14:00:28 <rafaelweingartner> #startmeeting cloudkitty
14:00:28 <opendevmeet> Meeting started Mon Aug 19 14:00:28 2024 UTC and is due to finish in 60 minutes.  The chair is rafaelweingartner. Information about MeetBot at http://wiki.debian.org/MeetBot.
14:00:28 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
14:00:28 <opendevmeet> The meeting name has been set to 'cloudkitty'
14:00:31 <rafaelweingartner> Hello guys!
14:00:36 <rafaelweingartner> Roll count
14:00:38 <rafaelweingartner> \o
14:01:12 <priteau> o/
14:03:20 <rafaelweingartner> I guess its just us today =)
14:03:34 <rafaelweingartner> #topic Target reviews
14:03:52 <rafaelweingartner> Pedro and I discusse the situation with the tempest tests #link https://review.opendev.org/c/openstack/cloudkitty-tempest-plugin/+/892382
14:04:05 <rafaelweingartner> and instead of creating new branches, and so on, which seemed way more work
14:04:30 <rafaelweingartner> we decided to adapt the tests to work with both, the current state of the APIs, and the new state of the patch in #link https://review.opendev.org/c/openstack/cloudkitty/+/876643
14:04:33 <rafaelweingartner> what do you think?
14:04:40 <rafaelweingartner> this way, we can merge the tempest tests
14:04:42 <rafaelweingartner> and then the code
14:05:43 <priteau> I am checking
14:07:49 <priteau> It looks OK testing-wise, I left some minor comments about typos/wording
14:08:26 <priteau> However this highlights to me that we shouldn't be making such changes in behaviour without bumping the API
14:08:47 <priteau> Other projects have microversions for this.
14:09:07 <rafaelweingartner> yes, but then we would need to introduce this kind of processes, which would take some good amount of energy
14:09:32 <priteau> That's true
14:09:59 <rafaelweingartner> I mean, I am in favor of it
14:10:10 <rafaelweingartner> but, I am not sure if it would be doable for ourside now
14:10:39 <rafaelweingartner> after Pedro updates the patch, can we merge it?
14:10:51 <rafaelweingartner> I mean, would you be ok with the merge?
14:12:56 <priteau> I can't say that I fully support it but I don't want to delay progress
14:13:12 <priteau> So I will approve.
14:15:49 <rafaelweingartner> Hmm
14:15:57 <rafaelweingartner> What would be the concens with it?
14:16:20 <rafaelweingartner> I mean, we are introducing a method to facilitate the life of people when scheduling/preparing rating rules
14:17:45 <priteau> It is still an API change
14:18:23 <priteau> It might break automation that people have in place
14:18:30 <rafaelweingartner> if they have :)
14:18:56 <rafaelweingartner> I see what you mean, but I have no idea on how to propose something like we are proposing, withotu changing the API
14:18:58 <priteau> https://specs.openstack.org/openstack/api-wg/guidelines/api_interoperability.html#evaluating-api-changes
14:19:10 <rafaelweingartner> I mean, we can introduce versioning
14:19:48 <priteau> That's the point of using microversions
14:19:49 <rafaelweingartner> but it would be so much work for so little gain
14:20:02 <rafaelweingartner> and right now CloudKitty has an even bigger issue
14:20:15 <rafaelweingartner> every reprocessing can generate different outcomes for unaware users
14:20:25 <rafaelweingartner> then, of course, we can blame the user
14:20:40 <priteau> This is why I am saying that I won't fight against it, just letting you know this is kind of against OpenStack guidelines / design.
14:20:41 <rafaelweingartner> but, if we face the situaiton, the software if limited
14:22:14 <rafaelweingartner> the change we are introducing requires a versiong change, but we do not need MV for that
14:22:19 <rafaelweingartner> We are already doing that
14:22:29 <rafaelweingartner> we will release this new behavior in a new major version, right?
14:23:07 <priteau> I see, you are talking about the software version
14:23:14 <priteau> Yes, there will be a major version bump
14:23:28 <priteau> But the client/user doesn't see this (it is not exposed to them via the API)
14:23:44 <priteau> Which is why there are also API versions (unrelated from software versions)
14:24:03 <priteau> e.g. Nova: https://docs.openstack.org/nova/latest/reference/api-microversion-history.html
14:24:47 <rafaelweingartner> yes, but that is not necessary a requirement, right?
14:26:24 <priteau> See the guideline link I posted above
14:28:10 <rafaelweingartner> I know about it
14:28:26 <rafaelweingartner> but that is what I am saying, we are adding/changing a new behavior in a new software version]
14:28:38 <rafaelweingartner> that is what the guideline is about
14:29:13 <rafaelweingartner> but anyways, I see that you are against it. I understand the decision, and I will see if we can (how to) overcome this kind of situation
14:29:39 <priteau> I am pretty sure the guideline is talking about API versioning, not software (git) version
14:30:01 <rafaelweingartner> that is up to interpretation
14:30:09 <rafaelweingartner> as many other things in openstack management
14:30:21 <rafaelweingartner> but I will discuss here how to acomodate what you suggested
14:31:14 <priteau> As I said earlier I don't want to delay features just because of this, since we have limited manpower
14:33:05 <rafaelweingartner> I see, but I understand what you mean. Therefore, let's follow the rule to the letter. It is better than to now follow the rule.
14:37:39 <rafaelweingartner> That was basically it from our side here. Do you have any priority or attention that is needed in some specific patch?
14:39:22 <priteau> Just a note that feature freeze is next week
14:39:38 <priteau> And if we need any client changes, that needs to be released before Thursday next week
14:41:00 <rafaelweingartner> I see
14:41:09 <rafaelweingartner> I guess we will not make it then with that feature
14:41:11 <rafaelweingartner> but that is fine
14:41:46 <priteau> Unrelated, there are issues with testing of some unmaintained branches.
14:41:59 <priteau> > keystoneauth1.exceptions.catalog.EndpointNotFound: internalURL endpoint for metric service in RegionOne region not found
14:42:11 <priteau> Could be some issue with gnocchi rather than cloudkitty
14:42:20 <rafaelweingartner> I am not sure
14:42:33 <rafaelweingartner> do you know what error they are receiving?
14:42:45 <rafaelweingartner> Gnocchi has not had a release in some good amount of time
14:42:59 <rafaelweingartner> we are working towards a new release right now
14:43:28 <priteau> I don't even see Gnocchi being deployed in https://dfcf198ef61f2cfa2f98-d07beda532036ff76a0dc6dba516b61f.ssl.cf5.rackcdn.com/911121/1/check/cloudkittyclient-devstack-functional-v1-client/613809a/
14:45:18 <rafaelweingartner> no idea
14:45:25 <rafaelweingartner> there is a build issue before that error
14:45:28 <rafaelweingartner> might be related
14:46:19 <priteau> I can take a look at it
14:46:48 <rafaelweingartner> ok
14:48:40 <priteau> Nevermind, this error happens in master too, which works
14:49:30 <priteau> I found the actual issue in the logs, I will check if I can fix it
14:49:51 <priteau> It's a tox issue
14:53:14 <rafaelweingartner> I see
14:53:45 <priteau> This might fix it: https://review.opendev.org/c/openstack/python-cloudkittyclient/+/926540
14:53:55 <priteau> Please check in one hour if Zuul has passed
14:54:29 <rafaelweingartner> I see
14:54:36 <rafaelweingartner> ok
14:54:45 <rafaelweingartner> Then I merge it
14:55:52 <rafaelweingartner> I guess that was it for today.
14:55:59 <rafaelweingartner> Do you have something else to add before we close?
14:56:17 <priteau> Nothing else.
14:56:32 <rafaelweingartner> Thank you guys for participating. Have a nice week.
14:56:37 <rafaelweingartner> #endmeeting
14:56:43 <rafaelweingartner> #endmeeting