*** bnemec has quit IRC | 00:28 | |
*** bnemec has joined #openstack-tc | 00:33 | |
*** bnemec has quit IRC | 00:40 | |
*** bnemec has joined #openstack-tc | 00:46 | |
*** bnemec has quit IRC | 00:57 | |
*** bnemec has joined #openstack-tc | 01:15 | |
*** bnemec has quit IRC | 01:45 | |
*** bnemec has joined #openstack-tc | 01:50 | |
*** bnemec has quit IRC | 01:57 | |
*** evrardjp has quit IRC | 04:33 | |
*** evrardjp has joined #openstack-tc | 04:33 | |
*** belmoreira has joined #openstack-tc | 05:17 | |
*** belmoreira has quit IRC | 05:17 | |
*** slaweq has joined #openstack-tc | 06:50 | |
*** belmoreira has joined #openstack-tc | 07:01 | |
*** e0ne has joined #openstack-tc | 07:07 | |
*** irclogbot_2 has quit IRC | 07:13 | |
*** e0ne has quit IRC | 07:15 | |
*** tosky has joined #openstack-tc | 07:35 | |
*** Luzi has joined #openstack-tc | 07:37 | |
*** ralonsoh has joined #openstack-tc | 07:59 | |
*** Luzi has quit IRC | 08:40 | |
*** ralonsoh_ has joined #openstack-tc | 09:21 | |
*** ralonsoh_ has quit IRC | 09:21 | |
*** e0ne has joined #openstack-tc | 09:51 | |
mnaser | evrardjp: :( | 11:26 |
---|---|---|
evrardjp | I am sorry | 11:27 |
mnaser | I’ve been at your position many months ago | 11:27 |
evrardjp | I tried to not speak | 11:27 |
mnaser | hence my seemingly ignorant vibe yesterday | 11:27 |
evrardjp | I just can't | 11:27 |
evrardjp | it's smarter not to engage, and just let our community | 11:28 |
evrardjp | just let our community grow* | 11:28 |
evrardjp | let's move on :) | 11:28 |
mnaser | I feel y’a, it hurts | 11:28 |
evrardjp | I have removed my comments on their thread full of misplaced comments and laughter. | 11:28 |
evrardjp | anyway | 11:29 |
evrardjp | Do the folks willing to fork have everything they need? | 11:29 |
mnaser | I think they have made another proposal but ttx has raised the good points of figuring out if it should be official eventually or not | 11:31 |
mnaser | (My opinion is no) | 11:31 |
evrardjp | is there an alternative review? | 11:32 |
evrardjp | I should vote on? | 11:32 |
mnaser | No, I posted a review on https://review.opendev.org/#/c/744592/2 asking them to reupload with that prefix | 11:33 |
evrardjp | got it. | 11:33 |
adriant | mnaser, if it isn't official we really need to make a long term plan to move away from Gnocchi (or whatever bloody name we give it) | 11:42 |
mnaser | adriant: I agree that we should have a long term plan! I am inclined to suggest the telemetry team to look at current TSDBs and see if they make sense to use :) | 11:43 |
adriant | but that will mean a new API in Ceilometer, and people will be unhappy with that | 11:44 |
adriant | I'm writing an email to the mailing list to get some opinions about that | 11:45 |
mnaser | adriant: perhaps it could mean that we expose the API of $tsdb in some way covered by keystone | 11:45 |
adriant | mnaser, hard, because something like Influx is the best candidate imo, but it isn't restful at all really | 11:46 |
adriant | and that's essentially the situation we are in with Gnocchi... | 11:46 |
adriant | we can't ditch the backend BECAUSE it's our bloody API :*( | 11:46 |
mnaser | I think my problem is gnocchi is both and API and a storage tsdb | 11:46 |
adriant | exactly | 11:47 |
mnaser | And no one here wants to actually maintain a tsdb | 11:47 |
mnaser | but we want to maintain an API | 11:47 |
adriant | contrary to what people on JD's twitter think, big projects like OpenStack need an abstraction layer | 11:47 |
adriant | we can make a solid consistent API, and provided the backend supports whatever we need, we can swap it and keep the API the same | 11:47 |
adriant | Gnocchi broke that :/ | 11:48 |
mnaser | Essentially, gnocchi/Prometheus or gnocchi/influxdb should be the combos we have | 11:48 |
mnaser | Just like we have nova/libvirt etc | 11:48 |
adriant | not go back to putting an API into Ceilometer? | 11:48 |
mnaser | All the processing should be put off to the backend, I agree | 11:48 |
mnaser | I mean, I’m torn on that | 11:49 |
adriant | because what you're suggesting is make Gnocchi official, but kill the backend part :P | 11:49 |
adriant | and replace it | 11:49 |
mnaser | Right because I think there is value in a project that provides multi tenant tsdb-aaS | 11:49 |
mnaser | which other openstack projects consume | 11:49 |
adriant | ceilometer used to be that | 11:50 |
adriant | before jd ripped out the API | 11:50 |
mnaser | my thought on where we should take this changes a lot as I discuss all these things | 11:50 |
adriant | well, correction, ceilometer was never a true tsdb | 11:50 |
mnaser | hence my suggestion to just start off in a different namespace, take the time to figure out how and why you want to accomplish it and then if it involves moving gnocchi to official we can do that and the rename might just take a while to finish | 11:51 |
mnaser | But in terms of governance it would be done | 11:51 |
adriant | ok, fair | 11:51 |
mnaser | so that’s why i think youll be able to start figuring out what you want to do with gnocchi | 11:52 |
adriant | I know renaming stuff on gerrit used to be a pain | 11:52 |
mnaser | so that the tc knows what it expects the project to do if it wants to be official | 11:52 |
adriant | mnaser, I personally want to do nothing with it, but I don't have a choice :P | 11:52 |
mnaser | aka is it just an API, API and tssb, etc | 11:52 |
mnaser | adriant: it’s just a pain in that it happens once every month or so — can’t remember how often | 11:53 |
mnaser | So we just have to wait for that time to have the repo renames | 11:53 |
adriant | it's a service restart from memory right? | 11:53 |
mnaser | adriant: I do really appreciate you putting in the time, that’s for sure | 11:53 |
mnaser | adriant: yes it involves restating gerrit hence why it’s once ever while | 11:53 |
adriant | Ok, then lets throw it in a non-os namespace, just to get the whole shit show started, and from there start planning our roadmap for what the hell we want to do with it | 11:55 |
adriant | mnaser, and it needs to be done. It's a pain, but for the health of OpenStack we have to deal with the stupid politics of situations like this | 11:56 |
mnaser | right — it positions us to get started with our tools and it doesn’t stop us from bringing it official later once we figure out the intentions of the project | 11:57 |
* adriant nods | 11:57 | |
adriant | and for the record, I chose "Farfalle" because you make that pasta by giving it a squeeze, much like Gnocchi shapes the data on input, but also because it's better than calling it Fartoffelkloesse (potato dumpling) | 12:00 |
adriant | shit kartoffelkloesse | 12:00 |
adriant | no F | 12:00 |
adriant | screwed up when capitalising it | 12:00 |
fungi | gerrit repository renames don't happen on any set schedule, people just let the opendev sysadmins know what needs to be renamed and then when there are some to batch up they pick a convenient date and time. but yes the process is mostly orchestrated these days and so the only major disruption is a brief gerrit outage | 12:04 |
adriant | fungi, good to know | 12:06 |
fungi | it used to be hours of downtime, until gerrit grew the ability to do its reindexing with the service up and running | 12:08 |
adriant | I'll admit, I do agree with JD (and his twitter followers) that gerrit is more than somewhat awful to work with until you get used to it (and even then it's not great). | 12:10 |
adriant | but github is still pretty bad and limited as well, just prettier | 12:10 |
fungi | it's not great, but still far, far, far less awful to work with than github, gitlab or bitbucket (in my opinion anyway) | 12:11 |
adriant | I like gitlab I'll be honest | 12:11 |
adriant | far more than github | 12:11 |
fungi | the workflow for both of them is very similar | 12:12 |
adriant | it is, but gitlab does some stuff a little nicer | 12:12 |
adriant | but the control and review process around gerrit is nicer | 12:12 |
fungi | but i'm used to the traditional code review on mailing lists workflow (a la lkml) and gerrit is much closer to that | 12:12 |
adriant | despite one of the worse interfaces ever | 12:12 |
fungi | oh, yeah i hardly touch gerrit's webui at all anyway | 12:13 |
fungi | but i'm not real big on doing things in web browsers to begin with | 12:13 |
fungi | console-based clients are far more convenient | 12:14 |
adriant | gerrit has a console interface?! | 12:15 |
adriant | Not that I'm likely to use it, but I'm curious | 12:15 |
fungi | https://pypi.org/project/gertty/ | 12:15 |
adriant | huh, I've only ever touched git-review | 12:16 |
fungi | gerttu is a compliment to git and git-review, it doesn't submit your changes or integrate its own editor, its focus is on reviewing changes | 12:17 |
fungi | s/gerttu/gertty/ | 12:17 |
fungi | though worth noting, the gerrit webui on review.opendev.org is old, we're getting ready for a major upgrade in coming weeks | 12:19 |
fungi | they've completely redesigned the webui in newer gerrit versions | 12:19 |
fungi | whether that's for better or for worse i can't say, as i tend not to find web interfaces convenient anyway | 12:20 |
adriant | we'll see I guess | 12:20 |
adriant | the last gerrit ui update was better while also making some old useful features worse :( | 12:21 |
openstackgerrit | Merged openstack/governance master: Add etcd3gw to Oslo https://review.opendev.org/747188 | 12:51 |
*** bnemec has joined #openstack-tc | 12:51 | |
tosky | adriant: I'd like to point out that "gnocchi" isn't exactly a kind of pasta :) | 12:52 |
adriant | tosky, true, but there is some inner library call carbonara in the code from memory as well | 12:53 |
tosky | or at least here in Italy I think not everyone may consider that as pasta | 12:53 |
adriant | so jd was already sort of mixing metaphors | 12:53 |
tosky | (and please consider this a Friday-inspired comment :) | 12:54 |
openstackgerrit | Merged openstack/governance master: Move towards dual office hours in diff TZ https://review.opendev.org/746167 | 12:54 |
adriant | tosky: https://gnocchi.xyz/operating.html?highlight=carbonara#how-to-plan-for-gnocchis-storage | 12:55 |
adriant | adriant, I am almost tempted to rename it to kartoffelkloesse just to truly annoy people :P | 12:55 |
adriant | ... | 12:55 |
adriant | tosky* | 12:55 |
tosky | :D | 12:55 |
* adriant should be asleep | 12:55 | |
adriant | Farfalle is at least fun to say | 12:56 |
tosky | I totally agree, but I'm biased | 12:56 |
openstackgerrit | Merged openstack/governance master: Update and simplify comparison of working groups https://review.opendev.org/746763 | 12:58 |
gmann | o/ | 13:48 |
fungi | fusilli would be more fun to say | 13:58 |
fungi | and possibly mostaccioli | 13:59 |
fungi | though if you just want to stick with the same basic food, austrians apparently call their version of gnocchi "nockerl" which benefits from being exactly the same letter count too | 14:02 |
*** bnemec has quit IRC | 14:15 | |
evrardjp | and is easy to pronounce! | 14:16 |
adriant | fungi, that's an option! | 14:17 |
adriant | but it may be a little too close for comfort to the existing name? | 14:17 |
* adriant shrugs | 14:17 | |
adriant | I don't care what it is called, as long as we move it back to where we can chose what to merge | 14:18 |
adriant | and I should go back to bed before I lose the ability to respond mostly politely to Julien... | 14:20 |
*** bnemec has joined #openstack-tc | 14:20 | |
zaneb | oof https://en.wikipedia.org/wiki/Eiernockerl#Alleged_tie_to_Hitler | 14:23 |
adriant | zaneb's invocation of Godwin's law is my signal to get to bed :P | 14:24 |
zaneb | rofl | 14:24 |
* adriant will look forward to all the responses about Gnocchi/Farfalle/fetucini/whateverpasta tomorrow morning | 14:25 | |
* zaneb votes for https://en.wikipedia.org/wiki/Kopytka | 14:26 | |
smcginnis | Let's just call it Pasta. :) | 14:26 |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/governance master: Add openstack-ansible/os_senlin role https://review.opendev.org/748677 | 14:26 |
*** dklyle has joined #openstack-tc | 14:41 | |
openstackgerrit | Dmitriy Rabotyagov (noonedeadpunk) proposed openstack/governance master: Add openstack-ansible/os_senlin role https://review.opendev.org/748677 | 15:11 |
*** belmoreira has quit IRC | 15:17 | |
*** e0ne has quit IRC | 15:49 | |
*** e0ne has joined #openstack-tc | 16:02 | |
*** e0ne has quit IRC | 16:10 | |
*** ralonsoh has quit IRC | 16:48 | |
*** tosky has quit IRC | 17:03 | |
*** tosky has joined #openstack-tc | 17:04 | |
*** tosky has quit IRC | 17:44 | |
openstackgerrit | Ken Giusti proposed openstack/governance master: Retire devstack-plugin-pika project https://review.opendev.org/748730 | 17:58 |
openstackgerrit | Ken Giusti proposed openstack/governance master: Retire the devstack-plugin-zmq project https://review.opendev.org/748731 | 18:02 |
*** bnemec has quit IRC | 18:43 | |
*** bnemec has joined #openstack-tc | 18:58 | |
gmann | mnaser updating new office hours info on meeting page - https://review.opendev.org/#/c/748740/1 | 19:01 |
mnaser | cc infra-root for the review ^ | 19:01 |
*** bnemec has quit IRC | 19:12 | |
*** bnemec has joined #openstack-tc | 19:29 | |
*** tosky has joined #openstack-tc | 21:33 | |
*** bnemec has quit IRC | 21:36 | |
-openstackstatus- NOTICE: A zuul server ended up with read only filesystems which caused many jobs to hit retry_limit. The server has been rebooted and appears happy. Jobs can be rechecked. | 22:14 | |
*** tosky has quit IRC | 23:13 | |
*** bnemec has joined #openstack-tc | 23:26 | |
*** bnemec has quit IRC | 23:35 | |
*** bnemec has joined #openstack-tc | 23:46 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!