Wednesday, 2015-06-10

*** dims has quit IRC00:03
*** smoriya has joined #openstack-cinder00:03
*** dims has joined #openstack-cinder00:03
*** alonmarx_ has quit IRC00:07
*** alonmarx has joined #openstack-cinder00:07
*** setmason has quit IRC00:10
*** vokt has quit IRC00:15
*** chlong-zzz has joined #openstack-cinder00:18
*** annashen has quit IRC00:18
*** tsekiyama has quit IRC00:21
*** markvoelker has joined #openstack-cinder00:21
*** Rockyg has quit IRC00:22
openstackgerritNikesh proposed openstack/cinder: Adds FC and ISCSI Cinder drivers for HPMSA Storage Arrays.
*** marcusvrn has quit IRC00:24
*** markvoelker has quit IRC00:26
openstackgerritNikesh proposed openstack/cinder: Adds FC and ISCSI Cinder drivers for Lenovo Storage Arrays.
*** mtanino has quit IRC00:27
*** daneyon_ has quit IRC00:30
*** leeantho has quit IRC00:37
openstackgerritxing-yang proposed openstack/cinder: Support SMI-S provider v8.0.3 in VMAX driver
*** heyun has joined #openstack-cinder00:47
*** zhenguo has joined #openstack-cinder00:47
*** vilobhmm has quit IRC00:52
*** simondodsley has joined #openstack-cinder00:52
*** simondodsley has quit IRC00:56
*** _cjones_ has quit IRC00:59
openstackgerritxing-yang proposed openstack/cinder: Add multiple pools support to VMAX driver
*** raunak has quit IRC01:03
*** mdbooth has quit IRC01:07
*** lixiaoy1 has joined #openstack-cinder01:09
*** patrickeast has quit IRC01:13
*** mdbooth has joined #openstack-cinder01:13
*** tobe has joined #openstack-cinder01:18
*** annashen has joined #openstack-cinder01:19
*** davechen_afk is now known as davechen01:21
*** Splitpeas has quit IRC01:22
*** markvoelker has joined #openstack-cinder01:22
*** dims has quit IRC01:23
*** akerr has joined #openstack-cinder01:24
*** ociuhandu has quit IRC01:24
*** annashen has quit IRC01:24
*** akerr_ has joined #openstack-cinder01:26
*** markvoelker has quit IRC01:27
*** Apoorva has quit IRC01:27
*** diogogmt has joined #openstack-cinder01:28
*** akerr has quit IRC01:29
*** Apoorva has joined #openstack-cinder01:32
*** Apoorva has quit IRC01:36
*** haigang has joined #openstack-cinder01:37
*** vincent_hou has joined #openstack-cinder01:38
*** gouthamr has joined #openstack-cinder01:39
vincent_houthingee: Ping.01:40
vincent_houjgriffith: Ping.01:40
*** vilobhmm has joined #openstack-cinder01:45
*** haigang has quit IRC01:47
*** marcusvrn has joined #openstack-cinder01:48
*** haigang has joined #openstack-cinder01:48
*** zhonghua-lee has joined #openstack-cinder01:54
*** zhonghua-li has joined #openstack-cinder01:54
*** zhonghua-lee has quit IRC01:55
*** zhonghua-li has quit IRC01:55
*** marcusvrn has quit IRC01:56
*** Zhongjun has quit IRC01:58
*** patrickeast has joined #openstack-cinder02:00
*** zhonghua-lee has joined #openstack-cinder02:00
*** asselin_ has joined #openstack-cinder02:00
*** vilobhmm has quit IRC02:02
*** Lee1092 has joined #openstack-cinder02:03
*** rushil has joined #openstack-cinder02:09
*** boris-42 has quit IRC02:12
*** dims has joined #openstack-cinder02:12
*** rongze has joined #openstack-cinder02:19
*** dims has quit IRC02:23
jgriffithvincent_hou: yo02:25
vincent_houjgriffith: Buddy, can you help me take a look at the spec I drafted:
jgriffithvincent_hou: silly question....02:28
jgriffithvincent_hou: I know you worked very hard on the spec, and that's great02:28
jgriffithvincent_hou: but why not just fix the issues?02:28
jgriffithvincent_hou: or log bugs instead :)02:28
vincent_houI am doing at the same time.02:29
jgriffithvincent_hou: smart man!! :)02:29
vincent_houcheck these up:
jgriffithvincent_hou: so I just have a minor nit picky suggestion02:30
vincent_houThere are still new ones coming.02:30
jgriffithvincent_hou: add some spacing between the line items02:30
jgriffithvincent_hou: if you do that I'll upvote the spec02:30
jgriffithvincent_hou: don't go crazy adding much more02:30
jgriffithvincent_hou: you're going to overload people like me :)02:31
*** haigang has quit IRC02:31
jgriffithvincent_hou: my little brain can only store, process and keep track of so much at once :)02:31
vincent_houIt won't. They are WIP.02:31
vincent_houJust let you know.02:31
jgriffithvincent_hou: sounds good... thanks for taking this one02:31
vincent_houSure. you are welcome.02:31
*** gouthamr has quit IRC02:36
*** gouthamr has joined #openstack-cinder02:46
*** gouthamr has quit IRC02:52
*** julim has joined #openstack-cinder02:54
*** gouthamr has joined #openstack-cinder02:56
*** julim has quit IRC02:59
*** akerr_ has quit IRC03:04
*** rushiagr_away is now known as rushiagr03:05
*** rushil has quit IRC03:06
*** changbl_ has joined #openstack-cinder03:09
*** markvoelker has joined #openstack-cinder03:11
openstackgerritwanghao proposed openstack/cinder: Fix weird change of volume status in re-scheduling
*** markvoelker has quit IRC03:16
*** gouthamr has quit IRC03:17
openstackgerritDave Chen proposed openstack/cinder: Policies based properties protection
*** annashen has joined #openstack-cinder03:21
*** rushiagr is now known as rushiagr_away03:25
*** annashen has quit IRC03:26
*** root has joined #openstack-cinder03:31
*** root is now known as Guest5040103:31
nikeshmis any cinder meeting tomorrow03:33
nikeshmno agenda yet03:33
*** vilobhmm has joined #openstack-cinder03:34
*** avishay_ has joined #openstack-cinder03:39
*** Guest50401 has quit IRC03:40
*** harlowja_at_home has joined #openstack-cinder03:41
*** coolsvap|afk is now known as coolsvap03:42
*** diogogmt has quit IRC03:45
*** sgotliv has joined #openstack-cinder03:59
*** vincent_hou has quit IRC04:01
openstackgerritwanghao proposed openstack/cinder: Add support for force-delete backups
*** sgotliv has quit IRC04:03
*** rushiagr_away is now known as rushiagr04:05
*** rushiagr is now known as rushiagr_away04:05
*** rushiagr_away is now known as rushiagr04:06
*** setmason has joined #openstack-cinder04:19
*** yamada-h has joined #openstack-cinder04:21
*** yamada-h has quit IRC04:25
*** heyun has quit IRC04:25
openstackgerritJohn Griffith proposed openstack/cinder: Add secondary account capability to SolidFire
*** BharatK has joined #openstack-cinder04:29
*** BharatK has quit IRC04:30
*** BharatK has joined #openstack-cinder04:31
*** yamada-h has joined #openstack-cinder04:31
*** chlong-zzz is now known as chlong04:32
*** setmason has quit IRC04:37
*** zhonghua-lee has quit IRC04:40
*** patrickeast has quit IRC04:52
*** swamireddy1 has joined #openstack-cinder04:53
*** setmason has joined #openstack-cinder04:54
*** setmason has quit IRC04:55
*** rushiagr is now known as rushiagr_away04:55
*** markvoelker has joined #openstack-cinder05:00
*** swamireddy1 has quit IRC05:00
*** deepakcs has joined #openstack-cinder05:02
*** markvoelker has quit IRC05:04
*** vilobhmm has quit IRC05:11
*** setmason has joined #openstack-cinder05:17
openstackgerritrakesh mishra proposed openstack/cinder: set-size-limit-for-volume
*** pradipta has joined #openstack-cinder05:23
*** rushiagr_away is now known as rushiagr05:25
*** setmason has quit IRC05:28
*** setmason has joined #openstack-cinder05:30
*** nihilifer_ has joined #openstack-cinder05:32
*** nkrinner has joined #openstack-cinder05:32
*** bswartz has quit IRC05:37
*** nlevinki has joined #openstack-cinder05:40
*** earlephilhower has quit IRC05:47
*** ankit_ag has joined #openstack-cinder05:47
*** yamada-h_ has joined #openstack-cinder05:47
*** yamada-h has quit IRC05:48
*** zhonghua-lee has joined #openstack-cinder05:52
*** haomaiwang has quit IRC05:54
*** rongze has quit IRC05:54
openstackgerritXi Yang proposed openstack/cinder: Provide snap copy feature in EMC VNX Cinder driver
*** belmoreira has joined #openstack-cinder06:01
*** setmason has quit IRC06:01
*** setmason has joined #openstack-cinder06:04
*** yamada-h_ has quit IRC06:13
*** setmason has quit IRC06:20
*** annashen has joined #openstack-cinder06:23
*** anshul has joined #openstack-cinder06:25
openstackgerritwanghao proposed openstack/cinder: Add support for force-delete backups
*** anshul has quit IRC06:27
*** anshul has joined #openstack-cinder06:27
*** yamada-h has joined #openstack-cinder06:28
*** annashen has quit IRC06:29
*** zhonghua-lee has quit IRC06:29
*** zhonghua-lee has joined #openstack-cinder06:32
*** mdbooth has quit IRC06:39
*** bswartz has joined #openstack-cinder06:40
*** harlowja_at_home has quit IRC06:41
*** mdbooth has joined #openstack-cinder06:45
*** jwcroppe_ has quit IRC06:46
*** markvoelker has joined #openstack-cinder06:49
*** markvoelker has quit IRC06:53
*** lpetrut has joined #openstack-cinder06:58
openstackgerritwanghao proposed openstack/cinder-specs: Incremental backup improvements for L
*** xiaohui has quit IRC07:09
*** jistr has joined #openstack-cinder07:37
*** BharatK has quit IRC07:38
*** heyun has joined #openstack-cinder07:39
openstackgerritZoltan Arnold Nagy proposed openstack/cinder: WIP: encrypted rbd support
*** BharatK has joined #openstack-cinder07:45
*** angela-s has quit IRC07:47
openstackgerritwanghao proposed openstack/cinder: Incremental backup improvements for L
*** chlong has quit IRC07:51
openstackgerritchenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
*** haomaiwang has joined #openstack-cinder07:56
flip214Please, give me a second pair of eyes.
flip214As I read that, it's a volume create with "volume_type": "d0724333-7fef-43da-93bf-f1e8aa4479d4"07:58
flip214but the result says "volume_type": "volume-type-2143297657"07:58
flip214is that okay, resp. expected?07:59
*** jordanP has joined #openstack-cinder07:59
*** leopoldj has joined #openstack-cinder08:00
*** nikeshm has quit IRC08:00
openstackgerritchenzongliang proposed openstack/cinder: Huawei Storage Multiple Pools Support
*** nikeshm has joined #openstack-cinder08:04
*** markus_z has joined #openstack-cinder08:16
*** zhonghua-lee has quit IRC08:17
openstackgerritwanghao proposed openstack/cinder: Fix weird change of volume status in re-scheduling
*** afazekas has joined #openstack-cinder08:20
*** smoriya has quit IRC08:21
openstackgerritchenzongliang proposed openstack/cinder: Add CHAP support for Huawei Storage
openstackgerritchenzongliang proposed openstack/cinder: Add CHAP support for Huawei Storage
*** aarefiev has quit IRC08:28
*** aarefiev has joined #openstack-cinder08:28
openstackgerritchenzongliang proposed openstack/cinder: Huawei Storage iSCSI multipath Support
*** markvoelker has joined #openstack-cinder08:37
openstackgerritCindy Pallares proposed openstack/cinder: Fix multi-line docstrings to meet hacking rules
openstackgerritAnton Arefiev proposed openstack/cinder: Fix lvm manage existing volume
*** markvoelker has quit IRC08:42
*** yamada-h has quit IRC08:42
*** lpetrut has quit IRC08:43
openstackgerritchenzongliang proposed openstack/cinder: Support SmartX for Huawei volume driver
*** lixiaoy1 has quit IRC08:55
openstackgerritPh. Marek proposed openstack/cinder: Re-add DRBD driver.
*** lpetrut has joined #openstack-cinder09:18
*** dims has joined #openstack-cinder09:27
*** sgotliv has joined #openstack-cinder09:27
*** dims has quit IRC09:31
*** aix has joined #openstack-cinder09:34
*** rakesh_mishra has joined #openstack-cinder09:37
*** xek has joined #openstack-cinder09:42
*** deepakcs has quit IRC09:45
*** rushiagr is now known as rushiagr_away09:49
*** dims has joined #openstack-cinder09:51
*** dims_ has joined #openstack-cinder09:52
*** e0ne has joined #openstack-cinder09:55
*** dims has quit IRC09:56
*** e0ne is now known as e0ne_10:01
*** fanyaohong has joined #openstack-cinder10:01
*** rakesh_mishra has quit IRC10:02
*** rakesh_mishra has joined #openstack-cinder10:02
openstackgerritZoltan Arnold Nagy proposed openstack/cinder: WIP: encrypted rbd support
*** e0ne_ has quit IRC10:07
openstackgerritGorka Eguileor proposed openstack/cinder: Use correct rtslib namespace for newer versions
*** boris-42 has joined #openstack-cinder10:09
*** rakesh_mishra has quit IRC10:11
*** Zhongjun has joined #openstack-cinder10:13
openstackgerritKyrylo Romanenko proposed openstack/python-cinderclient: Add test for cinder extra-specs-list command output Add test for cinder list-extension command output Fix pep8 issues "E128 continuation line under-indented for visual indent" in other functions Change-Id: I28266af7a6ff23b83899c50cc230f926a44592dc Partial
*** markvoelker has joined #openstack-cinder10:23
*** rakesh_mishra has joined #openstack-cinder10:24
*** annashen has joined #openstack-cinder10:27
*** markvoelker has quit IRC10:27
*** ociuhandu has joined #openstack-cinder10:28
*** avishay__ has joined #openstack-cinder10:29
*** avishay_ has quit IRC10:29
*** annashen has quit IRC10:31
flip214hemna: ping?10:33
*** cbader has quit IRC10:34
*** e0ne has joined #openstack-cinder10:35
flip214I get u'tgtadm: invalid request\\n'" with the DRBD cinder driver; google doesn't really help. do you have an idea for me?10:38
*** ociuhandu has quit IRC10:40
*** heyun has quit IRC10:40
*** kmartin has quit IRC10:44
*** coolsvap is now known as coolsvap|afk10:45
*** asselin has quit IRC10:50
flip214syslog says "kernel: [ 2135.066375]  connection6:0: detected conn error (1020)"10:53
*** sgotliv_ has joined #openstack-cinder10:53
*** sgotliv has quit IRC10:54
*** rakesh_mishra has quit IRC10:55
BharatKDuncanT, ping11:03
BharatKDuncanT, Hi, Regarding the patch
BharatKDuncanT, Related to your comment
BharatKDuncanT, Please let me know your thoughts, I will work based on that.11:05
DuncanTI can't see my comment there, just Mike's (to which I think your response is correct)11:05
BharatKDuncanT, Oh :( I am extremely sorry. Mistaken11:05
BharatKDuncanT, Please don't mind11:05
DuncanTNo probem :-)11:05
DuncanTI'll go and add a comment that I agree with you and finish my review11:06
BharatKDuncanT, Thanks11:06
BharatKDuncanT, My actual query related to the patch
*** rakesh_mishra has joined #openstack-cinder11:07
BharatKDuncanT, Regarding your comment "So this is going to print a warning for every create, and there's nothing the admin can/should do to fix it. Probably shouldn't be a warning IMO, unless we add a warn_once construct."11:07
BharatKDuncanT, Could you please tell me how can I make as warn_once?11:08
lifelessDuncanT: the warnings module does warn only once in a lifetime by default11:08
*** alexpilotti has joined #openstack-cinder11:08
lifelessDuncanT: IIRC; [but thats scoped - its once per callsite calling into warnings.warn]11:09
DuncanTlifeless: that sounds like what we need11:09
openstackgerritJessy Lee proposed openstack/cinder: Add volume drivers for Infortrend Storage
BharatKlifeless, DuncanT Thanks, is there any similar kind of patch/example?11:10
lifeless    key = (text, category, lineno)11:10
lifelessso a given message in a given category from a given lineno will only warn once11:10
BharatKlifeless, Thank you very much will look into this.11:11
lifelessweird but there you go11:11
DuncanTlifeless: That isn't the default behaviour of log.warning, is it?11:11
lifelessDuncanT: warnings.warn11:11
lifelessDuncanT: log.warning is every-time-you-call-it-logs11:11
DuncanTlifeless: Ah ha, got you, thanks11:11
lifelessDuncanT: /totally/ different11:11
DuncanTYup, I understand now11:12
lifelessdunno if that helps you11:12
DuncanTPython: batteries included, if you can find them down the back of the sofa11:12
BharatKDuncanT, lifeless, I will try that. Thanks.11:14
DuncanTBharatK: has some notes on warnings, and how to make them go to the log rather than stderr - we might have that code in cinder already, not sure11:17
BharatKDuncanT, Thanks will see.11:18
*** avishay__ has quit IRC11:21
*** Longgeek has joined #openstack-cinder11:22
*** rakesh_mishra has quit IRC11:22
*** markvoelker has joined #openstack-cinder11:24
*** Longgeek_ has joined #openstack-cinder11:25
*** markvoelker has quit IRC11:28
*** Longgeek has quit IRC11:28
*** rakesh_mishra has joined #openstack-cinder11:39
*** marcusvrn has joined #openstack-cinder11:43
*** marcusvrn1 has joined #openstack-cinder11:46
*** marcusvrn has quit IRC11:46
*** tobe has quit IRC11:47
*** marcusvrn1 has quit IRC11:51
*** marcusvrn has joined #openstack-cinder11:51
*** julim has joined #openstack-cinder11:52
*** markvoelker has joined #openstack-cinder11:54
*** timcl has joined #openstack-cinder12:01
*** pradipta has quit IRC12:02
openstackgerritDavanum Srinivas (dims) proposed openstack/python-cinderclient: cleanup openstack-common.conf and sync updated files
*** rushiagr_away is now known as rushiagr12:06
*** ganso_ has joined #openstack-cinder12:06
*** aix has quit IRC12:14
*** fanyaohong has quit IRC12:15
*** akerr has joined #openstack-cinder12:19
*** rakesh_mishra has quit IRC12:19
*** dulek has joined #openstack-cinder12:20
*** chlong has joined #openstack-cinder12:29
openstackgerritVictor Stinner proposed openstack/cinder: Drop L suffix from long integers
*** gouthamr has joined #openstack-cinder12:33
*** timcl has quit IRC12:33
openstackgerritVictor Stinner proposed openstack/cinder: Replace basestring with six.string_types
*** timcl has joined #openstack-cinder12:34
*** gouthamr has quit IRC12:36
*** gouthamr has joined #openstack-cinder12:36
*** eharney has joined #openstack-cinder12:37
openstackgerritVictor Stinner proposed openstack/cinder: Replace unicode with six.text_type
*** salv-orlando has quit IRC12:45
*** timcl has quit IRC12:45
*** eantysh9 is now known as eantyshev12:46
*** e0ne is now known as e0ne_12:47
openstackgerritVictor Stinner proposed openstack/cinder: Replace dict.itervalues() with six.itervalues(dict)
*** e0ne_ is now known as e0ne12:50
*** aix has joined #openstack-cinder12:50
*** ociuhandu has joined #openstack-cinder12:53
openstackgerritVictor Stinner proposed openstack/cinder: Use six.reraise() for Python 3 compatibility
*** ociuhandu has quit IRC12:54
openstackgerritVictor Stinner proposed openstack/cinder: Replace with next(it) for py3 compat
*** Zhongjun has quit IRC12:58
openstackgerritVictor Stinner proposed openstack/cinder: Replace unicode with six.text_type
*** cbader has joined #openstack-cinder13:03
*** nihilifer_ has quit IRC13:06
*** mriedem has joined #openstack-cinder13:13
*** timcl has joined #openstack-cinder13:15
*** Yogi11 has joined #openstack-cinder13:16
*** rushiagr is now known as rushiagr_away13:20
*** rakesh_mishra has joined #openstack-cinder13:22
openstackgerritAnton Arefiev proposed openstack/cinder: Add entry create and cast tasks to manage workflow
openstackgerritAnton Arefiev proposed openstack/cinder: Fix lvm manage existing volume
*** ociuhandu has joined #openstack-cinder13:25
*** setmason has joined #openstack-cinder13:25
mriedemis anyone aware of any documentation for the various vendor driver's connection_info keys that are returned from the initialize_connection call?13:26
mriedemi'm looking to model that with objects in nova and i was going to start with documenting what comes back today13:27
mriedembefore we can model what the common keys are, like driver_volume_type and data (and all of the gorp that's in the data key)13:27
*** changbl_ has quit IRC13:27
*** dustins has joined #openstack-cinder13:28
flip214asselin_: If the tempest.api.volume.admin.test_multi_backend.VolumeMultiBackendV2Test isn't skipped but fails, might the reason be the line 'CINDER_ENABLED_BACKENDS=,drbd:drbdmanage'? (the comma)13:28
asselin_flip214, not sure...never looked at that test case since ours skips13:29
*** annashen has joined #openstack-cinder13:30
flip214DuncanT: thingee: e0ne: jgriffith: anybody else knows this?? ^^13:30
jgriffithflip214: what's up?13:31
* jgriffith reading13:31
flip214jgriffith: hi, thanks for answering! If the tempest.api.volume.admin.test_multi_backend.VolumeMultiBackendV2Test isn't skipped but fails, might the reason be the line 'CINDER_ENABLED_BACKENDS=,drbd:drbdmanage'? (the comma)13:31
*** ociuhandu has quit IRC13:31
jgriffithflip214: well... did you try it without the comma ?13:32
e0neflip214: it should be 'CINDER_ENABLED_BACKENDS+=,drbd:drbdmanage', shouldn't it?13:32
*** teju has joined #openstack-cinder13:32
jgriffithflip214: and what did it fail for?13:32
jgriffithflip214: did it build a proper config file?13:32
jgriffithflip214: did it build the LVM  VG's?13:32
flip214this is already via -infra CI. DRBD did run fine (mostly), I'm now fixing the last few bugs.13:33
flip214And I believe that the multi-backend failure is due to that line - some part thinks multiple backends are available, but only one is properly initialized (is my guess).13:33
e0neflip214: accodding to you hace only one configured backend13:33
flip214jgriffith: that's the TestVolumeBootPatternV2 tests, I think I've fixed these in my local copy.13:34
*** annashen has quit IRC13:35
flip214e0ne: but why is the test not skipped, then?13:35
e0neflip214: good question:). actually, i don't know how it works13:35
jgriffithThat's the TestVolumeBootPattern Vol not found13:36
flip214jgriffith: yes, I believe to have that fixed locally.13:36
openstackgerritAbhishek Lekshmanan proposed openstack/cinder: Backups: add a force flag
flip214I'd like to know about the multibackend tests...13:36
*** ankit_ag has quit IRC13:36
*** ociuhandu has joined #openstack-cinder13:37
*** timcl has quit IRC13:37
jgriffithflip214: I'm looking13:43
jgriffithflip214: it may be a problem with the test13:45
jgriffithflip214: the volume the test fails on is ed860ce3-bdec-4e24-bf7a-021a4fd53b0413:45
jgriffithflip214: you can see the succesful create/delete sequence13:45
flip214okay... so what would you advise?13:46
jgriffithflip214: I would advise you dig in and figure it out :)13:46
*** rushil has joined #openstack-cinder13:47
jgriffithflip214: I can deploy multi-backend and test it on my side later today13:47
jgriffithflip214: ummm13:48
jgriffithto run multi-backend, wouldn't you expect to have multiple backends?13:49
jgriffithflip214: also not sure what your syntax is about in that conf file13:50
jgriffithflip214: have a look at
flip214jgriffith: I don't want to run multi-backend.13:50
jgriffithflip214: and if you add another, just use ",<backend-name>"13:51
flip214I'd like to know why the multi-backend tests are not skipped13:51
jgriffithflip214: because your setup in the local.conf is wrong13:51
*** setmason has quit IRC13:51
*** e0ne is now known as e0ne_13:51
flip214default_volume_type = drbdmanage13:53
flip214enabled_backends = drbdmanage13:53
jgriffithflip214: your cinder.conf ended up ok, but you have no tempest config13:53
flip214jgriffith: what's wrong there? This is in an -infra CI, so I can only have some small details wrong.13:53
*** rhagarty has quit IRC13:53
jgriffithflip214: oh... well, maybe I'm completely wrong13:53
*** rhagarty_ has quit IRC13:54
*** rhagarty has joined #openstack-cinder13:54
*** rhagarty__ has joined #openstack-cinder13:54
flip214 says13:54
flip214multi_backend = True13:55
*** setmason has joined #openstack-cinder13:55
flip214hmmm, where does that come from? asselin_, can you help me here, please?13:55
jgriffithflip214: It comes from the local.conf file13:55
jgriffithor localrc13:55
asselin_flip214, mtreinish usually helps us out...maybe he's around13:56
jgriffithalright... I give up13:56
flip214jgriffith: thanks, anyway!13:57
*** dustins_ has joined #openstack-cinder13:59
*** e0ne_ is now known as e0ne13:59
jgriffithflip214: Ok, I'm going to try ONE more time14:00
asselin_flip214, did you try to see how the other open source cinder backends are passing?14:00
jgriffithflip214: please look at your tempest_conf.txt file14:00
flip214asselin_: I'm trying to compare, yes.14:00
jgriffithflip214: You have multi_backend = True set14:00
jgriffithfrom volume-feature-enabled14:00
flip214jgriffith: just what I pasted 6 mins ago ... ;)14:00
flip214where does that come from? I guess the additional comma!14:00
jgriffiththen WTF is your question!14:01
*** Yogi12 has joined #openstack-cinder14:01
jgriffithget rid of the crappy formatting of your enabled-backends line14:01
*** juzuluag has joined #openstack-cinder14:01
jgriffithwhich is why I pointed you to a working localrc file14:01
jgriffithflip214: I even offered to test it for you later today :)14:01
flip214got it.14:02
flip214    if [[ -n "$CINDER_ENABLED_BACKENDS" ]] && [[ $CINDER_ENABLED_BACKENDS =~ .*,.* ]]; then14:02
flip214        iniset $TEMPEST_CONFIG volume-feature-enabled multi_backend "True"14:02
jgriffithflip214: but honestly you can do this easily yourself14:02
flip214jgriffith: yes, thanks!14:02
jgriffithflip214: just deploy devstack with that localrc on your own machine14:02
*** rushil_ has joined #openstack-cinder14:02
*** dustins has quit IRC14:02
flip214jgriffith: of course I can follow the various links ...14:02
jgriffithsee what the tempest conf ends up being14:02
flip214and/or deploy it.14:02
flip214but each of that takes quite some time.14:02
*** breitz has quit IRC14:02
jgriffithflip214: yeah... sorry, I know you can :)14:02
*** gouthamr has quit IRC14:02
flip214If I'm lucky, I'Ve got an answer here in 3 mins or so ... at least that's what I hoped!14:02
*** gouthamr has joined #openstack-cinder14:02
flip214never mind, thank you all!14:03
jgriffithflip214: LOL14:03
jgriffithflip214: some mornings you win, some morning I show up :)14:03
*** breitz has joined #openstack-cinder14:03
asselin_lol :)14:03
*** Yogi11 has quit IRC14:03
*** rushil has quit IRC14:05
jgriffithflip214: ^^ That's the answer you were looking for14:06
jgriffithflip214: and yes, back to the beginning, it's the syntax in your enabled_backends line14:06
*** lcurtis has joined #openstack-cinder14:07
*** setmason has quit IRC14:09
*** juzuluag has quit IRC14:09
*** amoturi has joined #openstack-cinder14:11
*** timcl has joined #openstack-cinder14:11
*** xyang has joined #openstack-cinder14:13
*** rmesta has joined #openstack-cinder14:13
flip214jgriffith: one more question... the VolumesTypeV1Test uses u'storage_protocol': u'drbd', but I can't see where that comes from.14:13
flip214my driver returns a storage_protocol as "iSCSI" (which isn't that clean, next iteration uses target_driver.protocol), but it's still different...14:14
*** rmesta has left #openstack-cinder14:14
jgriffithflip214: not sure off the top of my head, but personally I set that in my conf file usually14:15
jgriffithflip214: probably what I'd do if I were you14:15
jgriffithflip214: make sure you're not reporting it back in cinder stats14:16
jgriffithflip214: that's the other place that picks that up14:16
*** thangp has joined #openstack-cinder14:19
*** dguryanov has joined #openstack-cinder14:20
*** mtanino has joined #openstack-cinder14:21
openstackgerritxing-yang proposed openstack/cinder: Add multiple pools support to VMAX driver
*** Yogi12 has quit IRC14:22
*** zhonghua-lee has joined #openstack-cinder14:23
*** Yogi1 has joined #openstack-cinder14:26
*** nkrinner has quit IRC14:29
*** lpabon has joined #openstack-cinder14:31
*** timcl has quit IRC14:36
*** Yogi11 has joined #openstack-cinder14:36
*** timcl has joined #openstack-cinder14:37
*** julim has quit IRC14:38
*** simondodsley has joined #openstack-cinder14:39
*** Yogi1 has quit IRC14:40
*** openstackgerrit has quit IRC14:41
*** openstackgerrit has joined #openstack-cinder14:41
*** tsekiyama has joined #openstack-cinder14:47
*** diogogmt has joined #openstack-cinder14:50
*** zhonghua-li has joined #openstack-cinder14:50
*** nlevinki has quit IRC14:51
*** timcl has quit IRC14:53
*** zhonghua-lee has quit IRC14:53
*** zhipeng has joined #openstack-cinder14:55
*** Lee1092 has quit IRC14:55
*** belmoreira has quit IRC14:57
*** setmason has joined #openstack-cinder14:57
*** anshul has quit IRC14:57
*** patrickeast has joined #openstack-cinder14:58
flip214jgriffith: is 80x24 not big enough for you?14:58
jgriffithflip214: LOL14:58
jgriffithflip214: that's actually the size of this window funny enough :)14:59
jgriffithflip214: I was trying to get rid of my markers from another byobu session14:59
jgriffithflip214: but was doing it in the wrong window :(14:59
*** afazekas has quit IRC14:59
openstackgerritBharat Kumar Kobagana (BharatK) proposed openstack/cinder: Using 'fallocate' instead of 'dd' to create raw volumes
*** dims_ has quit IRC15:03
*** timcl has joined #openstack-cinder15:03
*** kmartin has joined #openstack-cinder15:05
*** crose has joined #openstack-cinder15:07
*** dims has joined #openstack-cinder15:08
*** dims has quit IRC15:08
*** patrickeast has quit IRC15:08
*** dims has joined #openstack-cinder15:08
*** patrickeast has joined #openstack-cinder15:09
*** bill_az has joined #openstack-cinder15:09
*** patrickeast has quit IRC15:10
*** patrickeast has joined #openstack-cinder15:10
*** BharatK has quit IRC15:12
jgriffithe0ne: upated as per your comment
*** Lee1092 has joined #openstack-cinder15:12
*** julim has joined #openstack-cinder15:12
e0nejgriffith: thanks!15:13
e0nejgriffith: may be I should wait to solidfire ci before +1:)15:15
*** jdurgin1 has joined #openstack-cinder15:15
*** timcl has quit IRC15:15
*** lpetrut has quit IRC15:15
e0nesorry, before +215:16
*** mikedillion has joined #openstack-cinder15:16
*** mikedillion has quit IRC15:18
e0nejgriffith: i answered to your comment for please, review when you'll have a time15:20
*** leopoldj has quit IRC15:26
*** jungleboyj has joined #openstack-cinder15:28
thingeepatrickeast: yay internal tenant spec was already approved :)15:28
patrickeastthingee: :D yea i saw15:29
patrickeastthings move fast once you threaten to approve things15:29
*** vokt has joined #openstack-cinder15:30
openstackgerritPh. Marek proposed openstack/cinder: Re-add DRBD driver.
thingeejgriffith: looks like some disagreements here at least feedback is coming in15:32
jgriffithe0ne: Is your last comment WRT the show command?15:32
jgriffithe0ne: or the list command?15:32
*** sileht has quit IRC15:33
*** pradipta has joined #openstack-cinder15:33
jgriffithe0ne: my point was I don't see value in options to list... if one would like details do a show; and even better make show take a list of ID's :)15:33
e0nejgriffith: it's about both og them15:33
jgriffithe0ne: yeah, ok... I know the patch was, just wasn't sure on your comment15:34
jgriffithe0ne: so I have my opinon, I'm certainly in the minority15:35
e0nejgriffith: but for now, we've got --details param (at least in REST API side) for list15:35
jgriffithe0ne: and it is "me" so take that FWIW :)15:35
jgriffithe0ne: yeah.. I hate that I have to say15:35
e0nejgriffith: :)15:35
jgriffithe0ne: I'm a firm beliver in simple API's with few options15:36
jgriffithe0ne: if you need a bunch of options hanging off of your API call, you likely need a new API call15:36
*** sileht has joined #openstack-cinder15:36
jgriffithe0ne: I'm not stongly enough opposed to fight about it, or get bummed out when it merges15:37
e0nejgriffith: i'm totally agree with you about simple API. but we need to support current api users:(15:37
jgriffithe0ne: at the same time I'm not likely to change my opinion :(15:37
*** timcl has joined #openstack-cinder15:37
jgriffithe0ne: I don't understand that part (support current users), so maybe I'm missing an important detail?15:37
thingeejgriffith: you mind if we throw replication v2 in the agenda for today?15:37
*** lpetrut has joined #openstack-cinder15:38
e0nejgriffith: i mean, e.g. we can't remove --details param for list now because it is used by somebody15:38
*** sileht has quit IRC15:39
jgriffithe0ne: Ahh... got ya15:40
jgriffiththingee: sure15:40
jgriffiththingee: I mean "no" I don't mind15:40
jgriffithe0ne: stupid call should've never been there :(15:40
*** sileht has joined #openstack-cinder15:40
e0nejgriffith: agree15:40
jgriffithe0ne: wait...15:41
e0nejgriffith: --details is used for list in client by default and we can't change it15:42
*** dulek has quit IRC15:42
thingeewinston-d: you around?15:43
*** zhipeng has quit IRC15:43
e0nejgriffith: at lease, for CLI. i'll check how it works in python api15:43
*** Daviey has quit IRC15:44
*** jdurgin1 has quit IRC15:44
e0nejgriffith: - detailed=True by default15:45
jgriffithe0ne: yeah, looking at that now15:45
jgriffithe0ne: but that's nothing more than a flag for the view builder15:45
e0nejgriffith: - we don't pass 'detailed' param15:45
jgriffithe0ne: Ok, what if we just updated list to return "more stuff"15:46
jgriffithe0ne: make the detailed flag obsolete15:46
e0nejgriffith: sound good for me15:47
jgriffithe0ne: either that or I just trust you and everyone else and remove my -1 :)15:47
jgriffithe0ne: I need to look at what that does15:47
jgriffithe0ne: ie how much more data15:47
*** garthb_ has joined #openstack-cinder15:49
*** garthb has joined #openstack-cinder15:49
e0nejgriffith: give me 2 mins, i'll find working env with cinder and compare data15:49
*** Daviey has joined #openstack-cinder15:50
e0nejgriffith: it's from our docs:
jgriffithe0ne: hly crap!15:51
jgriffithe0ne: not your link15:51
jgriffithe0ne: the view builder15:51
e0nejgriffith: :)15:51
jgriffithe0ne: so that's what I was afraid of15:52
*** earlephilhower has joined #openstack-cinder15:52
jgriffithe0ne: it bascily spews out the same info as show15:52
jgriffithe0ne: but for *every* volume :(15:53
e0nejgriffith: yes15:53
jgriffithe0ne: Well I certainly don't want that to be the default :)15:53
e0nejgriffith: and even if use we get all volumes data  from DB15:53
*** jnicklin has joined #openstack-cinder15:54
e0nejgriffith: looks lime my spec will be dropped in such case15:54
*** vokt has quit IRC15:54
*** BharatK has joined #openstack-cinder15:56
*** avishay has joined #openstack-cinder15:56
*** dannywilson has joined #openstack-cinder15:56
e0nejgriffith: so, we'll  leave default list action as is, am I correct?15:57
e0neit returns only id and name15:57
*** avishay has quit IRC15:58
*** lpetrut has quit IRC15:59
*** vokt has joined #openstack-cinder16:01
*** leeantho has joined #openstack-cinder16:01
*** dustins_ is now known as dustins16:04
*** adurbin_ has joined #openstack-cinder16:05
*** rushiagr_away is now known as rushiagr16:06
*** adurbin__ has joined #openstack-cinder16:06
*** daneyon has joined #openstack-cinder16:06
*** BharatK has quit IRC16:09
*** gouthamr has quit IRC16:10
*** dustins has quit IRC16:10
*** annegentle has joined #openstack-cinder16:11
*** gouthamr has joined #openstack-cinder16:11
*** Apoorva has joined #openstack-cinder16:11
*** _cjones_ has joined #openstack-cinder16:13
*** BharatK has joined #openstack-cinder16:13
*** gouthamr has quit IRC16:14
*** jistr has quit IRC16:14
*** rushil_ has quit IRC16:18
*** dustins has joined #openstack-cinder16:22
*** teju has left #openstack-cinder16:24
*** e0ne is now known as e0ne_16:25
*** e0ne_ is now known as e0ne16:25
*** vilobhmm has joined #openstack-cinder16:25
openstackgerritVictor Stinner proposed openstack/cinder: Replace with next(it) for py3 compat
*** rushil has joined #openstack-cinder16:30
*** zhonghua-li has quit IRC16:31
*** annashen has joined #openstack-cinder16:32
*** annashen has quit IRC16:37
*** jordanP has quit IRC16:39
*** timcl1 has joined #openstack-cinder16:39
*** gouthamr has joined #openstack-cinder16:41
*** timcl has quit IRC16:42
*** amoturi has quit IRC16:44
e0nejgriffith: need to head to home after the meeting. if you want we could continue to discuss 'detailed views' in ~2 hours16:45
openstackgerritEarle F. Philhower, III proposed openstack/cinder: Implement Cinder Volume driver for HGST Solutions
*** rakesh_mishra has quit IRC16:52
*** diemt has joined #openstack-cinder16:52
dguryanovWhen third party CI should run? On every patch to cinder?16:53
jgriffithe0ne: sure16:53
*** dannywilson has quit IRC16:53
thingeedguryanov: yes16:53
dguryanovI see16:53
e0nedguryanov: yes16:53
thingeedguryanov: make sure to read
diemtthingee: I'd really appreciate if you could let me know the soak time before the Oracle iSCSI driver is moved to "ready" section16:54
thingeedguryanov: which links to
thingeediemt: I've typically done an arbitrary number. I'll take a closer look when I have a chance.16:54
openstackgerritVictor Stinner proposed openstack/cinder: Replace urllib and urllib2 with six.moves.urllib
thingeenumber of days that is16:55
diemtthingee: Thank you!!! I really hope the driver can meet the L1 deadline16:56
*** annegentle has quit IRC16:56
*** gouthamr has quit IRC16:56
*** BharatK has quit IRC16:56
*** timcl1 has quit IRC16:56
*** asselin has joined #openstack-cinder16:56
*** sreedharv_ has joined #openstack-cinder16:57
*** gouthamr has joined #openstack-cinder16:57
vilobhmme0ne : regarding
e0nevilobhmm: hi16:59
vilobhmmwill remove the 'if' statement in next patch…the motive was to check if the column already exist or not16:59
vilobhmmbut since we are creating it for first time it should be fine16:59
*** BharatK has joined #openstack-cinder16:59
e0nevilobhmm: why it could exist?16:59
vilobhmmwill it be ok with you if i remove it in next patch16:59
vilobhmmnot needed to be frank…just defensive prog16:59
e0nevilobhmm: is if it's not needed - we should to remove it from the code17:00
e0neimo, defensive programming is good, but not need to use it everywhere17:01
e0neneed to go offline. bb in ~2 hours17:01
vilobhmmupdated the patch17:02
openstackgerritVilobh Meshram proposed openstack/cinder: Nested Quota : Create allocated column in cinder.quotas
vilobhmmplease check when you are back17:02
vilobhmme0ne : ^^17:03
*** vokt has quit IRC17:03
tbarronthingee: w.r.t., what are the "cleanup" standards?  Just the hacking/log guidelines?17:03
*** amoturi has joined #openstack-cinder17:03
*** timcl has joined #openstack-cinder17:04
thingeetbarron: jgriffith asked if we can keep stuff in his original etherpad ...sorry if I'll correct that.17:05
*** spradeepv_ has joined #openstack-cinder17:07
*** gouthamr has quit IRC17:08
tbarronthingee: thanks.  so the work here is to convert codebase to resource-tag style logging then as I understand it17:08
tbarronjgriffith: that right?17:09
tbarronjust need to understand what I would be signing up for :-)17:09
flip214thingee: I'm sorry, my kids kept me away from the meeting. Reading through the logs, do I understand correctly that when I get the CI for DRBD working I may move the block in from "Not ready" to "Ready"?17:09
*** gouthamr has joined #openstack-cinder17:11
*** e0ne has quit IRC17:11
*** lpetrut has joined #openstack-cinder17:11
*** timcl has quit IRC17:19
*** spradeepv_ has quit IRC17:21
*** daneyon has quit IRC17:21
*** annashen has joined #openstack-cinder17:23
tbarronjgriffith: OK, I see what to do in the etherpad now.17:23
tbarronjgriffith: I'm willing to help ...17:23
jgriffithtbarron: sweet17:24
*** pradipta has quit IRC17:24
tbarronjgriffith: are you going to break out the work into modules that we can sign up for?17:25
jgriffithtbarron: nope, you can do that on your own :)17:25
*** markus_z has quit IRC17:25
tbarronjgriffith: kk :-)17:26
jgriffithtbarron: just pick a file and go with it... and mark on the etherpad that you're working on it so somebody else doesn't come along and duplicate your efforts17:26
flip214thingee: thanks.17:28
openstackgerritEarle F. Philhower, III proposed openstack/cinder: Implement Cinder Volume driver for HGST Solutions
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
*** belmoreira has joined #openstack-cinder17:30
*** aix has quit IRC17:33
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
*** krtaylor has quit IRC17:37
*** zul has quit IRC17:38
*** changbl has joined #openstack-cinder17:40
*** daneyon has joined #openstack-cinder17:40
*** jdandrea has joined #openstack-cinder17:40
*** zul has joined #openstack-cinder17:40
*** Yogi11 has quit IRC17:41
*** IlyaG has joined #openstack-cinder17:41
jdandreaGiven a host with SSDs and HDDs, is there a way for a scheduler plugin to specify which of those types (or even which specific drive) should be used on that host? If this is in the docs, bear with me. Still looking for that info ...17:41
jdandreaRephrasing: Is there a way for a scheduler plugin to specify "Use a SSD on host X" or "Use SSD drive Y on host X."17:42
*** garthb_ has quit IRC17:44
*** garthb has quit IRC17:44
*** rushil has quit IRC17:45
*** agarciam has joined #openstack-cinder17:47
*** sgotliv_ has quit IRC17:49
*** dannywilson has joined #openstack-cinder17:56
*** annegentle has joined #openstack-cinder17:56
*** harlowja has quit IRC17:58
*** krtaylor has joined #openstack-cinder17:59
*** dannywil_ has joined #openstack-cinder18:00
*** dannywilson has quit IRC18:00
*** harlowja has joined #openstack-cinder18:02
*** bswartz has quit IRC18:02
*** bswartz has joined #openstack-cinder18:02
*** annegentle has quit IRC18:02
*** gouthamr has quit IRC18:09
*** annashen has quit IRC18:09
*** gouthamr has joined #openstack-cinder18:09
*** annegentle has joined #openstack-cinder18:12
*** daneyon has quit IRC18:14
*** daneyon has joined #openstack-cinder18:16
*** annashen has joined #openstack-cinder18:17
*** angela-s has joined #openstack-cinder18:18
*** rushiagr is now known as rushiagr_away18:23
*** timcl has joined #openstack-cinder18:26
*** daneyon has quit IRC18:26
openstackgerritJohn Griffith proposed openstack/cinder: Add a base FC target driver
*** ronenkat_ has joined #openstack-cinder18:34
*** agarciam has quit IRC18:34
*** gouthamr has quit IRC18:38
*** cbader has quit IRC18:39
*** gouthamr has joined #openstack-cinder18:40
*** dustins has quit IRC18:40
*** dustins has joined #openstack-cinder18:43
*** gouthamr has quit IRC18:46
*** julim has quit IRC18:49
*** e0ne has joined #openstack-cinder18:52
e0nevilobhmm: thanks for fix! let's wait for CI18:53
*** rhagarty_ has joined #openstack-cinder18:53
*** rhagarty_ has quit IRC18:54
*** Lee1092 has quit IRC18:55
*** gouthamr has joined #openstack-cinder18:57
*** e0ne has quit IRC18:57
*** zhonghua-lee has joined #openstack-cinder19:00
*** e0ne has joined #openstack-cinder19:00
*** e0ne has quit IRC19:05
vilobhmme0ne : thanks for reviewing it :) sure19:06
*** timcl has quit IRC19:11
*** timcl has joined #openstack-cinder19:14
*** timcl1 has joined #openstack-cinder19:17
*** vokt has joined #openstack-cinder19:19
*** Yogi1 has joined #openstack-cinder19:19
*** timcl has quit IRC19:19
*** zhonghua-lee has quit IRC19:20
openstackgerritJohn Griffith proposed openstack/cinder: Add secondary account capability to SolidFire
*** gouthamr has quit IRC19:31
*** gouthamr has joined #openstack-cinder19:39
*** rushil has joined #openstack-cinder19:42
*** BharatK has quit IRC19:43
*** sreedharv_ has left #openstack-cinder19:49
*** rushil has quit IRC19:51
*** lpabon has quit IRC19:53
*** aix has joined #openstack-cinder19:53
hemnajgriffith, thingee so the os-brick patches now are getting CI'd against cinder.19:53
hemnajust verified that it's running and is working19:53
hemnafor example,
thingeehemna: excellent!19:54
hemnathat one is an iSCSI change as well, which would directly affect most of Cinder drivers as well as lvm19:54
hemnathe logfile from the recheck against that patch that shows devstack pulling os-brick's patch and testing it against cinder, etc.19:54
hemnaasselin_, did all the heavy lifting for us.19:55
patrickeastawesome, so it should be ready to point our 3rd party ci’s at os-brick now, right?19:56
asselinpatrickeast, yes19:56
asselinpatrickeast, you'll need a slightly different job19:56
hemnapatrickeast, yah, we are going to do that next....once our network is back up19:56
asselinours is running on silent...but out network is down19:57
patrickeastasselin: what needs to change with the job?19:58
*** annashen has quit IRC19:59
earlephilhowerthingee:  I just added HGST driver discussed at cinder-meeting to the cinder-liberty-drivers etherpad in the Not Ready: section.  If I erred, apologies and I'll back it out post-haste.19:59
*** e0ne has joined #openstack-cinder20:00
*** timcl1 has quit IRC20:01
*** adurbin__ has quit IRC20:01
openstackgerritDmitry Guryanov proposed openstack/cinder: smbfs: fix invalid check for smbfs_user_ratio correctness
*** IlyaG has quit IRC20:03
*** rushil has joined #openstack-cinder20:04
jungleboyjasselin: There?20:04
asselinjungleboyj, i am20:04
*** julim has joined #openstack-cinder20:04
jungleboyjasselin: Hey man.  Have you seen people struggling with this on their CI systems:  "sudo: no tty present and no askpass program specified"20:05
asselinjungleboyj, yes20:05
jungleboyjI know that that can happen if you don't have sudo set up right.  Is there anything special needed for it to work with devstack/CI ?20:05
asselinasselin, well actually, no, I haven't seen ppl struggling, but yes I know the answer20:06
asselinyou need to make sure your sudoers file is setup right20:06
jungleboyjasselin: Cool.  That was what I thought.20:06
openstackgerritSean Chen proposed openstack/cinder: Tintri driver to manage existing backend storage objects
*** rushil has quit IRC20:09
jungleboyjasselin: Looks like it is failing trying to run Debian commands.20:10
jungleboyjsudo DEBIAN_FRONTEND=noninteractive20:10
asselinjungleboyj, what's failing? whcih user is running the sudo command?20:10
earlephilhowerjungleboyj:  If you're trying to use a static instance for running jobs, the devstack-gate may be eating the sudo permissions as part of its' devstack-gate scripts.  For static host I think I had to add stack as sudoer in another file that's not touched by the ansible/puppet stuff.20:10
*** gouthamr has quit IRC20:10
asselinjungleboyj, what earlephilhower said is true. the tests remove sudo permissions for the stack user20:11
jgriffith"HOLY I"m running automated six patching tools"!!!!,n,z20:11
jgriffithnote.. folks need to NOT just blindly + these20:11
asselinjungleboyj, but don't reuse nodes....20:12
jgriffithand frankly we need to decide if this crap is even worthwhile20:12
*** amoturi has quit IRC20:12
smcginnisjgriffith: DuncanT had challenged an earlier patch.20:12
smcginnisHe actually does have an overall plan.20:12
earlephilhowerjungleboyj, asselin: +1  Definitely don't reuse hosts, nothing but pain.  I spent 3 days and then gave in and had dynamic spinup running in a few hours...20:13
jgriffithsmcginnis: sure... that's not my point20:13
smcginnisjgriffith: Sure, I get it.20:13
*** rushil has joined #openstack-cinder20:13
jungleboyjasselin: Ok.  This was a question from Sasi who is trying to get the CI system reconfigured for GPFS.  Going to send a note back to him.20:13
hemnaholy carp20:14
jgriffithFor examples20:14
smcginnisjgriffith: Just wanted to point out it's not the case of someone finding a cool tool and just throwing up the output. :)20:14
hemnaare we in a rage to move to py3 ?20:14
jgriffithpersonally I think we should avoid the use of six if/when possible20:14
jgriffithit just introduces the next compatability thing to go through and scrub some day in the future20:14
DuncanTsmcginnis: the plan has already grown new legs, and more six... I'm not happy tbh20:14
jgriffithBesides... EVERYBODY knows python3 is dead!! :)20:14
jgriffithpython4 will be out soon and revert everything back :)20:15
DuncanTsmcginnis: I'm back to wanting to see a finished tree tbh20:15
jgriffithSo I'm fine with most of them, IFF there some testing/gating added20:15
jgriffithbecause a few of these we've already done once :)20:15
smcginnisDuncanT: So should we push that all get combined into a megapatch that gets to the end game?20:16
jgriffithsmcginnis: hush now20:16
smcginnisDuncanT: That doesn't seem ideal either.20:16
jgriffithsmcginnis: I just threw up in my mouth a little20:16
jungleboyjasselin: earlephilhower Thanks for confirming that it is a sudoers issue.20:16
jgriffithhemna: that's what I'm saying :)20:16
smcginnisjgriffith: :)20:16
jgriffithhemna:  but as long as it's taken for py3 adoption I probably won't see py4 in my lifetime anyway20:17
jgriffithBy then I'll have rewritten all of Cinder in GoLang anyway20:17
DuncanTsmcginnis: no, a GitHub tree that has all these patches on and works with python 320:17
jgriffithOr 'C'20:17
hemnaheh yah, the guy writing that article says py4 in 202320:17
smcginnisDuncanT: OK, I like.20:17
smcginnisDuncanT: That would be nice to see.20:18
DuncanTjgriffith: I've got a working cinder-backup in c, much better performance20:18
smcginnisHe's going to need a lot of help. He's trying to do this for several projects.20:18
DuncanT(For given values of working)20:18
smcginnisSo... should we block all of these patches and discuss a better plan in the next meeting?20:19
smcginnisSeems like a bigger thing to decide openstack-wide "screw py3, wait for py4" :)20:20  *sigh*20:20
*** gouthamr has joined #openstack-cinder20:21
smcginnisjgriffith: Just for my education - what's the issue with six.moves.range?20:21
smcginnisHe does call out in the commit why he's not going with the built in range.20:22
*** haypo has joined #openstack-cinder20:22
smcginnisThough I could see the argument for not using six.iteritems() applying here as well.20:22
haypojgriffith, hi. i'm the python3 guy :)20:22
jgriffithhaypo: Hi!! :)20:22
jgriffithNice to meet ya20:22
smcginnishaypo: Cool!20:23
haypojgriffith, i'm surprised that you approved whereas the patch has no vote yet20:23
haypojgriffith, i described a little bit my plan for py3 in cinder, in the latest comment of
jgriffithhaypo: I do that a lot :)20:24
jgriffithhaypo: I trust ya... and I cheated and looked at Zuul20:24
jgriffithhaypo: notice it's there now :)20:24
tbarronSo is there a place for oslo_log in unit tests?20:24
tbarronnot mocked out, but LOG.debug, etc. run from the unit test itself?20:25
jgriffithhaypo: and I actually only meant to +2, not a +A20:25
jgriffithtbarron: yeah, in oslo.log20:25
jgriffithwhere it belongs20:25
tbarronI see this e.g. in /cinder/tests/unit/api/contrib/test_backups.py20:25
tbarronjgriffith: :-)20:25
jgriffithhaypo: Ahh... cool20:26
jgriffiththat's great20:26
*** annegentle has quit IRC20:26
jgriffithhaypo: ^^20:26
haypojgriffith, that's what i expected (+2, not +A)20:26
jgriffithhaypo: can I offer some advice?20:26
jgriffithhaypo: and really.. you can say "no" if you like ;)20:26
jgriffithhaypo: haha... I'll take silence as a "yes" just because I like to hear what I have to say to myself :)20:27
jgriffithhaypo: rather than wait and batch the py34 tests until everything is fixed do it in segments20:28
tbarronjgriffith: my naive view was that, except for unit testing oslo.log itself, unit tests shouldn't actually be doing logging.20:28
haypojgriffith, i'm not interested to write hacking rules, i'm more interested to see tests which pass20:28
jgriffithhaypo: and ignore/regex the errors you've fixed20:28
jgriffithhaypo: I agree with you WRT the hacking rules20:28
smcginnistbarron: +120:28
jgriffithhaypo: that would be my last choice20:28
tbarronthey should likely be mocking it out, or just letting the code under test log, but should not do logging themselves.20:28
jgriffithhaypo: I think your plan of a py34 test if fantastic20:28
smcginnishaypo: Glad to see the tox -e py34 work.20:28
jgriffithand could be done similarly to what was tried with pylint20:29
tbarronsmcginnis: I'll likely submit some reviews with that cleanup then.20:29
jgriffithtbarron: that's not a naive view, that's a TRUE statement of fact!!!!20:29
smcginnistbarron: Cool, thanks Tom.20:29
*** rushil has quit IRC20:29
haypojgriffith, what do you mean by "do it in segments"?20:30
jgriffithhaypo: two options20:30
tbarronI know eharney has been looking for weird, unexplained external effects between unit tests20:30
jgriffithhaypo: run the test before 19016020:30
tbarronand this could be one candidate I guess.20:30
jgriffithhaypo: then run it after20:30
jgriffithhave a regex for everything that "still fails between them"20:31
*** rushil has joined #openstack-cinder20:31
tbarronjgriffith: glad to know the world is round as I thought :-)20:31
jgriffithyou know... like a white-list/black-list of errors20:31
jgriffithhaypo: or the other option...20:31
jgriffithhaypo: run the test against master, compare it to what you get on a patch20:31
jgriffithhaypo: verify the error count does not increase20:31
jgriffithhaypo: the second option is heavier but probably MUCH better20:32
*** lpetrut has quit IRC20:32
*** IlyaG has joined #openstack-cinder20:32
haypojgriffith, for nova, glance_store, neutron, etc. we are using a whitelist of tests which pass20:32
*** julim has quit IRC20:33
jgriffithhaypo: your call20:33
jgriffithhaypo: I think that's fine as well20:33
*** gouthamr has quit IRC20:33
*** akerr has quit IRC20:33
jgriffithhaypo: I was just saying I'd like to see that turned on in conjunction with the patches/fixes you're submitting20:33
haypojgriffith, currently, nova has exactly 1 test running on python 3.4 :-D
*** dustins has quit IRC20:34
jgriffithhaypo: because there are likely patches in the queue that will just introduce the issue you just fixed20:34
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
jgriffithhaypo: I can prove that by writing a few and submitting them myself20:34
jgriffithhaypo: yeah, I'm aware, and that's why I get a bit skeptical of the py3 stuff everytime it comes up (since Portland)20:34
jgriffithhaypo: but... you're here now :)20:35
jgriffithhaypo: and you're already off to a much better start than we've had in the past20:35
*** e0ne has quit IRC20:35
haypojgriffith, if you review patches, you should now be aware of python3 traps like ;-)20:36
jgriffithhaypo: "should" is the key word there20:36
haypojgriffith, it's not a big deal if python3 compatibility "regressions" are added, it's easy to fix them20:36
jgriffithhaypo: we're all human20:36
haypoi even have a tool for that ;)20:36
jgriffithhaypo: yes, but my point is that's what we've "been doing" for almost 2 years now20:37
jgriffithhaypo: your round of py3 patches is not the first20:37
*** dannywilson has joined #openstack-cinder20:37
jgriffithhaypo: and my point is that without something checking/monitoring it won't be the last20:37
jgriffithwe'll NEVER get there this way20:37
haypojgriffith, i'm not sure that it was feasible to port cinder 6 months ago. there were big dependency issues like eventlet which were not compatible with python 320:37
*** hemna is now known as hemnafk20:38
jgriffithSo basicly what you're saying to me is "thanks for your opinion, I don't care what you think.. so piss off"20:38
jgriffithwhich is fine20:38
haypojgriffith, the last dep not py3 compatible was  rtslib-fb. i ported it, the release 2.1.56 supports python 320:38
*** gouthamr has joined #openstack-cinder20:38
haypojgriffith, i'm not saying that20:39
jgriffithhaypo: Oh... ok, wasn't sure20:39
jgriffithhaypo: like I said, that's "ok too"20:39
hayposorry, i'm doing something else at the same time, i have to reread the discussion20:39
jgriffithI wouldn't blame you if you were; it's perfectly legit20:39
*** dannywil_ has quit IRC20:39
haypoi also helped to port many oslo code. my point is that we are close to being able to have a minimal py34 test which pass20:40
*** xyang has quit IRC20:40
*** IlyaG has quit IRC20:40
jgriffithhaypo: understood... and I think that's AWESOME20:41
jgriffithhaypo: I'm certainly not saying I don't like or appreciate the effort20:41
jgriffithhaypo: I'm just saying I *personally* would love to see something to keep us from just re-inroducing things you fix as much as possible20:41
haypoi'm running tox -e py34 to check :)20:41
jgriffithhaypo: and I stand by my votes/opinion on things like iter values20:41
haypowith you should be able to create the py34 venv20:44
jgriffithhaypo: Oh, crap!20:44
jgriffithI hadn't even seen that!20:45
haypojgriffith, it's the last dep not compatible with py3, mysql-python20:45
jgriffithhaypo: yeah.. thanks for pointing it out20:45
jgriffiththingee: smcginnis DuncanT ^^20:46
jgriffiththat's well worth landing IMHO20:46
*** diemt has quit IRC20:47
smcginnisI'd like to see vincent's question answered.20:47
smcginnisI'm assuming there will be some differences. No?20:47
smcginnisOr ideally all libraries are available on either and we're good. haypo, do you happen to know?20:48
*** ganso_ has quit IRC20:48
hayposmcginnis, rtslib-fb was supposed to be the latest dep not compatible with py3, and i just ported it20:49
hayposmcginnis, that's why i started to flood you with py3 patches :)20:50
smcginnishaypo: Works for me. Thanks!20:50
jgriffithhaypo: about the use of L20:52
jgriffithhaypo: does my comment/concern make any sense at all?20:52
haypojgriffith, like 123L?20:52
jgriffithhemnafk: ^^20:52
jgriffithhaypo: yes20:52
jgriffithseems only HP and EMC are using them20:53
jgriffithhaypo: and honestly the one HP has shouldn't be there anyway20:53
jgriffithhaypo: they should be using the unit lib20:53
haypojgriffith, "My point is, for systems running on 2.X still, they may have made these Longs for a reason and this patch breaks that." well, i should review my patch, but usually there is no reason for using a long20:53
*** thangp has quit IRC20:54
jgriffithhaypo: ok.. withdrew that one20:56
jgriffithhaypo: standing by the other ones still though :)20:56
haypojgriffith, hum, i don't see any modified code which looks to rely on the exact type20:56
*** daneyon has joined #openstack-cinder20:56
jgriffithhaypo: correct, that's why I just went ahead and change from -1 to +1 on that patch20:56
haypojgriffith, the exact type (int or long) matters when the type is used. i saw such code to choose a function depending on the type. i saw that in glance_store and i added a "if six.PY2: test.append(long)" for that (something like that)20:57
*** annashen has joined #openstack-cinder20:57
haypojgriffith, maybe i missed something. well, that's the purpose of reviews :)20:57
patrickeastasselin: hemnafk: ok, the pure storage ci should be watching brick changes now, i modified the jobs to add the export DEVSTACK_PROJECT_FROM_GIT='os-brick' as needed20:57
asselinpatrickeast, ok cool20:57
patrickeastasselin: hemnafk: i’ll keep an eye on it, but please do let me know if you see it doing something weird20:58
mriedemsdague: kind of thinking out loud question, if i want to add objects to os-brick for cinder to use, but need to test it all together, would the Depends-On tag handle that for me?20:58
mriedemso the cinder change to use the objects from os-brick would have depends-on the os-brick change that adds the objects20:59
mriedemthen iterate until it's working, release os-brick with the objects, pin min version of os-brick so cinder can use it and then cash in20:59
patrickeast:( now i need to figure out how to get more FC worker nodes21:00
patrickeastour FC tests are backed up almost 10 hours21:01
thingeepatrickeast: whoa21:01
patrickeasti’ve only got a single ‘live’ one running at a time21:01
patrickeastand like 8 iscsi21:01
patrickeastso the test results are all blocked21:01
*** jdurgin1 has joined #openstack-cinder21:01
*** gouthamr has quit IRC21:02
smcginnispatrickeast: Same.21:03
smcginnisWaiting for a gate outage so we can catch up. ;)21:03
haypoi tried tox -e py34 on cinder. i had to replace mysql-python with pymysql, run with operations six_moves,raise,basestring,iteritems,long. with 2 or 3 more changes, i'm able to run cinder.tests.unit.test_quota and only 6 tests failed on a total of 71 tests21:03
*** belmoreira has quit IRC21:04
patrickeasti’ve got some more hardware, but am having a hell of a time trying to get the networking correct adding them at my existing openstack installation21:04
*** jdurgin has quit IRC21:04
thingeepatrickeast: ditto21:06
openstackgerritJohn Griffith proposed openstack/cinder: Add a base FC target driver
*** Yogi1 has quit IRC21:13
asselinpatrickeast, thingee smcginnis so I though of a way to ease the issue you're having ^^21:13
asselinat least for those using zuul21:14
asselinbut not 100% kosher21:14
earlephilhowerpatrickeast: "having a hell of a time getting the networking correct"....makes me feel better about spending a XX late nights setting up my own boxen...seems even the experts have trouble w/OS networking.  :)21:14
jgriffithnova-network FTW!!!21:14
patrickeastasselin: sounds perfect21:14
asselinearlephilhower, I gave up a long time ago21:14
patrickeastjgriffith: i switched to nova-network and am still having problems!21:15
patrickeasthalf switched21:15
thingeejgriffith: yeah still running with nova network21:15
jgriffithpatrickeast: dang!21:15
patrickeasti left one by itself, and started a ‘new’ cloud21:15
*** rushil has quit IRC21:15
asselinearlephilhower, I stil have patrickeast's and jgriffith's suggestions saves somewhere, but cringe to think about it21:15
jgriffithOh... so it's the "other" half you're having trouble with21:15
patrickeastthat was using nova-network21:15
patrickeastno no21:15
*** ociuhandu has quit IRC21:15
patrickeastboth are broken21:15
jgriffithpatrickeast: you're having "half" the trouble :)21:15
patrickeastin their own special ways21:15
* jgriffith doesn't believe it!21:15
*** amoturi has joined #openstack-cinder21:15
patrickeastnova-network got me like 99% of the way there21:16
asselinpatrickeast, thingee smcginnis the idea is to run a 2nd pipeline like check to post results...similar to how jenkins uses check and gate.21:16
patrickeastbut the dhcp server refuses to let go of ips sometimes21:16
jgriffithnova-net is a champ!  As long as you're on Juno or earlier21:16
patrickeast:( crap21:16
patrickeasti did kilo21:16
jgriffithpatrickeast: hmm21:16
jgriffithpatrickeast: yeah, kilo and even Juno has a few "bugs" that are new21:16
asselinpatrickeast, thingee smcginnis this works since we're not voting, otherwise each pipepline result would overwrite the other21:17
jgriffithpatrickeast: because everybody "wants" neutron to be the future so bad21:17
jgriffithpatrickeast: lipstick on a pig IMO21:17
* asselin is on juno nova-networking21:17
patrickeastjgriffith: lol21:17
jgriffithSoon enough NONE of use will be on nova-net21:17
* jgriffith checks his git repo for that nova-network fork he started21:18
patrickeastthe problem i had with neutron is that it doesn’t really work for the model i want21:18
patrickeasti dont want to tunnel iscsi networks through my network node21:18
jgriffithpatrickeast: the problem I have is it's not very frikin helpful :)21:18
jgriffithIf I were a network guru I wouldn't need/use openstack :)21:18
jgriffithpatrickeast: yeah!  That's one that is CRAZY IMO21:19
patrickeasti’ve come to realize that there is a certain level of complexity in your network configuration where neutron probably makes it easier and is the best option21:19
patrickeastbut my ci system is not that21:19
jgriffithpatrickeast: I mean.. really? Funnel traffic through a node?  Oh... well use lbalancer.21:19
jgriffithWTF... I don't want to!  Why do I need a bunch of nodes just for networking :)21:19
jgriffithcrazy crazy21:19
patrickeastdefinitely crazy21:19
smcginnisIf you're networking needs are simple, your networking management should be able to be simple.21:20
jgriffithpatrickeast: yeah.. in all seriousness, I know there are people that need/want that21:20
patrickeastsmcginnis: +121:20
jgriffithI'm just not one of them :)21:20
*** anshul has joined #openstack-cinder21:20
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
jgriffithsmcginnis: well said21:20
patrickeastnova network with multi-host and flat dhcp is almost perfect for what i want21:20
patrickeastget my vm’s bridged directly to the 10Gb iscsi21:20
jgriffithsmcginnis: patrickeast now that there's big-tent I see no reason not to spend some time making a simple networking project21:21
jgriffithwith ZERO plugins21:21
patrickeastwe could call it proton21:21
jgriffithoh... nice21:21
smcginnisI'm in! :)21:21
jgriffithI was thinking SNFSP21:22
earlephilhowerProton == positive charge! +121:22
jgriffithbut I like proton.. A LOT21:22
smcginnisOr SNFSM21:22
smcginnis/people/minds/ :)21:22
smcginnis+1 for proton.21:22
* jgriffith resembles that remark21:22
*** jungleboyj has quit IRC21:24
patrickeastso i’ve asked in a bunch of places, but not in here yet, any of you guys know why this might happen?
smcginnispatrickeast: I had something funky like that.21:25
smcginnisTurned out it was servicing external DHCP requests (yikes!)21:26
patrickeastsmcginnis: :O and were able to fix it?!21:26
patrickeastoh interesting21:26
smcginnisI never figured out what the root cause/misconfiguration was, but I just used ebtables to block it.21:26
smcginnisNot sure if it's all related to this instance.21:26
haypojgriffith, back. i attended a swift meeting to discuss py3 too ;)21:27
jgriffithhaypo: you're going to be the py3 goto person!21:27
jgriffithpatrickeast: interesting... the case smcginnis has come up before21:28
jgriffithpatrickeast: might be worth trying to find that ID21:28
jgriffithsmcginnis: I did the same thing... bridged the wrong interface... took down our entire lab21:28
jgriffithsmcginnis: it was AWESOME!!!!!!!!!!!!!!!!!!!!!!!!!!!!21:28
haypojgriffith, for me, it's fine if some parts of cinder are not compatible with py3 immediatly, the idea is more to port code files one by one, but always ensure that there is no regression21:29
smcginnisjgriffith: Mine just limped along filling the syslog for weeks until I noticed.21:29
haypojgriffith, a whitelist of tests in tox.ini is convenient for that21:29
smcginnisjgriffith: But there is a guy here who also took down the lab.21:29
jgriffithhaypo: I think you and I agree21:29
haypojgriffith, if we compare the number of errors, it may block patches whereas nobody would have to bandwith to take care of py3 on specific parts of the code21:30
jgriffithhaypo: I know I had mentioned the diff approach, but the more I thought about that the worse it seemed21:30
jgriffithhaypo: meh... non-voting to start, get some stability;  then at some point... that's life21:30
*** simondodsley has quit IRC21:31
*** bnemec has quit IRC21:31
*** smcginnis is now known as smcginnisafk21:31
haypojgriffith, to be more factual, i expect a voting gate in 1 month21:32
*** jdurgin has joined #openstack-cinder21:32
haypoi don't really care if tox -e py34 only runs a single test21:32
*** jdurgin1 has quit IRC21:32
haypothe point is just to ensure that we cannot go back (as you wrote, i already happened in cinder)21:32
jgriffithhaypo: perfect21:33
jgriffithhaypo: that makes me happy21:33
haypojgriffith, for nova, i wrote a spec. it takes a lot of time to write a spec, so i didn't propose a spec for cinder, ceilometer, glance, swift, etc. (i'm porting all these app to py3)21:34
*** akerr has joined #openstack-cinder21:35
haypobut i understand that the plan can be non obvious if you don't spend you whole day on python 3 :-D21:35
haypojgriffith, is there something lighter than spec?21:35
*** vokt has quit IRC21:36
*** vokt has joined #openstack-cinder21:37
jgriffithhaypo: yeah, blueprint21:38
jgriffithhaypo: that would be more appropriate IMO, and much lighter weight21:38
haypook, i may do that21:40
jgriffithhaypo: cool, let me know if you need help21:41
*** akerr has quit IRC21:44
*** akerr has joined #openstack-cinder21:45
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
*** amoturi has quit IRC21:51
*** marcusvrn has quit IRC21:54
*** mriedem is now known as mriedem_away21:54
*** crose has quit IRC21:55
haypojgriffith, ok. i have to go. i will read comments on my patches tomorrow and try to reply. i hope that you know have a better view of the overall plan for cinder21:56
jgriffithhaypo: I do thanks21:56
jgriffithhaypo: have a good evening21:56
*** anshul has quit IRC22:01
openstackgerritEarle F. Philhower, III proposed openstack/os-brick: Add HGST Solutions connector
*** pschaef has joined #openstack-cinder22:03
*** pschaef has quit IRC22:04
*** changbl has quit IRC22:05
* patrickeast looking22:05
*** __afazekas has quit IRC22:05
*** salv-orlando has joined #openstack-cinder22:05
patrickeastasselin: nice, thats a good addition22:06
*** daneyon has quit IRC22:06
patrickeastasselin: i had been wondering how to segragate some hba’s for dev/playing around with stuff22:06
patrickeastnow i can just change the variable22:06
asselinpatrickeast, thanks...really liking having this code shared. big thanks for doing that22:06
asselinyup :)22:06
asselinwas a bit tricky to do it...probably a better way, but this works22:07
patrickeastyou want i should merge it or wait for other folks?22:07
patrickeastim not sure who all is watching/using these scripts right now22:07
asselinpatrickeast, I have mine setup to use a specific commit, so merging won't break my setup22:07
patrickeastmerge it is :D22:08
asselinyeah...manual testing. will do the official testing after you merge (to save time)22:08
patrickeastasselin: hehe yea, speaking which i put some tasks on my backlog to set up some automated tests for these, would be nice to test just doing the FC passthrough and stuff on a VM real quick22:10
patrickeastshouldn’t be too hard to set up i don’t think22:10
patrickeastbut need to get this other hardware online first :)22:10
asselinpatrickeast, yeah....on the back burner22:11
asselinshould do it sooner than later....22:11
*** alexpilotti has quit IRC22:13
*** akl_ has joined #openstack-cinder22:15
*** alexpilotti has joined #openstack-cinder22:17
*** salv-orlando has quit IRC22:18
*** akerr has quit IRC22:23
openstackgerritEarle F. Philhower, III proposed openstack/cinder: Implement Cinder Volume driver for HGST Solutions
*** haypo has left #openstack-cinder22:26
*** harlowja has quit IRC22:27
*** akerr has joined #openstack-cinder22:28
*** harlowja has joined #openstack-cinder22:32
*** rushil has joined #openstack-cinder22:33
*** jnicklin has quit IRC22:34
*** rushil has quit IRC22:35
*** salv-orlando has joined #openstack-cinder22:36
*** akerr has quit IRC22:40
*** leeantho has quit IRC22:41
*** setmason has quit IRC22:43
*** daneyon has joined #openstack-cinder22:44
*** daneyon has quit IRC22:45
*** daneyon has joined #openstack-cinder22:45
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
*** leeantho has joined #openstack-cinder22:52
*** krtaylor has quit IRC22:54
*** akl__ has joined #openstack-cinder22:58
*** Adriano__ has quit IRC23:01
*** Adriano_ has quit IRC23:01
*** daneyon has quit IRC23:04
*** vokt has quit IRC23:06
*** daneyon has joined #openstack-cinder23:07
jgriffithasselin: thingee DuncanT smcginnisafk hemnafk tbarron patrickeast anteaya
jgriffith^^ Mabye we should think about getting a collective thing going with a bunch of us in the coming weeks to start hammering this stuff out23:08
tbarronjgriffith: agreed.  And FWIW I also find it annoying when logs download when clicked when one wants to view them.23:09
* patrickeast gets his pitchfork23:09
patrickeastjgriffith: yea its gotten pretty bad23:09
anteayawell the logs can't download when clicked23:09
anteayaso here is my suggestion23:10
anteayatake one system on the list23:10
patrickeastjgriffith: the number of ci systems that are voting in cinder has gone down by almost half since the deadline in Kilo23:10
anteayaand have someone agree to approach the operators23:10
asselinanteaya, +123:10
jgriffithpatrickeast: you noticed that as well :)23:10
jgriffithanteaya: +123:10
asselinI did this for 2 operators with positve success23:10
anteayastart off by ensuring they know to access the third party requirements23:10
anteayait is possible that they want to comply and don't know how23:11
anteayahelp them find the answer or find someone who knows the answer23:11
jgriffithasselin: I'd also like us to at least somewhat standardize on the directory structure and contents of the logs23:11
asselinand offered to help, by e.g. linking to FAQ which shows how to make logs downloadable23:11
anteayainvite them to attend a third party meeting23:11
asselinjgriffith, agree..I vote for ....drum roll please....what jenkins does23:11
anteayaif you get no response or a dismissal then note that23:11
jgriffithasselin: yeah23:11
anteayaI bet at least 50% of these issues will be addressed through a friendly conversation23:12
anteayaand clean up is good23:12
jgriffithanteaya: hmmm... you sure you wanna do that?23:12
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
patrickeastso, whos going to run point and contact the operators?23:12
asselinI would also like ot see everyone use os-loganalyzer to filter log files23:12
asselinby log level23:12
patrickeastasselin: +123:12
jgriffithasselin: let's walk before we run :)23:12
vilobhmmjgriffith : ping..wheneven you are free23:12
asselinjgriffith, yea...add it to the end of the list :)23:12
anteayajgriffith: no I won't be approaching operators23:13
jgriffithvilobhmm: gimmie just a minute23:13
anteayaI suggest everyone take one account and try to talk to them23:13
vilobhmmsure sure np23:13
tbarrongotta run out but will check the backlog later tonight23:13
vilobhmmjgtiffith : ^^23:13
jgriffithanteaya: sounds good, divide it up23:13
anteayaI support anyone willing to help another operator23:13
anteayajgriffith: yes23:13
jgriffithOk, thanks everyone23:13
jgriffithno need for me to sit here and rant23:13
asselinjgriffith, one more thing to add to the list:23:13
asselinmake sure they're testing cinder patches23:14
patrickeasthaha yea23:14
jgriffithbut I do think the whole thing is somewhat broken/flawed and we should fix it23:14
asselinI've notified a few who were testing master on every cinder patch set23:14
jgriffithasselin: YES!!!23:14
patrickeasti think we got all of those fixed by now23:14
jgriffithI noticed that too!23:14
jgriffithI also found one testing LVM... but I can't find it again23:14
jgriffithvilobhmm: what's up?23:14
anteayajgriffith: I agree and am glad you spoke up23:15
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 iSCSI backend driver.
anteayajgriffith: if I can suggest, take you paste and transfer to an etherpad23:15
anteayathen everyone assign themselves one ci23:15
openstackgerritAlexey Khodos proposed openstack/cinder: NexentaStor 5 NFS backend driver.
vilobhmmjgriffith : who keeps track of instance snapshot ?23:15
asselinanteaya, +123:16
vilobhmmis it us ?23:16
anteayaand try to approach the operators, media of your choice23:16
anteayaand post results23:16
jgriffithanteaya: Oh...sorry.. yeah, already did23:16
anteayaah okay cool23:16
jgriffithI had to use paste as a "hop"23:16
jgriffithMy \n's in vim won't work direct from vim to ehterpad23:17
asselinpatrickeast, fc passthrough update passed my ci. I've enabled it for future patches23:17
anteayaha ha ha23:17
jgriffithdon't know why?23:17
patrickeastasselin: awesome23:17
anteayaI have no idea23:17
jgriffithso I hope... vim --> paste -->etherpad23:17
jgriffithstupd VIM23:17
anteayaha ha ha23:17
jgriffitherrr... s/hope/hop/23:17
patrickeastjgriffith: so you mention collecting stats on ci systems, if thats something we can automate we could integrate it with my scoreboard thing23:18
jgriffithpatrickeast: Yes!23:18
jgriffithpatrickeast: so there's two approaches I've been messing with23:18
patrickeastjgriffith: i think it was tenatively decided we are going to try and put that thing up on an infra host23:18
*** daneyon has quit IRC23:19
patrickeastjgriffith: while the dashboard/radar thing keeps getting worked on23:19
jgriffithpatrickeast: one uses Joe Gordons last-comment tool (with some mods) and gets data23:19
*** daneyon has joined #openstack-cinder23:19
jgriffithpatrickeast: the other is have everybody implement subunit and db23:19
jgriffithpatrickeast: I think to start going down the last-comment path is the way to go23:19
anteayapatrickeast: my hope is yes, to offer a spec to have it on infra, and then any new features can get added _after_ it has been stood up23:19
jgriffithpatrickeast: getting everybody to setup db and do everything is going to be near impossible23:20
vilobhmmjgriffith : who keeps track of instance snapshot QUOTA ? is it us ?23:20
anteayapatrickeast: is anyone willing to write that spec yet?23:20
jgriffithlsat-comment just queries gerrit23:20
patrickeastanteaya: have you and asselin talked since the ci working group meeting this morning?23:20
asselinanteaya, I volunteered23:20
patrickeastoh ok23:20
jgriffithvilobhmm: yes23:20
anteayaasselin: okay good23:20
jgriffithvilobhmm: if by "us" you mean Cinder23:20
anteayaasselin: do toss up a draft and I can review it23:20
vilobhmmus = Cinder23:20
asselinanteaya, will do23:20
vilobhmmjgriffith : ^^23:21
jgriffithvilobhmm: yep, that's us23:21
patrickeastjgriffith: so the last comment thing sounds pretty cool, i’m not sure i know what you mean by implement subunit and db23:21
jgriffithvilobhmm: we track all of our quotas23:21
anteayaasselin: thank you23:21
jgriffithvilobhmm: it's a bit of a brain bender23:21
jgriffithvilobhmm: but you'll see all the reserve/update/commit etc in cinder.volume.api23:21
jgriffithand a few other places23:21
vilobhmmso the snapshot quota is overloaded with volume-snapshot, instance-snapshot ?23:21
vilobhmmjgriffith : ^^23:21
jgriffithand there's some loopie abstractions in to deal with it all23:22
vilobhmmyes i do see that23:22
jgriffithno.. we shouldn't have anything like that (instance-snapshot stuff)23:22
* asselin bookmarks
vilobhmmjgriffith : then how do you distinguish between volume snapshot vs instance snapshot ? because in code it just says *volume snapshot*
vilobhmmcould not get this so thought will check with you23:24
jgriffithvilobhmm: we don't need to.. all of the quota stuff is done in "Cinder" or "Nova" for each set of resources respectively23:24
jgriffithvilobhmm: theyr'e "different"23:24
jgriffithvilobhmm: so cinder volume-snapshots for example are ONLY managed in Cinder and kept track of via cinder database23:24
jgriffithvilobhmm: same holds true for Nova and compute snapshots23:25
jgriffithor.. "instance" snapshots23:25
*** daneyon has quit IRC23:25
vilobhmmnova just manages these resources no *instance snapshot* quota mentioned
vilobhmmthis is the list of resources that nova manages from nova quota-show —tenant <id>23:25
jgriffithwell.. so that's one of my faviroite things :)23:25
jgriffithThat's not necessarily everything23:26
vilobhmmoh okay23:26
jgriffithjust like this isn't everything for Cinder23:26
jgriffithvilobhmm: but the other thing is I *think* the snapshot quota in Nova is actually an image thing23:27
*** akl_ has quit IRC23:27
jgriffithvilobhmm: because that's actually what it is23:27
vilobhmmyes it should be right but i could not even see it in the code …
jgriffithvilobhmm: and snap of an Instance that's bootfrom voluem is actually a cinder-snapshot23:28
jgriffithwhich I find highly annoying and kinda stupid the way it is now23:28
jgriffithvilobhmm: welcome to OpenStack Quota code23:28
jgriffithvilobhmm: I spent wayyy too much time on it a long time ago and swore if I ever had to work on it again it would be to rewrite it all23:29
vilobhmmyes it is confusing …so finally who owns *instance snapshot quota* becasue i couldn't find relevant code anywhere jgriffith23:29
jgriffithvilobhmm: the thing is I don't pretend that I could do better23:29
jgriffithvilobhmm: check glance23:29
jgriffithvilobhmm: but frankly I'm not sure... I would have to dig in the nova code to try and find that info23:29
vilobhmmjgriffith : you can make it better…I m sure abt it…glance doesn not maintain quota jgriffith23:30
vilobhmmso the mystery remains a mystery23:30
jgriffithvilobhmm: sometimes it's easiest to start at the "end"23:30
jgriffithvilobhmm: lookin around in nova/nova/api/openstack23:30
vilobhmmwill trace it23:30
vilobhmmthanks for your input23:30
vilobhmmone last thing since we are talking about quotas here23:31
vilobhmmjgriffith : for nested quota23:31
jgriffithvilobhmm: I warned you in Vancouver :)23:31
jgriffithvilobhmm: some people think I'm negative, I think I'm realistic :)23:31
vilobhmm:P hey but its an interestring topic to work on ; I think you are realistic …So i am adding a new Driver called as NestedQuotaDriver23:32
vilobhmmsimilar to above : ^^23:32
jgriffithIt's an AWESOME topic23:32
vilobhmmso i realized that in NestedQuotaDriver except setting quota values everything else will be same23:32
jgriffithsure, but that ONLY covers your db interactions23:33
jgriffithall the spaghetti and mess is still outside of that23:33
vilobhmmso I was thinking either to inherit it from DbQuotaDriver….or have an abstract class called Quota with all common functionality and let DbQuotaDriver and NestedQuotaDriver inherit from it23:33
jgriffithvilobhmm: Oh?23:33
vilobhmmwhats your take on it23:33
jgriffithvilobhmm: well... first I didn't think that would work23:33
vilobhmmyes spagheti and meat will be in the extension or contrib stuff23:34
jgriffithbut obviously I must be wrong23:34
vilobhmmmore specifically cinder/api/contrib/quotas.py23:34
jgriffithvilobhmm: so I think either of the approaches you mention could be ok23:34
jgriffithvilobhmm: it comes down to how much commonality you can extract as to which is better23:35
vilobhmmok…want to minimise code duplication23:35
jgriffithvilobhmm: yeah, exactly23:35
jgriffithvilobhmm: I just don't know what it looks like (code wise)23:35
*** BharatK has joined #openstack-cinder23:35
jgriffithvilobhmm: if I had to guess I would lean towards a base class with two options under it23:35
*** mdenny has quit IRC23:35
jgriffithbut that's just off of memory of the quota db code23:36
vilobhmmi also want to take that route23:36
jgriffithI could be completely wrong :)23:36
jgriffithvilobhmm: it makes sense, at least in concept23:36
vilobhmmreason bein tommorow if we get rid of something X or plan to add something Z we can inherit from the abstract base anyhow23:36
jgriffithvilobhmm: but also be warned, I'm of the opinion that inheritance is the most abused concept in software23:36
jgriffithvilobhmm: yep23:37
vilobhmmthat was helpful23:37
jgriffithit's more flexible for sure23:37
vilobhmmthanks jgriffith23:37
jgriffithvilobhmm: you bet, thank you23:37
jgriffithLooking forward to you having this all finished and working tomorrow :)23:37
*** krtaylor has joined #openstack-cinder23:38
jgriffithvilobhmm: ^^ that was a joke of course :)23:38
*** david-lyle has quit IRC23:40
vilobhmmjgriffith : sureeeee :) …but will have it up soon…working on uni test23:40
jgriffithvilobhmm: Oh, dang!!  :)  See.. I try and be funny, but I forget you're an over-achiever!23:42
jgriffithOk... well, I've completely broken the code I was working on :(23:42
jgriffithThink I'll take a walk and come back later23:42
jgriffithCatch ya all later23:42
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
vilobhmmsure :) ..good luck !23:44
*** IlyaG has joined #openstack-cinder23:53
*** harlowja has quit IRC23:54
*** harlowja has joined #openstack-cinder23:55
*** setmason has joined #openstack-cinder23:55
asselinpatrickeast, hemnafk thingee ok we're posting to os-brick:
patrickeasti’ve got one for that patchset on the way for mine… its #12 in line for FC tests : /23:58
patrickeastbut in the zuul queue!23:59
*** IlyaG has quit IRC23:59

Generated by 2.14.0 by Marius Gedminas - find it at!