Friday, 2020-03-06

*** mattw4 has quit IRC00:00
*** Defolos has quit IRC00:23
*** erbarr has quit IRC00:49
*** zxiiro has quit IRC01:18
*** jamesmcarthur has joined #zuul01:18
*** igordc has quit IRC01:18
*** tosky has quit IRC01:39
*** jamesmcarthur has quit IRC01:41
*** jamesmcarthur has joined #zuul01:42
*** jamesmcarthur has quit IRC01:47
*** jamesmcarthur has joined #zuul02:12
*** Goneri has quit IRC02:18
*** swest has quit IRC02:20
*** jamesmcarthur has quit IRC02:21
*** jamesmcarthur has joined #zuul02:21
*** jamesmcarthur has quit IRC02:24
*** igordc has joined #zuul02:25
*** jamesmcarthur has joined #zuul02:26
*** jamesmcarthur has quit IRC02:32
*** swest has joined #zuul02:35
*** jamesmcarthur has joined #zuul02:54
*** bhavikdbavishi has joined #zuul03:18
*** zxiiro has joined #zuul03:20
*** jamesmcarthur has quit IRC03:50
*** jamesmcarthur has joined #zuul03:51
*** rlandy has quit IRC03:56
*** jamesmcarthur has quit IRC03:56
*** michael-beaver has quit IRC04:11
*** jamesmcarthur has joined #zuul04:21
*** hashar has joined #zuul04:23
*** jamesmcarthur has quit IRC04:26
*** decimuscorvinus_ has quit IRC04:59
*** decimuscorvinus has joined #zuul04:59
*** saneax has joined #zuul05:08
*** jamesmcarthur has joined #zuul05:22
*** raukadah is now known as chandankumar05:24
*** jamesmcarthur has quit IRC05:26
*** evrardjp has quit IRC05:35
*** evrardjp has joined #zuul05:35
*** jamesmcarthur has joined #zuul05:53
*** jamesmcarthur has quit IRC05:58
*** hashar has quit IRC06:40
*** igordc has quit IRC06:41
*** jamesmcarthur has joined #zuul06:54
*** jamesmcarthur has quit IRC06:59
openstackgerritFelix Edel proposed zuul/zuul master: Allow check runs to be configured as required status in pipeline config  https://review.opendev.org/71124107:17
*** armstrongs has joined #zuul07:23
*** armstrongs has quit IRC07:27
openstackgerritFelix Edel proposed zuul/zuul master: Make github file annotation levels configurable via zuul return  https://review.opendev.org/71117907:28
*** jamesmcarthur has joined #zuul07:28
*** jamesmcarthur has quit IRC07:33
*** jcapitao_off has joined #zuul07:46
*** Defolos has joined #zuul07:52
*** jamesmcarthur has joined #zuul07:53
*** hashar has joined #zuul07:55
*** jcapitao_off is now known as jcapitao07:55
*** sshnaidm_ has joined #zuul07:57
*** sshnaidm|afk has quit IRC07:58
*** jamesmcarthur has quit IRC07:58
*** jamesmcarthur has joined #zuul08:29
*** jamesmcarthur has quit IRC08:35
*** tosky has joined #zuul08:40
*** avass has joined #zuul08:42
*** jpena|off is now known as jpena08:47
openstackgerritFelix Edel proposed zuul/zuul master: Report canceled changes via Github checks API  https://review.opendev.org/71102308:49
openstackgerritFelix Edel proposed zuul/zuul master: Provide some documentation for the checks API implementation  https://review.opendev.org/71149309:05
openstackgerritAntoine Musso proposed zuul/zuul master: Enhance some logging messages  https://review.opendev.org/70433109:07
*** tflink has quit IRC09:07
*** tflink has joined #zuul09:07
*** avass has quit IRC09:08
*** danpawlik has joined #zuul09:09
openstackgerritFelix Edel proposed zuul/zuul master: Provide some documentation for the checks API implementation  https://review.opendev.org/71149309:12
*** zxiiro has quit IRC09:25
*** yolanda has quit IRC09:29
*** jamesmcarthur has joined #zuul09:31
*** yolanda has joined #zuul09:33
*** jamesmcarthur has quit IRC09:35
*** jamesmcarthur has joined #zuul11:11
*** sshnaidm_ is now known as sshnaidm|afk11:13
*** jamesmcarthur has quit IRC11:16
*** harrymichal has joined #zuul11:30
*** jcapitao is now known as jcapitao_lunch11:35
*** hashar has quit IRC11:38
*** jamesmcarthur has joined #zuul11:43
*** jamesmcarthur has quit IRC11:48
*** irclogbot_2 has quit IRC12:05
*** amotoki has quit IRC12:05
*** jlk has quit IRC12:06
*** amotoki has joined #zuul12:09
*** irclogbot_3 has joined #zuul12:09
*** jpena is now known as jpena|lunch12:37
*** jamesmcarthur has joined #zuul12:45
*** hashar has joined #zuul12:46
*** jamesmcarthur has quit IRC12:49
*** rlandy has joined #zuul12:55
*** jcapitao_lunch is now known as jcapitao13:03
*** jamesmcarthur has joined #zuul13:16
*** sgw has quit IRC13:20
*** jamesmcarthur has quit IRC13:23
*** jamesmcarthur_ has joined #zuul13:23
tristanCcorvus: Shrews: while refactoring the zk auth code a bit to get a better interface, i found some code inconsistencies, which makes me wonder how this work at all... let me explain, kazoo client connect auth_data argument needs to be a list of (scheme, cred) tuple, like so: https://review.opendev.org/#/c/619155/32/nodepool/zk.py@968 or https://review.opendev.org/#/c/619155/32/nodepool/tests/__init__.py@7813:26
tristanCwhich match with the current kazoo client code here: https://github.com/python-zk/kazoo/blob/2.6.1/kazoo/client.py#L30813:27
tristanCarg, nevermind, i thought we were not using a list of tuple in some places, but that's ok because nodepool.zk takes care of the transformation. sorry for the noise13:30
*** jpena|lunch is now known as jpena13:33
*** avass has joined #zuul13:40
*** jamesmcarthur_ has quit IRC13:49
*** zenkuro has quit IRC13:56
mordredtristanC: my favorite way of debugging - explaining the bug to someone else until you realize it isn't a bug14:18
fungicommonly known as "rubber duck debugging"14:26
*** sreejithp has joined #zuul15:11
*** avass has quit IRC15:33
*** Goneri has joined #zuul15:33
*** Goneri has quit IRC15:38
*** jcapitao is now known as jcapitao_afk15:50
*** hashar has quit IRC15:55
*** mattw4 has joined #zuul16:12
tobiashcorvus, mordred: shall we merge 578557? (match tag items against containing branches) it has now +2 from both of you and myself (and we're running this in prod since 1.5 years)16:13
clarkbtobiash: I just volunteered to look at it if you want to wait just a few minutes longer16:15
clarkbslow start this morning but happy to review it soon16:15
tobiashk16:15
*** jcapitao_afk is now known as jcapitao16:16
*** jamesmcarthur has joined #zuul16:19
corvustobiash, clarkb, mordred: i actually just started looking into that again, give me a bit more time as well16:23
tobiashk16:25
*** chandankumar is now known as raukadah16:38
clarkbrgr16:39
*** danpawlik has quit IRC16:41
openstackgerritMerged zuul/zuul master: Support pausing merge jobs  https://review.opendev.org/70719216:48
*** harrymichal has quit IRC16:54
*** harrymichal_ has joined #zuul16:54
openstackgerritMerged zuul/zuul master: Test that retries don't trigger fail-fast  https://review.opendev.org/70498316:54
*** harrymichal_ is now known as harrymichal16:54
clarkbcorvus: I +2'd but did leave several comments that may be worthy of followups17:00
corvusclarkb: ack, thx17:01
*** jamesmcarthur has quit IRC17:05
*** hashar has joined #zuul17:06
*** Defolos has quit IRC17:07
openstackgerritMerged zuul/zuul master: Don't reset file list to None in updateChange  https://review.opendev.org/70432817:20
openstackgerritMerged zuul/zuul master: Protect prime_installation_map with mutex  https://review.opendev.org/70493217:20
*** dtroyer has quit IRC17:33
*** evrardjp has quit IRC17:35
*** evrardjp has joined #zuul17:35
*** harrymichal_ has joined #zuul17:36
*** jlk has joined #zuul17:36
*** harrymichal has quit IRC17:37
*** harrymichal_ is now known as harrymichal17:37
*** harrymichal has quit IRC17:47
openstackgerritMerged zuul/zuul master: Centralize merge handling  https://review.opendev.org/70769217:56
*** jpena is now known as jpena|off17:57
*** bhavikdbavishi has quit IRC17:59
openstackgerritMerged zuul/zuul master: Uncap GitPython  https://review.opendev.org/70859318:03
openstackgerritMerged zuul/zuul master: Log duration of tenant reconfigurations  https://review.opendev.org/70840318:04
openstackgerritMerged zuul/zuul master: Reduce gearman logging in tests  https://review.opendev.org/70758718:04
openstackgerritMerged zuul/zuul master: Increase timeout in timeout test  https://review.opendev.org/71014618:04
openstackgerritMerged zuul/zuul master: web ui: fix buildset display when no builds  https://review.opendev.org/71050818:04
openstackgerritMerged zuul/zuul master: Optimize canMerge using graphql  https://review.opendev.org/70983618:08
*** jcapitao has quit IRC18:10
openstackgerritMerged zuul/zuul master: Increase zookeeper timeout during tests  https://review.opendev.org/70794518:12
openstackgerritMerged zuul/zuul master: Authorization rules: add templating  https://review.opendev.org/70519318:12
*** rlandy is now known as rlandy|brb18:13
*** dtroyer has joined #zuul18:14
*** saneax has quit IRC18:15
*** igordc has joined #zuul18:20
*** igordc has quit IRC18:20
*** gmann is now known as gmann_lunch18:27
*** gmann_lunch is now known as gmann18:32
*** hashar has quit IRC18:33
*** rlandy|brb is now known as rlandy18:33
openstackgerritMerged zuul/zuul master: Add reference pipelines file for Github driver  https://review.opendev.org/67271218:40
*** jamesmcarthur has joined #zuul18:41
openstackgerritJames E. Blair proposed zuul/zuul master: Add a test for retroactive branch tag application  https://review.opendev.org/71174118:57
corvusmordred, clarkb, tobiash: i replied (extensively) to the tags-on-branches change and wrote a followup patch with a test case showing the behavior that Niklas was asking about18:59
corvusi believe the original patch is now in merge conflict; i'm going to take a break and work on rebasing it, probably will have something after lunch.19:00
*** Defolos has joined #zuul19:00
tobiashcorvus: wow, great answer19:04
openstackgerritMerged zuul/zuul master: Improve error reporting when pr merge fails  https://review.opendev.org/70763719:12
*** erbarr has joined #zuul19:15
*** jamesmcarthur has quit IRC19:24
mordredcorvus: have I mentioned I really wish there was a magical way to have speculative tags that can be gated?19:27
openstackgerritTristan Cacqueray proposed zuul/nodepool master: Implement zookeeper-auth  https://review.opendev.org/61915519:28
openstackgerritTristan Cacqueray proposed zuul/nodepool master: Add mypy to linter test  https://review.opendev.org/71175019:28
ShrewstristanC: that release note still references zookeeper-update-zk-auth  :(19:33
tristanCcorvus: clarkb: in ^ PS33 i suggest we only support a login/password digest sasl scheme, let me know if that's ok for you19:34
tristanCShrews: arg oops, it's half fixed then :) I'll update after ci results19:35
tristanCin the meantime, perhaps we should consider a common tool for both zuul and nodepool?19:35
ShrewstristanC: i *think* it's true that zuul is aware of all znodes nodepool creates, but the reverse is not true. So maybe it could just be moved to zuul/zuul?19:36
tristanCwhat do you think about a creating a kazoo wrapper library to share the zk module between zuul and nodepool. then we could have a generic update-zk-auth script provided by that new library?19:36
tristanCfor example, a new zuul/zklib project19:37
Shrewsi proposed that at the beginning of the v3 work19:37
tristanCsuch project does rings a bell :)19:38
tristanCor, as you suggested, let's drop the script from nodepool and let's implement it as part of the zuul change?19:38
Shrewsi think just moving it to zuul is best at this point19:38
Shrewsrather than a major code extraction  :)19:39
openstackgerritMerged zuul/zuul master: Add load-branch to tenant configuration  https://review.opendev.org/70566419:41
corvustristanC: i think supporting sasl only is okay, but we should support both kerberos or digest with sasl.19:41
tristanCShrews: well, i think that's a worthy operation, and I think i can work on the implementation19:42
corvustristanC: (that should be pretty easy, the acl is the same in both cases, the only difference is the credential passed to auth info)19:42
tristanCcorvus: i linked an upcoming change in kazoo about kerberos19:44
tristanCwhich should be easy to support in the new zk_auth module i proposed. but we would need an extra toggle in the user configuration19:44
tristanCe.g. (scheme: sasl-digest, username: text, password: text)  and (scheme: sasl-gssapi, service: text, principal: text)19:45
ShrewstristanC: oh, i suppose there are znodes created by nodepool that zuul doesn't know about (e.g., image builds). That being said, I think just having that script in a single place is fine (either nodepool or zuul), but it needs to consider znodes from both.19:45
tristanCShrews: understood. I suggested a new zuul/zk lib project to also reduce code duplication between zuul and nodepool.19:47
ShrewstristanC: that can (and I think should) be considered independently of the zk-auth change, IMO19:48
corvusthere's very little overlap between the two, and there will be even less over time as more of the zk action is in zuul.  i'm not opposed but i'm not sure it's worth the effort.19:49
corvus(and, tbh, i'd suggest if we're going to do that, do it after zuul v5)19:49
Shrewsthe zk.py files are different enough in some of the things they do, it's going to be more work than you think19:50
tristanCcorvus: well, it's not very little: https://opendev.org/zuul/zuul/src/branch/master/zuul/zk.py#L466-L73619:50
tristanCShrews: the difference is actually what worries me19:51
corvustristanC: that comment is a lie19:51
corvusnodepool doesn't have hold requests19:51
corvushttps://opendev.org/zuul/zuul/src/branch/master/zuul/zk.py#L466-L540  is closer to the functional overlap19:52
tristanCcorvus: that's fair, and i'm the one to blame as I added the comment :)19:52
corvustristanC: well, it was true when you wrote it, but i guess we forgot to add an "End copy of..." :)19:53
tristanCanyway, i'm also happy to not do the code extraction.19:54
corvustristanC: back to the other thing, yeah, i think maybe a small abstraction layer for zk auth so we can be forward compatible with the gssapi change if it lands would be good.  so maybe we can implement something that accepts a "principal" now and encodes it as kazoo currently needs, and can encode it differently later19:54
corvustristanC: yeah, i think if we go ahead and accept "user/password/service/principal/mechanism" as inputs as appropriate, we can support current and future kazoo19:56
ShrewstristanC: looking at update_zk_auth.py, it supports a configurable chroot, so maybe all we need is a note to run it with "chroot=/zuul" ?19:57
tristanCcorvus: i'm not entirely sure how current kazoo works with kerberos, it seems like it tries to import the kerberos module but i don't see principal setting beside an hardcoded self.sasl_server_principal = "zk-sasl-md5"19:59
tristanCShrews: yes, i'll do that when moving the script to zuul19:59
corvustristanC: i think that's the server principal, or the hostname of the server, and i suspect it's just ignored in most cases.  that should be different than the client principal, which is == username for our purposes (and is the thing that should be in the acl).20:02
openstackgerritTristan Cacqueray proposed zuul/nodepool master: Implement zookeeper-auth  https://review.opendev.org/61915520:02
tristanCcorvus: isn't the username the `service` in gssapi?20:05
tristanCi'm not sure we can make such assumption, and without an integration test it's hard to tell if that would work.20:07
corvustristanC: this is relevant https://github.com/thobbs/pure-sasl/blob/master/puresasl/client.py#L7320:07
corvustristanC: so it seems that for zk, host is not relevant for sasl-digest (ie, there's no third-party gssapi server, it's zk itself that handles the sasl requests)20:09
corvustristanC: based on that patch, it does look like only digest-md5 is supported by the current kazoo code?20:14
corvustristanC: specifically https://github.com/python-zk/kazoo/commit/cd49b3fa01136848c5e6bfafb4c241b9704f249d#diff-7c89c6cfc6a591cef0ed3be6bbfb375eL703 ?20:14
corvustristanC: if that's the case, then i agree, there's not much point to our supporting krb now, but we can plan our config syntax to be forward-compatible with the kazoo gssapi change, so if/when it lands we can support both20:16
tristanCcorvus: i find zk auth quite confusing. it seems like the sasl scheme used in ci does work (and it fails when auth is missing as we noticed yesterday).20:23
corvuswell, sasl is confusing20:23
tristanCi think we should assume zuul operator are also zookeeper operator, and thus we should provide an easy to follow documentation to setup the zk auth.20:24
corvusno objection there, i think that would be the typical case.  we just shouldn't assume that, so that enterprises with externally managed zk can still be used20:25
tristanCin that situation, how about we support this initial schemas (username, password), and when gssapi lands (or when enterprise user reports auth compat issue), then we support an optional scheme that enables setting extra parameter (for example if scheme == gssapi, then uses (service, principal) instead of (username, password) )20:29
tristanCand when scheme is not defined, we default to sasl-digest which only requires an (username, password) product20:30
corvustristanC: sure, but let's just make the syntax forward compatible.  like, use a dictionary so we can have 'username' or 'principal' as keys20:33
tristanCcorvus: the way zk_auth is designed is that the ZkAuth is opaque for the rest of nodepool code, only the read method can convert a nodepool.yaml dict into a ZkAuth, then only the `acl` and `kazoo_args` accepts the ZkAuth type. thus it should be relatively safe to extend it for new scheme20:37
*** erbarr has quit IRC21:11
*** Shrews has quit IRC21:11
*** portdirect has quit IRC21:13
*** ChanServ has quit IRC21:13
*** adam_g has quit IRC21:13
*** clayg has quit IRC21:13
*** johnsom has quit IRC21:13
*** ironfoot has quit IRC21:13
*** aspiers has quit IRC21:13
*** jkt has quit IRC21:13
*** corvus has quit IRC21:13
*** mugsie has quit IRC21:13
*** klindgren has quit IRC21:13
*** donnyd has quit IRC21:13
*** kklimonda has quit IRC21:13
*** maxamillion has quit IRC21:13
*** logan- has quit IRC21:13
*** stevthedev has quit IRC21:13
*** evgenyl has quit IRC21:13
*** dustinc has quit IRC21:13
*** samccann has quit IRC21:13
*** pots has quit IRC21:13
*** mgoddard has quit IRC21:13
*** SpamapS has quit IRC21:13
*** arxcruz|rover has quit IRC21:13
*** jpena|off has quit IRC21:13
*** adamw has quit IRC21:13
*** zbr|pto has quit IRC21:13
*** gmann has quit IRC21:13
*** bstinson has quit IRC21:13
*** tristanC has quit IRC21:13
*** smcginnis has quit IRC21:13
*** iamweswilson has quit IRC21:13
*** persia has quit IRC21:13
*** lseki has quit IRC21:13
*** yolanda has quit IRC21:13
*** nhicher has quit IRC21:13
*** fdegir has quit IRC21:13
*** jhesketh has quit IRC21:13
*** yoctozepto has quit IRC21:13
*** fbo has quit IRC21:13
*** dmsimard|off has quit IRC21:13
*** ianw has quit IRC21:13
*** tobberydberg has quit IRC21:13
*** EmilienM has quit IRC21:13
*** dtroyer has quit IRC21:13
*** irclogbot_3 has quit IRC21:13
*** reiterative has quit IRC21:13
*** dmellado has quit IRC21:13
*** evrardjp has quit IRC21:13
*** swest has quit IRC21:13
*** webknjaz has quit IRC21:13
*** dcastellani has quit IRC21:13
*** jbryce has quit IRC21:13
*** mnasiadka has quit IRC21:13
*** fungi has quit IRC21:13
*** cloudnull has quit IRC21:13
*** Defolos has quit IRC21:13
*** flaper87 has quit IRC21:13
*** toabctl has quit IRC21:13
*** mhu has quit IRC21:13
*** tobiash has quit IRC21:13
*** johanssone has quit IRC21:13
*** timburke has quit IRC21:13
*** clarkb has quit IRC21:13
*** ianychoi has quit IRC21:13
*** aluria has quit IRC21:13
*** guilhermesp has quit IRC21:13
*** wxy-xiyuan has quit IRC21:13
*** tdasilva has quit IRC21:13
*** gouthamr has quit IRC21:13
*** frickler has quit IRC21:13
*** Diabelko has quit IRC21:13
*** mordred has quit IRC21:13
*** ChrisShort has quit IRC21:13
*** andreaf has quit IRC21:13
*** mattw4 has quit IRC21:13
*** sreejithp has quit IRC21:13
*** jlk has quit IRC21:13
*** tosky has quit IRC21:13
*** sshnaidm|afk has quit IRC21:13
*** decimuscorvinus has quit IRC21:13
*** pabelanger has quit IRC21:13
*** etp has quit IRC21:13
*** raukadah has quit IRC21:13
*** tflink has quit IRC21:13
*** sugaar has quit IRC21:13
*** smyers has quit IRC21:13
*** amotoki has quit IRC21:13
*** openstackgerrit has quit IRC21:13
*** Miouge has quit IRC21:13
*** marvs has quit IRC21:13
*** lennyb has quit IRC21:13
*** shanemcd has quit IRC21:13
*** mmedvede has quit IRC21:13
*** gothicmindfood has quit IRC21:13
*** masterpe has quit IRC21:13
*** jtanner has quit IRC21:13
*** mnaser has quit IRC21:13
*** tributarian has quit IRC21:13
*** gundalow has quit IRC21:13
*** ttx has quit IRC21:13
*** kgz has quit IRC21:13
*** erbarr has joined #zuul21:13
*** Defolos has joined #zuul21:13
*** dtroyer has joined #zuul21:13
*** jlk has joined #zuul21:13
*** evrardjp has joined #zuul21:13
*** mattw4 has joined #zuul21:13
*** sreejithp has joined #zuul21:13
*** irclogbot_3 has joined #zuul21:13
*** amotoki has joined #zuul21:13
*** yolanda has joined #zuul21:13
*** tflink has joined #zuul21:13
*** tosky has joined #zuul21:13
*** sshnaidm|afk has joined #zuul21:13
*** decimuscorvinus has joined #zuul21:13
*** swest has joined #zuul21:13
*** flaper87 has joined #zuul21:13
*** pabelanger has joined #zuul21:13
*** portdirect has joined #zuul21:13
*** Shrews has joined #zuul21:13
*** sugaar has joined #zuul21:13
*** marvs has joined #zuul21:13
*** reiterative has joined #zuul21:13
*** nhicher has joined #zuul21:13
*** toabctl has joined #zuul21:13
*** mgoddard has joined #zuul21:13
*** dmellado has joined #zuul21:13
*** openstackgerrit has joined #zuul21:13
*** Miouge has joined #zuul21:13
*** mhu has joined #zuul21:13
*** masterpe has joined #zuul21:13
*** tobiash has joined #zuul21:13
*** ChanServ has joined #zuul21:13
*** smyers has joined #zuul21:13
*** lennyb has joined #zuul21:13
*** etp has joined #zuul21:13
*** raukadah has joined #zuul21:13
*** andreaf has joined #zuul21:13
*** tdasilva has joined #zuul21:13
*** gouthamr has joined #zuul21:13
*** frickler has joined #zuul21:13
*** Diabelko has joined #zuul21:13
*** aluria has joined #zuul21:13
*** ttx has joined #zuul21:13
*** bstinson has joined #zuul21:13
*** orwell.freenode.net sets mode: +o ChanServ21:13
*** tristanC has joined #zuul21:13
*** smcginnis has joined #zuul21:13
*** iamweswilson has joined #zuul21:13
*** persia has joined #zuul21:13
*** dmsimard|off has joined #zuul21:13
*** ianw has joined #zuul21:13
*** tobberydberg has joined #zuul21:13
*** EmilienM has joined #zuul21:13
*** arxcruz|rover has joined #zuul21:13
*** jpena|off has joined #zuul21:13
*** clayg has joined #zuul21:13
*** johnsom has joined #zuul21:13
*** ironfoot has joined #zuul21:13
*** aspiers has joined #zuul21:13
*** jkt has joined #zuul21:13
*** corvus has joined #zuul21:13
*** logan- has joined #zuul21:13
*** stevthedev has joined #zuul21:13
*** evgenyl has joined #zuul21:13
*** dustinc has joined #zuul21:13
*** samccann has joined #zuul21:13
*** pots has joined #zuul21:13
*** zbr|pto has joined #zuul21:13
*** gmann has joined #zuul21:13
*** maxamillion has joined #zuul21:13
*** gundalow has joined #zuul21:13
*** kklimonda has joined #zuul21:13
*** tributarian has joined #zuul21:13
*** donnyd has joined #zuul21:13
*** mnaser has joined #zuul21:13
*** cloudnull has joined #zuul21:13
*** fungi has joined #zuul21:13
*** yoctozepto has joined #zuul21:13
*** fbo has joined #zuul21:13
*** wxy-xiyuan has joined #zuul21:13
*** adamw has joined #zuul21:13
*** mnasiadka has joined #zuul21:13
*** guilhermesp has joined #zuul21:13
*** jhesketh has joined #zuul21:13
*** SpamapS has joined #zuul21:13
*** fdegir has joined #zuul21:13
*** gothicmindfood has joined #zuul21:13
*** lseki has joined #zuul21:13
*** klindgren has joined #zuul21:13
*** jbryce has joined #zuul21:13
*** jtanner has joined #zuul21:13
*** ChrisShort has joined #zuul21:13
*** mmedvede has joined #zuul21:13
*** dcastellani has joined #zuul21:13
*** mordred has joined #zuul21:13
*** webknjaz has joined #zuul21:13
*** ianychoi has joined #zuul21:13
*** clarkb has joined #zuul21:13
*** timburke has joined #zuul21:13
*** shanemcd has joined #zuul21:13
*** johanssone has joined #zuul21:13
*** mugsie has joined #zuul21:13
*** adam_g has joined #zuul21:13
*** kgz has joined #zuul21:16
openstackgerritJames E. Blair proposed zuul/zuul master: Match tag items against containing branches  https://review.opendev.org/57855721:29
openstackgerritJames E. Blair proposed zuul/zuul master: Add a test for retroactive branch tag application  https://review.opendev.org/71174121:29
corvustristanC: i left some comments on the zk_auth portion of the nodepool change21:36
*** hashar has joined #zuul21:46
tristanCcorvus: thanks, i replied21:47
tristanCi don't mind using a class, though using a newtype seems like a good example of opaque data type21:48
*** mattw4 has quit IRC21:52
corvusthis just looks like a class with methods (the file actually has the same structure as a C file implementing object oriented programming (eg, glib)) so since the rest of nodepool is oop, let's stick with that paradigm.  :)21:53
fungii assume the jit compiler optimizes both to the same underlying structure regardless21:55
*** mattw4 has joined #zuul21:56
tristanCcorvus: when adding support for gssapi/service/principal it will looks odd without subclass, which i find it more cumbersome to use than pure fp with an opaque data type21:58
clarkbI think the issue is that without some carried state you are forced to parse and unparse the string?21:59
corvustristanC: then use a subclass21:59
clarkboh I see that consistency is the other concern21:59
corvusit's an object oriented program22:00
corvuslet's not change that now22:00
corvustristanC: all of the methods you wrote in zk_auth.py should just be methods of the class22:01
tristanCcorvus: alright, i'll rewrite in oop22:02
corvuslike "ZKAuth.getACL()" should return the acl22:02
*** mattw4 has quit IRC22:02
*** mattw4 has joined #zuul22:02
tristanCcorvus: i didn't meant to change nodepool model, i just find it more efficient to use fp in such situation.22:06
tristanCto properly use subclass i now need to define abstract static method22:08
corvustristanC: right, but please include consistency and the comfort of others when you make decisions like that.  it's a big cost to other maintainers if every file in the program looks different.  plus, in this particular case, you did not use a functional model.  you literally just made a class without using the word "class" or self.22:08
corvustristanC: you know what, i'll just do it22:09
tristanCcorvus: the zk_auth module act like a class, but without mutable self. to make it extra safe i used the typing.NewType function, hoping that such style would not be a big deal.22:14
tristanCthough if you think that is too awkward i'm happy to rewrite it in oop22:17
*** mattw4 has quit IRC22:17
*** sreejithp has quit IRC22:26
openstackgerritJames E. Blair proposed zuul/nodepool master: Implement zookeeper-auth  https://review.opendev.org/61915522:27
*** hashar has quit IRC22:31
*** rlandy has quit IRC22:33
*** Defolos has quit IRC23:01
*** mattw4 has joined #zuul23:05
*** igordc has joined #zuul23:17
*** erbarr has quit IRC23:17
openstackgerritJames E. Blair proposed zuul/nodepool master: Implement zookeeper-auth  https://review.opendev.org/61915523:32
*** mattw4 has quit IRC23:38
*** mattw4 has joined #zuul23:41
*** tosky has quit IRC23:51

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