem with importing classes instead of modules. | 15:41 |
| DuncanT | smcginnis: I'd rather get all reference to the rule scrubbed from the docs and educate reviewers than to have worse code | 15:41 |
| smcginnis | DuncanT: But that seems to be something others are enforcing, so I do agree it would be good for consistency. | 15:41 |
| smcginnis | DuncanT: +1 | 15:41 |
| smcginnis | DuncanT: Good luck with that. :) | 15:42 |
| DuncanT | smcginnis: I'll start a mailing list thread on the subject, see if anybody has any real justification for the rule | 15:42 |
| smcginnis | DuncanT: Oh, good idea. It would be good to open that up to some discussion to get some view points. | 15:42 |
| smcginnis | DuncanT: I just think if we are going to go with that, then we should enforce it. | 15:42 |
| xyang | jungleboyj: oh, ok:) He's going to compare the two answers then | 15:42 |
| DuncanT | smcginnis: I'll start a thread, see what happens | 15:43 |
| *** tsekiyama has quit IRC | 15:43 | |
| smcginnis | DuncanT: Thanks! | 15:44 |
| *** emagana has joined #openstack-cinder | 15:45 | |
| openstackgerrit | Walter A. Boring IV (hemna) proposed openstack/cinder: WIP: Remove Brick from cinder codebase https://review.openstack.org/155552 | 15:45 |
| jungleboyj | xyang: Ok. Thanks for answering. | 15:45 |
| *** tsekiyam_ has joined #openstack-cinder | 15:46 | |
| *** rushiagr is now known as rushiagr_away | 15:46 | |
| *** rmesta has joined #openstack-cinder | 15:47 | |
| *** thingee has joined #openstack-cinder | 15:48 | |
| *** IanGovett1 has joined #openstack-cinder | 15:49 | |
| *** nellysmitt has joined #openstack-cinder | 15:49 | |
| *** BharatK has joined #openstack-cinder | 15:51 | |
| *** IanGovett has quit IRC | 15:51 | |
| *** Lee1092 has joined #openstack-cinder | 15:52 | |
| *** dulek has joined #openstack-cinder | 15:54 | |
| *** nellysmitt has quit IRC | 15:54 | |
| *** dulek has quit IRC | 15:54 | |
| *** dulek has joined #openstack-cinder | 15:54 | |
| flip214 | in 3 mins: /join #openstack-meeting | 15:56 |
| *** MasterPiece has joined #openstack-cinder | 15:56 | |
| *** mdenny has joined #openstack-cinder | 15:56 | |
| *** ronis has quit IRC | 15:57 | |
| *** kallebe has joined #openstack-cinder | 15:57 | |
| *** ebalduf has joined #openstack-cinder | 15:57 | |
| *** TobiasE has quit IRC | 15:58 | |
| flip214 | /join #openstack-meeting | 15:59 |
| *** TobiasE has joined #openstack-cinder | 15:59 | |
| *** asselin has joined #openstack-cinder | 16:00 | |
| *** TobiasE has quit IRC | 16:01 | |
| openstackgerrit | Shay Halsband proposed openstack/cinder: XtreamIO version 4.0 support https://review.openstack.org/152973 | 16:01 |
| aarefiev | smcginnis: thanks for your opinion DuncanT: Why do think it makes code less readable, I think it become more clear? Also as Sean mentioned some developer don't follow style guidelines | 16:01 |
| *** kmartin has joined #openstack-cinder | 16:01 | |
| *** ebalduf has quit IRC | 16:02 | |
| aarefiev | I think it's better to automate it | 16:02 |
| e0ne | aarefiev, DuncanT: imo, it makes code less readable only if we got a_very_long_modules_names | 16:03 |
| *** lpetrut has quit IRC | 16:04 | |
| *** annegent_ has quit IRC | 16:04 | |
| *** anshul has quit IRC | 16:07 | |
| DuncanT | aarefiev: You're very much right that if we're going to have the rule then it should be automated. I'm just not convinced it is a good rule. I've started a post for the mailing list, with examples, so if you don't mind I'll point you at that once it is up (should be straight after the meeting) | 16:08 |
| *** cnesa has joined #openstack-cinder | 16:08 | |
| aarefiev | DuncanT: ok | 16:09 |
| openstackgerrit | Sean McGinnis proposed openstack/cinder: WIP: Dell Storage Center: Address GetList failure https://review.openstack.org/159171 | 16:11 |
| *** annegent_ has joined #openstack-cinder | 16:12 | |
| *** annegent_ has quit IRC | 16:12 | |
| sigmavirus24 | thingee++ on looking to the API-WG for guidance in earnest | 16:14 |
| *** annegent_ has joined #openstack-cinder | 16:14 | |
| *** dannywilson has joined #openstack-cinder | 16:16 | |
| *** jcru has joined #openstack-cinder | 16:17 | |
| *** jcru has quit IRC | 16:17 | |
| *** jcru has joined #openstack-cinder | 16:18 | |
| *** Mandell has joined #openstack-cinder | 16:18 | |
| openstackgerrit | Mikhail Khodos proposed openstack/cinder: Nexenta Edge iSCSI backend driver. https://review.openstack.org/158745 | 16:24 |
| *** dannywilson has quit IRC | 16:26 | |
| *** dannywilson has joined #openstack-cinder | 16:27 | |
| *** coolsvap is now known as coolsvap_ | 16:28 | |
| *** akerr has joined #openstack-cinder | 16:30 | |
| *** rwsu-afk has quit IRC | 16:31 | |
| *** cpallares has joined #openstack-cinder | 16:32 | |
| *** rushiagr_away is now known as rushiagr | 16:33 | |
| jaypipes | DuncanT: lol "The client calls to do multi-attach won't allow multiattach." | 16:36 |
| jaypipes | DuncanT: I don't think that came across quite as you intended :) | 16:36 |
| DuncanT | jaypipes: Possibly not. 'It won't work. Not even a little bit' covered it nicely though, I thought | 16:38 |
| *** jdurgin1 has joined #openstack-cinder | 16:38 | |
| *** jdurgin1 has quit IRC | 16:38 | |
| jaypipes | DuncanT: :) so, my point was that you can't say "we cannot do that cinderclient changes because Nova doesn't support it yet" and at the same time, Nova *needs* the cinderclient changes in order to support multi-attach. :( | 16:40 |
| DuncanT | jaypipes: Nova is definitely not going to do those changes until L | 16:41 |
| jaypipes | DuncanT: yes, agreed. | 16:41 |
| jaypipes | DuncanT: but that is no reason to prevent the Cinder and cinderclient changes from going in. | 16:41 |
| *** openstackstatus has joined #openstack-cinder | 16:42 | |
| *** ChanServ sets mode: +v openstackstatus | 16:42 | |
| openstackgerrit | Sean McGinnis proposed openstack/cinder: WIP: Dell Storage Center: Address GetList failure https://review.openstack.org/159171 | 16:43 |
| *** Mandell has quit IRC | 16:44 | |
| anteaya | thingee: I do have something for open discussion if we get there | 16:46 |
| DuncanT | jaypipes: yes, it is | 16:48 |
| jaypipes | DuncanT: why? | 16:48 |
| DuncanT | jaypipes: Having a feature that doesn't work, because the nova changes aren't merged, in the client, for a release, means people pop up on this channel, oon the mailing list, on ask.openstack, etc, asking how they get multiattach to work | 16:48 |
| jaypipes | DuncanT: so delaying all of this another X months is the solution? :( | 16:49 |
| jaypipes | DuncanT: and, FYI, it's not like the feature itself doesn't work on the Cinder side... it's just the integration of the feature support in Nova isn't ready yet. | 16:50 |
| DuncanT | jaypipes: Put it in the client as soon as the K release is out of the door | 16:50 |
| DuncanT | jaypipes: Then people who just consume the release don't see it | 16:50 |
| jaypipes | meh, ok | 16:50 |
| DuncanT | jaypipes: Which reduces support load, a lot | 16:51 |
| DuncanT | jaypipes: People who take point releases tend to be more clueful | 16:51 |
| jaypipes | DuncanT: not really :) instead, you just have a shit-ton of people battering down *my* door asking why multi-attach isn't in Kilo. :) | 16:51 |
| DuncanT | jaypipes: We went through this pain point with replication that only worked with one backend... | 16:51 |
| DuncanT | jaypipes: 'Because it doesn't work because nova didn't take the patch' | 16:51 |
| *** rwsu has joined #openstack-cinder | 16:52 | |
| jaypipes | DuncanT: how can nova take a patch without the support in cinder? | 16:52 |
| DuncanT | jaypipes: If we release the client, you still can't attach the volume to two VMs | 16:52 |
| jaypipes | just seems to me delaying the feature for now particular reason. | 16:53 |
| DuncanT | jaypipes: If nova had any plans over even maybe merging it, I'd be happy to take the support hit, but they aren't. Merge the client support as the first patch in L | 16:54 |
| DuncanT | jaypipes: It is delaying nothing. Nova are definitely not, no way no how, not going to happen, take the patch in Kilo | 16:54 |
| DuncanT | jaypipes: Given we know that, releasing client support for a feature that you can't use, definitely in no way not even a little bit can you use it, is dumb | 16:55 |
| hemna | DuncanT, FWIW, this is one of the main reasons why multiattach hasn't landed in the previous 3 attempts | 16:56 |
| jaypipes | yes, that ^^ | 16:56 |
| hemna | due to 1) Nova's very limited time window for landing features combined with 2) Cinder's requirement that nova has to use a cinder feature before cinder lands it. chicken/egg | 16:56 |
| DuncanT | hemna: You saw my comment on the server side patch... I'm happy to land that, though I'd rather the API rejected multiattach=True unless you flip a config option | 16:56 |
| hemna | so cinder will forever be not allowed to land features that nova wants | 16:57 |
| jaypipes | and with this pushed yet another release, hemna likely won't be able to work on it furtther. | 16:57 |
| DuncanT | hemna: Put the patch to get rid of that limitation as soon as the release is out of the way | 16:57 |
| thingee | jaypipes: are you testing this patch as well? | 16:57 |
| hemna | thingee, that's part of the issue | 16:57 |
| *** annegent_ has quit IRC | 16:57 | |
| hemna | the tempest tests fail, because of the cross project deps | 16:58 |
| DuncanT | jaypipes: Anything that claims that openstack supports multiattach in kilo, even with all the cinder stuff landed, is a lie | 16:58 |
| hemna | cinder has to land first for any of the nova side/tempest tests can possibly pass. | 16:58 |
| thingee | hemna: you can test something in your own lab | 16:58 |
| DuncanT | hemna: I though we could stack reviews now? | 16:58 |
| hemna | thingee, well sure. I have been. | 16:58 |
| hemna | but any patches in nova in gerrit will always fail | 16:58 |
| thingee | hemna: well that was my question to jaypipes | 16:58 |
| *** ebalduf has joined #openstack-cinder | 16:58 | |
| hemna | unless cinder lands | 16:58 |
| hemna | chicken/egg | 16:58 |
| jaypipes | right. | 16:58 |
| hemna | DuncanT, we can do cross project deps | 16:59 |
| DuncanT | Land it as athe first thing in L. It won't get the nova part landed any slower! | 16:59 |
| hemna | but that doesn't install the new client. | 16:59 |
| jaypipes | but now that (yet again) we arepast the nova feature freeze, we are going to kill this feature again. | 16:59 |
| DuncanT | (The client part) | 16:59 |
| hemna | the deps don't work like that unfortunatel. | 16:59 |
| DuncanT | The server part I'm less bothered by, though I'd far rather the API gave a sensible error for a create with multi-attach=True | 16:59 |
| hemna | DuncanT, so don't land the client changes ? | 17:00 |
| hemna | and update the cinder patch to report fail on seeing multiattach flag ? | 17:00 |
| DuncanT | Actually Hemna, can we key the API response off the volume type, and require multi-attach to have it noted in the type? At least then normal deployments will get a 'not supported' response | 17:00 |
| hemna | ugh | 17:01 |
| DuncanT | hemna: Then even if they use the new client, they still get a good error | 17:01 |
| DuncanT | hemna: We probably need that anyway, right, since many drivers won't do multi-attach? | 17:01 |
| DuncanT | s/many/some maybe | 17:01 |
| hemna | well the create should always work | 17:01 |
| DuncanT | hemna: Create of a multi-attach=true volume in a type that doesn't support it? I think that sould always fail | 17:02 |
| hemna | if we want to filter on the multiattach capability, then drivers need time to add the capability reporting | 17:02 |
| hemna | again delaying the cinder side from landing. | 17:02 |
| hemna | anyway, I don't know what to do. I've tried 3 releases now. | 17:02 |
| * hemna is frustrated | 17:02 | |
| DuncanT | hemna: We need some way of stopping multi-attach requests going to backends that don't support it | 17:02 |
| *** ebalduf has quit IRC | 17:02 | |
| thingee | I was already ignored on this, but I will repeat again, I'm not going to approve this if it's not tested. hemna I appreciate you testing your work, but if there are interested parties, I hope they brought all deps in their own test area to play around with it. | 17:03 |
| hemna | DuncanT, drivers need to report the capability | 17:03 |
| hemna | and the scheduler honors it. | 17:03 |
| *** asselin has quit IRC | 17:03 | |
| thingee | I just need a simple sign off from people I trust. | 17:03 |
| *** pwehrle has quit IRC | 17:03 | |
| hemna | thingee, yah that's fine. it should get tested and gated | 17:03 |
| hemna | but the gating part is a chicken/egg issue right now. | 17:04 |
| thingee | I don't need gated right now. just "I played with this, cinder didn't break, it's good!" | 17:04 |
| DuncanT | hemna: Right, so if we check the type has some multi-attach flag in it, and if it rejects the create if the type doesn't have that flag, what is the harm? | 17:04 |
| DuncanT | hemna: As drivers test multi-attach, then they can add the capability\ | 17:04 |
| hemna | well the flag is on the volume itself, not in the volume type. | 17:04 |
| openstackgerrit | Bob Callaway proposed openstack/cinder: Add Manage/Unmanage support to NetApp NFS drivers https://review.openstack.org/158864 | 17:04 |
| DuncanT | hemna: We just add it to LVM for the gate & for people to play with | 17:04 |
| hemna | winston-d, can the scheduler look at the volume multiattach flag to compare against reported capabilities? | 17:05 |
| DuncanT | hemna: Means the API gives a sensible error unless the admin turned it on, which solves my client complaints | 17:05 |
| hemna | I can modify every driver to report multiattach=False | 17:06 |
| *** Yogi1 has quit IRC | 17:06 | |
| hemna | and/or modify our drivers and lvm to report true. | 17:06 |
| *** atan8 has joined #openstack-cinder | 17:06 | |
| hemna | but the scheduler needs to be able to filter on the capability vs. volume attribute | 17:07 |
| thingee | DuncanT, dulek, winston-d: can we continue to the scheduler discussion after the multiattach stuff here? | 17:07 |
| dulek | thingee: sure | 17:07 |
| hemna | that part I'm unsure of | 17:07 |
| DuncanT | thingee: Sure | 17:07 |
| *** markvoelker has joined #openstack-cinder | 17:07 | |
| dulek | thingee: How long will the current one last? I wonder if I can get 30 minutes to drive home. | 17:07 |
| DuncanT | hemna: Would require an extra filter in the scheduler I think, but that isn't hard AFIACT | 17:08 |
| winston-d | hemna: yes, that's doable | 17:08 |
| thingee | dulek: not sure. I need to leave soon myself. | 17:08 |
| winston-d | 'scheduler look at multi-attach flag' | 17:08 |
| DuncanT | hemna: I can try to knock up a patch to your patch now, if you want? | 17:08 |
| dulek | thingee: ok, I'll stay then | 17:09 |
| hemna | DuncanT, that would be great. | 17:09 |
| hemna | winston-d, is it a new filter ? | 17:09 |
| *** atan8 has quit IRC | 17:09 | |
| hemna | I guess that'd have to be enabled by default | 17:09 |
| winston-d | hemna: no, doesn't have to be. just pass that as a 'scheduler_hint' | 17:09 |
| *** _cjones_ has joined #openstack-cinder | 17:09 | |
| hemna | winston-d, ok that's new to me. | 17:10 |
| hemna | that can be done automatically in the API -> scheduler message ? | 17:10 |
| winston-d | hemna: 'cinder create --scheduler_hint 'multi-attach_support=true', sth like that | 17:10 |
| hemna | gah | 17:10 |
| hemna | we already have a multiattach flag on the volume itself. | 17:10 |
| hemna | can I inject the scheduler hint automatically? | 17:11 |
| hemna | if the flag is on | 17:11 |
| winston-d | hemna: ah, ok. yes, you can do that in API. | 17:11 |
| hemna | ok that would be better than adding a new filter | 17:11 |
| hemna | I can research how to do that. | 17:12 |
| thingee | DuncanT, dulek, winston-d : if you're fine with continuing discussions tomorrow, we can do that too. | 17:12 |
| hemna | so the follow up, is the drivers aren't reporting up the capability now | 17:12 |
| hemna | so does that mean it defaults to false ? | 17:12 |
| hemna | and then the create will fail? | 17:12 |
| thingee | or...maybe multi-attach is just about done :) | 17:12 |
| hemna | until drivers are updated to report the capability as true? | 17:12 |
| winston-d | scheduler hint will eventually be part of 'filter_properties'. | 17:12 |
| winston-d | hemna: you just need to inject a dict to 'filter_properties' and done | 17:13 |
| dulek | thingee, DuncanT, winston-d: I would prefer to do it today as I'm already staying late. ;) | 17:13 |
| thingee | dulek: ok | 17:13 |
| DuncanT | hemna: It should default to false and the attach fail if the driver doesn't advertise, yes | 17:13 |
| dulek | thingee, DuncanT, winston-d: Maybe we can meet in one hour? | 17:13 |
| DuncanT | dulek: that's too late for me, sorry, it is already 7:15 pm | 17:14 |
| jgriffith | thingee: yo... what'up | 17:14 |
| hemna | winston-d, trying to find where we are adding filter_properties in the api | 17:14 |
| thingee | dulek: I need to disappear for a little bit. might not be around. You're all welcome to carry on discussing without me though | 17:14 |
| winston-d | hemna: actuall, put it into request_specs | 17:15 |
| DuncanT | dulek: I vote for the spin/first hit approach and don't care about fairness | 17:15 |
| hemna | winston-d, can you send me a gist snippet ? | 17:15 |
| winston-d | DuncanT: too late for me too, it's laready 1:15 am | 17:15 |
| *** Manshoon has quit IRC | 17:15 | |
| winston-d | dulek: ^^ | 17:15 |
| DuncanT | winston-d: You're insane ;-) | 17:15 |
| winston-d | hemna: sure, few secs. | 17:15 |
| hemna | winston-d, thank you. | 17:15 |
| dulek | DuncanT, winston-d, thingee: Cool, let's continue tomorrow morning! | 17:15 |
| winston-d | dulek: which timezone r u in? | 17:16 |
| *** diemt has joined #openstack-cinder | 17:16 | |
| dulek | UTC+1 | 17:16 |
| DuncanT | dulek: Please. I don't actually see anybody complaining about spin/first hit though | 17:16 |
| dulek | And thingee will have the problem solved before his morning | 17:16 |
| anteaya | so what I was thinking was that I could work with about 5 or so people to help them understand how to trace back the jobs listed on cinder repos in zuul/layout.yaml with the agreement that the folks I teach will share with others | 17:16 |
| dulek | DuncanT: Hm, you're right. So you suggest to just propose the patch with your idea and gather reviews? | 17:17 |
| anteaya | then when I ask in cinder, who knows about this job or proposed job, folks will say yes, I know about it | 17:17 |
| anteaya | more welcome but I don't see how I can get everyone together at teh same time | 17:17 |
| *** Manshoon has joined #openstack-cinder | 17:17 | |
| *** Apoorva has joined #openstack-cinder | 17:17 | |
| *** annegent_ has joined #openstack-cinder | 17:18 | |
| *** Manshoon has quit IRC | 17:18 | |
| cebruns | DuncanT: +1 - I like the idea of finding "good enough" in this situation too. | 17:18 |
| anteaya | so we could work in here or -dev or a meeting channel | 17:18 |
| anteaya | whatever folks want as long as they have scroll back to share with others | 17:18 |
| *** Manshoon has joined #openstack-cinder | 17:18 | |
| *** annegent_ has quit IRC | 17:18 | |
| dulek | thingee, DuncanT, winston-d: Okay, so I'll try to come up with a patch tomorrow. | 17:19 |
| anteaya | does anyone have an opinion? | 17:20 |
| *** annegent_ has joined #openstack-cinder | 17:21 | |
| winston-d | dulek: thx, looking forward to it. | 17:21 |
| *** asselin has joined #openstack-cinder | 17:21 | |
| *** ronis has joined #openstack-cinder | 17:21 | |
| *** jamielennox is now known as jamielennox|away | 17:22 | |
| winston-d | hemna: https://github.com/openstack/cinder/blob/master/cinder/volume/flows/api/create_volume.py#L47-L448 this is the part, where cinder vol API extracts information that needs to a create request. | 17:22 |
| *** jistr has quit IRC | 17:22 | |
| hemna | winston-d, looking for the scheduler_hints | 17:23 |
| anteaya | because I have to admit I do have an expectation when a patch is up for a new job that more than one cinder person knows what is going on with the proposed job | 17:23 |
| winston-d | hemna: for example: this is how 'availablity_zone' is extracted from the original create request. | 17:23 |
| anteaya | so if I have that expectation I am willing to do the work to fill in the gaps to meet that expectation | 17:23 |
| winston-d | hemna: https://github.com/openstack/cinder/blob/master/cinder/volume/flows/api/create_volume.py#L399-L401 | 17:23 |
| hemna | winston-d, but isn't that too late? | 17:24 |
| *** nellysmitt has joined #openstack-cinder | 17:24 | |
| hemna | this is after the scheduler has picked the host | 17:24 |
| hemna | oh this is a taskflow on the API side ? | 17:24 |
| winston-d | hemna: nono, this is before the request being sent out to scheduler | 17:24 |
| hemna | ok | 17:25 |
| *** Manshoon_ has joined #openstack-cinder | 17:25 | |
| winston-d | hemna: that's how the request being sent to scheduler is constructed/composed | 17:25 |
| *** Manshoon_ has quit IRC | 17:25 | |
| hemna | so add a new multiattach: (True/False) to the return of execute? | 17:25 |
| *** dulek has quit IRC | 17:25 | |
| hemna | still unsure if that's sufficient | 17:25 |
| hemna | as I don't see any hints in there | 17:26 |
| *** Manshoon_ has joined #openstack-cinder | 17:26 | |
| winston-d | hemna: you can add a static method in this ExtractVolumeRequestTask, say, _extract_multiattach_flag, from vol metadata, and put it into request spec | 17:26 |
| winston-d | hemna: scheduler_hint was extracted even earlier, at C API | 17:27 |
| winston-d | hemna: I suggest do multiattach thing heree, scheduler hint is a bit differnt, it has it's own API extension. | 17:27 |
| *** Yogi1 has joined #openstack-cinder | 17:28 | |
| *** Apoorva_ has joined #openstack-cinder | 17:28 | |
| hemna | so I'm looking in there to see where the request_spec is used | 17:28 |
| *** Manshoon has quit IRC | 17:28 | |
| hemna | the VolumeCastTask pulls values from a request_spec dict. | 17:29 |
| hemna | *sigh* taskflow | 17:29 |
| *** rwsu has quit IRC | 17:29 | |
| *** jamielennox|away is now known as jamielennox | 17:30 | |
| *** Apoorva has quit IRC | 17:31 | |
| hemna | I don't see any request_specs being used or returned in the ExtractVolumeRequestTask | 17:32 |
| hemna | winston-d, the scheduler_hints are set here: https://github.com/openstack/cinder/blob/master/cinder/volume/flows/api/create_volume.py#L747 | 17:33 |
| hemna | can't I just inject scheduler_hints['multiattach'] = True/False there ? | 17:34 |
| *** annegent_ has quit IRC | 17:34 | |
| *** rwsu has joined #openstack-cinder | 17:34 | |
| *** annegent_ has joined #openstack-cinder | 17:35 | |
| winston-d | hemna: my understanding is every task's output is the next task's input, so what ExtractVolumeRequestTask returns is consumed by EntryCreateTask. | 17:35 |
| hemna | I see the that VolumeCastTask requires scheduler_hints | 17:35 |
| hemna | the scheduler_hints must be set somewhere else | 17:37 |
| e0ne | winston-d: +1 | 17:37 |
| *** Ilja has quit IRC | 17:37 | |
| winston-d | hemna: problem with scheduler_hints is it only be consumed in filter, if you put it in there, you will have to add a new Multiattach filter | 17:38 |
| winston-d | hemna: which i guess you don't | 17:38 |
| hemna | ok | 17:38 |
| hemna | that's what I was trying to avoid | 17:38 |
| hemna | as the filter can be removed at deployment time no? | 17:39 |
| hemna | I need to inject something to test against a capability | 17:39 |
| hemna | which is what I thought the scheduler_hints did | 17:39 |
| hemna | ? | 17:39 |
| hemna | no? | 17:39 |
| *** dannywilson has quit IRC | 17:39 | |
| winston-d | hemna: scheduler_hint relies one an API extension to work, which can also be turned off in a deployment. | 17:40 |
| hemna | sure | 17:40 |
| *** annegent_ has quit IRC | 17:40 | |
| hemna | but we can also inject it in the API flow if it doesn't exist | 17:41 |
| *** Manshoon_ has quit IRC | 17:41 | |
| hemna | I guess I'm trying to understand how I can get the volume multiattach flag to be compared to a capability reporting by the drivers | 17:41 |
| winston-d | hemna: ok, let me try again. | 17:42 |
| hemna | winston-d, so fwiw, here is where the multiattach flag is set https://review.openstack.org/#/c/85847/35/cinder/api/v2/volumes.py | 17:43 |
| *** annegent_ has joined #openstack-cinder | 17:43 | |
| openstackgerrit | Steven Kaufer proposed openstack/python-cinderclient: Adopt CLI sorting argument guidelines https://review.openstack.org/145913 | 17:43 |
| *** leeantho has joined #openstack-cinder | 17:44 | |
| *** btran has joined #openstack-cinder | 17:46 | |
| winston-d | hemna: ahha, you already deal with that flag in API create flow: https://review.openstack.org/#/c/85847/35/cinder/volume/flows/api/create_volume.py | 17:46 |
| *** jordanP has quit IRC | 17:46 | |
| winston-d | hemna: which is good | 17:47 |
| hemna | yah, but can the scheduler compare that against capabilities ? | 17:47 |
| hemna | I think that's just passed through to the manager to set the flag in the DB, currently | 17:48 |
| *** EmilienM is now known as EmilienM|afk | 17:49 | |
| winston-d | hemna: not yet, you forgot the modify the ExtractVolumeReuquestTask to let it return what EntryCreateTask need (the additional multiattach flag). | 17:49 |
| *** leeantho has quit IRC | 17:49 | |
| hemna | winston-d, it's working though | 17:49 |
| winston-d | hemna: hmm | 17:50 |
| hemna | that flag is getting set correctly and the volume manager sees it | 17:50 |
| winston-d | hemna: i'm new to taskflow. :) | 17:50 |
| hemna | yah, same here | 17:50 |
| hemna | it's hard to follow | 17:50 |
| *** karimb has quit IRC | 17:51 | |
| *** leeantho has joined #openstack-cinder | 17:53 | |
| winston-d | hemna: sorry, i need some time to check the code to make sure I don't mislead you. | 17:55 |
| hemna | winston-d, ok np | 17:55 |
| hemna | winston-d, chew on it while you sleep :) | 17:55 |
| winston-d | hemna: no, i'm pretty close now | 17:55 |
| *** kallebe has quit IRC | 17:56 | |
| *** jcru has quit IRC | 17:57 | |
| winston-d | hemna: first thing, CapablitiesFilter only looks at 'extra_specs' in request_spec['resource_type'], that comes from volume type extra specs: https://github.com/openstack/cinder/blob/master/cinder/openstack/common/scheduler/filters/capabilities_filter.py#L33 | 17:57 |
| hemna | yah | 17:58 |
| hemna | that's what I was afraid of | 17:58 |
| winston-d | hemna: so in order to make scheduler check multiattach flag against backend capabilities, we need to find a way to inject that into extra specs, which is doable | 17:58 |
| *** ebalduf has joined #openstack-cinder | 17:59 | |
| winston-d | hemna: 'resource_type' was added to 'filter_properties' here: https://github.com/openstack/cinder/blob/master/cinder/scheduler/filter_scheduler.py#L268 | 17:59 |
| winston-d | hemna: this can be the point to inject multiattach flag into extra specs | 18:00 |
| *** cnesa has quit IRC | 18:00 | |
| hemna | so that's just the volume type id no ? | 18:00 |
| *** annegent_ has quit IRC | 18:00 | |
| hemna | not the extra specs | 18:00 |
| hemna | huh | 18:01 |
| hemna | line 267 and 268 do the same thing? | 18:01 |
| hemna | volume_type and resource_type | 18:01 |
| winston-d | hemna: yeah, that's becasuse Nova uses 'instance_type' insetad of 'volume_type', so to be able to put CapFilter to oslo, I did some dupication here. | 18:02 |
| hemna | oh I see | 18:02 |
| hemna | got it | 18:02 |
| *** jamielennox is now known as jamielennox|away | 18:02 | |
| hemna | so if there is no volume_type specified that will be None | 18:03 |
| hemna | or it will have the volume_type uuid no ? | 18:03 |
| hemna | ick | 18:03 |
| *** kallebe has joined #openstack-cinder | 18:03 | |
| hemna | I'm guessing something else extracts extra_specs later on ? | 18:04 |
| *** ebalduf has quit IRC | 18:04 | |
| *** yuriy_n17 has quit IRC | 18:04 | |
| *** Mandell has joined #openstack-cinder | 18:04 | |
| hemna | winston-d, https://github.com/openstack/cinder/blob/master/cinder/scheduler/filter_scheduler.py#L340 | 18:05 |
| winston-d | hemna: volume type was added here: https://github.com/openstack/cinder/blob/master/cinder/volume/flows/api/create_volume.py#L410 | 18:05 |
| hemna | doesn't that look like it's injecting something into extra specs ? | 18:05 |
| hemna | xyang, | 18:05 |
| winston-d | hemna: yes, she did | 18:05 |
| hemna | so can't I just do the same thing there? | 18:06 |
| winston-d | hemna: yes, you can. :) | 18:06 |
| winston-d | hemna: glad you find that out youself. | 18:07 |
| hemna | resource_type['extra_specs'].update(multiattach='<is> True') | 18:07 |
| hemna | ? | 18:07 |
| hemna | if the multiattach flag is set to true on the volume of course | 18:07 |
| hemna | I guess I can try that | 18:08 |
| hemna | no driver is reporting that capability today | 18:08 |
| winston-d | hemna: you can do the simliar check like xyang has done, in case some admin put 'multiattach' into type extra sepcs already | 18:08 |
| hemna | so as soon as I add that, my create requests should fail with no host found right ? | 18:08 |
| hemna | yah | 18:08 |
| winston-d | hemna: yes | 18:08 |
| hemna | ok cool. | 18:08 |
| hemna | :) | 18:08 |
| hemna | winston-d, thanks for your help. | 18:08 |
| hemna | :) | 18:08 |
| winston-d | hemna: the reporting part can be a separate patch against vol manage and sch host manager | 18:09 |
| winston-d | hemna: let it be a 'standard'/manditory capablities that every backend shouuld report | 18:09 |
| hemna | which is just updates to drivers capabilityes right ? | 18:09 |
| hemna | winston-d, there is something that enforces those 'standard' capabilities ? | 18:10 |
| winston-d | hemna: yes, it can be like that, but since this is a generic enough capabilties, I suggeest adding it to host manager like these: https://github.com/openstack/cinder/blob/master/cinder/scheduler/host_manager.py#L100-L120 | 18:11 |
| *** annashen has joined #openstack-cinder | 18:11 | |
| hemna | ok | 18:12 |
| winston-d | hemna: we can take care of that part later. | 18:12 |
| hemna | ok cool :) | 18:12 |
| *** Anticimex has quit IRC | 18:12 | |
| hemna | so for now I'll update my existing patch to inject the extra spec | 18:12 |
| *** vilobhmm has joined #openstack-cinder | 18:12 | |
| *** vilobhmm1 has joined #openstack-cinder | 18:13 | |
| hemna | hopefully that'll make DuncanT happy | 18:13 |
| winston-d | hemna: yes, correct | 18:13 |
| winston-d | hemna: sorry I havn't reviewed your multiattach patch for quite some time | 18:13 |
| hemna | it's ok. it hasn't gotten much love from anyone in K. | 18:13 |
| *** deepakcs has quit IRC | 18:14 | |
| hemna | jdg looked at it a few times early on | 18:14 |
| *** deepakcs has joined #openstack-cinder | 18:14 | |
| *** _cjones_ has quit IRC | 18:15 | |
| winston-d | hemna: i have been occupied more than b4 in K, so short on reviews. sorry about that, hopefully can do better in L | 18:15 |
| *** jamielennox|away is now known as jamielennox | 18:15 | |
| hemna | heh work gets in the way :) | 18:15 |
| hemna | winston-d, ok I'm off to a meeting. thanks for the help! | 18:15 |
| winston-d | hemna: ok, let me know if you have any question/issues with scheduler, i'd be happy to help | 18:16 |
| hemna | winston-d, great. thanks man. you rock! | 18:16 |
| winston-d | hemna: take care, i'm getting off to get some sleep now. | 18:16 |
| *** winston-d is now known as winston-d_zZZ | 18:16 | |
| *** nikesh_vedams has joined #openstack-cinder | 18:16 | |
| nikesh_vedams | hi | 18:17 |
| *** vilobhmm has quit IRC | 18:17 | |
| nikesh_vedams | something wrong in review.openstack.org | 18:18 |
| nikesh_vedams | is server down | 18:18 |
| *** Anticimex has joined #openstack-cinder | 18:18 | |
| *** deepakcs has quit IRC | 18:18 | |
| nikesh_vedams | i was trying to see uploaded cinder patch | 18:19 |
| nikesh_vedams | but its not opening from half an hour | 18:19 |
| e0ne | nikesh_vedams: hi. it's working for me | 18:22 |
| *** aix has quit IRC | 18:25 | |
| *** patrickeast has joined #openstack-cinder | 18:26 | |
| nikesh_vedams | e0ne: hmm i m getting "The server at review.openstack.org is taking too long to respond." | 18:27 |
| nikesh_vedams | i m from india | 18:27 |
| bswartz | nikesh_vedams: are you using the right port number? | 18:28 |
| *** Miouge has quit IRC | 18:28 | |
| e0ne | nikesh_vedams: http://www.downforeveryoneorjustme.com/review.openstack.org | 18:28 |
| *** alecv has quit IRC | 18:28 | |
| e0ne | nikesh_vedams: maybe your provider blocks it | 18:28 |
| bswartz | Make sure you're SSHing to port 29418, and make sure there's not a firewall blocking that port | 18:29 |
| *** _cjones_ has joined #openstack-cinder | 18:29 | |
| bswartz | eharney: ping | 18:29 |
| eharney | bswartz: hi | 18:30 |
| kallebe | nikesh_vedams you can test if the port is available to you with command | 18:30 |
| kallebe | nc -zw3 review.openstack.org 29418 && echo "opened" || echo "closed" | 18:30 |
| *** Adriano_ has joined #openstack-cinder | 18:30 | |
| bswartz | kallebe: +1 | 18:32 |
| nikesh_vedams | kallebe: i am using windows | 18:32 |
| nikesh_vedams | kallebe : on my linux vm its giving opened | 18:33 |
| *** MasterPiece has quit IRC | 18:33 | |
| nikesh_vedams | kallebe :but my desktop is windows | 18:33 |
| kallebe | nikesh_vedams so it is open for you. Maybe the problem is not the port | 18:33 |
| bswartz | maybe windows is firewalling the outgoing connection? | 18:34 |
| bswartz | submit the patch from a linux box | 18:34 |
| *** MasterPiece has joined #openstack-cinder | 18:35 | |
| bswartz | nikesh_vedams: one other thing to check is to make sure that you've added you SSH pubkey to gerrit | 18:35 |
| *** rushiagr is now known as rushiagr_away | 18:35 | |
| bswartz | if you forget that step then git review won't work | 18:36 |
| kallebe | yes. you paste your ssh key here https://review.openstack.org/#/settings/ssh-keys | 18:36 |
| nikesh_vedams | bswartz : hmm i just wanted to see web version of review,i already submiteed patch | 18:37 |
| nikesh_vedams | ssh key added for linux | 18:37 |
| *** afazekas_ has quit IRC | 18:40 | |
| *** avishay has joined #openstack-cinder | 18:40 | |
| nikesh_vedams | nc -zw3 review.openstack.org 29418 && echo "opened" || echo "closed" | 18:41 |
| *** BharatK has quit IRC | 18:42 | |
| nikesh_vedams | i tried on another linux desktop ,its giving opened but on browser of linux desktop review.openstack.org not opening | 18:42 |
| nikesh_vedams | was trying to open "https://review.openstack.org/#/c/159156/" in linux desktop browser | 18:43 |
| kallebe | maybe it is a DNS problem for your browser case? | 18:45 |
| kallebe | try on linux | 18:45 |
| kallebe | nslookup review.openstack.org | 18:45 |
| *** Manshoon has joined #openstack-cinder | 18:47 | |
| nikesh_vedams | hmm upto evening everything was working fine | 18:47 |
| kallebe | If it is a DNS problem, you can try access the IP directly. nslookup returns the direct IP | 18:47 |
| kallebe | for example, here it gives 23.253.232.87 for review.openstack.org | 18:47 |
| openstackgerrit | Mikhail Khodos proposed openstack/cinder: Nexenta Edge iSCSI backend driver. https://review.openstack.org/158745 | 18:49 |
| *** dannywilson has joined #openstack-cinder | 18:49 | |
| *** jcru has joined #openstack-cinder | 18:50 | |
| *** jkremer has joined #openstack-cinder | 18:51 | |
| openstackgerrit | Steven Kaufer proposed openstack/python-cinderclient: Adopt CLI sorting argument guidelines https://review.openstack.org/145913 | 18:52 |
| *** nellysmitt has quit IRC | 18:53 | |
| *** dannywilson has quit IRC | 18:54 | |
| *** dannywilson has joined #openstack-cinder | 18:54 | |
| *** jgravel_ has joined #openstack-cinder | 18:57 | |
| *** ebalduf has joined #openstack-cinder | 19:00 | |
| *** rhagarty has joined #openstack-cinder | 19:00 | |
| *** annegent_ has joined #openstack-cinder | 19:01 | |
| nikesh_vedams | kallebe bswartz : looks like some problem in my ISP | 19:01 |
| kallebe | nikesh_vedams yes, maybe it is indeed. Have you tried with the direct IP? | 19:02 |
| nikesh_vedams | yes | 19:02 |
| nikesh_vedams | its not pinging continuously | 19:02 |
| kallebe | it must be losing packets then | 19:03 |
| nikesh_vedams | might be | 19:04 |
| nikesh_vedams | i will talk to my network admin | 19:04 |
| nikesh_vedams | thanks | 19:04 |
| nikesh_vedams | all | 19:04 |
| *** ebalduf has quit IRC | 19:04 | |
| bswartz | nikesh_vedams: from your VM you were able to contact gerrit, so transfer your code to that VM and submit it from there | 19:05 |
| *** annegent_ has quit IRC | 19:06 | |
| kallebe | I agree with bswartz. You can use WinSCP to transfer the files from windows to your VM | 19:06 |
| bswartz | or better yet, commit the changes to a remote git repo from the windows box, and pull them down from the git repo on the other box | 19:07 |
| bswartz | accounts on github are free... | 19:07 |
| nikesh_vedams | hmm i able to submit patch but to see the reviews | 19:10 |
| nikesh_vedams | i wanted to open in browser | 19:11 |
| nikesh_vedams | :) | 19:11 |
| bswartz | nikesh_vedams: presumably your VM can connect to review.openstack.org | 19:11 |
| nikesh_vedams | yes | 19:11 |
| bswartz | tunnel your browser session over an SSH connection to the VM | 19:11 |
| nikesh_vedams | ok | 19:11 |
| bswartz | ssh -D 1080 user@host | 19:11 |
| *** Manshoon has quit IRC | 19:11 | |
| nikesh_vedams | thanks | 19:11 |
| bswartz | configure your browser to use a SOCKS proxy of localhost:1080 | 19:11 |
| bswartz | and you're good to go | 19:11 |
| *** Manshoon has joined #openstack-cinder | 19:12 | |
| bswartz | if you're on windows and using PuTTY, putty has a feature called "dynamic port forwarding" which does the same thing | 19:12 |
| nikesh_vedams | ok | 19:13 |
| bswartz | I have the opposite situation -- I have Linux on my laptop and I run Windows in a VM for those time when I need to open a powerpoint doc | 19:14 |
| *** annegent_ has joined #openstack-cinder | 19:18 | |
| *** nikesh_vedams has quit IRC | 19:18 | |
| *** Manshoon_ has joined #openstack-cinder | 19:18 | |
| jgriffith | bswartz: what's this "power point" thing that you speak of? | 19:18 |
| jgriffith | :) | 19:18 |
| *** Manshoon_ has quit IRC | 19:19 | |
| bswartz | jgriffith: something I avoid whenever possible | 19:19 |
| *** Manshoon_ has joined #openstack-cinder | 19:19 | |
| jgriffith | bswartz: haha... good answer | 19:19 |
| bswartz | actually I'm pretty happy -- I've only needed to create 1 powerpoint in the last 6 months | 19:20 |
| *** jcru has quit IRC | 19:20 | |
| *** Mandell has quit IRC | 19:21 | |
| *** timcl has quit IRC | 19:22 | |
| *** Manshoon has quit IRC | 19:22 | |
| *** avishay has quit IRC | 19:23 | |
| *** avishay has joined #openstack-cinder | 19:24 | |
| *** nikesh_vedams has joined #openstack-cinder | 19:24 | |
| *** nikesh_vedams has quit IRC | 19:24 | |
| openstackgerrit | arkady kanevsky proposed openstack/cinder: fix for infinity capacity reporting in EQL driver https://review.openstack.org/159231 | 19:24 |
| *** nikesh_vedams has joined #openstack-cinder | 19:25 | |
| *** PaulCuzner has quit IRC | 19:25 | |
| *** timcl has joined #openstack-cinder | 19:26 | |
| *** timcl has quit IRC | 19:27 | |
| *** jcru has joined #openstack-cinder | 19:27 | |
| *** Manshoon has joined #openstack-cinder | 19:31 | |
| *** timcl has joined #openstack-cinder | 19:31 | |
| *** annashen has quit IRC | 19:32 | |
| *** asmith_brcd has joined #openstack-cinder | 19:32 | |
| *** timcl has quit IRC | 19:33 | |
| *** Manshoon has quit IRC | 19:34 | |
| *** Manshoon_ has quit IRC | 19:34 | |
| *** Manshoon has joined #openstack-cinder | 19:35 | |
| *** tsekiyam_ has quit IRC | 19:35 | |
| *** jgravel_ has quit IRC | 19:36 | |
| *** annashen has joined #openstack-cinder | 19:39 | |
| *** annegent_ has quit IRC | 19:42 | |
| nikesh_vedams | https://review.openstack.org/#/c/157356/ | 19:44 |
| nikesh_vedams | https://review.openstack.org/#/c/159156/ | 19:44 |
| *** annegent_ has joined #openstack-cinder | 19:44 | |
| nikesh_vedams | please provide your valuable reviews | 19:44 |
| nikesh_vedams | :) | 19:44 |
| nikesh_vedams | going to bed, | 19:46 |
| openstackgerrit | arkady kanevsky proposed openstack/cinder: fix for infinity capacity reporting in EQL driver https://review.openstack.org/159231 | 19:49 |
| *** e0ne has quit IRC | 19:50 | |
| *** tsekiyama has joined #openstack-cinder | 19:50 | |
| *** tsekiyama has quit IRC | 19:50 | |
| openstackgerrit | arkady kanevsky proposed openstack/cinder: Fix for infinity capacity reporting in EQL driver https://review.openstack.org/159231 | 19:51 |
| *** kallebe has left #openstack-cinder | 19:52 | |
| *** tsekiyama has joined #openstack-cinder | 19:56 | |
| *** tsekiyama has quit IRC | 19:56 | |
| *** tsekiyama has joined #openstack-cinder | 19:57 | |
| *** ebalduf has joined #openstack-cinder | 20:01 | |
| *** annegent_ has quit IRC | 20:01 | |
| *** tsekiyama has quit IRC | 20:01 | |
| *** timcl has joined #openstack-cinder | 20:03 | |
| *** ebalduf has quit IRC | 20:05 | |
| *** gugl has joined #openstack-cinder | 20:06 | |
| *** emagana has quit IRC | 20:08 | |
| *** emagana has joined #openstack-cinder | 20:11 | |
| *** annegent_ has joined #openstack-cinder | 20:13 | |
| *** Manshoon has quit IRC | 20:13 | |
| *** mriedem has quit IRC | 20:14 | |
| *** emagana has quit IRC | 20:16 | |
| patrickeast | thingee: ping | 20:18 |
| *** Apoorva_ has quit IRC | 20:19 | |
| *** Apoorva has joined #openstack-cinder | 20:20 | |
| jungleboyj | patrickeast: Sounded like he was going to be away for a while after this morning's meeting. | 20:20 |
| patrickeast | jungleboyj: ok, thanks! | 20:20 |
| *** annegent_ has quit IRC | 20:21 | |
| jungleboyj | patrickeast: Not my day to watch him, but that was what I got out of discussion this morning. ;-) | 20:21 |
| patrickeast | haha | 20:21 |
| *** annegent_ has joined #openstack-cinder | 20:21 | |
| *** Manshoon has joined #openstack-cinder | 20:21 | |
| *** Manshoon has quit IRC | 20:24 | |
| *** Manshoon has joined #openstack-cinder | 20:24 | |
| *** Mandell has joined #openstack-cinder | 20:25 | |
| *** mriedem has joined #openstack-cinder | 20:25 | |
| *** avishay has quit IRC | 20:27 | |
| *** annegent_ has quit IRC | 20:27 | |
| *** avishay has joined #openstack-cinder | 20:28 | |
| *** diemt has quit IRC | 20:30 | |
| *** emagana has joined #openstack-cinder | 20:30 | |
| *** lpetrut has joined #openstack-cinder | 20:32 | |
| *** cpallares has quit IRC | 20:32 | |
| *** cpallares has joined #openstack-cinder | 20:32 | |
| *** avishay has quit IRC | 20:35 | |
| *** Lee1092 has quit IRC | 20:40 | |
| openstackgerrit | Patrick East proposed openstack/cinder: PureISCSIDriver consistency group updates. https://review.openstack.org/157637 | 20:45 |
| *** IanGovett1 has quit IRC | 20:47 | |
| *** PaulCuzner has joined #openstack-cinder | 20:47 | |
| *** IanGovett has joined #openstack-cinder | 20:49 | |
| *** Manshoon_ has joined #openstack-cinder | 20:50 | |
| *** dustins has quit IRC | 20:51 | |
| *** Manshoon has quit IRC | 20:51 | |
| *** MasterPiece has quit IRC | 20:55 | |
| *** IanGovett1 has joined #openstack-cinder | 20:55 | |
| *** IanGovett has quit IRC | 20:56 | |
| *** MasterPiece has joined #openstack-cinder | 20:57 | |
| anish | jgriffith: thingee should I be closing this bug out https://bugs.launchpad.net/cinder/+bug/1417499 ? | 20:58 |
| openstack | Launchpad bug 1417499 in Cinder "Cert test of devstack setup with LVM and chiscsi iscsi_helper" [Undecided,New] | 20:58 |
| *** lpetrut has quit IRC | 20:58 | |
| *** timcl has left #openstack-cinder | 20:59 | |
| openstackgerrit | Mitsuhiro Tanino proposed openstack/cinder: Clear migration_status from a destination volume if migration fails https://review.openstack.org/154739 | 21:01 |
| *** ebalduf has joined #openstack-cinder | 21:01 | |
| *** karimb has joined #openstack-cinder | 21:02 | |
| *** Ilja has joined #openstack-cinder | 21:04 | |
| *** Manshoon has joined #openstack-cinder | 21:04 | |
| *** Manshoon has quit IRC | 21:05 | |
| *** Manshoon has joined #openstack-cinder | 21:05 | |
| *** ebalduf has quit IRC | 21:06 | |
| *** Manshoon_ has quit IRC | 21:08 | |
| *** annashen has quit IRC | 21:08 | |
| openstackgerrit | Bob Callaway proposed openstack/cinder: Add Manage/Unmanage support to NetApp NFS drivers https://review.openstack.org/158864 | 21:08 |
| *** sseverson has quit IRC | 21:09 | |
| *** nikesh_vedams has quit IRC | 21:11 | |
| *** gugl has quit IRC | 21:13 | |
| *** cdelatte has quit IRC | 21:15 | |
| *** nikesh_vedams has joined #openstack-cinder | 21:15 | |
| *** dannywil_ has joined #openstack-cinder | 21:19 | |
| *** Yogi1 has quit IRC | 21:22 | |
| *** dannywilson has quit IRC | 21:23 | |
| *** takedakn has joined #openstack-cinder | 21:23 | |
| *** annashen has joined #openstack-cinder | 21:23 | |
| *** dannywil_ has quit IRC | 21:26 | |
| *** ronis has quit IRC | 21:26 | |
| *** jcru has quit IRC | 21:26 | |
| *** dannywilson has joined #openstack-cinder | 21:27 | |
| *** tbarron has quit IRC | 21:30 | |
| *** akerr has quit IRC | 21:31 | |
| openstackgerrit | Bob Callaway proposed openstack/cinder: NetApp eseries implementation for manage/unmanage https://review.openstack.org/138093 | 21:31 |
| *** dannywilson has quit IRC | 21:31 | |
| *** bswartz has quit IRC | 21:33 | |
| *** markvoelker has quit IRC | 21:35 | |
| *** markvoelker has joined #openstack-cinder | 21:36 | |
| *** jcru has joined #openstack-cinder | 21:39 | |
| *** markvoelker has quit IRC | 21:40 | |
| *** markvoelker has joined #openstack-cinder | 21:42 | |
| guitarzan | hey folks, does anyone know anything about scheduler hints? | 21:44 |
| guitarzan | I guess it's in the client, just no help text maybe? | 21:45 |
| *** tbarron has joined #openstack-cinder | 21:46 | |
| openstackgerrit | Thang Pham proposed openstack/cinder: VMware: Add vmware:adapter_type extra spec https://review.openstack.org/159260 | 21:48 |
| *** takedakn has quit IRC | 21:48 | |
| guitarzan | nevermind, even simpler than that, have to specify v2 in the client... | 21:50 |
| openstackgerrit | Thang Pham proposed openstack/cinder: VMware: Add vmware:adapter_type extra spec https://review.openstack.org/159260 | 21:51 |
| *** thangp has quit IRC | 21:52 | |
| jungleboyj | hemna and/or xyang ping | 21:56 |
| hemna | jungleboyj, sup | 21:56 |
| jungleboyj | Need to run something by you. | 21:57 |
| *** jcru has quit IRC | 21:57 | |
| jungleboyj | hemna: Trying to move over to using the new config generator so we can remove the old oslo-config from oslo-incubator. | 21:57 |
| hemna | k | 21:57 |
| *** Mandell has quit IRC | 21:58 | |
| jungleboyj | The bad thing, is the new generator uses entry points to get the config options instead of being sent a list of files. | 21:58 |
| hemna | ew | 21:58 |
| hemna | so the existing file that has our requires doesn't work ? | 21:59 |
| jungleboyj | I am trying to figure out the best way to implement this. Talked to Doug and he said we can either put the list in the setup.cfg . | 21:59 |
| jungleboyj | hemna: Which file? | 21:59 |
| hemna | openstack-common.conf | 21:59 |
| jungleboyj | hemna: Different issue I am talking about. | 21:59 |
| hemna | oh | 22:00 |
| * hemna is confused then | 22:00 | |
| jungleboyj | hemna: I am talking about telling oslo-config-generator where all our config options are to create a cinder.conf.sample | 22:00 |
| hemna | oh that. | 22:00 |
| jungleboyj | There are two ways I can do this. Like keystone: https://github.com/openstack/keystone/blob/master/setup.cfg#L66 | 22:01 |
| jungleboyj | though we would have a list of like 70 files in there. | 22:01 |
| jungleboyj | Or, like oslo.concurrency: https://github.com/openstack/oslo.concurrency/blob/master/oslo_concurrency/opts.py#L26 | 22:01 |
| *** jcru has joined #openstack-cinder | 22:02 | |
| *** jdurgin has quit IRC | 22:02 | |
| jungleboyj | Where I would add an options file that basically pulls all the options out of all of the other files as a single entry point for the config generator. | 22:02 |
| *** ebalduf has joined #openstack-cinder | 22:02 | |
| jungleboyj | Either way, if new files that are adding configuration options come along we need to make sure that whatever place we choose is updated accordingly. | 22:03 |
| hemna | dang | 22:03 |
| hemna | that really sux | 22:03 |
| hemna | so we'll have to catch that in reviews. | 22:03 |
| jungleboyj | hemna: Agreed. Yes. | 22:03 |
| hemna | unless we have a hacking check | 22:03 |
| jungleboyj | Oooh, Hmmm ... | 22:03 |
| hemna | puts more of a burden on reviewers | 22:03 |
| hemna | because we don't have enough to watch for already. | 22:04 |
| jungleboyj | Lets see, I think I could do a hacking check. | 22:04 |
| jungleboyj | If a file has 'cfg' in it, make sure it is listed where it needs to be. | 22:04 |
| hemna | yah that might be doable | 22:05 |
| jungleboyj | So, do you think it is better to add cinder/options.py that basically pulls everything in as a single entry point? | 22:06 |
| jungleboyj | Then I just have to check that file to make sure that it has the right module listed if there are opts included? | 22:06 |
| hemna | yah that seems like a simpler thing to manage no? | 22:07 |
| *** ebalduf has quit IRC | 22:07 | |
| jungleboyj | Yeah, probably better than a big ugly setup.cfg file. | 22:07 |
| *** Manshoon has quit IRC | 22:07 | |
| *** jdurgin has joined #openstack-cinder | 22:07 | |
| *** Manshoon has joined #openstack-cinder | 22:08 | |
| *** Manshoon has quit IRC | 22:08 | |
| *** tbarron has quit IRC | 22:08 | |
| *** Manshoon has joined #openstack-cinder | 22:08 | |
| jungleboyj | hemna: Thanks for the input. I will work on getting that coded up. | 22:09 |
| hemna | good luck :) | 22:09 |
| jungleboyj | See if I can figure out a way to do the hacking check. | 22:09 |
| jungleboyj | hemna: Bout time I had to write some real code. | 22:09 |
| hemna | :) | 22:10 |
| *** annashen has quit IRC | 22:14 | |
| *** markstur_ has joined #openstack-cinder | 22:15 | |
| *** markstur_ has quit IRC | 22:16 | |
| *** markstur has quit IRC | 22:16 | |
| *** annashen has joined #openstack-cinder | 22:17 | |
| *** markstur has joined #openstack-cinder | 22:20 | |
| *** mriedem has quit IRC | 22:26 | |
| *** jungleboyj has quit IRC | 22:26 | |
| *** scottda has quit IRC | 22:27 | |
| *** Mandell has joined #openstack-cinder | 22:28 | |
| *** rushil has quit IRC | 22:30 | |
| *** esker has quit IRC | 22:33 | |
| *** amoturi has quit IRC | 22:36 | |
| *** cdelatte has joined #openstack-cinder | 22:40 | |
| *** delattec has joined #openstack-cinder | 22:40 | |
| *** emagana has quit IRC | 22:40 | |
| *** emagana has joined #openstack-cinder | 22:41 | |
| *** karimb has quit IRC | 22:45 | |
| *** emagana has quit IRC | 22:45 | |
| *** bswartz has joined #openstack-cinder | 22:46 | |
| *** eharney has quit IRC | 22:47 | |
| *** Ilja has quit IRC | 22:48 | |
| *** kaufer has quit IRC | 22:49 | |
| *** eharney has joined #openstack-cinder | 22:50 | |
| *** Manshoon has quit IRC | 22:50 | |
| *** jcru has quit IRC | 22:57 | |
| *** Apoorva_ has joined #openstack-cinder | 22:58 | |
| *** asselin_ has joined #openstack-cinder | 23:00 | |
| *** Apoorva has quit IRC | 23:01 | |
| *** Adriano_ has quit IRC | 23:02 | |
| *** ebalduf has joined #openstack-cinder | 23:03 | |
| *** asselin_ has quit IRC | 23:04 | |
| *** dannywilson has joined #openstack-cinder | 23:05 | |
| *** dannywilson has quit IRC | 23:05 | |
| *** dannywilson has joined #openstack-cinder | 23:06 | |
| *** diegows has joined #openstack-cinder | 23:06 | |
| *** ebalduf has quit IRC | 23:07 | |
| *** jcru has joined #openstack-cinder | 23:09 | |
| *** erlon has quit IRC | 23:11 | |
| *** jkremer has quit IRC | 23:12 | |
| *** ganso_ has quit IRC | 23:17 | |
| thingee | patrickeast: hey | 23:23 |
| thingee | anish: sure | 23:23 |
| *** jungleboyj has joined #openstack-cinder | 23:24 | |
| thingee | jungleboyj: thanks for taking my pings today ;) | 23:24 |
| patrickeast | thingee: hey, just wanted to check in on the initiator data table thing now that the spec is merged if the bp’s https://blueprints.launchpad.net/cinder/+spec/driver-private-data & https://blueprints.launchpad.net/cinder/+spec/pure-iscsi-chap-support should get targeted for k-3 | 23:24 |
| anish | thingee: ty ! | 23:25 |
| jungleboyj | thingee: Welcome. I think there was just the one. | 23:25 |
| jungleboyj | thingee: Even thought it wasn't my day to watch you. | 23:26 |
| patrickeast | thingee: thanks! | 23:26 |
| thingee | patrickeast: done and added to https://etherpad.openstack.org/p/cinder-k3-priorities | 23:28 |
| patrickeast | awesome | 23:28 |
| jungleboyj | thingee: Did you see the conversation that hemna and I had about oslo-config-generator earlier? | 23:28 |
| thingee | jungleboyj: it wasn't you're day? are there are other days that you watch me? | 23:28 |
| thingee | your* | 23:28 |
| jungleboyj | ;-) I plead the 5th. | 23:29 |
| thingee | I can't type today | 23:32 |
| thingee | anyways | 23:32 |
| thingee | jungleboyj: regarding your convo with hemna, either way you're going to have update the setup.cfg entrypoints or thing options.py file with the options it deep copies, right? | 23:33 |
| *** dannywilson has quit IRC | 23:33 | |
| thingee | jungleboyj: I didn't immediately see the benfit of the options.py solution. | 23:33 |
| thingee | rather, I still don't. | 23:34 |
| thingee | I'm sure I'm missing something though | 23:34 |
| jungleboyj | Either way, yes, we will have to update one of the files. I think having a giant setup.cfg is ugly and I think it will be easier to write a hacking check against the options.py file. | 23:37 |
| *** afazekas has quit IRC | 23:37 | |
| jungleboyj | thingee: ^^ | 23:37 |
| jungleboyj | thingee: Unless you have a strong aversion to that. | 23:37 |
| *** markvoelker has quit IRC | 23:42 | |
| *** markvoelker has joined #openstack-cinder | 23:42 | |
| *** chlong has quit IRC | 23:43 | |
| *** chlong_ has quit IRC | 23:44 | |
| thingee | jungleboyj: Ah I see now. Well configparser should definitely allow you to read the cfg file for hacking purposes. | 23:46 |
| *** Apoorva_ has quit IRC | 23:46 | |
| * thingee looks through other projects to see what they're doing | 23:46 | |
| *** Apoorva has joined #openstack-cinder | 23:46 | |
| jungleboyj | thingee: I looked at Nova and I don't think they have moved yet. Keystone does the setup.cfg wile oslo uses the options file. | 23:47 |
| thingee | glance does the same as keystone | 23:47 |
| thingee | same with heat | 23:47 |
| *** markvoelker has quit IRC | 23:47 | |
| *** chlong has joined #openstack-cinder | 23:48 | |
| *** jcru has quit IRC | 23:48 | |
| *** asselin_ has joined #openstack-cinder | 23:48 | |
| thingee | ceilometer too | 23:48 |
| jungleboyj | Ok, so you prefer all the entry points in setup.cfg then? | 23:48 |
| jungleboyj | hemna: Concerns with me going that route? | 23:49 |
| *** markvoelker has joined #openstack-cinder | 23:49 | |
| thingee | I prefer being consistent. I don't have a strong opinion where it should go myself | 23:49 |
| hemna | well if others are doing it | 23:49 |
| hemna | then there is safety in numbers I suppose | 23:49 |
| hemna | it'll suck for all projects equally. | 23:49 |
| thingee | jungleboyj: can you ask in the oslo meeting this friday? | 23:49 |
| thingee | what liasons plan to do who haven't moved | 23:50 |
| jungleboyj | thingee: I suppose so. dhellmann didn't indicate a preference when I talked to him. | 23:50 |
| thingee | regardless, I trust you to make a good decision. I'd prefer OpenStack projects to be similar with this, but maybe not everyone agrees. if there is no consensus, we can just do our own thing | 23:51 |
| thingee | of what we think is better looking | 23:52 |
| hemna | thingee, re: _LI, _LW, _LE > | 23:52 |
| hemna | ? | 23:52 |
| hemna | :P | 23:52 |
| hemna | ignore me. I'm just stirring a pot | 23:52 |
| jungleboyj | brb, need to eat dinner with family. | 23:52 |
| thingee | doh | 23:52 |
| hemna | ImportError: No module named oslo_vmware | 23:53 |
| hemna | ugh | 23:53 |
| thingee | :( | 23:53 |
| thingee | where are you seeing that | 23:53 |
| hemna | my local dev | 23:53 |
| hemna | running tox | 23:53 |
| hemna | maybe I need to rebase, dunno | 23:53 |
| thingee | or recreate the environment | 23:53 |
| thingee | tox -r | 23:54 |
| hemna | thingee, so fwiw, I have the scheduler change locally for multiattach | 23:54 |
| hemna | and it's workin | 23:54 |
| hemna | so, just going through the motions with tox to ensure no tests break and then I'll push up a new patchset for cinder multiattac | 23:54 |
| thingee | hemna: is jaypipes trying this out at all locally, or any other interested parties | 23:54 |
| hemna | hopefully that'll make Duncan hapy. | 23:54 |
| hemna | if only I could type | 23:55 |
| hemna | I think he and Tobias had tested it out locally | 23:55 |
| hemna | afaik jaypipes wrote the nova tempest tests | 23:55 |
| thingee | if he can sign off it's working locally with the tempest tests, and you can sign off testing it, that's good to me | 23:56 |
| thingee | jaypipes: ^ | 23:56 |
| hemna | but for K, nova isn't landing so any cinder/cinderclient changes have to work against existing nova. which it already does afaik. | 23:56 |
| hemna | I'm just adding in a change to the filter scheduler to allow testing multiattach flag setting against driver capabilities | 23:56 |
| jaypipes | thingee: I have not yet, no :( | 23:56 |
| hemna | doh | 23:57 |
| thingee | jaypipes: even manually testing works for me. I just need to people to try it. It's pretty late, but I'll review it if people I trust can sign off. | 23:57 |
| *** tbarron has joined #openstack-cinder | 23:57 | |
| jaypipes | thingee: DuncanT was pretty adamant about waiting until L... | 23:58 |
| openstackgerrit | Anish Bhatt proposed openstack/cinder: Add partial unit tests for IET driver https://review.openstack.org/158925 | 23:58 |
| * thingee puts on sunglasses | 23:58 | |
| anish | mtanino: done | 23:58 |
| mtanino | anish: very quick :) | 23:59 |
| anish | it was a one line change :) | 23:59 |
| mtanino | anish: I need cross review with my coworker :) | 23:59 |
| *** asmith_brcd has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!