14:00:26 <rafaelweingartner> #startmeeting cloudkitty
14:00:26 <opendevmeet> Meeting started Mon Jan  9 14:00:26 2023 UTC and is due to finish in 60 minutes.  The chair is rafaelweingartner. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:26 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:26 <opendevmeet> The meeting name has been set to 'cloudkitty'
14:00:33 <rafaelweingartner> Hello guys!
14:00:48 <mkarpiarz> Hi!
14:01:37 <rafaelweingartner> I hope you all had a safe and happy hollyday season.
14:02:08 <mkarpiarz> Thank you!
14:02:23 <rafaelweingartner> Roll count
14:02:27 <rafaelweingartner> \O
14:03:27 <mkarpiarz> I'm here.
14:07:43 <rafaelweingartner> Pierre: are you going to join us as well?
14:08:31 <rafaelweingartner> priteau: :)
14:10:25 <rafaelweingartner> Let's move on, and then if priteau join us, we can sync with him
14:10:26 <priteau> Hello
14:10:32 <rafaelweingartner> hello :)
14:10:40 <rafaelweingartner> #topic OpenSearch support
14:11:12 <rafaelweingartner> We need to crete the OpenSearch option in the Storage backend and deprecate elastic search. Are you priteau going to take a lead on that one?
14:11:21 <priteau> Sorry, I missed the notification
14:11:42 <priteau> Yes, I can work on it with some colleagues
14:11:59 <rafaelweingartner> No worries. Thanks for taking this one up!
14:12:40 <rafaelweingartner> #topic Monasca deprecation
14:12:51 <rafaelweingartner> Monasca has been deprecated, and now we can remove it in the B release
14:12:56 <rafaelweingartner> Are we all on the same page?
14:13:22 <priteau> +1
14:13:34 <rafaelweingartner> ok, I will take this one when B release arrives
14:13:39 <mkarpiarz> Yes!
14:13:40 <rafaelweingartner> I mean, when we prepare for B release
14:13:52 <mkarpiarz> OK
14:13:56 <rafaelweingartner> #topic add CloudKitty API reference docs
14:14:16 <rafaelweingartner> mkarpiarz: is taking this one
14:14:23 <rafaelweingartner> it is almost finished, I guess
14:14:25 <rafaelweingartner> right?
14:14:31 <mkarpiarz> Correct.
14:14:35 <rafaelweingartner> there is only that issue with Tox that we are addressing right now
14:15:04 <rafaelweingartner> Thanks Mariusz!
14:15:12 <mkarpiarz> All patches we need for this have been submitted and some of them reviewed.
14:15:30 <rafaelweingartner> Talking about Tox, we have the following: https://review.opendev.org/c/openstack/cloudkitty/+/869557
14:15:33 <mkarpiarz> But yes, #link https://review.opendev.org/c/openstack/cloudkitty/+/867628 is a blocker
14:15:41 <rafaelweingartner> It is a bit odd, but the same tox test passes here
14:15:44 <rafaelweingartner> but not in Zuul
14:15:53 <rafaelweingartner> do you guys know how to check the Tox version of Zuul?
14:15:57 <priteau> Which version of tox are you using?
14:16:04 <priteau> It should be in the logs
14:16:08 <rafaelweingartner> 4.2.6
14:16:14 <priteau> I can reproduce Zuul's failure
14:16:29 <priteau> I have 4.2.6 too
14:16:38 <priteau> Did you clean your tox envs?
14:16:50 <rafaelweingartner> yes
14:16:56 <mkarpiarz> Right, I'm still on 3.21.4...
14:16:57 <rafaelweingartner> I deleted the .tox folder
14:17:17 <rafaelweingartner> What is your O.S.?
14:17:35 <rafaelweingartner> maybe there is some other library affecting us as well then
14:18:05 <mkarpiarz> Debian 11. I'm using the "python:3.9" Docker image.
14:18:08 <priteau> I am on macOS
14:18:37 <priteau> Python 3.10.9
14:18:58 <rafaelweingartner> hmm
14:19:02 <rafaelweingartner> I am on Ubuntu 22
14:19:05 <rafaelweingartner> Python 3.8.10
14:19:13 <rafaelweingartner> let me check with python 3.10
14:19:17 <mkarpiarz> OK, I'll switch to the 3.10 image then. :)
14:19:49 <rafaelweingartner> but the weird part is that Zuul fails with python 3.8.10
14:20:05 <mkarpiarz> Do we have a pin for a specific version of tox or does Zuul use whatever version is in the repos?
14:20:45 <priteau> Zuul uses the latest tox 4.x, which has caused breakage.
14:20:54 <priteau> Stable branches should be pinned to 3.x
14:21:04 <priteau> We have to be compatible with tox4
14:21:16 <priteau> Since I can reproduce locally I will check
14:21:39 <rafaelweingartner> are you using a clean tox env?
14:22:42 <priteau> Fixed it :)
14:23:22 <priteau> I will push
14:23:34 <mkarpiarz> Sure!
14:23:39 <rafaelweingartner> ok, thanks
14:23:51 <rafaelweingartner> it is funny, I tested with Python 3.10 here
14:23:55 <rafaelweingartner> and it also works just fine
14:24:00 <rafaelweingartner> there must be something else in play
14:24:07 <mkarpiarz> So I take it for the time being we need to test against both latest tox3 and tox4?
14:24:32 <priteau> No, just tox4 for master
14:24:36 <priteau> tox3 for stable
14:24:50 <rafaelweingartner> for the stable branches, we do not need to do anything I guess
14:24:57 <rafaelweingartner> maybe, just pinning the library if needed
14:25:06 <mkarpiarz> OK
14:26:56 <priteau> tox is not a requirement
14:27:05 <priteau> It's something that already lives inside the env
14:27:43 <rafaelweingartner> priteau: so you are going to add a new patch in the open patchset?
14:27:52 <priteau> I will update your patch with the fix
14:27:58 <rafaelweingartner> ah ok
14:27:59 <rafaelweingartner> thanks!
14:28:08 <rafaelweingartner> So, we can move on with the target reviews
14:28:18 <rafaelweingartner> #topic Target reviews
14:28:40 <rafaelweingartner> Have you guys evaluated the spec #link ttps://review.opendev.org/c/openstack/cloudkitty-specs/+/866198?
14:29:00 <rafaelweingartner> do you think it needs some further adjusts or changes?
14:29:11 <priteau> I started reading yesterday but haven't been through it yet
14:29:30 <rafaelweingartner> I see, thanks!
14:29:33 <mkarpiarz> I must admit I have forgotten about this one. :|
14:29:58 <opendevreview> Pierre Riteau proposed openstack/cloudkitty master: Make tox.ini tox 4.0.0 compatible  https://review.opendev.org/c/openstack/cloudkitty/+/869557
14:30:02 <mkarpiarz> I'll read it as soon as this meeting ends.
14:30:10 <rafaelweingartner> ok, thanks!
14:31:46 <rafaelweingartner> So, moving on we have this optimization patch #link https://review.opendev.org/c/openstack/cloudkitty/+/864269
14:31:54 <rafaelweingartner> It is an important one, can you guys review it?
14:32:33 <priteau> Looking
14:33:04 <opendevreview> Merged openstack/cloudkitty master: Fix typo in docs command sample  https://review.opendev.org/c/openstack/cloudkitty/+/867559
14:33:12 <mkarpiarz> Like I said, I can't test this one at scale
14:33:28 <mkarpiarz> but it's working in my test environment, hence the +2.
14:34:04 <rafaelweingartner> This was a complicated one to catch
14:34:38 <rafaelweingartner> I guess we can even do more optimizations there, but will leave it for other round
14:35:42 <priteau> I left some minor comments
14:35:44 <mkarpiarz> Sounds good!
14:35:53 <priteau> Cosmetic mostly
14:37:22 <rafaelweingartner> Thanks!
14:37:27 <rafaelweingartner> I will address them
14:37:46 <rafaelweingartner> So, moving on, we have the following one: https://review.opendev.org/c/openstack/cloudkitty/+/862824
14:37:58 <rafaelweingartner> #link https://review.opendev.org/c/openstack/cloudkitty/+/862824 is a new feature that we are introducing
14:38:24 <rafaelweingartner> There was already a round of review there, and we might need your eyes again
14:39:55 <mkarpiarz> Does this change work for you with PyScripts enabled?
14:39:55 <priteau> Well, I am still worried about the fact that we store this description for every dataframe
14:40:36 <mkarpiarz> Yeah, we've had a discussion about this...
14:43:11 <opendevreview> Rafael Weingartner proposed openstack/cloudkitty master: Optimize Gnocchi fetcher  https://review.opendev.org/c/openstack/cloudkitty/+/864269
14:43:41 <rafaelweingartner> Yes, we too worry about that
14:43:46 <rafaelweingartner> but that is already an issue that we have
14:43:58 <rafaelweingartner> we are storing over and over again alt_name, units, and metadata
14:44:14 <rafaelweingartner> we have noted here a task to find an optimization for that
14:44:33 <rafaelweingartner> so, it would optimiza all other situations and not just for descriptions
14:44:48 <mkarpiarz> That's a fair point.
14:45:32 <rafaelweingartner> I am not sure what you meant regarding PyScript mkarpiarz
14:46:22 <rafaelweingartner> The fix for PyScript is the following patch #link https://review.opendev.org/c/openstack/cloudkitty/+/864578
14:46:44 <mkarpiarz> If you check my last comment, I said I had to disable PyScripts for 862824 to work.
14:47:00 <rafaelweingartner> hmm
14:47:04 <mkarpiarz> Otherwise I'd get "TypeError: 'DataFrame' object is not subscriptable".
14:47:16 <mkarpiarz> But this could be something with my test environment.
14:47:27 <rafaelweingartner> yes, we have it running here with PyScript
14:47:45 <priteau> I guess it will make the storage backend more bloated only for the users of this feature
14:48:09 <rafaelweingartner> exactly
14:48:17 <rafaelweingartner> if you do not use it, everything will remain the same
14:48:57 <mkarpiarz> OK, I'll try this patch again in a clean environment.
14:49:23 <priteau> rafaelweingartner: on the Gnocchi optimisation patch, please make sure you've seen my comment: https://review.opendev.org/c/openstack/cloudkitty/+/864269/8/cloudkitty/fetcher/gnocchi.py
14:49:24 <rafaelweingartner> ok, thanks!
14:51:47 <rafaelweingartner> priteau: that is not going to happen
14:52:01 <rafaelweingartner> as we check the chunk_len before using it in the expression
14:52:10 <rafaelweingartner> chunk_len and resources_chunk[chunk_len - 1]['id'] == marker
14:52:11 <priteau> Ah yes, of course
14:52:20 <rafaelweingartner> If, it is 0, then it evaluates to False already
14:52:46 <priteau> Can you make that chunk_len > 0
14:52:57 <rafaelweingartner> hmm
14:52:59 <priteau> Just for clarity
14:53:20 <rafaelweingartner> then it will make it hard to format the line :)
14:53:25 <priteau> Nevermind then
14:53:30 <rafaelweingartner> but I will do work around it
14:53:46 <priteau> Also you could rename your var to just is_last or is_last_marker
14:54:24 <rafaelweingartner> I do prefer the readability of the variable
14:54:53 <priteau> We have different readability priorities :D
14:55:10 <opendevreview> Rafael Weingartner proposed openstack/cloudkitty master: Optimize Gnocchi fetcher  https://review.opendev.org/c/openstack/cloudkitty/+/864269
14:55:11 <rafaelweingartner> exactly
14:55:22 <rafaelweingartner> I tried to accomodate both
14:55:59 <priteau> OK. I just saw two small nits, otherwise I will +2
14:56:35 <opendevreview> Rafael Weingartner proposed openstack/cloudkitty master: Optimize Gnocchi fetcher  https://review.opendev.org/c/openstack/cloudkitty/+/864269
14:56:58 <rafaelweingartner> Awesome, thanks!
14:57:29 <rafaelweingartner> So, moving on to the last patch #link: https://review.opendev.org/c/openstack/cloudkitty/+/864578
14:57:42 <rafaelweingartner> This is also an important patch for us to move on and merge it as soon as possible
14:57:50 <rafaelweingartner> PyScripts were broken for a long time
14:58:10 <mkarpiarz> I agree!
14:58:56 <mkarpiarz> We should merge this one as soon as possible.
15:00:40 <rafaelweingartner> Guys, we are running out of time
15:00:43 <priteau> you could be just a bit more verbose in the reno
15:00:54 <rafaelweingartner> I guess we continue in our Async mode to review the patches
15:01:00 <rafaelweingartner> Do you guys have something else to add?
15:01:01 <priteau> ok
15:01:04 <rafaelweingartner> To our meeting
15:01:43 <mkarpiarz> I want to add that I was able to build PDF docs
15:01:50 <rafaelweingartner> awesome!
15:02:06 <mkarpiarz> but some lines go out of the frame and the v2 API refs are completely missing.
15:02:37 <mkarpiarz> So this will need some work to fix. :(
15:02:53 <priteau> It doesn't seem worth it
15:03:47 <mkarpiarz> True, unless there is an easy way to automatically break long listings like this one:
15:04:08 <mkarpiarz> https://docs.openstack.org/cloudkitty/latest/admin/configuration/samples/cloudkitty-conf.html
15:05:36 <mkarpiarz> Either way, I wouldn't have time to investigate any further this week.
15:06:13 <priteau> How many people will read our API docs in PDF? Probably none
15:06:13 <priteau> Can we add a note that rendering is not complete?
15:06:43 <rafaelweingartner> I agree that is is not worth it
15:06:54 <mkarpiarz> OK
15:06:54 <rafaelweingartner> I mean, I do not see a problem in the listing being in multiple pages
15:07:12 <rafaelweingartner> Thank you guys for participating. Have a nice week.
15:07:38 <rafaelweingartner> #endmeeting