Wednesday, 2018-07-18

*** dalgaaf has quit IRC00:05
*** dalgaaf has joined #openstack-glance00:07
*** btully has joined #openstack-glance00:10
*** btully has quit IRC00:14
*** threestrands has joined #openstack-glance00:40
*** threestrands has quit IRC00:40
*** threestrands has joined #openstack-glance00:40
*** itlinux has quit IRC00:59
*** harlowja has quit IRC01:13
*** abhishekk|mbl has joined #openstack-glance01:26
*** mfedosin has quit IRC02:03
openstackgerritBrian Rosmaita proposed openstack/glance_store master: Multihash Implementation for Glance  https://review.openstack.org/52879002:11
*** rosmaita has quit IRC02:19
*** gcb has joined #openstack-glance02:22
*** mfedosin has joined #openstack-glance02:27
*** dpawlik has joined #openstack-glance02:41
*** dpawlik has quit IRC02:45
*** udesale has joined #openstack-glance03:24
*** abhishekk|mbl has quit IRC03:35
*** btully has joined #openstack-glance03:48
*** btully has quit IRC03:52
*** udesale has quit IRC04:55
*** udesale has joined #openstack-glance04:55
*** abhishekk has joined #openstack-glance04:59
*** pdeore has joined #openstack-glance05:04
*** MrWatson has joined #openstack-glance05:10
*** openstackgerrit has quit IRC05:10
*** links has joined #openstack-glance05:15
*** jmlowe has quit IRC05:16
*** wxy has quit IRC05:16
*** NostawRm has quit IRC05:16
*** serverascode has quit IRC05:16
*** ameade has quit IRC05:16
*** pas-ha has quit IRC05:16
*** DuncanT has quit IRC05:16
*** Roamer` has quit IRC05:16
*** jmccrory has quit IRC05:16
*** harlowja has joined #openstack-glance05:28
*** harlowja has quit IRC05:32
*** jmlowe has joined #openstack-glance05:35
*** wxy has joined #openstack-glance05:35
*** serverascode has joined #openstack-glance05:35
*** ameade has joined #openstack-glance05:35
*** pas-ha has joined #openstack-glance05:35
*** DuncanT has joined #openstack-glance05:35
*** Roamer` has joined #openstack-glance05:35
*** jmccrory has joined #openstack-glance05:35
*** ratailor has joined #openstack-glance05:35
*** Luzi has joined #openstack-glance05:45
*** bkopilov has joined #openstack-glance06:03
*** tshefi has joined #openstack-glance06:03
*** pcaruana has joined #openstack-glance06:35
*** alexchadin has joined #openstack-glance06:36
*** dpawlik has joined #openstack-glance06:49
*** openstackgerrit has joined #openstack-glance06:52
openstackgerritwangxiyuan proposed openstack/python-glanceclient master: Do not quote token  https://review.openstack.org/58346806:52
*** peereb has joined #openstack-glance07:09
*** tesseract has joined #openstack-glance07:18
*** btully has joined #openstack-glance07:25
*** rcernin has quit IRC07:28
*** btully has quit IRC07:30
*** AlexeyAbashkin has joined #openstack-glance07:45
openstackgerritwangxiyuan proposed openstack/python-glanceclient master: Do not quote token  https://review.openstack.org/58346808:01
*** markvoelker_ has quit IRC08:07
jokke_need to run an errand ... back at the afternoon08:35
*** bkopilov has quit IRC08:44
abhishekkjokke_, ok08:49
*** threestrands has quit IRC08:51
*** dtantsur|afk is now known as dtantsur09:20
*** hoonetorg has quit IRC09:34
*** hoonetorg has joined #openstack-glance09:48
*** abhishekk has quit IRC10:03
*** markvoelker has joined #openstack-glance10:08
*** alexchadin has quit IRC10:08
*** kukacz_ has quit IRC10:12
*** kukacz_ has joined #openstack-glance10:13
*** alexchadin has joined #openstack-glance10:22
*** dpawlik has quit IRC10:39
*** dpawlik has joined #openstack-glance10:41
*** markvoelker has quit IRC10:42
*** priteau has joined #openstack-glance10:42
*** bkopilov has joined #openstack-glance10:54
*** dpawlik has quit IRC10:55
*** dpawlik has joined #openstack-glance10:56
*** btully has joined #openstack-glance11:03
*** btully has quit IRC11:07
smcginniso/11:23
*** pdeore has quit IRC11:25
*** ratailor has quit IRC11:38
*** markvoelker has joined #openstack-glance11:39
*** abhishekk|mbl has joined #openstack-glance11:44
*** markvoelker has quit IRC12:05
*** markvoelker has joined #openstack-glance12:05
*** MattMan has quit IRC12:06
*** MattMan has joined #openstack-glance12:07
*** edmondsw has joined #openstack-glance12:11
*** dpawlik has quit IRC12:30
*** rosmaita has joined #openstack-glance12:36
rosmaitaabhishekk|mbl : thanks for your multihash review, and especially for the testing you've done12:46
rosmaitagood point about the docstrings, i'll fix that now12:46
*** dpawlik has joined #openstack-glance12:47
*** btully has joined #openstack-glance12:51
*** dpawlik has quit IRC12:52
*** btully has quit IRC12:56
openstackgerritBrian Rosmaita proposed openstack/glance_store master: Multihash Implementation for Glance  https://review.openstack.org/52879013:02
*** lbragstad has quit IRC13:19
openstackgerritBrian Rosmaita proposed openstack/glance_store master: Multihash Implementation for Glance  https://review.openstack.org/52879013:28
rosmaitasmcginnis ^^ when you get a chance later, i added one more test (no code changes, though)13:29
*** gcb has quit IRC13:31
smcginnisrosmaita: Awesome, more tests are always a good thing.13:32
smcginnisWell, almost always. ;)13:32
*** alexchadin has quit IRC13:34
*** alexchadin has joined #openstack-glance13:34
*** alexchadin has quit IRC13:35
*** alexchadin has joined #openstack-glance13:35
*** lbragstad has joined #openstack-glance13:46
*** r-daneel has joined #openstack-glance13:48
*** dpawlik has joined #openstack-glance13:51
abhishekk|mblrosmaita: no issues14:09
rosmaitaabhishekk|mbl ty14:11
rosmaitaabhishekk|mbl : how did you like the rewritten back-compat wrapper?14:12
abhishekk|mblrosmaita: its less complex and mor flexoble14:14
rosmaitai tried to use the inspect module like erno suggested, but the problem is the way the compatability check wrapper is tested -- it relies on the name of of the function returned to verify that each driver is wrapped, so i had to wrap the add() with capabilities_check first, then my wrapper14:14
rosmaitai didn't want to get into trying to refactor that stuff14:15
abhishekk|mblSorry, i didn't checked what erno has suggested but the wrapper wriiten was very easy than the previous one14:23
*** udesale_ has joined #openstack-glance14:26
*** udesale has quit IRC14:27
*** alexchadin has quit IRC14:35
*** alexchadin has joined #openstack-glance14:41
jokke_rosmaita: so did the capabilities check wrapper break the inspect?14:50
*** peereb has quit IRC14:51
rosmaitayeah, i didn't investigate further, was getting an error something like hashing_algo unknown argument14:53
rosmaitawhen i did the inspect.getargspec14:53
rosmaitathere's probably a way to get around that but i didnt' want to get into it14:54
*** pcaruana has quit IRC14:56
rosmaitaactually, it was when i tried to get the position of hashing_algo from the argspec14:57
*** jmlowe has quit IRC15:00
jokke_I think literally only thing what was needed was that @wraps(...) wrapper for the inner function of that capabilities check15:00
jokke_or decorator I should say ... as decorating function without it basically hides the original function details otherwise ands the capabilities.check function (which is decorator) does not have it15:02
jokke_if this works as it is I'm fine with that for now ... i might play around with it some day if I get super bored15:04
jokke_but thanks, the back_compat_add looks already much cleaner and easier to follow15:05
rosmaitayeah, i tried that, but there's a test for compatability check that relies on the name of the wrapped function being compat_check or something, so when i used @wraps on compat-check those tests all failed15:08
jokke_Oh FFS15:08
jokke_;)15:08
rosmaitathat's why i had to use @wraps so that the name of the function you get back from my wrapper is actually named compatabilitiy_check15:08
rosmaitayeah15:08
rosmaitathere's definitely a better way to do that stuff, but i figure that's a different patch15:09
jokke_ohh yeah15:09
*** tshefi has quit IRC15:09
rosmaitain any case, your suggestion was really good, simplified that crazy thing i wrote by a lot!15:09
* jokke_ makes a mental note of the next set of shitty tests to throw away15:09
rosmaitawell, they're good tests in the sense that they make sure the capabilities wrapper is being used, it's just the way they do it that sucks15:10
*** alexchadin has quit IRC15:10
rosmaita... and where i'm saying compatability_check above, i of course mean capabilities_check15:11
rosmaita:)15:11
jokke_yeah15:11
rosmaitai have backward compatability stuck in my head!15:11
jokke_;)15:11
jokke_yeah you know the problem with python is that it's super relaxed how you can call the functions15:12
jokke_which means that any decorators like this gets easily super complex without either hardcoding stuff (like you had to do now, and possibly causing grey hair to someone in future) or the inspect15:13
rosmaitayeah, the inspect module is cool15:13
jokke_I've really learned to love that feature when using decorators15:13
jokke_'cause you can have pretty muc any combo of positional and kw args as long as you don't call with keyword anything that is at/before the positionals and it will be just fine function call15:15
jokke_also totally unrelated to your change, but I think we should throw that _check_metadata away from production code and move it to tests ;)15:18
jokke_not trusting our own driver code in production is extremely wasteful thing to do (bit like schema validating every field in the image-list call) ;)15:18
rosmaitayeah, i was wondering about that, but figured it was not a related change15:22
rosmaitayes, would be better in some kind of test15:22
*** jmlowe has joined #openstack-glance15:24
jokke_so instead of extending the wrapping to store_add_to_backend you just plan to use the store_add_to_backend_with_multihash in glance internally to maintain the backwards compatibility on that layer of the glance_store?15:25
*** jose-phillips has quit IRC15:25
rosmaitayeah, i figured that the store_add_to_backend is supposed to be the entry point for consumers, so best to leave that one as is15:26
rosmaitabut i don't know, it's really just us and glare at this point15:27
jokke_I'm fine with that, could you please make a follow up patch to deprecate it so we can throw it away on the 1.0.0 release as well. We wanted to be backwards compatible with muti back-ends and the multihash as we don't have time to warn anyone we're changing things right now, but I'd like to get that cleaned out as well15:27
jokke_we Glare we know of15:27
jokke_then there is the possibility that X other things, perhaps outside of the OpenStack ecosystem is using it15:28
jokke_*well Glare15:28
rosmaitaright15:28
rosmaitaok, i can put up a deprecation patch15:29
jokke_We've been asked to provide secure hashing for so long that I really don't want to maintain non-secure path any longer we absolutely have to15:29
rosmaitai'll do it now, meeting i was supposed to have got cancelled15:29
jokke_coolcool15:30
*** links has quit IRC15:55
*** pcaruana has joined #openstack-glance16:00
*** lbragstad has quit IRC16:02
abhishekk|mblI guess infra has some problem16:02
abhishekk|mblGlanceStore 0.25.0 is available on pypi but its not published on mirror used by infra16:03
abhishekk|mblThey are out of space as of now16:03
rosmaitabummer16:07
jokke_recheck the tips job16:08
jokke_that will pull the current state of the repo into the test run16:08
rosmaitaarrgh ... the multihash patch had passed everything but tempest, guess it will have to start over16:08
rosmaitathough maybe they are only restarting the zuul interface, not zuul itself16:09
openstackgerritBrian Rosmaita proposed openstack/glance_store master: Deprecate store_add_to_backend()  https://review.openstack.org/58365716:11
*** udesale_ has quit IRC16:13
*** udesale has joined #openstack-glance16:13
*** udesale has quit IRC16:18
*** Luzi has quit IRC16:21
*** btully has joined #openstack-glance16:29
*** btully has quit IRC16:33
*** AlexeyAbashkin has quit IRC16:36
*** AlexeyAbashkin has joined #openstack-glance16:39
*** AlexeyAbashkin has quit IRC16:43
*** abhishekk has joined #openstack-glance16:49
*** lbragstad has joined #openstack-glance16:50
*** abhishekk|mbl has quit IRC16:55
*** prometheanfire has left #openstack-glance16:57
abhishekkjokke_, rosmaita when are we targeting to release 0.26.0?17:03
rosmaitaabhishekk : tomorrow17:23
*** jose-phillips has joined #openstack-glance17:24
*** dtantsur is now known as dtantsur|afk17:26
rosmaitaabhishekk actually, today, but i think it is going to take all day to get multihash merged17:27
abhishekkrosmaita, yes, gate is quiet slow17:33
*** tesseract has quit IRC17:35
*** r-daneel has quit IRC17:45
*** r-daneel has joined #openstack-glance17:47
*** abhishekk has quit IRC17:51
*** r-daneel_ has joined #openstack-glance17:53
*** r-daneel has quit IRC17:53
*** r-daneel_ is now known as r-daneel17:53
*** abhishekk|mbl has joined #openstack-glance18:00
*** pcaruana has quit IRC18:08
rosmaitajokke_ abhishekk|mbl  smcginnis : https://review.openstack.org/#/c/528790/11 just got +1 from zuul if someone would like to do the honors18:09
rosmaitaalso https://review.openstack.org/#/c/583657/18:11
*** itlinux has joined #openstack-glance18:13
*** btully has joined #openstack-glance18:18
*** btully has quit IRC18:22
abhishekk|mblrosmaita: i just have one question18:30
*** itlinux has quit IRC18:31
rosmaitaabhishekk|mbl sure18:31
abhishekk|mblWe will be deprecating capabilities related config parameter right?18:32
abhishekk|mblSo later we also need to remove capabilities_check decorator?18:32
rosmaitanot exactly18:33
rosmaitathe deprecated config param has to do with dynamic capabiities updates18:33
rosmaitaalthough18:33
rosmaitai get your point, why bother checking each time for static capabilities?18:33
rosmaitai have not thought that far ahead, tbh18:34
abhishekk|mblYes18:34
abhishekk|mblOk18:34
*** itlinux has joined #openstack-glance18:34
abhishekk|mblI guess I will +w to multihash patch now, just curious to hear from jokke_18:35
abhishekk|mblIf he has any questions, thought on this18:35
rosmaitayeah, the capabilities stuff looks like it could be useful, but we really don't use it at all, i don't think18:35
smcginnisrosmaita, abhishekk|mbl: I just approved. Should I hold that one?18:36
abhishekk|mblNo, let it be18:36
abhishekk|mblsmcginnis: have a look at release notes patch as well18:37
rosmaitahttps://review.openstack.org/#/c/583657/18:37
rosmaitaonce those merge i will put up releases patch with a -W and we can make sure the glance_store-tips job passes before doing a release18:38
abhishekk|mblrosmaita, could you please have look at modified hidden images specs18:38
smcginnisDone18:38
rosmaitaabhishekk|mbl will do18:38
abhishekk|mblsmcginnis, jokke_  ^^18:39
smcginnisI was thinking it was a little too low level of detail for a release note, but since this is a lib that makes sense.18:39
abhishekk|mblThank you18:39
rosmaitasmcginnis ty18:39
abhishekk|mblhttps://review.openstack.org/#/c/580568/18:40
abhishekk|mblspecs ^^18:40
rosmaitaabhishekk|mbl ty18:41
rosmaitaabhishekk|mbl : can a user specify "hidden": true for image-create?18:44
abhishekk|mblrosmaita: need to check, will confirm tomorrow is that ok?18:45
rosmaitaoh yeah, no problem18:46
rosmaitai forgot that it is late for you18:46
rosmaitai will check myself if i have some time18:46
abhishekk|mblI think he can, so that need to restrict as well18:46
rosmaitathanks for your reviews today, have a good evening18:46
abhishekk|mblIf you able to verify, please comment on the specs18:46
abhishekk|mblThank you, have a good day ahead18:47
*** harlowja has joined #openstack-glance18:54
*** priteau has quit IRC18:58
abhishekk|mblrosmaita: uese can specify -- hidden 'true' at the time of creation but glance does not consider it it sets hidden=False to newly created image as described in the specs19:04
abhishekk|mbls/uses/user19:04
rosmaitaok, thanks19:04
rosmaitai think the spec says false unless specified otherwise19:05
rosmaitabut we can change the spec19:05
rosmaita(or you can, rather)19:05
rosmaitafor the use case we have, i don't think it matters if the image must be created with hidden==false19:05
rosmaitaalso, if you have a minute, i'm wondering about the 409 for pre-rocky19:06
abhishekk|mblI guess we should not allow --hidden at the time of creation?19:06
rosmaitawon't pre-rocky glance treat 'hidden' as user custom metadata?19:06
rosmaitaabhishekk|mbl i think we could allow it ... maybe we should discuss at glance meeting tomorrow19:07
abhishekk|mblMeans user should able to create hidden image?19:07
rosmaitai don't know, it would be kind of dumb to create an image that disappears immediately19:08
rosmaitai think not allowing at creation would be OK19:08
abhishekk|mblYeah19:08
rosmaitai think pre-rocky glance will be ok with a user creating19:09
rosmaitai mean updating19:09
rosmaitaan image "hidden":"19:09
rosmaitashoot, sorry, bad fingers today19:09
abhishekk|mblAnyway good thing to discuss in tomorrow's meeting ( please remind me :))19:10
abhishekk|mblYeah that will be ok to update19:10
rosmaitai think it will allow "hidden": "true" in JSON (note that true is a string)19:10
rosmaitai think it will not allow "hidden": true where true is a JSON boolean19:10
abhishekk|mblYes19:10
rosmaitaok19:11
rosmaitai will put this on tomorrow's agenda, ttyl -- go get some sleep!19:11
abhishekk|mblYeah, thank you19:13
*** abhishekk|mbl has quit IRC19:20
*** dpawlik has quit IRC19:47
*** arvindn05 has quit IRC19:48
*** arvindn05 has joined #openstack-glance19:52
*** itlinux has quit IRC20:09
openstackgerritMerged openstack/glance_store master: Multihash Implementation for Glance  https://review.openstack.org/52879020:10
openstackgerritMerged openstack/glance_store master: Deprecate store_add_to_backend()  https://review.openstack.org/58365720:10
rosmaitawoot!20:10
*** r-daneel has quit IRC20:31
*** jmlowe has quit IRC20:36
*** itlinux has joined #openstack-glance20:37
*** r-daneel has joined #openstack-glance20:40
*** itlinux has quit IRC20:42
*** dpawlik has joined #openstack-glance20:56
*** dpawlik has quit IRC21:00
*** edmondsw has quit IRC21:54
*** btully has joined #openstack-glance21:56
*** btully has quit IRC22:00
*** linkmark has joined #openstack-glance22:03
*** MrWatson has quit IRC22:12
*** NostawRm has joined #openstack-glance22:12
*** jmlowe has joined #openstack-glance22:24
*** jmlowe has quit IRC22:25
*** jmlowe has joined #openstack-glance22:25
*** lbragstad has quit IRC22:29
*** rcernin has joined #openstack-glance22:32
*** imacdonn has quit IRC22:44
*** imacdonn has joined #openstack-glance22:45
*** dpawlik has joined #openstack-glance22:57
*** jmlowe has quit IRC22:58
*** jmlowe has joined #openstack-glance22:58
*** dpawlik has quit IRC23:01
*** itlinux has joined #openstack-glance23:04
*** jmlowe has quit IRC23:05
*** itlinux has quit IRC23:16
*** itlinux has joined #openstack-glance23:28
*** jmlowe has joined #openstack-glance23:37
*** btully has joined #openstack-glance23:44
*** btully has quit IRC23:48
*** threestrands has joined #openstack-glance23:50
*** threestrands has quit IRC23:50
*** threestrands has joined #openstack-glance23:50
*** rosmaita has quit IRC23:58

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!