Wednesday, 2015-06-17

*** dannywilson has quit IRC00:00
*** haomai___ has joined #openstack-cinder00:02
openstackgerritNikesh proposed openstack/cinder: Adds FC and ISCSI Cinder drivers for Lenovo Storage Arrays
*** haomaiwa_ has quit IRC00:03
*** rmesta has quit IRC00:04
*** takedakn has joined #openstack-cinder00:11
*** takedakn has quit IRC00:16
*** mtanino has quit IRC00:22
*** _cjones_ has quit IRC00:23
*** _cjones_ has joined #openstack-cinder00:23
diogogmti’m trying to do a simple hack to test some changes to my existing cinder installation, i can see that the files are installed in this dir: /usr/lib/python2.7/dist-packages/cinder/api/v1/ — after i modify the file by adding some log statements i don’t see anything being logged to the log files, /var/log/cinder/api.log/cinder-volumes.log — do i need to recompile cinder or shouldn’t the changes be picked up right way?00:23
*** _cjones_ has quit IRC00:28
*** smoriya has joined #openstack-cinder00:31
*** haomai___ has quit IRC00:34
openstackgerritOpenStack Proposal Bot proposed openstack/cinder: Updated from global requirements
*** leeantho has quit IRC00:36
*** jasondotstar has quit IRC00:42
*** chlong has joined #openstack-cinder00:42
nikeshmdiogogmt: try to restart cinder services00:50
diogogmtnikeshm: why would that make a difference? are the files cached or something?00:51
nikeshmdiogogmt: not sure,but if yu restart like service cinder-api restart and using LOG.debug in the file,it should work.i think it will update /usr/lib/python2.7/dist-packages/cinder/api/v1/volumes.pyc with your changes00:53
nikeshmassuming yu r not using devstack00:54
*** tobe has joined #openstack-cinder01:05
*** jdurgin1 has quit IRC01:07
*** diogogmt has quit IRC01:07
*** lixiaoy1 has joined #openstack-cinder01:11
*** patrickeast has joined #openstack-cinder01:12
*** Yogi1 has joined #openstack-cinder01:13
*** rakesh_mishra has joined #openstack-cinder01:15
*** rakesh_mishra has quit IRC01:15
*** rakesh_mishra has joined #openstack-cinder01:15
*** patrickeast has left #openstack-cinder01:19
*** rakesh_mishra has quit IRC01:21
*** rakesh_mishra_ has joined #openstack-cinder01:21
*** patrickeast has joined #openstack-cinder01:21
*** Yogi1 has quit IRC01:24
*** Lee1092 has joined #openstack-cinder01:29
openstackgerritDave Chen proposed openstack/cinder-specs: Update of `support-modify-volume-image-metadata.rst`
*** patrickeast_ has joined #openstack-cinder01:32
*** patrickeast has quit IRC01:32
*** patrickeast_ has quit IRC01:33
*** patrickeast has joined #openstack-cinder01:33
*** mtecer has quit IRC01:34
openstackgerritThang Pham proposed openstack/cinder: Conversion to volume object
openstackgerritThang Pham proposed openstack/cinder: WIP: RPC and object versioning
*** Apoorva has quit IRC01:43
*** Apoorva has joined #openstack-cinder01:45
*** ociuhandu has quit IRC01:46
*** diogogmt has joined #openstack-cinder01:49
*** pboldin has quit IRC01:50
*** Apoorva has quit IRC01:55
*** vilobhmm has joined #openstack-cinder02:02
*** rakesh_mishra_ has quit IRC02:04
*** dims has quit IRC02:04
*** vilobhmm has quit IRC02:07
*** dims has joined #openstack-cinder02:08
*** jasondotstar has joined #openstack-cinder02:09
*** jasondotstar has quit IRC02:09
*** rhefner has joined #openstack-cinder02:10
*** jwcroppe has quit IRC02:10
*** garthb_ has joined #openstack-cinder02:10
*** garthb has joined #openstack-cinder02:10
*** jasondotstar has joined #openstack-cinder02:10
*** Yogi1 has joined #openstack-cinder02:11
*** davechen is now known as davechen_afk02:13
*** dims has quit IRC02:19
*** jasondotstar has quit IRC02:24
*** Yogi1 has quit IRC02:32
*** changbl has joined #openstack-cinder02:32
*** haomaiwang has joined #openstack-cinder02:38
*** dannywilson has joined #openstack-cinder02:39
*** Yogi1 has joined #openstack-cinder02:52
*** patrickeast has quit IRC02:59
*** sgotliv has joined #openstack-cinder03:01
*** heyun has joined #openstack-cinder03:01
*** lidan has joined #openstack-cinder03:02
*** lidan has left #openstack-cinder03:04
*** rushil has joined #openstack-cinder03:05
*** lidan has joined #openstack-cinder03:05
*** rushil has quit IRC03:06
lidanhi, Does someone meet this problem about "Patch in Merger conflict"?03:08
*** Yogi1 has quit IRC03:12
*** Apoorva has joined #openstack-cinder03:15
lidanRecently when submitting code, I always meet code conflict problem.03:20
winston-dwho reported that issue? Is it Jenkins or other 3rd party CI?03:20
lidanAfter Jekins+103:21
lidanThis happened After Jekins+1.03:22
winston-ddid you rebase your code to tip?03:22
lidanWhat's the tip?03:23
winston-dlatest change of master branch03:24
winston-dor whatever branch you are trying to commit to03:24
*** yamada-h has joined #openstack-cinder03:25
*** yamada-h has quit IRC03:29
*** markvoelker has quit IRC03:31
*** coolsvap|afk is now known as coolsvap03:34
*** aswadr has joined #openstack-cinder03:40
*** vilobhmm has joined #openstack-cinder03:53
openstackgerritJessy Lee proposed openstack/cinder: Add volume drivers for Infortrend Storage
*** rushiagr_away is now known as rushiagr04:10
*** patrickeast has joined #openstack-cinder04:15
*** patrickeast has quit IRC04:23
*** lidan has quit IRC04:25
*** markvoelker has joined #openstack-cinder04:32
*** markvoelker has quit IRC04:37
*** lidan has joined #openstack-cinder04:55
*** lidan has quit IRC04:56
openstackgerritchenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
*** Apoorva has quit IRC04:59
*** lidan has joined #openstack-cinder04:59
*** IlyaG has quit IRC05:02
*** fifieldt has joined #openstack-cinder05:10
*** IanGovett has joined #openstack-cinder05:12
*** nihilifer has joined #openstack-cinder05:12
*** shausy has joined #openstack-cinder05:14
*** haomaiwang has quit IRC05:14
*** dannywilson has quit IRC05:14
*** dannywilson has joined #openstack-cinder05:15
*** BharatK has joined #openstack-cinder05:16
*** Maike has joined #openstack-cinder05:18
*** nkrinner has joined #openstack-cinder05:31
*** nlevinki has joined #openstack-cinder05:32
*** lpetrut has joined #openstack-cinder05:33
*** tobe has quit IRC05:39
openstackgerritJessy Lee proposed openstack/cinder: Add volume drivers for Infortrend Storage
*** sudipto has joined #openstack-cinder05:44
*** belmoreira has joined #openstack-cinder05:44
*** IanGovett has quit IRC05:55
*** markvoelker has joined #openstack-cinder05:55
*** garthb_ has quit IRC05:56
*** garthb has quit IRC05:57
*** haomaiwa_ has joined #openstack-cinder05:59
*** markvoelker has quit IRC06:00
*** IanGovett has joined #openstack-cinder06:05
*** IanGovett has quit IRC06:09
*** shausy has quit IRC06:09
*** shausy has joined #openstack-cinder06:10
*** yamada-h has joined #openstack-cinder06:14
*** tobe has joined #openstack-cinder06:20
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** abhishekk has joined #openstack-cinder06:21
abhishekkhi all, is there any meeting for cinder nova volume discussion?06:22
*** dulek has joined #openstack-cinder06:22
*** rakesh_mishra has joined #openstack-cinder06:23
*** afazekas has joined #openstack-cinder06:23
*** dulek has quit IRC06:25
*** sgotliv has quit IRC06:32
*** sgotliv has joined #openstack-cinder06:32
*** ianbrown has quit IRC06:39
openstackgerritrakesh mishra proposed openstack/cinder: Add config option to set max_volume_size_limit
*** salv-orl_ has joined #openstack-cinder06:41
*** ronenkat has joined #openstack-cinder06:41
*** boris-42 has quit IRC06:42
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** salv-orlando has quit IRC06:44
*** anshul has joined #openstack-cinder06:44
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** dulek has joined #openstack-cinder06:49
*** Longgeek has joined #openstack-cinder06:49
*** lpetrut has quit IRC06:50
*** lixiaoy1 has quit IRC06:50
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** Longgeek_ has joined #openstack-cinder06:53
*** Longgeek has quit IRC06:57
*** vilobhmm has quit IRC06:58
*** laughterwym has joined #openstack-cinder07:04
lanhi,  do anybody know this option CONF.keymgr.fixed_key , what is it used for ? and  what value should I specified (seems that there is no default value for this option)?07:05
*** laughterwym has quit IRC07:10
*** laughterwym has joined #openstack-cinder07:10
*** ronis has joined #openstack-cinder07:10
*** laughterwym has quit IRC07:11
*** laughterwym has joined #openstack-cinder07:11
openstackgerritchenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
lidan<openstackgerrit> chenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
*** lixiaoy11 has joined #openstack-cinder07:20
*** Longgeek_ has quit IRC07:26
*** lixiaoy1 has joined #openstack-cinder07:27
*** dannywilson has quit IRC07:28
*** dannywilson has joined #openstack-cinder07:30
*** lixiaoy11 has quit IRC07:30
*** dannywilson has quit IRC07:34
openstackgerritAbhijeet Malawade proposed openstack/python-cinderclient: Add support to return request_id of last request
*** ociuhandu has joined #openstack-cinder07:42
*** markvoelker has joined #openstack-cinder07:44
*** ociuhandu has quit IRC07:45
*** jordanP has joined #openstack-cinder07:46
*** Longgeek has joined #openstack-cinder07:47
*** Longgeek has quit IRC07:48
*** Longgeek has joined #openstack-cinder07:48
*** markvoelker has quit IRC07:49
*** markus_z has joined #openstack-cinder07:56
*** ociuhandu has joined #openstack-cinder08:04
*** ronenkat_ has joined #openstack-cinder08:08
*** alonmarx_ has joined #openstack-cinder08:08
*** alonmarx has quit IRC08:11
*** ronenkat has quit IRC08:11
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** chlong has quit IRC08:18
*** aswadr has quit IRC08:18
*** c0m0 has joined #openstack-cinder08:19
*** aswadr has joined #openstack-cinder08:19
*** asrangne has joined #openstack-cinder08:21
*** yamada-h_ has joined #openstack-cinder08:22
*** aswadr has quit IRC08:24
*** yamada-h has quit IRC08:26
*** ronenkat__ has joined #openstack-cinder08:30
*** deepakcs has joined #openstack-cinder08:31
*** belmoreira has quit IRC08:33
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** belmoreira has joined #openstack-cinder08:33
*** ronenkat_ has quit IRC08:34
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** laughterwym has quit IRC08:41
*** mdbooth has quit IRC08:42
*** mdbooth has joined #openstack-cinder08:45
openstackgerritAbhijeet Malawade proposed openstack/cinder: Remove unused context parameter
*** ociuhandu has quit IRC08:52
*** coolsvap is now known as coolsvap|afk08:52
*** shyama has joined #openstack-cinder08:53
DuncanTIan: It is the fixed key for encrypted volumes, if you're using encrypted volumes and no key manager08:56
*** ociuhandu has joined #openstack-cinder08:57
lidan chenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
openstackgerritShay Halsband proposed openstack/cinder: XtremIO Volume driver requests, multipath
*** tobe has quit IRC08:58
*** c0m0 has quit IRC08:59
*** tobe has joined #openstack-cinder08:59
openstackgerritchenzongliang proposed openstack/cinder: Add CHAP support for Huawei Storage
*** tobe_ has joined #openstack-cinder09:01
*** tobe_ has quit IRC09:01
openstackgerritchenzongliang proposed openstack/cinder: Huawei Storage iSCSI multipath Support
*** coolsvap|afk is now known as coolsvap09:03
*** tobe has quit IRC09:03
openstackgerritchenzongliang proposed openstack/cinder: Support SmartX for Huawei volume driver
*** daneyon has joined #openstack-cinder09:09
*** daneyon has quit IRC09:14
openstackgerritStefan Amann proposed openstack/os-brick: Support host type specific block volume attachment
*** ndipanov has joined #openstack-cinder09:21
openstackgerritVincent Hou proposed openstack/cinder: Add migration progress indication
*** tobe has joined #openstack-cinder09:23
*** rakesh_mishra has quit IRC09:24
*** rakesh_mishra has joined #openstack-cinder09:25
openstackgerritDeepak C Shetty (deepakcs) proposed openstack/cinder: Cinder part of volume snapshot improvements
openstackgerritrakesh mishra proposed openstack/cinder: Add config option to set max_volume_size_limit
openstackgerritShay Halsband proposed openstack/cinder: XtremIO Volume driver requests, multipath
*** markvoelker has joined #openstack-cinder09:33
*** e0ne has joined #openstack-cinder09:36
*** rakesh_mishra has quit IRC09:36
*** rakesh_mishra has joined #openstack-cinder09:37
*** markvoelker has quit IRC09:37
*** laughterwym has joined #openstack-cinder09:38
*** haomaiwa_ has quit IRC09:41
*** e0ne is now known as e0ne_09:43
*** rakesh_mishra has quit IRC09:46
*** e0ne_ is now known as e0ne09:47
*** yamada-h_ has quit IRC09:48
*** rakesh_mishra has joined #openstack-cinder09:51
*** lixiaoy1 has quit IRC09:51
*** sgotliv_ has joined #openstack-cinder09:54
openstackgerritchenzongliang proposed openstack/cinder: Huawei Storage Multiple Pools Support
openstackgerritRajesh Tailor proposed openstack/cinder: Validate bool value using strutils.bool_from_string method
*** sgotliv has quit IRC09:58
*** sgotliv__ has joined #openstack-cinder10:01
*** sgotliv_ has quit IRC10:04
*** sgotliv__ has quit IRC10:07
*** ianbrown has joined #openstack-cinder10:07
*** lpetrut has joined #openstack-cinder10:10
*** lpetrut has quit IRC10:10
*** heyun has quit IRC10:16
*** dims has joined #openstack-cinder10:17
*** tobe has quit IRC10:17
*** dims has quit IRC10:17
*** dims has joined #openstack-cinder10:18
*** ianbrown has quit IRC10:19
*** cebruns has quit IRC10:19
*** sgotliv has joined #openstack-cinder10:20
*** cebruns has joined #openstack-cinder10:21
*** lpetrut has joined #openstack-cinder10:21
*** ianbrown has joined #openstack-cinder10:26
*** shyama_ has joined #openstack-cinder10:28
*** shyama has quit IRC10:28
*** shyama_ is now known as shyama10:28
*** haomaiwang has joined #openstack-cinder10:28
*** shyama_ has joined #openstack-cinder10:33
*** shyama has quit IRC10:35
*** shyama_ is now known as shyama10:35
*** timcl has joined #openstack-cinder10:36
*** jasondotstar has joined #openstack-cinder10:41
*** lihit has joined #openstack-cinder10:47
*** lihit has quit IRC10:47
*** e0ne is now known as e0ne_10:49
*** aix has joined #openstack-cinder10:49
*** e0ne_ has quit IRC10:54
*** timcl has quit IRC10:55
*** e0ne has joined #openstack-cinder10:56
*** daneyon has joined #openstack-cinder10:58
*** sgotliv_ has joined #openstack-cinder10:59
*** daneyon has quit IRC11:02
*** sgotliv has quit IRC11:02
openstackgerritShay Halsband proposed openstack/cinder: XtremIO Volume driver requests, multipath
*** jasondotstar has quit IRC11:06
*** sgotliv_ has quit IRC11:07
*** sgotliv has joined #openstack-cinder11:07
dulekDuncanT: ping11:08
dulekDuncanT: Can you provide an update on work in Nova for c-vol A/A? Is there anything I can help with now?11:09
*** markvoelker has joined #openstack-cinder11:22
*** coolsvap is now known as coolsvap|away11:22
*** ianbrown has quit IRC11:22
*** rushiagr is now known as rushiagr_away11:24
*** markvoelker has quit IRC11:27
*** e0ne is now known as e0ne_11:29
*** dannywilson has joined #openstack-cinder11:30
openstackgerritAlessandro Pilotti proposed openstack/cinder: Update guru meditation reports from oslo.incubator
*** shyama has quit IRC11:34
*** dannywilson has quit IRC11:34
*** e0ne_ has quit IRC11:34
*** laughterwym has quit IRC11:36
*** laughterwym has joined #openstack-cinder11:36
*** marcusvrn has joined #openstack-cinder11:39
openstackgerritStefan Amann proposed openstack/os-brick: Support host type specific block volume attachment
*** ganso_ has joined #openstack-cinder11:45
*** thangp has joined #openstack-cinder11:46
openstackgerritVipin Balachandran proposed openstack/cinder: Get updated volume status in begin_detaching
openstackgerritVipin Balachandran proposed openstack/cinder: Get updated volume status in begin_detaching
*** dulek_ has joined #openstack-cinder11:57
*** timcl has joined #openstack-cinder11:58
*** dulek has quit IRC12:01
*** markvoelker has joined #openstack-cinder12:05
*** julim has joined #openstack-cinder12:08
*** rushiagr_away is now known as rushiagr12:11
*** rakesh_mishra has quit IRC12:13
openstackgerritPeter Penchev proposed openstack/cinder-specs: Add the StorPool Brick driver specification.
openstackgerritPeter Penchev proposed openstack/os-brick: Add the StorPool brick connector.
*** shausy has quit IRC12:18
*** shausy has joined #openstack-cinder12:19
*** e0ne has joined #openstack-cinder12:19
*** yrabl has quit IRC12:20
*** yrabl has joined #openstack-cinder12:21
Liuchenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
Liu chenzongliang proposed openstack/cinder: Refactor Huawei Volume driver
*** lpetrut has quit IRC12:22
*** lpetrut has joined #openstack-cinder12:22
*** akerr has joined #openstack-cinder12:23
*** ociuhandu has quit IRC12:26
*** dulek_ is now known as dulek12:34
*** dims has quit IRC12:36
*** dims has joined #openstack-cinder12:36
*** kjnelson has joined #openstack-cinder12:37
openstackgerritDmitry Guryanov proposed openstack/os-brick: remotefs: add virtuozzo storage support
*** dulek has quit IRC12:39
*** marcusvrn1 has joined #openstack-cinder12:40
*** dulek has joined #openstack-cinder12:40
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Ceph driver support retries on rados_connect_timeout
openstackgerritJessy Lee proposed openstack/cinder: Add volume drivers for Infortrend Storage
*** marcusvrn has quit IRC12:44
*** rushil has joined #openstack-cinder12:45
*** zhipeng has joined #openstack-cinder12:46
*** daneyon has joined #openstack-cinder12:46
*** bswartz has quit IRC12:49
*** Yogi1 has joined #openstack-cinder12:49
*** smoriya has quit IRC12:49
*** amoturi has joined #openstack-cinder12:50
*** deepakcs has quit IRC12:51
*** daneyon has quit IRC12:51
*** merooney has joined #openstack-cinder12:51
*** Ctina_ has joined #openstack-cinder12:52
*** e0ne is now known as e0ne_12:55
*** eharney has joined #openstack-cinder12:55
*** IanGovett has joined #openstack-cinder12:56
*** e0ne_ is now known as e0ne12:57
*** edmondsw has joined #openstack-cinder12:57
*** lixiaoy1 has joined #openstack-cinder12:59
*** lixiaoy1 has left #openstack-cinder13:00
*** julim has quit IRC13:01
*** merooney has quit IRC13:04
*** julim has joined #openstack-cinder13:05
*** sudipto has quit IRC13:06
*** diogogmt has quit IRC13:07
*** rakesh_mishra has joined #openstack-cinder13:08
openstackgerritPranali Deore proposed openstack/cinder: Fix useless-else-on-loop pylint issues
*** jamielennox is now known as jamielennox|away13:13
*** ociuhandu has joined #openstack-cinder13:13
*** nihilifer has quit IRC13:14
*** rakesh_mishra has quit IRC13:19
*** rakesh_mishra has joined #openstack-cinder13:20
*** jamielennox|away is now known as jamielennox13:21
*** xyang1 has joined #openstack-cinder13:23
*** gouthamr has joined #openstack-cinder13:24
*** IanGovett has quit IRC13:25
*** timcl has quit IRC13:32
*** jamielennox is now known as jamielennox|away13:33
*** erlon has joined #openstack-cinder13:36
*** jskinner has joined #openstack-cinder13:37
*** timcl has joined #openstack-cinder13:40
*** jamielennox|away is now known as jamielennox13:42
*** mriedem_away is now known as mriedem13:42
*** nlevinki has quit IRC13:43
*** bswartz has joined #openstack-cinder13:45
*** rushil has quit IRC13:46
*** Liu has quit IRC13:46
*** kbyrne has quit IRC13:46
*** Liu has joined #openstack-cinder13:47
DuncanTdulek: I'm in broken-IT hell at the moment, so I've nothing useful to say :-(13:47
DuncanTdulek: I'm on a wet string internet connection, and my main workstation is down13:47
dulekDuncanT: That's fine, just tell me where I can start helping. :)13:48
*** Longgeek has quit IRC13:48
dulekDuncanT: I've spent last days in the hell of understanding how Nova actually calls Cinder, and I hope I'm ready to make some changes.13:48
*** aswadr has joined #openstack-cinder13:49
*** Longgeek has joined #openstack-cinder13:49
dulekDuncanT: I just need an information what's the first priority.13:49
dulekDuncanT: Not to overlap with others people work.13:49
*** gouthamr has quit IRC13:50
*** asrangne has quit IRC13:52
DuncanTdulek: First priority IMO is to make minimal changes to nova to make it handle errors returned by initialise_connection etc gracefully13:52
DuncanTdulek: Easiest thing to do is nobble cinder to always return an error (e.g. volume busy) then make nova handle that13:53
DuncanTdulek: We can't fix cinder to return better errors until nova can handle them13:53
DuncanTdulek: Once that's done, we can get cinder to return the errors, then remove all/most of the volume state checks in nova13:53
*** BharatK has quit IRC13:54
DuncanTdulek: I think that workflow gives us our biggest win quickly13:54
*** rushil has joined #openstack-cinder13:57
dulekDuncanT: Okay, I start that tomorrow morning. vilobhmm has a lot on his plate right now so hopefully he won't mind me helping on that. ;)13:58
*** gouthamr has joined #openstack-cinder13:58
dulekDuncanT: Thanks for help!13:58
*** haypo has joined #openstack-cinder13:58
openstackgerritIvan Kolodyazhny proposed openstack/cinder-specs: Cinder API WSGI application under Apache/Nginx
haypojgriffith, hi. could you please review "Replace basestring with six.string_types"? it's required by my other my tox -e py34 patch13:59
haypojgriffith, i'm trying to reproduce your tox issue, i installed ubuntu 14.04 in a VM just to reproduce this bug :-p13:59
*** timcl has quit IRC14:00
*** skraynev has quit IRC14:02
*** zhipeng has quit IRC14:02
*** crose has joined #openstack-cinder14:02
*** skraynev has joined #openstack-cinder14:02
*** timcl has joined #openstack-cinder14:04
*** jdurgin1 has joined #openstack-cinder14:05
*** jdandrea has left #openstack-cinder14:06
*** jdandrea has joined #openstack-cinder14:06
hayposmcginnis, jgriffith : for "tox: add a basic py34 venv running a single test", Anton Arefiev wrote "looks good, working for me on Ubuntu 14.04.2 LTS with virtualenv (13.0.3)". a colleague just tried on Ubuntu 14.04.2 LTS, it also worked for him14:07
haypodid you cherry-picked the basestring patch?14:07
*** merooney has joined #openstack-cinder14:07
smcginnishaypo: No, I did not. Didn't realize there wasn't an explicit dependency on the patch.14:09
openstackgerritStefan Amann proposed openstack/os-brick: Support host type specific block volume attachment
smcginnishaypo: Would really rather have it be dependent on any dependent patches, rather than using the Depends-On bit.14:10
hayposmcginnis, the missing cherry-pick doesn't explain the "got a ton of errors while it was trying to install requirements" failure14:10
jgriffithhaypo: if it depends on a patch, that patch needs to be listed as dependent; OR it should be wip til the deps merge14:10
smcginnishaypo: Willing to accept it's something wrong in my environment for sure. :)14:10
jgriffithhaypo: I'm not14:11
jgriffithhaypo: not at all14:11
jgriffithhaypo: and frankly I'm not sure how much more clear I need to be14:11
jgriffithhaypo: fresh ubuntu 14.04, git clone openstack/cinder, cd cinder,  git review -d 192118, tox -epy3414:12
haypojgriffith, "if it depends on a patch, that patch needs to be listed as dependent" it's the case. : "Depends-On: Ie1aedf1cbb9d3e54a996321cd586b875e69ac85a"14:12
haypojgriffith, i'm doing the same (trying to run tox -e py34 on a fresh ubuntu), i'm looking for the ubuntu package for postgresql developement (pg_config) :-p14:13
jgriffithhaypo: I'm looking right at your patch and the "Depends On" block is empty?14:13
jgriffithhaypo: Oh14:13
jgriffithhaypo: yeah... so FYI14:13
jgriffithhaypo: listing "Depends-On" in the commit message isn't quite enough14:14
jgriffithhaypo: anywho... I'd recommend either getting all those to merge, or making them all formal dependencies for 19211814:14
jgriffithhaypo: that's likely the problem smcginnis and I were having14:14
*** kbyrne has joined #openstack-cinder14:15
jgriffithhaypo: FWIW, it's the same problem Jenkins will have :)14:15
eharneydoesn't the Jenkins gate respect Depends-On in the commit message?14:15
jgriffitheharney: It looks like maybe it will (by not running against it)14:16
jgriffitheharney: but I thought typically sort of SOP was adding a dep in Gerrit, so everything gets pulled/tested14:16
eharneyyeah, the Depends-On bit is mostly for cross-project dependencies i think, using direct git dependencies is simpler when in the same project14:17
smcginnisI agree. Within the same project I would rather not see that used much.14:17
eharneybut it does work because Jenkins passes, so that's kinda neat14:17
openstackgerritYuriy Nesenenko proposed openstack/cinder: Filter snapshots data on the DB side
jgriffitheharney: it is pretty cool14:20
smcginnisI would be fine with it if14:21
smcginnis"git review -d [number]" would take care of it for me. :)14:21
haypojgriffith, jenkins understands well depends-on, i used that on other projects14:21
jgriffithhaypo: indeed14:21
haypojgriffith, it requires manual action if a patch is accepted whereas dependencies are not accepted yet, and other patches are for other projects14:21
jgriffithhaypo: fine by me now that I realize it's there and works14:22
haypojgriffith, well, please review the dependency :-)
jgriffithfrankly I don't care that much... but using git dependencies in the project is a bit easier for the masses IMO14:22
smcginniseharney just pushed it along.14:22
jgriffithhaypo: I've reviewed so many of your patches at this point :)14:23
*** nkrinner has quit IRC14:23
haypojgriffith, yeah, by the way thank you for that :)14:23
haypoi'm going to add a non voting py34 gate for cinder, so the gate will tell me if my tox patch works or not :)14:23
*** merooney has quit IRC14:24
jgriffithhaypo: yes please :)14:24
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Ceph driver support retries on rados_connect_timeout
jgriffithhaypo: if you need help or want somebody else to do it let me know14:24
jgriffithhaypo: but seems like this is a pretty good mission for you14:24
haypojgriffith, my next goal is to have a working tox -e py34 and a voting py34 gate. since this point, following reviews will be simpler because we will be able to rely on the py34 gate14:27
*** mdbooth has quit IRC14:27
eharneysounds awesome14:27
*** sudipto has joined #openstack-cinder14:28
thingeejgriffith: you may want to comment on,cm14:29
*** sudipto has quit IRC14:30
thingeejgriffith: since you've been doing some clean ups in this area14:30
*** xyang1 has quit IRC14:31
*** shausy has quit IRC14:31
*** merooney has joined #openstack-cinder14:32
openstackgerritIvan Kolodyazhny proposed openstack/cinder-specs: Cinder API WSGI application under Apache/Nginx
haypo "Add a non-voting py34 check job to Cinder" here you have ;)14:34
*** ronenkat__ has quit IRC14:34
jgriffiththingee: good call, thanks14:34
*** rmesta has joined #openstack-cinder14:35
jgriffithhaypo: one problem14:35
jgriffithI don't think you want stable branches there14:35
*** daneyon has joined #openstack-cinder14:35
haypojgriffith, my patch is only for master: "Only add the check job to the master branch, ignore Kilo branches and older since they will not get Python 3 support."14:36
haypojgriffith, am i wrong?14:36
jgriffithhaypo: haha... no, I interpretted the inverse of the regex )14:36
*** ronis has quit IRC14:36
haypojgriffith, i didn't write the regex, i copied it from other projects. (?!...) means "Matches if ... *doesn’t* match next."14:37
haypoit would be a real issue if the gate becomes voting, tests will obvisouly fail on python 3.4 for stable branches :-)14:37
jgriffithindeed it would14:38
*** daneyon has quit IRC14:40
*** anshul has quit IRC14:40
*** gouthamr has quit IRC14:43
*** belmoreira has quit IRC14:45
thingeeaarefiev, DuncanT: retargeting this for l-2 ... not sure if we can resolve things here
openstackLaunchpad bug 1307491 in Cinder "quota-update should error out if input provided is non-existent tenant id" [Medium,In progress] - Assigned to Anton Arefiev (aarefiev)14:47
*** cdelatte has joined #openstack-cinder14:47
hayposmcginnis, jgriffith : FYI on a fresh ubuntu 14.04.2, tox -e py34 succeded, same result than Cyril Roelandtand Anton Arefiev. i guess that you just missed a dependency and/or had issues with your setup14:48
*** diogogmt has joined #openstack-cinder14:50
*** xylan_kong has left #openstack-cinder14:50
*** gouthamr has joined #openstack-cinder14:51
lidanhi all ,can you take some to review our driver,
aarefievthingee: unfortunately, also I think nested quota  should touch this since we need info about tenants14:51
*** xyang has joined #openstack-cinder14:51
DuncanTthingee: I still worry about cross-api chained calls... in the case of keystone though, if keystone is down you don't get any calls anyway, so maybe this is ok?14:52
openstackgerritIvan Kolodyazhny proposed openstack/cinder-specs: Cinder API WSGI application under Apache/Nginx
*** dulek has quit IRC14:53
*** sudipto has joined #openstack-cinder14:54
e0nethingee: hi. i guess you leave +A for Jay?14:54
thingeee0ne: yeah, although the last issue was addressed14:55
e0nethingee: ok, thanks for review14:55
aarefievDuncanT: why do you worry about keystone down, if it is so authentication request fail too14:58
DuncanTaarefiev: Slow is worse than down14:59
DuncanTaarefiev: Though I guess that will also throttle incoming requests too15:00
DuncanTaarefiev: keystone might be a special case here, hence the above comment15:00
*** rakesh_mishra has quit IRC15:00
DuncanTaarefiev: I'll take off the -2 I think, and just leave a comment for others to consider15:00
*** rhagarty__ has quit IRC15:01
*** rhagarty has quit IRC15:02
*** rhagarty__ has joined #openstack-cinder15:02
*** markstur has quit IRC15:02
aarefievDuncanT: but we don't have another way to verify tenant15:02
*** rhagarty has joined #openstack-cinder15:02
*** markstur has joined #openstack-cinder15:03
DuncanTaarefiev: There are lots of things we can't verify without cross-proejct calls, and sometimes the cost is too high and it is better not to verify15:03
DuncanTaarefiev: There was something added to call glance during a common API call - that was definitely a no-no15:04
*** xyang1 has joined #openstack-cinder15:04
DuncanTaarefiev: This call is admin-only and not common, and keystone is self-rate-limitting, so maybe it is ok, I haven't made my mind up, thinking about it now though15:04
*** smoriya has joined #openstack-cinder15:05
*** openstackgerrit has quit IRC15:05
*** openstackgerrit has joined #openstack-cinder15:05
*** zongliang has joined #openstack-cinder15:08
*** xyang has quit IRC15:08
*** xyang has joined #openstack-cinder15:08
*** afazekas has quit IRC15:10
*** adurbin__ has joined #openstack-cinder15:11
*** jungleboyj has joined #openstack-cinder15:12
*** haypo has left #openstack-cinder15:12
*** xyang has quit IRC15:13
aarefievDuncanT: agree, we must be careful with cross project communication, but we can't without it to keep consistency15:13
*** xyang has joined #openstack-cinder15:14
DuncanTaarefiev: In many cases, I'd rather have loose consistency than tight coupling. I get woken up at 4am when the Cinder API goes down, even if the root cause is some other service. It gives me a certain prospective.15:15
*** xyang has quit IRC15:16
*** xyang has joined #openstack-cinder15:16
aarefievDuncanT: yea, it ts known problem :)15:17
*** anshul has joined #openstack-cinder15:17
DuncanTaarefiev: I'll continue to strongly resist anything that makes me more likely to get woken up, just like I try to resist things that break upgrades, since that also comes back to bite my team hard15:18
*** mtanino has joined #openstack-cinder15:18
*** jungleboyj has quit IRC15:19
*** xyang has quit IRC15:20
*** xyang has joined #openstack-cinder15:21
*** afazekas has joined #openstack-cinder15:23
*** anshul has quit IRC15:24
*** jnicklin has joined #openstack-cinder15:25
aarefievDuncan: got your point, fair enough15:27
*** afazekas has quit IRC15:28
*** alonmarx has joined #openstack-cinder15:31
*** Maike has quit IRC15:32
*** alonmarx_ has quit IRC15:33
*** timcl has quit IRC15:34
*** timcl has joined #openstack-cinder15:34
*** patrickeast has joined #openstack-cinder15:35
*** timcl has quit IRC15:36
*** leeantho has joined #openstack-cinder15:36
*** jamielennox is now known as jamielennox|away15:38
*** jdurgin1 has quit IRC15:39
*** vokt has joined #openstack-cinder15:39
*** timcl has joined #openstack-cinder15:40
jordanPeharney, could you re-review (Scality SOFS: enhance how the remoteFS mount is detected) plz ?15:40
*** patrickeast has quit IRC15:40
eharneyjordanP: sure15:40
*** lpetrut has quit IRC15:42
*** harlowja_at_home has joined #openstack-cinder15:42
*** timcl1 has joined #openstack-cinder15:42
*** BharatK has joined #openstack-cinder15:43
*** dannywilson has joined #openstack-cinder15:44
eharneyjordanP: also, i'm happy to see the blockrebase patch15:44
*** timcl has quit IRC15:44
eharneyjordanP: someone else was just asking about that this week15:44
jordanPeharney, I see you follow closely the ML. Yeah, it"s an attempt but it seems to work and it's not that intrusive imo15:45
*** merooney has quit IRC15:45
jordanPeharney, I need to write some unit tests but I'd like some feedback in the mean time15:45
*** smoriya has quit IRC15:46
thingeejgriffith: should we revisit this in the meeting ?15:46
*** dannywilson has quit IRC15:46
*** jungleboyj has joined #openstack-cinder15:46
*** garthb has joined #openstack-cinder15:46
*** garthb_ has joined #openstack-cinder15:46
*** jamielennox|away is now known as jamielennox15:47
*** dannywilson has joined #openstack-cinder15:47
jordanPmy long term goal is to refactor our driver (Scality) to reuse the remoteFS code but I need the full tempest tests to pass and currently the rewritten version fails the cert test because of the boot_from_volume scenario15:47
*** hemnafk is now known as hemna15:47
flip214jgriffith: thingee: duncan: DRBD driver looks fine. I'll keep watching it for a while, but eg. on patch set 11 says SUCCESS15:48
*** sreedharv has joined #openstack-cinder15:48
*** lcurtis has joined #openstack-cinder15:49
thingeeflip214: excellent15:50
*** rakesh_mishra has joined #openstack-cinder15:50
*** cs_0x6373 has quit IRC15:51
e0nehemna: evening:)15:51
smcginnishemna: Shouldn't you be fishing?15:52
hemnaI'd love another few days out there.15:52
hemnabut today, it's back to the grind15:52
smcginnishemna: The fish probably need the break. ;)15:52
hemnaI'm good and sunburned15:52
hemnawas about 20 fish a day, so yah, I'm sure they are tired as well15:52
jungleboyjhemna: Wow, that is impressive.15:54
xyanghemna: what do you do with that many fish?  Do you put them back or consume them or sell them?:)15:54
hemnaI ate all of them15:54
hemnalol, jk15:54
hemnacatch and release for me15:55
xyangthen you should have endless supply15:55
smcginnisSame dumb fish over and over. ;)15:55
hemnaheh yah.   it was a good few days.   I forgot what a day off work was like15:55
*** adurbin_ has joined #openstack-cinder15:55
kmartinI only saw a picture of one fish, so this 20 might be one of those fish stories :)15:56
*** patrickeast has joined #openstack-cinder15:57
*** alau2 has joined #openstack-cinder15:57
*** jamielennox is now known as jamielennox|away15:58
*** tsekiyama has joined #openstack-cinder15:58
e0nejungleboyj: hi. we need your final vot for, please15:58
*** cs_0x6373 has joined #openstack-cinder15:58
*** krtaylor has quit IRC15:59
*** jordanP has quit IRC15:59
*** gouthamr has quit IRC16:00
*** dulek_home has joined #openstack-cinder16:02
*** daneyon has joined #openstack-cinder16:04
*** vokt has quit IRC16:04
*** daneyon_ has joined #openstack-cinder16:04
*** jamielennox|away is now known as jamielennox16:06
*** vilobhmm has joined #openstack-cinder16:06
*** BharatK has quit IRC16:06
*** gouthamr has joined #openstack-cinder16:08
*** daneyon has quit IRC16:08
*** Apoorva has joined #openstack-cinder16:09
*** Adriano__ has joined #openstack-cinder16:11
*** dims_ has joined #openstack-cinder16:11
*** krtaylor has joined #openstack-cinder16:11
openstackgerritAnton Arefiev proposed openstack/cinder: WIP: add driver list generator
*** dims has quit IRC16:13
*** Yogi1 has quit IRC16:14
openstackgerritWalter A. Boring IV (hemna) proposed openstack/cinder: LVM add multiattach flag capability
flip214hemna: sorry about my (stupid?) questions ... coming from the HA side I'm burned by people running DRBD in Dual-Primary, with GFS2/OCFS2 or even multipoint iSCSI, and as soon as they crush their data they come to us whining...16:16
hemnaflip214, no worries man.  not stupid questions at all.  :)16:17
*** ronis has joined #openstack-cinder16:18
*** Yogi1 has joined #openstack-cinder16:18
flip214hemna: and so I fear that multi-attach against multi-homed storage volumes => one client writes *only* to A, the other *only* to B, and everything goes to hell.16:18
flip214yeah, that didn't match the topic there. Sorry.16:18
openstackgerritAnton Arefiev proposed openstack/cinder: WIP: add driver list generator
hemnano problemo16:20
*** timcl1 has quit IRC16:22
*** harlowja_at_home has quit IRC16:22
*** amoturi has quit IRC16:22
erlonguys, is there a way to find old blueprints? Doing a search on launch pad only shows current blueprints.16:26
*** markus_z has quit IRC16:27
*** afazekas has joined #openstack-cinder16:28
*** vilobhmm has quit IRC16:28
*** cs_0x6373 has quit IRC16:29
*** lpetrut has joined #openstack-cinder16:30
*** rakesh_mishra has quit IRC16:30
*** openstackgerrit has quit IRC16:33
*** openstackgerrit has joined #openstack-cinder16:34
*** jamielennox is now known as jamielennox|away16:35
*** cs_0x6373 has joined #openstack-cinder16:35
*** krtaylor has quit IRC16:37
*** gouthamr has quit IRC16:37
*** sudipto has quit IRC16:38
*** gouthamr has joined #openstack-cinder16:42
*** Tross1 has joined #openstack-cinder16:43
*** thangp_ has joined #openstack-cinder16:43
*** rhagarty_ has joined #openstack-cinder16:43
*** med_` has joined #openstack-cinder16:43
*** Tross has quit IRC16:43
*** changbl_ has joined #openstack-cinder16:44
*** fifieldt_ has joined #openstack-cinder16:44
*** mtanino has quit IRC16:45
*** rhagarty has quit IRC16:45
*** rhagarty__ has quit IRC16:45
*** rushil has quit IRC16:45
*** thangp has quit IRC16:45
*** fifieldt has quit IRC16:45
*** changbl has quit IRC16:45
*** med_ has quit IRC16:45
*** a| has joined #openstack-cinder16:46
*** a| has quit IRC16:46
*** timcl has joined #openstack-cinder16:47
*** mfisch has quit IRC16:47
*** logan2 has joined #openstack-cinder16:47
*** krtaylor has joined #openstack-cinder16:48
*** mtanino has joined #openstack-cinder16:49
*** jamielennox|away is now known as jamielennox16:49
*** mfisch has joined #openstack-cinder16:49
*** mfisch has quit IRC16:50
*** mfisch has joined #openstack-cinder16:50
openstackgerritAlessandro Pilotti proposed openstack/cinder: Update guru meditation reports from oslo.incubator
*** laughterwym has quit IRC16:50
*** alexpilotti has joined #openstack-cinder16:51
alexpilottijgriffith: morning16:54
alexpilottijgriffith: we have a blocking issue with a patch that merged recently and broke all the Windows code:,n,z16:54
alexpilottijgriffith: this patch includes some very old oslo code, which is strange since the patch is recent16:55
*** belmoreira has joined #openstack-cinder16:57
*** abhishekk has quit IRC16:58
thingeejungleboyj: was going to ask about these and
openstackLaunchpad bug 1409733 in Cinder "adopt namespace-less oslo imports" [High,In progress] - Assigned to Jay Bryant (jsbryant)16:58
openstackLaunchpad bug 1381563 in Cinder "cinder using obsolete oslo modules" [Medium,In progress] - Assigned to Jay Bryant (jsbryant)16:58
*** vilobhmm has joined #openstack-cinder16:58
thingeeI think one of them we need to move to L-216:58
erlonthingee: quick question, is there a way to find old blueprints?16:59
erlonthingee: I only  can see the new ones16:59
vilobhmmDuncanT : ping16:59
thingeeerlon: yeaaa, I think they get archived into a blackhole after marked for release16:59
*** _cjones_ has joined #openstack-cinder16:59
thingeeerlon: not sure why that happens16:59
eharneythingee: could you put the stable branch tasks on this bug into the right state?  (or jgriffith)16:59
openstackLaunchpad bug 1415087 in OpenStack Compute (nova) "[OSSA 2015-011] Format-guessing and file disclosure in image convert (CVE-2015-1850, CVE-2015-1851)" [High,Triaged] - Assigned to Tony Breeds (o-tony)16:59
*** dulek_home has quit IRC17:00
*** e0ne has quit IRC17:00
jungleboyjthingee: Oh, cool.  That is easy to deal with.  :-)  I know there is other stuff that I need to get merged in.  Want to force time on that yet this week.17:00
erlonthingee: too bad17:00
jungleboyjI can easily get the status on those bugs udpated after I go grab some brain food.17:00
*** aswadr has quit IRC17:02
*** asrangne has joined #openstack-cinder17:02
*** dulek_home has joined #openstack-cinder17:02
*** dulek_home has quit IRC17:03
thingeeeharney: done17:03
thingeeeharney tristanC  and thanks to you both on this work17:03
xyangthingee: I finally have some good news to report on EMC XIO CI.  It reported success here today:
eharneythingee: thanks, i'm not sure why the process of gerrit updating those seems to break down17:03
*** haomaiw__ has joined #openstack-cinder17:04
thingeexyang: oh good17:04
thingeexyang: thanks17:04
*** rushiagr is now known as rushiagr_away17:04
xyangthingee: sorry it took so long17:04
*** haomaiwang has quit IRC17:04
*** asrangne__ has joined #openstack-cinder17:05
thingeezongliang: I'm definitely not accepting any new patches from Huawei until we can see that changes will pass with your CI. I know you're all working on it though17:05
zongliangWe have try our best to fix the network problems of the ci,we want to target the Bo to l-217:08
*** asrangne has quit IRC17:08
zongliangthingee: 1:08:32 <zongliang> We have try our best to fix the network problems of the ci,we want to target the Bo to l-217:09
*** ronenkat has joined #openstack-cinder17:11
DuncanTvilobhmm: Hi17:11
*** ndipanov has quit IRC17:12
*** jamielennox is now known as jamielennox|away17:13
vilobhmmas part of nested quota driver one of the action item was to Add tracing to include request_id, volume_id for easy tracing the magic performed by quota operation.17:15
vilobhmmmy approach is to include request_id as a field in cinder.quotas does it sound ok to you17:15
*** dims_ has quit IRC17:16
vilobhmmDuncanT : ^^17:16
*** dims has joined #openstack-cinder17:17
*** patrickeast has quit IRC17:18
DuncanTvilobhmm: I think it needed adding to the reservation, it doesn't make sense to add it to the quota object I don't think?17:18
*** BharatK has joined #openstack-cinder17:20
*** gouthamr has quit IRC17:23
*** gouthamr has joined #openstack-cinder17:24
DuncanTOther than that, yes, seems right17:25
vilobhmmsorry someone stopped by17:32
vilobhmmbut i don;t think that we get volume_id details when the reservation_uuid is created or a reference for reservation is created in reservation table….volume_id is created after cinder.volume.flows.api.create_volume.EntryCreateTask is executed….17:32
*** hodos has joined #openstack-cinder17:32
vilobhmmso as per the feedback in the summit17:32
vilobhmmwill add both volume_id as well as the request_id in cinder.reservations table so that we can trace the quota request for a specific volume17:33
vilobhmmthanks DuncanT17:34
*** sreedharv has quit IRC17:35
hodosthingee: hi there, last 24 nexenta ci has about 20 successfull comments, so i think it's getting stable17:35
*** anshul has joined #openstack-cinder17:35
*** belmoreira has quit IRC17:38
thingeehodos: it's not really running that often17:38
*** afazekas has quit IRC17:41
*** breitz has joined #openstack-cinder17:42
hodosthingee: ok, last 24 hours it 19 from that paste, let's see how it goes from here17:42
*** aix has quit IRC17:42
thingeehodos: I would like to point out that Nexenta at this point is way past the deadline of having their CI together.
thingeehodos: so I'm wondering what makes Nexenta special here from the other vendors that have been cut for the same reason17:43
hodosthingee: it said 12th for successfull run (a single one) and ci for ns4 was running even before that but not stable17:44
hodosthingee: that paper says nothing about stability17:45
thingeehodos: if your driver isn't stable, that's no good17:45
thingeehodos: same standards we've held other vendors too17:45
hodosthingee: i didn't change the driver since then17:45
*** gouthamr has quit IRC17:45
thingeeyour backend + openstack is not stable17:45
thingeeI don't care about the python code.17:46
*** gouthamr has joined #openstack-cinder17:46
hodosthingee: i didn't change the backend either, just better network in the lab17:46
hodosthingee: it was really a bad ci infra17:47
hodosthingee: not a backend/driver issue17:47
thingeehodos: we have a scheduled time to have things done. That's the point of the deadlines. Nexenta has known about CI's since last year. They were cut in March from Kilo. Having your CI only now to get up to speed on stability is not acceptable17:47
thingeehodos: we're not going by a vendor's timeline17:47
*** Longgeek has quit IRC17:48
hodosthingee: i perfectly understand that, but that deadline didn't say how stable ci must be, that's my only point, and it's pretty good now, i don't see why we can't get our driver back since it was there before; if ci continues to run of course17:49
thingeehodos: it's common sense. if we can't see your CI passing, its not stable in an OpenStack environment from our eyes. If you're not ready for us to see things in passing in the time that we allocated to do vendor driver reviews, we can't drop our other priorities because you dropped the ball.17:51
xyangthingee: can we merge a patch that uses taskflow?  The manage/unmanage snapshot patch uses taskflow:  I know we haven't made decision on taskflow yet, so the question is do we accept a patch that introduces a new feature but uses taskflow to implement17:51
xyang it17:51
*** e0ne has joined #openstack-cinder17:51
*** IlyaG has joined #openstack-cinder17:52
*** timcl has quit IRC17:56
*** merooney has joined #openstack-cinder17:56
thingeexyang: I'd really like the people who oppose going further with taskflow to sign off on this. (e.g. jgriffith and DuncanT) ... we only today merged the first patch of refactoring17:56
thingeexyang: and the additional refactoring work is happening in l-217:56
xyangthingee: sure.  I'll ping jgriffith and DuncanT to look at the patch then17:57
*** dannywilson has quit IRC17:57
e0nethingee: fyi, we've got a very big performance impact with persistance taswflow. but may be it's not only taskflow issue17:58
thingeee0ne: oh?17:59
thingeexyang: I can retarget that to L-217:59
thingeexyang: I think we need the refactor stuff to land first, which will be in L-217:59
xyangthingee: sure, that makes sense17:59
*** gouthamr has quit IRC18:00
e0nethingee: but aarefiev found issue in dulek's patch, so performance impact could be not so big. i hope, we'll get results tomorrow18:01
*** hodos has quit IRC18:02
openstackgerritDmitry Guryanov proposed openstack/cinder: Add Virtuozzo Storage Volume Driver
*** ociuhandu has quit IRC18:06
*** timcl has joined #openstack-cinder18:06
vilobhmme0ne : nice work !18:06
e0nevilobhmm: thanks!18:07
*** annashen has joined #openstack-cinder18:07
rluciojust a general FYI, the VMEM v6000_iscsi CI is going to be disabled temporarily, the iscsi connections are acting up, and going to be replaced18:08
*** IanGovett has joined #openstack-cinder18:11
hemnaearlephilhower, ping18:11
mtaninojungleboyj: Hi,18:11
mtaninojungleboyj: May I ask a question about your oslo work?18:12
jungleboyjmtanino: Sure.18:12
*** merooney has quit IRC18:12
mtaninojungleboyj: I think oslo.service graduated from oslo-incubator recently. Do you have a plan to import it into Cinder?18:13
*** annashen has quit IRC18:13
jungleboyjYep, I saw that bug/patch you had and that was the first Oslo change I was looking at.18:13
jungleboyjHopefully today, don't know how much work it will be.  Hopefully not a lot.18:13
*** amoturi has joined #openstack-cinder18:14
mtaninojungleboyj: OK. I understand. After introducing the oslo.service, I need to more patch to fix a problem of graceful shutdown.18:14
mtaninoone more patch.18:14
jungleboyjmtanino: Right.18:14
mtaninojungleboyj: thanks :)18:14
jungleboyjmtanino: I will ping you as soon as I have a review up for you to look at.18:15
*** merooney has joined #openstack-cinder18:15
*** annashen has joined #openstack-cinder18:15
mtaninojungleboyj: sure. I understand.18:15
*** gouthamr has joined #openstack-cinder18:16
jungleboyjThanks for bringing that up.  I caught your comment on that last night and had it as a todo to look at today.18:16
mtaninojungleboyj: Also in order to fix stable/kilo, we  need to fix openstack/common/** directly, because we don't backport oslo.service for stable branch, right?18:18
mtaninoyou are welcome :)18:18
* jungleboyj 's mind is blown18:18
jungleboyjmtanino: So, do you have a pointer to what needs to change in openstack/common/**18:19
mtaninojungleboyj: openstack/common/service.py18:19
jungleboyjmtanino: Right, but what is the fix?18:19
mtaninojungleboyj: I think we need backport this fix
mtaninoAm I missing something?...18:23
jungleboyjmtanino: Looking.18:24
*** markvoelker has quit IRC18:25
openstackgerritTom Swanson proposed openstack/cinder: Dell SC Removed _find_domain and associated tests
*** markvoelker has joined #openstack-cinder18:25
jungleboyjmtanino: So, looking at bug 1464822, that is ugly looking.18:26
openstackbug 1464822 in Cinder "Cinder does not shutdown processes gracefully" [Low,In progress] - Assigned to Mitsuhiro Tanino (mitsuhiro-tanino)18:26
jungleboyjSeems like it would be good to fix that for Kilo.18:26
*** ericksonsantos has joined #openstack-cinder18:26
mtaninoyes. I agree on it.18:27
mtaninoOK. so I will propose backport patch for stable/kilo. Thanks.18:27
jungleboyjInteresting, looks like is actually still in the incubator.18:28
*** akerr has quit IRC18:28
jungleboyjmtanino: Doing some more looking.18:28
*** akerr has joined #openstack-cinder18:29
alexpilottithingee jgriffith: any chance you could help in getting this patch merged to have the Windows CI resume?
jungleboyjmtanino so there is this fix that went in to kilo for
openstackLaunchpad bug 1446583 in Keystone "services no longer reliably stop in stable/kilo" [Critical,Fix committed] - Assigned to Brant Knudson (blk-u)18:31
jungleboyjThat is not the same thing though.18:31
jungleboyjSince service is still in oslo-incubator we should propose the change there first and then pull it back into cinder ...18:32
mtaninoThis 1446583 destroyed graceful shutdown...18:32
*** mriedem has quit IRC18:32
jungleboyjThat is silly though if we are just going to move off of using openstack/common/service.py18:33
jungleboyjI am ok with doing an exception to push this back into Kilo given the situation if thingee and jgriffith don't oppose.18:33
mtaninojungleboyj: ok. let me try to propose stable/kilo and get feedback from John and Mike.18:34
jungleboyjmtanino: Works for me.18:35
jungleboyjmtanino: Thanks for your help.  Greatly appreciated!18:36
mtaninojungleboyj: You too!18:37
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Ceph driver support retries on rados_connect_timeout
openstackgerritIvan Kolodyazhny proposed openstack/cinder: Ceph driver support retries on rados_connect_timeout
*** afazekas has joined #openstack-cinder18:41
e0neam i alone not happy with code architecture like
e0neI mean a lot of drivers have something like "def create_volume(self, volume):        self.common._create_volume(volume)" instead of adding 'common' class to parents?18:43
eharneye0ne: i actually was a bit puzzled by this same thing... self.common seems wrong18:44
*** daneyon_ has quit IRC18:45
*** daneyon has joined #openstack-cinder18:46
e0neeharney: i asked this question in review and got answer like 'because we use ABC and need to inherit only from VolumeXXX classes'18:46
e0nesounds wrong18:46
eharneye0ne: yes, it does18:46
earlephilhowerhemna:  Pong.  Was doing a webinar.  Anything I can do for you?18:47
*** markvoelker_ has joined #openstack-cinder18:47
*** markvoelker has quit IRC18:47
hemnaearlephilhower, just wanted to see if you can update the HGST connector in os-brick18:48
hemnaso we can get it to land18:48
e0neeharney: i feel good that not only me is confused... is openstack-dev a good place to start conversation about it?18:48
*** Longgeek has joined #openstack-cinder18:48
* earlephilhower checks the gerrit18:49
*** daneyon_ has joined #openstack-cinder18:49
eharneye0ne: i dunno... it seems like basic code review debate to me.  The main point of common classes is so you can inherit from them, not instantiate them again...18:50
earlephilhowerhemna, sure.  I always get expected/actual backwards on asserts.18:50
*** mtecer has joined #openstack-cinder18:50
*** akerr_ has joined #openstack-cinder18:50
e0neeharney: +1. but we already have a lot of such code18:50
eharneye0ne: well, we have a lot of messy code... but we don't need to add more :)18:51
e0neeharney: :)18:51
*** akerr has quit IRC18:51
eharneye0ne: in this instance it mostly depends on how hard it is to fix imo18:51
*** daneyon has quit IRC18:52
*** Longgeek has quit IRC18:53
eharneye0ne: but looking at what's in the common class, i think it's just structured wrong18:53
e0neeharney: do you mean that one from my link above?18:54
eharneye0ne: yes18:54
*** mriedem has joined #openstack-cinder18:54
*** ociuhandu has joined #openstack-cinder18:54
*** akerr_ has quit IRC18:55
*** krtaylor has quit IRC18:57
*** mtecer has quit IRC19:02
hemnathingee, when you have a moment, can you update this guy?
hemnathingee, the cinderclient change has landed for that BP19:03
*** timcl has quit IRC19:04
openstackgerritEarle F. Philhower, III proposed openstack/os-brick: Add HGST Solutions connector
*** timcl has joined #openstack-cinder19:06
jungleboyjmtanino: You still around?19:07
mtaninojungleboyj: yes.19:07
jungleboyjhemna: Congrats!19:07
hemnajungleboyj, thanks :)19:07
hemnaNova has 3 different patches in flight still.....19:07
jungleboyjHey, I was just talking to our Oslo guy about that service change.19:07
mtaninojungleboyj: oh, nice!19:08
mtaninoso the plan is?19:08
jungleboyjHe wanted to make sure it went through incubator.  So, he is going to work on getting it into and backporting to stable so we can just do a sync the way we should.19:09
jungleboyjThere is another guy on stable that will, most likely, reject what we discussed.  :-)19:10
jungleboyjSo, let me work the process here.19:11
mtaninoso backporting oslo.service to stable/kilo is a little bit large fix, I think. but they will do it?19:11
dimsdon't think that's a good idea19:12
jungleboyjYes, the risk we run doing it the way you and I talked is that if someone else tries to sync, we lose the patch.19:12
jungleboyjdims: Ok.19:12
dimswe still haven't released or stabilized oslo.service19:12
mtaninoI understand.19:12
*** fifieldt_ has quit IRC19:12
*** bswartz has quit IRC19:12
dimssorry for jumping in :)19:12
mtaninodims: welcome :)19:13
*** IanGovett has quit IRC19:13
*** ronis has quit IRC19:14
*** BharatK has quit IRC19:14
jungleboyjdims: No problem.  We are just talking about backporting the fix for this:
openstackLaunchpad bug 1464822 in Cinder "Cinder does not shutdown processes gracefully" [Low,In progress] - Assigned to Mitsuhiro Tanino (mitsuhiro-tanino)19:14
jungleboyjfrom oslo.service to oslo-incubator's version of service.19:14
*** openstackgerrit has quit IRC19:16
jungleboyjdims: You are not ok with that?19:16
*** openstackgerrit has joined #openstack-cinder19:17
*** jecarey has joined #openstack-cinder19:20
*** afazekas has quit IRC19:21
dimsjungleboyj: when we start cutting releases, we remove the incubated code19:22
*** zongliang has quit IRC19:22
ociuhanduthingee: can you or anyone else help us with reviews on so we can bring back asap the Microsoft CI?19:22
dimslet's talk on oslo channel, as we can include dhellmann as well19:22
jungleboyjdims: Ok.19:24
*** mtecer has joined #openstack-cinder19:27
*** merooney has quit IRC19:28
*** krtaylor has joined #openstack-cinder19:29
*** fifieldt_ has joined #openstack-cinder19:30
*** merooney has joined #openstack-cinder19:32
*** jungleboyj has quit IRC19:32
*** ronis has joined #openstack-cinder19:38
*** IanGovett has joined #openstack-cinder19:42
*** nkrinner has joined #openstack-cinder19:46
*** Longgeek has joined #openstack-cinder19:49
*** jungleboyj has joined #openstack-cinder19:50
*** Yogi1 has quit IRC19:51
*** sgordon has joined #openstack-cinder19:56
thingee hemna done19:56
*** Rockyg has joined #openstack-cinder19:57
*** Rockyg has quit IRC19:57
*** timcl has quit IRC19:58
*** Ctina_ has quit IRC19:59
*** Rockyg has joined #openstack-cinder19:59
*** Yogi1 has joined #openstack-cinder19:59
rlucioe0ne: the primary issue with the inheriting from common class is that we are arequired to inherit from the driver.ISCSIDriver or driver.FibreChannelDriver20:00
rlucioe0ne: I got dinged pretty hard in kilo for that on the other vmem drivers20:01
jungleboyjmtanino: Talking over in the oslo channel.  Maybe my first idea was right.  Will let you kow.20:01
rlucioe0ne: so i restructured to use the common class more like a utility library20:01
e0nerlucio: you can inherrit from driver.ISCSIDriver and your DriverNCommonClass20:01
mtaninojungleboyj: waiting the discussion result :)20:01
rlucioe0ne: is that acceptable for coding standards?20:02
e0nerlucio: why not? we use multiple inheritance in Cinder20:02
rlucioe0ne: I personally think it should be OK, but I don't want to restructure everything and then be told NO later20:03
rlucioe0ne: cause that sucked last time20:04
e0nerlucio: hm... did somebody told you that you should not inherit from you common class?20:05
*** kjnelson has quit IRC20:05
*** Longgeek has quit IRC20:06
rlucioe0ne: yea, back in kilo when i submitted the v6000_fcp and v6000_iscsi drivers20:07
*** marcusvrn1 has quit IRC20:07
e0nerlucio: i'm very surprised20:07
rlucioe0ne: i don't recall exactly who it was unfortunately, do we keep reviews that old around?  I could try to dig it up20:07
e0nerlucio: is it hard for you to find that review request?20:07
rlucioe0ne: gimme a min, i'll try20:08
e0nerlucio: sure, thanks20:08
hemnathingee, thank you20:08
e0nerlucio: btw, i found my first commit in gerrit made in Dec, 2011:)20:09
rlucioe0one: looks like it was hemna who mentioned it,
rlucioe0ne: patch set 5 comments20:11
*** hodos has joined #openstack-cinder20:11
hemnarlucio, well the driver wasn't extended from the FC parent class20:12
rlucioe0ne: here is direct link:
e0nerlucio: i believe, you misunderstood each other20:13
rlucioe0ne: how so?20:13
e0nerlucio: 1min, please. trying to draw ascii in paste20:14
rlucioe0ne: lol, k20:14
*** sgotliv_ has joined #openstack-cinder20:14
*** Lee1092 has quit IRC20:15
e0nei failed with ascii:(20:15
geguileoAnybody knows how capacity filter works?20:16
e0nehemna: did you mean something like class V6000FCDriver(driver.FibreChannelDriver, v6000_common.V6000CommonDriver)? for rlucio's patch?20:16
geguileoI'm trying to figure out a patch and I must be missing something with free_capacity_gb, total_capacity_gb, etc20:16
geguileoIt's related to thin provisioning20:17
hemnae0ne, yes20:17
*** sgotliv has quit IRC20:17
e0nerlucio: so, looks like now we understand what we want, do we?20:18
rlucioe0ne, hemna: huh.  well i did indeed misunderstand him20:18
rlucioe0ne: yea20:18
e0nerlucio: great!20:18
*** amoturi has quit IRC20:18
rlucioe0ne: gonna be a bit of work but i'll fix it up20:19
e0nerlucio: thanks!20:19
*** dannywilson has joined #openstack-cinder20:19
*** dannywilson has quit IRC20:19
e0nerlucio: i'll review your patch tomorrow morning if new version will be available20:19
rlucioe0ne: sounds good, thanks a bunch20:19
*** dannywilson has joined #openstack-cinder20:20
e0nerlucio: np20:20
thingeeameade: hey question about the eseries FC CI20:20
ameadethingee: what's up?20:20
e0nebb in ~12 hours. time to sleep:)20:21
thingeeameade: you have zero failures in your last 200 comments. that's impossible20:21
ameadethingee: I dont allow failures to post until they are analyzed20:22
thingeeameade: uhh ok, we need to change that20:22
ameadethingee: I can do that, why exactly though?20:23
*** merooney has quit IRC20:23
thingeeameade: I highly doubt someone is paid full time by netapp to analyze these failures and give them at a reasonable time for reviewers to make decisions.20:23
thingeeameade: and also, there have been valid failures in the last 200 patches.20:24
*** thangp_ has quit IRC20:24
thingeeameade: so this is not working20:24
ameadethingee: I look at them, and they only run on patches that have Jenkins +120:24
ameadethingee: the alternative is failures posted on patches that could be false failures and nobody looks at them but me anyways20:25
thingeeameade: my point is in your last 200 comments, none have been marked as failures for valid reasons of the patch not working. There have been patches that should've been marked as failure by your CI20:25
thingeeameade: I will say it again, this is not working20:25
thingeeameade: I will post this to third party announce list as no one should be doing this.20:25
ericksonsantosvilobhmm, ping20:25
thingeeameade: This even comes across shady. You have some vendors who aren't even making it into the release because of the amount of false failures they have been posting20:26
thingeeameade: does this make sense?20:26
*** thangp has joined #openstack-cinder20:27
ameadekind of? I still think it's better that if a CI says failure on a patch that it actually means failure, if it says thumbs up on a patch that it shouldnt then that is a real problem. I will go ahead and tweak it to post no matter what though20:28
*** lpetrut has quit IRC20:29
thingeeameade: wow, I need someone else at netapp to talk to then.20:29
thingeepreferably someone who understands the point of the CI's here.20:29
ameadethingee: what? i'm trying to discuss what is best here, as a member of the community20:29
ameadeI am changing the netapp ci right now20:30
ameadei am not a vendor, i run a vendors CI20:30
thingeeameade: I pointed out shadiness here and you're slapping me in the face here by saying you "kind of" understand?20:31
hemnaall failures should post20:32
ameadethingee: as someone with pure intentions i'm not thinking with shadiness in the forefront of my mind20:32
hemnahiding failures is bad mmmkay20:32
*** e0ne has quit IRC20:32
thingeeameade: I will type it again, you have vendors like Nexenta who are not making it in the release because of their false  CI failures. You don't see anything wrong with hiding *all* of your failures here?20:32
ameadei'm not hiding failures, i'm just not posting false failures20:33
thingeeameade: and what's the point of us automating this if you can't tell me if this patch breaks your integration?20:33
ameadenothing i have seen it run against has broken the integration20:33
thingeeameade: in the last 200 runs there should be failures being posted, and your CI is claiming 100% passed20:33
*** e0ne has joined #openstack-cinder20:33
*** mtecer has quit IRC20:33
thingeethe best part of all of this is it's just the CI with the separate account for the driver that needed to be accepted in liberty according to stats.20:35
thingeethe other netapp ci's are posting failures20:36
ameadeit's all the FC drivers20:36
ameadethey have a seperate account because they are a separate CI system20:36
ameadei'm an open book man, ask me anything you want20:36
ameadejust updated it to post it's failure20:36
ameadei'm not trying to be dissenting here20:37
ameadeit's a little offensive to assume something shady20:37
hemnanot reporting failures is less than open.20:39
*** akerr has joined #openstack-cinder20:39
hemnawe report failures20:40
hemnaand those failures on the hp CI are failures of our own doing here.20:40
hemnaand yet, we are analyzing them here.20:40
hemnabut still reporting them.20:40
ameadehemna: I originally have this system let me review failures before they post so that incase the fabric or something gets out of whack i dont spam a bunch of patches with false negatives20:40
hemna(internal network problems)20:40
*** e0ne has quit IRC20:41
openstackgerritEric Harney proposed openstack/cinder: Sync 'report' from oslo-incubator
thingeeameade: I don't expect you to keep up with discussions, but we had a previous vendor take this approach and it was clear that this defeats the purpose of the CI.20:41
hemnaI guess we are 'spamming' then20:41
ameadedoesnt that desensitize submitters?20:41
thingeeameade: the point is you have pointed *zero*, again *zero* failures. You're system does not work.20:41
thingeethere are correct times to report failures20:42
openstackgerritEric Harney proposed openstack/cinder: ScaleIO: Fix broken format string
openstackgerritVictor Stinner proposed openstack/cinder: tox: add a basic py34 venv running a single test
*** ronis has quit IRC20:46
*** lan has quit IRC20:47
hemnaspeaking of reporting failures20:50
jgriffithhemna: ping20:50
hemnaeven Datera reports failures20:50
hemnajgriffith, hey20:50
jgriffithhemna: thingee Yo.. just reading back the meeting logs20:50
jgriffithhemna: thingee so there was some misunderstanding of my comment I think;20:50
ameadei've already turned it on and said why i had it that way20:50
ameadei can see how that could look shady20:50
jgriffithhemna: thingee what I meant was set "multi-attach=True" in the base so it's inherited rathr than chase it through every driver20:51
jgriffithhemna: thingee but I'm good either way20:51
ameadenot sure i agree with the policy of posting false failures though20:51
hemnaposting of failures in general20:51
hemnayou don't know they are a false failure until it's analyzed20:51
hemnapost results first20:51
asselinameade, if your ci posts too many false failures, you need to turn it off altogether20:51
hemnaanalyze next.20:51
hemnathat's what we do20:51
*** lan has joined #openstack-cinder20:51
*** cknight has joined #openstack-cinder20:52
eharneyi would like if people could stop posting false "Merge Conflict" failures... those are just bothersome20:52
jgriffitheharney: +120:52
asselinameade, otherwise that's something that needs to be fixed through maintenance and robustness20:52
jgriffitheharney: and they're not supposed to be reporting any of that anyway :)20:52
jgriffitheharney: but that may be just my "interpretation"20:52
ameadeasselin: I think the problem here is it's too robust20:52
ameadehemna: yeah that makes sense, i was just doing it in a different order20:53
eharneyjgriffith: well, merge conflict = i did something messy in my CI (extra patches etc) and broke it20:53
hemnaameade, then usually we issue a recheck after we fix the failure on our end.20:53
*** jskinner has quit IRC20:54
jgriffitheharney: I know what it means, I was just saying third party shouldn't be saying anything about it to begin with.20:55
akerreharney: I'm not so sure they're always false.  We wait until after jenkins votes to even start running so in the 40-50 minutes jenkins takes to run its very possible for the code base to change20:55
jgriffitheharney: but I guess maybe I'm wrong there20:55
jgriffitheharney: and that's what I meant by it's "my interpretatoin"20:55
jgriffitheharney: but obviously I must be wrong, so forget I said anything :)20:55
eharneyakerr: i don't know how you can get a merge conflict... it merged into gerrit... if it can't merge into your CI then something is happening wrong20:56
akerrbut if the merges happen at different times then the state of the repository is different20:56
jgriffithakerr: why are you merging anything?20:56
akerrjgriffith: because zuul uses merger20:57
eharneyit shouldn't be, you should be testing it on top of the same parent it has in gerrit20:57
jgriffithfor check?20:57
eharneyi dunno how this makes sense otherwise20:57
eharneycan't have everyone's CI testing different code for the same patchset20:57
*** nkrinner has quit IRC20:58
*** nikeshm_ has joined #openstack-cinder20:58
*** delattec has joined #openstack-cinder20:59
*** delatte has joined #openstack-cinder20:59
eharneyeven scarier what that means for rechecks20:59
*** gouthamr has quit IRC20:59
jgriffithakerr: I'm still trying to figure out why you care?20:59
jgriffithakerr: the check is "supposed" to be: download the patch in gerrit and run it20:59
jgriffithakerr: done20:59
jgriffithakerr: what's in gerrit is what you get20:59
*** nikeshm has quit IRC21:00
jgriffithakerr: no merging, no funny business merging change sets etc21:00
*** dulek_home has joined #openstack-cinder21:00
jgriffithakerr: that's why I'm confused (sorry)21:00
*** cdelatte has quit IRC21:00
jgriffithakerr: and you're *wrong* WRT the comment about "different states of the repo"21:00
jgriffiththat has nothing to do with gate checks21:00
*** nikeshm has joined #openstack-cinder21:01
*** dulek_home has quit IRC21:01
jgriffiththat's a part of the verify/merge process in the gate, and the gate is the ONLY one that has the ability or intelligence to do that21:01
akerrjgriffith: we built our system from the tools infra had out about 1.5 years ago (before all the work was done to try and make it easier to consume).  So we are using zuul and the gerrit triggers.  Internally zuul then uses merger to put the patch on the latest code in the repo for whichever branch its on.  I suppose we could go do surgery to zuul to change that21:01
jgriffithakerr: well.. you're using it wrong :)21:01
*** dulek has joined #openstack-cinder21:02
eharneyyeah... that just doesn't make sense conceptually21:02
eharneyi wonder if others are doing this21:02
*** jungleboyj has quit IRC21:02
asselinakerr, I think the issue is that the patch is in the pipeline, but there's another patch that causes a merge conflict that merges.21:02
jgriffithakerr: sounds like you implemented verify instead of check maybe?21:02
asselinakerr, upstream, they -1 the patch and remove it from the queue.21:02
akerrcheck and verify just have different triggers21:02
jgriffithasselin: that's the verify stage, check stage doesn't care about that21:02
asselinakerr, I think you leave it in the queue, and it fails when it gets there21:03
eharneybut this scheme means that even "recheck" doesn't recheck the same code, right?21:03
*** nikeshm_ has quit IRC21:03
jgriffithasselin: akerr please see:
jgriffithasselin: akerr notice check VS gate21:03
asselinjgriffith, akerr no I'm talking about the "merger-check" pipeline:
jgriffithasselin: akerr of course currently it's kinda quiet so doesn't illustrate the dependencies and merge actions21:03
akerri think what happens is that a patch merges while jenkins is running.  we then update our repo to be latest, when jenkins finishes we see the +1 and grab the patch, attempt to put it on a newer version of the repo that it now conflicts with.  Jenkins will have the same issue when it runs next21:03
jgriffithasselin: I think you and I are mostly on the same page, I'm saying akerr 's system is doing things it shouldn't be doing ;)21:04
eharneyakerr: but why are you updating the repo to the latest?21:04
*** annashen has quit IRC21:05
akerrto stay up-to-date21:05
jgriffithakerr: why do you update your repo to be the laest though... that's my point.  You shouldn't be doing that21:05
jgriffithakerr: you don't WANT to be up to date, you want to run the patch under test21:05
jgriffithakerr: so all of your testing over the last however long is basically bogus it sounds like21:05
akerri don't see why this is an issue.  If it conflicts in our system it will conflict in upstream too, we just catch it first21:05
ameadethen how do you know it'll work with your drivers when it does merge?21:05
jgriffithakerr: cuz you're doing some crazy internal update/merge scheme21:06
jgriffiththat isn't even testing the submitted patch21:06
eharneyit's still valid testing, it just means that there's a lot more inconsistency between CIs21:06
*** sgotliv_ has quit IRC21:06
ameadethis sounds like bikeshedding to me21:06
eharneyseems bad21:06
jgriffithameade: it's most certainly not bikeshedding.. .not in the least21:06
*** edmondsw has quit IRC21:06
jgriffithameade: it's one system testing something different than what it says it's testing21:06
jgriffithameade: akerr anyway... whatever.  Just quit reporting it :)21:07
ameadeit all depends on what you expect i suppose, if you assume it's not doing merges then yeah its lying to you21:07
* jgriffith walks away whistling theme to Andy Griffith21:07
akerrjgriffith: fair enough, i'll update the pipeline right now to not report merge conflicts21:07
ameademaybe we should make this clear as a req?21:07
eharneyit could at least report them as "unstable" instead of "failure" if reporting them21:08
eharneybut i dunno what the best plan is really21:08
akerrugh, can only modify the merge-failure-message, not the actual reporting of the failure21:11
asselinsorry...what's the issue with reporting merge failures?21:11
akerrasselin: that's my question :)21:11
ameadecosmetic :)21:12
eharneyasselin: they don't make any sense and basically indicate that the CI system is broken (IMO)21:12
eharneybut apparently there are differing theories on this that i had no idea about21:12
ameadeor working? if we dont want folks merging we gotta make that a rule21:12
ameadei do think there is value to it we are overlooking21:12
eharneyi never thought that that was what these CIs were supposed to report on though21:12
eharneyi mean, it sounds useful... but... very different from what jenkins does and what i thought the scheme was21:12
*** rushil has joined #openstack-cinder21:12
akerrit's exactly what jenkins does, we built it from jenkins21:13
ameadethe difference is these CIs dont run in the gate21:13
ameadeso if you are gonna try with the merge you gotta do it on check21:13
*** haomaiwang has joined #openstack-cinder21:13
asselinmany of us are using zuul which is reporting the merge failure. So it's not something specific anyone is doing21:13
eharneyasselin: maybe i've missed previous discussion on this (i haven't heard any before)21:13
*** porrua has joined #openstack-cinder21:14
asselinbut I think there are ways to work around it21:14
akerrthe difference between check and gate is that gate is a dependent pipeline, so it merges several patchsets, in order, when it runs.  check only merges the one patch onto the latest repo version21:14
eharneyit seems weird to have everyone's CI running against different code depending on when they start21:14
asselinby e.g. look at what infra does in the merge-check pipeline21:14
*** haomaiw__ has quit IRC21:15
eharneyand when you do a recheck, you may recheck against different code?  that doesn't seem useful21:15
jgriffithakerr: ding ding ding ding :)21:15
asselineharney, that's the case regardless of merge failures, b/c every project is on master and master is constantly changing21:15
asselinakerr, no, check doesn't merge21:15
eharneyasselin: not for the check queue...?21:15
asselinakerr, check does the equivalent of checkout in the gerrit UI21:16
asselineharney, it uses the patch as-is for cinder, and master for everything else21:16
asselinmaster for everything else is different b/w different ci runs21:16
asselinbut cinder should be the same21:16
asselinexactly what that patch set says it should be21:17
eharneyi didn't know that we were special in that regard... why?21:17
asselineharney, in my example, I'm assuming you're testing a cinder patch21:17
asselineharney, if it were a nova patch, then nova is fixed, everything else is master21:17
eharneyoh, yeah, i get that part21:17
*** dulek has quit IRC21:18
akerrpretty sure it does the merge.  it uses the ZUUL_REF to do the checkout for cinder, which go through zuul-merger21:19
eharneyi may just have no idea how anything works21:19
vilobhmmericksonsantos : ack21:20
openstackgerritThang Pham proposed openstack/cinder: Conversion to volume object
akerrif it never did the merge, how would a patch EVER be in merge conflict?21:21
*** haomaiw__ has joined #openstack-cinder21:21
eharneybut that's my original point.  i'm getting "merge conflict" failures from CI systems when there isn't a conflict in Jenkins upstream (i think... don't have an example at the moment)21:21
*** haomaiwang has quit IRC21:21
*** Yogi1 has quit IRC21:22
asselineharney, which patch?21:23
eharneyasselin: well... i was just saying i don't have an example at the moment, but i'm fairly sure i've seen a handful of them21:24
*** openstackgerrit has quit IRC21:24
akerreharney: I think that's where my point about the fact that we're waiting until after jenkins votes is coming into play.  If the base repo changes while jenkins is running then subsequent CI system runs could hit the merge conflict while jenkins was still able to vote.  I'll bet if you try a recheck after you get a merge conflict from a CI (that's not messed up, i'll concede it's possible for the system to also be in error) you'll probably hit a merge21:24
akerr conflict in jenkins too21:24
ameadeheck if anything it's helpful in that case lol21:24
eharneyyeah, i understand how it can happen now21:24
SwansonIs there any reason I want to inherit from one of the san. drivers rather than just driver.iscsiyada and driver.fibrechannelyada?21:24
*** openstackgerrit has joined #openstack-cinder21:24
*** dulek_home has joined #openstack-cinder21:25
*** kjnelson has joined #openstack-cinder21:25
eharneythis was not included in the "Important CI Details for the Cinder Core Reviewer" Handbook21:25
*** thangp has quit IRC21:25
*** dulek_home has quit IRC21:25
akerrheh, it's taken me a year to finally wrap my head around how this system really works21:25
eharneywell i thought i had done that, but, apparently not21:26
*** dulek has joined #openstack-cinder21:26
eharneyat any rate, i'll consider that less totally wrong now based on hearing the explanation21:28
*** erlon has quit IRC21:30
*** jungleboyj has joined #openstack-cinder21:34
*** Rockyg has quit IRC21:38
*** annashen has joined #openstack-cinder21:42
*** porrua has quit IRC21:45
earlephilhowerasselin, hemna:  Should 3rd party CIs fire jobs on simple "recheck" (as opposed to "recheck <vendor-specific-tag>"?  The os-ext-testing-data default regex doesn't fire on naked "recheck"s, just "recheck no bug" or "recheck bug ###".21:49
asselinearlephilhower, yeah that was the old way21:49
asselinearlephilhower, we currently fire off of "recheck"21:50
asselinearlephilhower, you can choose your own regex, I'm not familiar with an specifici requirements other than the -infra team perfers 3rd party ci fire off of "recheck .*"21:50
*** adurbin__ has quit IRC21:51
*** rushil has quit IRC21:52
earlephilhowerasselin, Thanks.  Makes the regex less painful to drop the add'l conditions.21:52
*** ianbrown has joined #openstack-cinder21:53
*** xyang has quit IRC21:57
*** mriedem has quit IRC21:58
*** kjnelson has quit IRC22:03
*** Longgeek has joined #openstack-cinder22:03
akerrasselin: just a suggestion for other CI systems, we take both "recheck netapp" and "run netapp" as valid triggers, since recheck netapp will also trigger jenkins and run will not.  Lets us recheck just our own system and not bother upstream's system22:03
*** bswartz has joined #openstack-cinder22:05
*** jungleboyj has quit IRC22:09
tbarronakerr: +122:10
*** delatte has quit IRC22:10
*** delattec has quit IRC22:10
*** leeantho has quit IRC22:13
*** dannywilson has quit IRC22:14
*** rmesta has quit IRC22:16
*** rmesta has joined #openstack-cinder22:16
openstackgerritxing-yang proposed openstack/cinder: Add multiple pools support to VMAX driver
*** delattec has joined #openstack-cinder22:18
*** delatte has joined #openstack-cinder22:18
openstackgerritDoug Hellmann proposed openstack/cinder: Update version for Liberty
*** julim has quit IRC22:19
*** dims has quit IRC22:26
*** rmesta has quit IRC22:29
*** rmesta has joined #openstack-cinder22:29
*** cknight has quit IRC22:30
*** jecarey has quit IRC22:30
*** Longgeek has quit IRC22:36
*** lcurtis has quit IRC22:39
asselinakerr, yeah...we might do something like that22:42
*** alau2 has quit IRC22:43
asselinakerr, was just talking to hemna about it. if we do that, it would be good to standardize it to e.g. check <CI Reviewer Name>22:43
*** krtaylor has quit IRC22:44
*** jasondotstar has joined #openstack-cinder22:45
*** IanGovett has quit IRC22:45
*** ociuhandu has quit IRC22:49
*** IanGovett has joined #openstack-cinder22:56
*** ganso_ has quit IRC22:58
jgriffithakerr: IIRC the standard is "recheck-<ciName>"22:59
jgriffithakerr: the '-' makes it so the gate won't pick it up23:00
jgriffithakerr: it also lets *anybody* trigger a run on any specific system without having to "know* anything about it23:00
asselinjgriffith, thanks good to know23:02
*** IanGovett has quit IRC23:06
*** crose has quit IRC23:09
*** Adriano__ has quit IRC23:11
*** dannywilson has joined #openstack-cinder23:16
hemnaproblem solved :)23:21
*** dims has joined #openstack-cinder23:24
* tbarron likes having a standard way since lots of CIs don't announce their way to recheck23:26
tbarronjgriffith: I've been thinking a bit about your idea, expressed during the scaling backup services session in vancouver23:27
tbarronof using service VMs (dunno if that was your phrase) for running backup services23:27
tbarronI like the idea23:27
tbarronor even potentially kolla containers23:28
tbarronany precedent for this with cinder services?23:28
*** dims has quit IRC23:28
*** dims has joined #openstack-cinder23:28
tbarronor magnum containers, maybe I meant, gotta learn about this stuff.23:29
*** hodos has quit IRC23:29
tbarrondon't tie down physical infra, scale out backup servers on demand23:29
*** dannywilson has quit IRC23:31
*** diemt has quit IRC23:31
jgriffithtbarron: nope, it's brand new concept for cinder23:31
*** diemt has joined #openstack-cinder23:31
*** krtaylor has joined #openstack-cinder23:32
*** dims has quit IRC23:33
tbarronjgriffith: that's what I figured23:35
tbarronjgriffith: well, we'll start as planned, enabling backup service to run on multiple "nodes", and on separate nodes than the volume services23:36
*** rmesta has quit IRC23:37
tbarronjgriffith: then we can look at service vms, or container/minions23:37
tbarrononce things are decoupled23:37
jgriffithtbarron: You're call... I don't know how the multiple physical nodes idea would fly23:37
jgriffithtbarron: but it's a good way to step through things I suppose23:37
tbarronif you think of other cinder services that might fit that model, lemee know23:37
tbarronjgriffith: dedicating multiple physical nodes would be bad, but enabling it proves the decoupling23:38
*** chlong has joined #openstack-cinder23:38
tbarronjgriffith: make sense?23:38
jgriffithtbarron: yeah, sure23:38
tbarronjgriffith: i'm just setting up multinode devstack for the first time :-) to see what I can figure out on this front23:39
tbarronjgriffith: haven't forgotten about helping with the resource-tag logging either, just was away for a few days23:39
tbarronkk - i'm AFK for a while, ty for listening23:40
jgriffithtbarron: :)  NO worries at all23:40
jgriffithtbarron: and starting w/ multi-node devstack isn't a bad idea at all23:41
*** adurbin_ has quit IRC23:42
*** mriedem has joined #openstack-cinder23:42
*** diogogmt has quit IRC23:45
*** mriedem1 has joined #openstack-cinder23:45
*** jskinner has joined #openstack-cinder23:46
*** mriedem has quit IRC23:47
*** earlephilhower has quit IRC23:54
*** laughterwym has joined #openstack-cinder23:55
*** garthb_ has quit IRC23:57
*** garthb has quit IRC23:57

Generated by 2.14.0 by Marius Gedminas - find it at!