Thursday, 2019-01-17

*** ttsiouts has quit IRC00:02
*** tosky has quit IRC00:29
*** dave-mccowan has joined #openstack-sdks00:51
*** Qiming has quit IRC00:59
*** Qiming has joined #openstack-sdks01:00
openstackgerritDuc Truong proposed openstack/openstacksdk master: Change Resource location type to Header
*** slaweq has joined #openstack-sdks01:11
*** dave-mccowan has quit IRC01:15
*** slaweq has quit IRC01:16
*** mriedem has quit IRC01:39
*** bobh has joined #openstack-sdks01:57
*** adriant has quit IRC02:17
*** edleafe has quit IRC02:19
*** adriant has joined #openstack-sdks02:20
*** bobh has quit IRC02:37
*** manfred has joined #openstack-sdks02:40
*** lifeless has quit IRC02:40
*** lifeless has joined #openstack-sdks02:41
*** gtmanfred has quit IRC02:42
*** manfred is now known as gtmanfred02:42
*** gtmanfred is now known as excitebike02:42
*** excitebike is now known as dontcare02:42
*** dontcare is now known as gtmanfred02:42
*** toabctl has quit IRC02:46
*** nicholas has quit IRC02:46
*** odyssey4me has quit IRC02:46
*** larsks has quit IRC02:46
*** larsks has joined #openstack-sdks02:58
*** slaweq has joined #openstack-sdks03:11
*** slaweq has quit IRC03:15
*** lifeless has quit IRC04:30
*** slaweq has joined #openstack-sdks05:11
*** slaweq has quit IRC05:16
*** lifeless has joined #openstack-sdks05:51
*** slaweq has joined #openstack-sdks06:11
*** radeks has joined #openstack-sdks06:12
*** radeks_ has joined #openstack-sdks06:14
*** slaweq has quit IRC06:16
*** radeks has quit IRC06:17
*** rcarrillocruz has joined #openstack-sdks06:27
*** Luzi has joined #openstack-sdks06:43
*** slaweq has joined #openstack-sdks06:56
*** radeks_ has quit IRC07:16
*** radeks_ has joined #openstack-sdks07:32
*** e0ne has joined #openstack-sdks07:52
*** gtema has joined #openstack-sdks08:14
*** radeks__ has joined #openstack-sdks08:36
*** jpich has joined #openstack-sdks08:38
*** radeks_ has quit IRC08:39
*** jpena|off is now known as jpena08:39
*** tosky has joined #openstack-sdks08:40
*** dtantsur|afk is now known as dtantsur09:25
*** odyssey4me has joined #openstack-sdks09:44
*** odyssey4me has quit IRC10:49
*** gtema has quit IRC10:51
*** cdent has joined #openstack-sdks10:54
*** e0ne has quit IRC11:44
*** e0ne has joined #openstack-sdks11:52
*** jpena is now known as jpena|lunch12:45
*** irclogbot_2 has quit IRC12:48
*** bobh has joined #openstack-sdks12:55
*** irclogbot_2 has joined #openstack-sdks12:58
*** gtema has joined #openstack-sdks13:11
*** irclogbot_2 has quit IRC13:12
*** bobh has quit IRC13:19
*** irclogbot_2 has joined #openstack-sdks13:27
*** edleafe- has joined #openstack-sdks13:30
*** edleafe- is now known as edleafe13:31
*** irclogbot_2 has quit IRC13:35
*** radeks__ has quit IRC13:36
*** jpena|lunch is now known as jpena13:41
*** irclogbot_2 has joined #openstack-sdks13:49
*** e0ne has quit IRC13:53
*** e0ne has joined #openstack-sdks13:59
*** radeks__ has joined #openstack-sdks14:08
*** gtema has quit IRC14:31
*** mriedem has joined #openstack-sdks14:58
*** radeks__ has quit IRC15:21
*** radeks__ has joined #openstack-sdks15:21
openstackgerritMonty Taylor proposed openstack/openstacksdk master: Add class to override clustering delete responses
*** ignatenkobrain has quit IRC15:32
*** Luzi has quit IRC15:33
*** ignatenkobrain has joined #openstack-sdks15:35
*** radeks__ has quit IRC15:55
edleafeAPI-SIG Office Hour starts in 2 minutes. Buckle your seat belts!15:58
dtantsurdo they serve drinks in this flight?15:59
edleafeYes, but they're cash-only15:59
toskythose low-cost companies16:00
dtantsurah this low-cost world16:00
dtantsurgreat minds think alike, yeah?16:00
edleafeYou'd think moving from a WG to a SIG would be a step up, right?16:00
elmikosomeone say drinks?16:01
*** e0ne has quit IRC16:25
elmikoheads up edleafe dtantsur, i'll be on CET next week so i will most likely show up for the early session (gonna be at a conference)16:56
edleafeelmiko: ack16:56
*** aspiers has joined #openstack-sdks17:02
*** evrardjp has joined #openstack-sdks17:02
*** witek has joined #openstack-sdks17:02
aspiershi... we have a question about what the path for the new version of the oslo.middleware healthcheck API should be17:03
aspiersIIUC the current one is /healthcheck17:03
aspiersno doubt I'm about to ask a stupid question as never worked with OpenStack API code before17:04
aspiersbut I guess /v2/healthcheck is a bad idea since that stomps on the whole microversion thing?17:04
edleafeaspiers: I'm not familiar with the usage. Do users specify a version for the healthcheck call?17:04
evrardjpedleafe: not right nwo17:04
aspiersI think the idea was to specify it in the path17:04
aspiersbut we're open to recommendations17:04
aspiersis the WIP prototype from mugsie17:05
edleafeGenerally a header is preferred over mangling the path with version info17:05
aspiersI'm just polishing it up now17:05
dtantsurelmiko: cool!17:05
aspiersedleafe: OK, but this would be a different header to the normal microversion?17:05
dtantsur"normal microversion"? :D17:05
elmikodtantsur: you gonna be at devconf?17:05
dtantsurelmiko: - yep17:05
evrardjpaspiers: it could technically be the same header, it's a different path17:06
mugsiein this case, there is a pre-existing URL, so it should be something like /status - move away from the /healthcheck that v1 uses17:06
dtantsurOpenStack-Api-Version: compute 2.42, healthcheck 1.1\17:06
dtantsuror something17:06
elmikodtantsur: \o/17:06
aspiersright, that one17:06
edleafeaspiers: microversions are per-call17:06
evrardjpmugsie: I like that idea17:06
aspiersedleafe: sure17:06
aspiersevrardjp: me too17:06
aspiersstatus can be about more than just health17:07
mugsieand microversions make it unusable for a lot of monitoring application uses - so we should avodi it if possible17:07
cdentI think if we want the middleware to be super easy, all it should be is a path17:07
dtantsurjust a word of caution: obvious things like /status may clash with services that do not use major versions17:07
evrardjpmugsie: agreed again17:07
mugsiedtantsur: true17:07
evrardjpcdent: ok17:07
aspiersah ok17:07
aspiersApache has which uses /status IIRC17:07
dtantsurso if we declare a standard endpoint, we may want it verbose. something like /openstack-status17:07
edleafemugsie: my only concern is when you run out of synonyms for 'healthcheck' :)17:07
aspiersalthough that is optional17:08
mugsie /statuscheck /superhealthcheck17:08
dtantsurjust to make sure e.g. Placement never wants to introduce its own /status :)17:08
evrardjpdtantsur: thanks for the advice here , I would totally have forgotten that17:08
dtantsurnow, if you do insist on major versions (we you shouldn't), it could be /openstack-status/v117:09
* dtantsur is not sure what he meant by "we"17:09
aspiersthere is some benefit to being somewhat consistent with the existing /healthcheck17:09
aspiers/healthcheck-v2 ?17:09
dtantsurwhat is this benefit?17:10
* evrardjp takes his bike to the shed17:10
aspiersand we want to leave room for v3 somehow :)17:10
dtantsurif you want to leave room for large changes, use microversions17:10
aspiersdtantsur: so that people aren't confused about two totally different endpoints coming from the same oslo.middleware code17:10
elmikomugsie: ++ for /superhealthcheck XD17:10
mugsiemicroversions will make it really hard for a lot of monitoring tools17:10
evrardjpI think it would be fine to later add different path for different /superhealthchecks17:10
evrardjpmugsie: I agree with you there17:11
dtantsurmugsie: microversions are hard for everything, essentially17:11
* dtantsur hides17:11
evrardjpwe need the simplest thing for monitoring17:11
mugsiepersonally, I am OK with /v<number>, but thats just me17:11
aspiersI don't want to have to write a doc which says "for v1, use /healthcheck, for v2 use /openstack-status, and for v3 use /superawesomebunchofchecks"17:11
dtantsurI don't see a problem with ^^ tbh17:11
dtantsurespecially if openstack-status is broader than healthcheck17:12
aspiersOK I can try to privately deal with my OCD about inconsistency then :)17:12
edleafeThe problem is the repeated change of the API.17:12
dtantsurI hear you, it can be painful17:12
elmikoi'm more in favor of a commonly known path that could live outside of any path-based versioning scheme, but that's just my opinion17:12
dtantsurhmm, we can do /healthcheck/v2 even while keeping /healthcheck17:12
elmikoi don't have a strong technical argument17:12
aspiersdtantsur: if that works, it sounds good17:13
evrardjpelmiko: /healthcheck-v2.0 then /healthcheck-v3.0 ?17:13
mugsieelmiko: this should be independant of the API version17:13
evrardjpshould that happen later?17:13
dtantsurjust make /healthcheck/(?!v2) redirect to /healthcheck/v1 and move the old stuff to /v117:13
*** masayukig[m] has joined #openstack-sdks17:13
elmikohonestly, we should get this stuff written into the proposed spec/guideline for health checks17:13
elmikomugsie: right, that's why i like it to live on a consistent endpoint17:13
edleafedtantsur: that's the least terrible so far17:13
dtantsurso e.g. /healthcheck/foo/bar redirects to /healthcheck/v1/foo/bar17:13
evrardjpelmiko: it sounds like we're gonna get 10 different answers17:13
dtantsurbut /healthcheck/v2/foo/bar is just v217:13
aspiersLOL, we're back where we were in the same bikeshed as 2 years ago XD17:13
elmikoevrardjp: imo, /healthcheck/v{whatever} is preferable17:14
elmikoevrardjp: yup, pretty much17:14
aspiersI like /healthcheck/v$NUMBER17:14
dtantsurI think we actually do it in ironic with our only major version (/nodes equivalent to /v1/nodes)17:14
aspiersmugsie: that work for you?17:14
edleafeNot to divert the conversation, but what is driving the need for new healthcheck versions?17:14
dtantsuredleafe: it's a good question actually17:14
mugsieedleafe: the old one misses some information, and the body changes17:14
aspiersplease read that first :)17:15
aspiersthere's a lot of history here17:15
aspiersa *lot*17:15
*** mattoliverau has joined #openstack-sdks17:15
aspiersabout 2+ years of bikeshedding17:15
dtantsur"please read these 2 years of bikeshedding history first", thanks, so nice of you :D17:15
edleafedtantsur: it'll cure your insomnia :)17:15
dtantsurso, breaking changes?17:15
evrardjpedleafe: also check L105 around
evrardjpeven if it's not enough for the context17:16
elmikoimo aspiers, whoever is writing the initial draft spec/guideline you just pick a path and we can debate the wording in the pr17:17
elmikoat least get a vote on record at that point17:18
elmikoi'm fine to see my opinion lose in the end, but at least some sort of roll call on the pr would help drive the discussion17:18
elmiko(unless this has already happened and i missed it all XD)17:18
edleafeMy feelings on API changes is to not release it until there is very little liklihood of it needing to be changed in the foreseeable future. IOW, it should not be an iterative process17:18
elmiko++ agreed17:19
elmikomind you, i am not proposing the release but the spec/guideline that proposes the usage17:19
*** jpich has quit IRC17:19
*** radeks__ has joined #openstack-sdks17:23
dtantsuryeah, a spec would be nice (I thought we had one?)17:24
*** witek has quit IRC17:24
elmikoi thought we did too, but maybe mugsie did such a good job presenting it that i just didn't notice XD17:27
cdentclearly the healthcheck can be written, tested, and nearly packaged without ever needing to choose its official name, yes?17:28
elmikooh definitely17:28
elmiko /s17:28
cdentin fact, ideally the middleware would be written with the name being a parameter17:28
cdent(with a default)17:28
mugsiecdent: I did think about that17:28
mugsieBut I like the idea that we have a single well known url17:28
elmikoi do too mugsie17:29
mugsiejust make it part of the refstack tests with the good URL, and that should fix it :P17:29
cdentmugsie: oh, yeah, I want there to be a single well known url too17:29
cdentI'm just saying that the coding doesn't require that17:29
mugsiein 2-5 years17:29
elmikoheh, ++17:30
elmikowell see, we just add a new entry in the service catalog for healthcheck uris..... (kidding!)17:30
* cdent glares at elmiko 17:31
* elmiko runs17:31
dtantsurit's kind of these additional fields in DNS, right?17:31
elmikometadata all the things! \o/17:31
dtantsurso we probably could do it.. but then the monitoring tools will have to learn about the catalog17:32
elmikook, i'm headed out for lunch. take care all o/17:35
dtantsuryeah, I should go as well17:36
*** dtantsur is now known as dtantsur|afk17:37
*** jpena is now known as jpena|off17:43
*** e0ne has joined #openstack-sdks18:14
*** e0ne has quit IRC18:16
*** bobh has joined #openstack-sdks18:23
*** e0ne has joined #openstack-sdks18:36
*** bobh has quit IRC18:39
*** radeks__ has quit IRC18:56
*** ttsiouts has joined #openstack-sdks19:11
*** ttsiouts has quit IRC19:11
*** bobh has joined #openstack-sdks19:18
*** radeks__ has joined #openstack-sdks19:33
*** e0ne has quit IRC19:44
*** toabctl has joined #openstack-sdks20:05
*** gtema has joined #openstack-sdks20:12
*** irclogbot_2 has quit IRC20:37
*** irclogbot_2 has joined #openstack-sdks20:48
*** openstackgerrit has quit IRC20:56
*** bobh has quit IRC21:30
*** cdent has quit IRC21:48
*** mriedem has quit IRC21:49
*** mriedem has joined #openstack-sdks21:56
*** radeks__ has quit IRC22:05
*** mriedem has quit IRC23:17
*** tosky has quit IRC23:57

Generated by 2.15.3 by Marius Gedminas - find it at!