Thursday, 2017-11-30

*** sdague has quit IRC00:01
*** david-lyle has quit IRC00:03
*** dklyle has joined #openstack-tc00:03
*** kumarmn has joined #openstack-tc00:04
*** david-lyle has joined #openstack-tc00:07
*** dklyle has quit IRC00:08
*** hongbin has quit IRC00:08
*** kumarmn has quit IRC00:09
*** dklyle has joined #openstack-tc00:29
*** david-lyle has quit IRC00:30
*** david-lyle has joined #openstack-tc00:55
*** dklyle has quit IRC00:55
*** david-lyle has quit IRC01:01
*** kumarmn has joined #openstack-tc01:05
*** kumarmn has quit IRC01:09
*** kumarmn has joined #openstack-tc01:15
*** liujiong has joined #openstack-tc01:20
*** mriedem has quit IRC01:21
*** mriedem has joined #openstack-tc01:23
*** kumarmn has quit IRC01:32
*** mriedem has quit IRC02:11
*** gcb has joined #openstack-tc02:40
*** diablo_rojo has quit IRC03:13
*** gcb has quit IRC03:25
*** gcb has joined #openstack-tc03:25
*** kumarmn has joined #openstack-tc03:33
*** kumarmn has quit IRC03:38
*** gcb has quit IRC04:09
*** kumarmn has joined #openstack-tc04:31
*** rosmaita has quit IRC04:43
*** kumarmn has quit IRC05:40
*** david-lyle has joined #openstack-tc07:19
*** kumarmn has joined #openstack-tc07:41
*** kumarmn has quit IRC07:46
ttxEmilienM: also the infra team is using StoryBoard (was first migrated)08:01
*** jpich has joined #openstack-tc08:55
*** cdent has joined #openstack-tc09:29
*** liujiong has quit IRC09:33
*** liujiong has joined #openstack-tc09:34
*** mfedosin_ has joined #openstack-tc09:38
*** mfedosin has quit IRC09:39
*** EmilienM has quit IRC09:39
*** mfedosin_ is now known as mfedosin09:39
*** EmilienM has joined #openstack-tc09:41
*** purplerbot has quit IRC09:42
*** liujiong has quit IRC09:42
*** liujiong has joined #openstack-tc09:43
*** liujiong has quit IRC09:43
*** david-lyle has quit IRC09:46
*** purplerbot has joined #openstack-tc10:03
*** dtantsur|afk has quit IRC10:09
*** dtantsur has joined #openstack-tc10:15
*** dtantsur has quit IRC10:15
*** dtantsur has joined #openstack-tc10:15
openstackgerritFlavio Percoco proposed openstack/governance master: Add ansible-role-k8s-glance
*** sdague has joined #openstack-tc11:02
*** cdent has quit IRC11:04
*** zaneb has quit IRC11:09
*** efried has quit IRC11:10
*** efried has joined #openstack-tc11:14
*** zaneb has joined #openstack-tc11:15
*** eumel8 has joined #openstack-tc11:35
*** cdent has joined #openstack-tc11:35
*** kumarmn has joined #openstack-tc12:42
*** cdent has quit IRC12:46
*** kumarmn has quit IRC12:46
*** dtantsur is now known as dtantsur|brb13:13
dhellmanndiablo_rojo: if you're still looking for an example of using the LP API to  leave comments, check
openstackgerritEric Fried proposed openstack/project-team-guide master: Trivial change to trigger publish job
efrieddhellmann FYI ^13:43
smcginnisefried, dhellmann: Got it.13:44
*** mriedem has joined #openstack-tc13:47
efriedsmcginnis Thanks!13:53
openstackgerritMerged openstack/project-team-guide master: Trivial change to trigger publish job
*** rosmaita has joined #openstack-tc14:03
*** cdent has joined #openstack-tc14:06
*** dtantsur|brb is now known as dtantsur14:42
*** kumarmn has joined #openstack-tc14:53
*** hongbin has joined #openstack-tc14:59
cdenttc-members, to arms, to arms15:00
cdentoffice hours time15:00
* mugsie lurks15:00
* EmilienM in a meeting already but around15:00
* ttx looks up his agenda15:00
* fungi finds something to call an office15:00
ttxdhellmann: wanted to ask about docs:follows-policy, which is a tag without any projects applied to it15:01
cdentI was wondering if there's any more hashing out on the interop testing review that we can or should do:
cdentit feels stuck, for multiple reasons15:03
* ttx catches up on that15:03
cdentwe had quite a bit of discussion about here yesterday15:03
cdentabout it15:03
mugsieThe interop people said that they would put some comments on it as well yesterday15:04
smcginnisBefore I forget, I wanted to mention that Freezer might be in bad shape.15:05
smcginnisI checked in with the PTL because they missed Q-1 and found out he's pretty much the only active core, and part time at that.15:05
ttxcdent: yes i tend to agree that reusing the same thing for multiple purposes is what brings those weird discussions up. A situation created in the first place by the lack of a dedicated body of contributors working on interop tests, hence the reuse of existing tests15:07
eumel8hello tc o/15:08
ttxA lot of people more directly involved with maintaining those tests (inside and outside tempest) have much stronger opinions than me on this (as they should)15:08
ttxwhich is why we need to hear more from the various stakeholders15:09
ttxsmcginnis: so last time I saw the freezer folks they were contemplating a merge witht he karbor folks15:09
smcginniseumel8: o/15:10
fungiyeah, the interop effort lacked sufficient humans to build a new validation framework and write all the tests necessary to confirm every capability, and so it was recommended that they partner with the qa team on using an existing test framework and body of tests to accomplish that goal15:10
ttx"folks" being a bit of an overstatement here15:10
smcginnisttx: I haven't heard of any activity there, but I stopped following the karbor activity once that got off the ground.15:10
cdentsmcginnis: was that vitaliy?15:10
ttxI think having two slightly-overlapping projects struggling is a bit stupid there15:10
smcginniscdent: Saad Zaher15:11
smcginnisttx: Agreed. Though at the time Smaug was starting up, they had some significant differences in approach.15:11
ttxsmcginnis: I can send an email to karbor/freezer see what future may hold15:12
ttxif they miss q2 we'll pull them off anyway15:12
smcginnisttx: ++15:12
cmurphypull them off?15:13
ttxcmurphy: from release content15:13
cmurphyah ok15:13
ttxwhich is usually a slippery slope to being made unofficial, too15:13
ttxbut that's just release team policy, if you can't hit the milestones, we prefer to tell people not to expect it in the final release15:14
ttxsmcginnis: we should really also be doing that for intermediary-released stuff that doesn't do intermediary releases15:14
ttxsince those were supposed to be "more active"15:15
* ttx types email to freezer/karbor15:15
smcginnisTrue. Though I think some of those are more "stable" now, so less frequently released.15:15
cmurphywe're either using or planning on using (i forget) freezer in our product, i don't think anyone realized it was in such bad shape15:16
cdentis it in bad shape, or stable?15:17
cmurphygood question15:17
*** marst has joined #openstack-tc15:17
cmurphyif it was really in bad shape i think we would have noticed15:17
cdentthat's what I was thinking15:17
mugsieI know I have run into issues with it before, but I think they were more config issues than software issues15:18
fungiit's certainly unlikely to introduce a lot of new incompatibilities if there's nobody writing/reviewing changes for it ;)15:18
ttxcdent: from that session there was a lot of work on the table, but it was ~working15:18
smcginnisI think it's stable, but also at risk of destabilization with lack of eyes on it.15:19
cdentYeah, I'm not suggesting there's not a problem, just grinding my axe that sometimes software doesn't have to change much to be good.15:21
*** rosmaita has quit IRC15:21
smcginniscdent: Yep, absolutely.15:22
smcginnisJust not sure one part time person can keep the code rot at bay.15:22
cdentyup, def worth investigation15:22
cdentAnybody know anything about who is thinking about running for the board?15:23
* dtroyer takes a quick step backward15:24
ttxnope. I usually stay away from it since it's weird to be a director and a member of the staff15:24
smcginnisI actually thought about it for 30 seconds or so given the previous discussions.15:25
ttxbut please consider running, since individual directors can end up weird otherwise15:26
dtroyersmcginnis: ++  I'll support you there15:26
cdent"end up weird"?15:26
smcginniscdent: Maybe that's what it would be if I ran. :D15:26
cdentToo late.15:27
dtroyerto me that means "not enough techincally-knowledgeable representation"15:27
smcginnisdtroyer: Thanks. Want to be my campaign manager? :)15:27
ttxcdent: sometimes they end up wearing their company hat instead of being representative of all contributors15:27
* dtroyer sets up the smcginnis-for-board SuperPAC15:27
cdentYeah, you have to run a full court social media presss smcginnis and we know that dtroyer is _so_ your person for that.15:27
ttx(contributors in a very large sense)15:27
ttxdtroyer: you could pressure him by nominating him15:28
dtroyercdent: hey!  That'll be my tweet for this year!15:28
ttxthen it's just a click away15:28
smcginnisTweet in 2017: check15:28
ttx"dtroyer tweeted after a long absence"15:28
ttxI don't tweet much. I made a promise when I joined twitter to always have less tweets than followers.15:29
fungioh, of likely interest to the tc (related to discussions at the joint leadership meeting in sydney), infra spec about hosting new communities under the foundation's umbrella:
* cdent adds to list15:30
ttxfungi: wow, that came in fast15:31
fungiit was made a priority15:31
ttxfungi: I'll read, does it establish a dependency on the git flattening exercise or not ?15:31
fungino, but it does take it into account15:31
ttxok, will review, thanks for the pointer15:32
dimsvery interesting fungi15:32
fungiit's a bit of a priority for the infra team because we'd like to de-openstack zuul/nodepool before we officially release/announce zuul v3 general availability15:32
fungiand there are at least two efforts (one already has a mailing list and did an opendev event, the other slated to be announced next week) which will want the same15:33
ttxanyone interested in being CC-ed on that exchange with Freezer/Karbor PTLs ? smcginnis maybe ?15:34
dhellmannttx: pkovar was going to rewrite the definition of that tag to make it more relevant for today. Maybe we should remove it for now and let the docs team propose a new tag?15:34
smcginnisttx: Yep. I already spoke to Freezer, so I can be part of that.15:35
ttxdhellmann: yeah, that might be better than leaving a policy around that nobody follows15:35
dimssounds fair dhellmann15:35
* smcginnis notes to take another look through tags15:35
fungidhellmann: seems simple to remove/repropose since nobody applied it15:36
dtroyerI was going to suggest that we don't want to re-use tag labels, but fungi mooted that worry15:36
dhellmanndtroyer : any new docs tag would likely have a different name anyway15:37
dhellmannI'll talk to pkovar about removing it15:37
cdentdtroyer: I would like to borrow this mooter15:37
cdenthelp with anxiety15:37
dtroyerI let him be my proxy for yelling at the kids on the lawn15:38
-openstackstatus- NOTICE: if you receieved a result of "RETRY_LIMIT" after 14:15 UTC, it was likely due to an error since corrected. please "recheck"15:38
fungii definitely do a lot of yelling15:38
* smcginnis does a lot of mumbling to himself15:39
dtroyersmcginnis: I can tell how much of that I am doing based on how close (or not) the dog is sleeping15:40
fungicanine-based mubmle indicator15:40
ttxok, freezer/karbor email sent. I will summarize the discussion when over here15:43
openstackgerritDoug Hellmann proposed openstack/governance master: remove docs:follows-policy tag
efriedpabelanger dhellmann smcginnis Docs have published.  Thanks for your help!15:43
dhellmannefried : good!15:44
smcginnisefried: ++15:45
ttxwhile speaking of tags -- in the same way the stable policy apply better to the "cloud components" than to the SDks or deployment tools buckets of the map, the upgrade tags  also don't make much sense for anything except cloud components with an API and all (openstack and openstack-operations buckets in the map)_15:46
ttxso I might propose the same kind of application limits. Currently only cloud components projects actually apply those tags anyway15:46
ttxwe should just not expect things like openstackclient to assert them15:47
pabelangerefried: \o/15:47
smcginnisttx: That makes sense to me.15:47
* cdent concurs15:48
dtroyerthings like SDKs and OSC do provide an interface, but yes, the assumptions around them and upgrades are different.  I'm not sure if a parallel set of tags would be useful though15:48
dhellmannmaybe someone wants to volunteer to make those updates so ttx doesn't have to15:48
ttxdhellmann: I'm doing them as part of the project navigator revamp exercise anyway15:49
dhellmannah, ok15:49
dhellmannABD - always be delegating15:49
ttxAh! you know how bad I am with that15:49
dhellmannhence the nudge15:50
ttxsays the dude who spent most of yesterday writing ptgbot new code15:50
dhellmannttx: does it tweet yet?15:51
ttxNo, I think what was over the top was submitting irc-meetings changes to defragment the meeting space. Also yesterday15:51
ttxI blame yak shavnig15:52
dhellmannyesterday I finally wrote some validation code to check releases with dependency changes as a way to put off writing a talk15:53
dhellmannleading by example, we are15:53
ttxI wrote an op-ed for a newspaper today !15:53
ttxLike one that gets printed and all15:53
ttxin French though15:53
ttxnow let's see how much it changes between now and publication, if it ever gets published.15:54
fungimy over-the-top devwork lately is probably the pull requests to implement pep 345 project-url support in setuptools15:54
dhellmannoh, even better15:54
fungiwhich i need to push up some new iterations of now that i have feedback15:54
fungii have pbr support already written, but wip until the setuptools implementation is released15:55
ttxAlso good progress with the SIG governance website, almost there15:55
ttxfungi: needs review love to finalize it15:55
fungithanks for teh heads-up, checking15:56
ttxthen I can push the publish job15:56
ttxOH now I remember how I shaved that yak. Involved a broken publication job for irc-meetings15:57
ttxsomething something trailing slash something15:57
fungirsync is really picky about those trailing slashes15:57
*** kumarmn has quit IRC16:28
*** kumarmn has joined #openstack-tc16:29
*** cdent has quit IRC16:34
*** rosmaita has joined #openstack-tc16:35
*** cdent has joined #openstack-tc16:39
openstackgerritDean Troyer proposed openstack/governance master: Move os-client-config to OpenStackSDK from OpenStackClient team
*** david-lyle has joined #openstack-tc16:59
*** dtantsur is now known as dtantsur|afk17:03
*** kmalloc has quit IRC17:11
*** jpich has quit IRC17:44
*** diablo_rojo has joined #openstack-tc18:22
*** openstackgerrit has quit IRC18:48
*** cdent has quit IRC19:31
*** kmalloc has joined #openstack-tc20:07
*** mguiney has quit IRC20:10
*** mguiney has joined #openstack-tc20:11
zanebcan I just mention that in a sane world Freezer would be like 2 Mistral workbooks20:44
zanebbut nobody wants to install Mistral (another service! too complicated!)20:45
zanebso instead we get a whole new service project (or 2, if you count Karbor as the same thing) to handle in a hard-coded way a small subset of what Mistral can do20:46
zanebthis is a classic example of why I think it's crazy to think we can _reduce_ the complexity of OpenStack deployment by sidelining general purpose tools like Mistral that put control of other openstack services in the hands of the user rather than locking it down in python code20:48
* zaneb dismounts soapbox20:49
* smcginnis thinks zaneb has a Summit session proposal - "How to backup your cloud with duct tape." :)20:59
zanebI mean, we have APIs to make backups. You just need something to call them on a schedule21:01
fungithe case was made to include castellan as something services can expect to be present and directly depend on. sounds like a similar case can (should?) be made for mistral i guess?21:05
zanebfungi: so... I wouldn't go that far21:08
zanebit's not that other services need a hard dependency on it21:09
zanebit's more that if every cloud had it then it would be easy for users who need something to create a cinder snapshot once a day to create a workflow that creates a cinder snapshot and have Mistral's cloud cron feature trigger it once a day21:10
zanebnot to mention all the other things users can do with it21:10
zaneband in that world we'd never be creating a separate service (Freezer) to just create Cinder snapshots on a schedule21:11
smcginniszaneb: Well, to be serious, that would be a nice basic "batteries included" thing for really basic needs. But services like Freezer and Karbor do a little bit more than kick off a backup.21:12
smcginnisMost actual user needs would not be met by cron.21:12
zanebI'm not actually familiar with Karbor. I read about it just now and it looks more complex21:13
zanebI did review the Heat resources for Freezer though, and I didn't see anything it does that couldn't be implemented by the user using Mistral in under half an hour21:14
* zaneb could be missing something though21:14
smcginnisI'm more familiar with Karbor. Just based on experience, users need retention policies, business logic for perform recovery when the environment has been changed since time of backup, etc.21:15
smcginnisSo I think there certainly are pieces that could be made better/easier with having Mistral, but not a full solution for larger environments.21:15
smcginnisBTW, looking back, I was not really calling Mistral "duct tape". You just need catchy Summit titles, right? :)21:16
zanebI wouldn't know ;)21:24
fungiwho needs catchy summit titles? just liberally sprinkle with buzzwords: "serverless container network" (doesn't even need to make sense)22:05
fungi"bare metal distributed lock management drivers cage match"22:07
dtroyer"converged software defined quantum nova v3 extensions" is so 201222:15
*** marst has quit IRC22:19
*** kumarmn has quit IRC23:33

Generated by 2.15.3 by Marius Gedminas - find it at!