*** vishalmanchanda has joined #cloudkitty | 04:24 | |
zigo | Hi. I'm getthing this in the cloudkitty-processor.log: | 12:06 |
---|---|---|
zigo | http://paste.openstack.org/show/802883/ | 12:06 |
zigo | Then it stops processing completely... | 12:06 |
zigo | Any idea how can this be fixed? | 12:06 |
priteau | Hi zigo. There was a bug reported a few days ago: https://storyboard.openstack.org/#!/story/2008598 | 12:10 |
priteau | I'll bring it up at today's IRC meeting | 12:11 |
zigo | priteau: This looks similar to what I'm experiencing with the CPU. | 12:11 |
priteau | I've added it to today's meeting topics: https://etherpad.opendev.org/p/cloudkitty-meeting-topics | 12:13 |
priteau | It's at 1400 UTC here, if you want to attend. | 12:13 |
*** mkarpiarz has joined #cloudkitty | 12:17 | |
zigo | priteau: In this channel ? | 12:19 |
zigo | I'll be there, ping me then ... | 12:20 |
*** rafaelweingartne has joined #cloudkitty | 13:49 | |
rafaelweingartne | Hello guys | 13:59 |
rafaelweingartne | #startmeeting cloudkitty | 14:00 |
openstack | Meeting started Mon Feb 22 14:00:04 2021 UTC and is due to finish in 60 minutes. The chair is rafaelweingartne. Information about MeetBot at http://wiki.debian.org/MeetBot. | 14:00 |
openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 14:00 |
*** openstack changes topic to " (Meeting topic: cloudkitty)" | 14:00 | |
openstack | The meeting name has been set to 'cloudkitty' | 14:00 |
rafaelweingartne | Roll count | 14:00 |
priteau | o/ | 14:00 |
mkarpiarz | Hi! | 14:01 |
rafaelweingartne | #topic New core reviewers | 14:03 |
*** openstack changes topic to "New core reviewers (Meeting topic: cloudkitty)" | 14:03 | |
rafaelweingartne | I proposed Marius as a new core reviewers. I guess that we will approve this proposal. I am not sure now about the next steps. | 14:03 |
rafaelweingartne | Maybe priteau could guide me through the process? | 14:03 |
priteau | On Gerrit, go to: | 14:04 |
priteau | Browse > Groups | 14:04 |
priteau | Search for cloudkitty-core | 14:04 |
priteau | Click on it | 14:04 |
priteau | Click members on the left | 14:05 |
priteau | There is an option to add a new member | 14:05 |
mkarpiarz | Thank you guys, I'm honoured! :) | 14:05 |
rafaelweingartne | hmm | 14:06 |
rafaelweingartne | I do not see this option | 14:06 |
rafaelweingartne | I see all of the different CloudKitty groups, but this one | 14:06 |
rafaelweingartne | I am logged in* | 14:06 |
rafaelweingartne | ah, I found it | 14:06 |
rafaelweingartne | never mind | 14:06 |
rafaelweingartne | So, I just add there? | 14:07 |
rafaelweingartne | mkarpiarz, can you confirm if this is your user? | 14:07 |
rafaelweingartne | Mariusz Karpiarz <mariusz.karpiarz@vscaler.com> | 14:07 |
mkarpiarz | Yes, that's me. | 14:07 |
rafaelweingartne | Ok, added | 14:15 |
rafaelweingartne | priteau: do I need to do anything else? | 14:16 |
rafaelweingartne | Welcome to the core team mkarpiarz | 14:16 |
mkarpiarz | :) | 14:17 |
rafaelweingartne | mkarpiarz, Can you test it? | 14:17 |
priteau | Hooray! | 14:17 |
rafaelweingartne | to see if you have the permissions now | 14:17 |
mkarpiarz | Are there any extra responsibilities that come with being part of the core? | 14:17 |
mkarpiarz | (aside for what I've been asked to do so far, so reviewing changes, submitting patches, etc) | 14:18 |
rafaelweingartne | I guess nothing much besides helping on reviews, and merging things sometimes | 14:18 |
priteau | mkarpiarz: You can now merge patches by voting with W+1. So don't merge stupid or buggy patches ;-) | 14:18 |
mkarpiarz | Checking now. | 14:18 |
mkarpiarz | Yes, I can see +-2s in code review and +-1 for workflow. | 14:19 |
rafaelweingartne | Awesome | 14:19 |
rafaelweingartne | Moving on... | 14:20 |
rafaelweingartne | #topic Review priorities | 14:20 |
*** openstack changes topic to "Review priorities (Meeting topic: cloudkitty)" | 14:20 | |
rafaelweingartne | Do you guys have some review priorities to ask for? | 14:20 |
rafaelweingartne | I see that one of my specs was merged last week. Therefore, I will be working on the patch to submit this week | 14:20 |
mkarpiarz | I didn't have much time for Cloudkitty last week but this week should be better. | 14:21 |
mkarpiarz | I'll be testing https://review.opendev.org/c/openstack/cloudkitty/+/774520, unless there are other priorities. | 14:21 |
rafaelweingartne | ah, that is a cool one to test | 14:22 |
rafaelweingartne | we have in production already here | 14:22 |
priteau | There are two low-impact patches we can merge to clear the queue | 14:22 |
priteau | https://review.opendev.org/c/openstack/cloudkitty/+/770236 | 14:22 |
priteau | https://review.opendev.org/c/openstack/cloudkitty/+/768506 | 14:22 |
rafaelweingartne | agreed | 14:23 |
rafaelweingartne | mkarpiarz: would you do the honors? | 14:23 |
mkarpiarz | Both look good to me. | 14:24 |
rafaelweingartne | They can be your first merge :) | 14:25 |
priteau | mkarpiarz: We didn't say explicitly earlier, it is customary to require two +2 votes before merging | 14:25 |
priteau | These two patches have my +2 so you're good to go. | 14:25 |
priteau | mkarpiarz: You can give the +2 and W+1 in a single vote | 14:29 |
mkarpiarz | Ok, so what are we missing on https://review.opendev.org/c/openstack/cloudkitty/+/770236? | 14:29 |
mkarpiarz | (to be able to merge) | 14:29 |
priteau | Nothing missing. | 14:30 |
priteau | It's in the Zuul gate now | 14:30 |
priteau | Zuul need to run gate jobs, it will take ~40 minutes | 14:31 |
mkarpiarz | Cool. How can you check what's in the gate? | 14:31 |
priteau | https://zuul.openstack.org/status#cloudkitty | 14:31 |
rafaelweingartne | Awesom! | 14:31 |
priteau | Should we move to the next topic? | 14:32 |
mkarpiarz | Thanks, https://review.opendev.org/c/openstack/cloudkitty/+/768506 should be good to go soon as well. | 14:33 |
rafaelweingartne | Moving on... | 14:33 |
rafaelweingartne | #topic Help request | 14:33 |
*** openstack changes topic to "Help request (Meeting topic: cloudkitty)" | 14:33 | |
rafaelweingartne | There was a help request, but I am not sure if the reported is here in the channel | 14:33 |
priteau | ping zigo | 14:34 |
rafaelweingartne | I guess that should be something to be discussed in the mailing list right? | 14:34 |
priteau | (not the first reporter, but someone else who saw the bug) | 14:34 |
priteau | I don't think it has to be on the mailing list. We have a bug tracker for this. | 14:34 |
rafaelweingartne | I put there some of the details we would need to help them, but I am not sure if they will provide a follow-up | 14:34 |
zigo | Yeah, got the bug with Victoria. | 14:34 |
priteau | Mailing list is good for more open-ended questions, when you don't really know who to ask, cross-project issues, etc. | 14:35 |
zigo | paste.openstack.org/show/802883/ | 14:35 |
rafaelweingartne | priteau, I mentioned the mailing list because it seems easier, at least to me, to request data, inspect code and so on before replying | 14:35 |
priteau | Well, you can do the same thing with Storyboard | 14:36 |
zigo | I'm available if you need me to investigate some data... | 14:36 |
priteau | For context, we are discussing https://storyboard.openstack.org/#!/story/2008598 | 14:36 |
rafaelweingartne | zigo: we would need more inputs, such as the gnocchi resource show for the problematic metric | 14:36 |
zigo | Hard to tell which is the problematic one... | 14:37 |
priteau | I wonder why we haven't seen this in CI. | 14:37 |
priteau | Could it be a regression from new gnocchi or gnocchiclient releases? | 14:37 |
rafaelweingartne | I am suspecting that the aggregation method `mean` is not configured to the CPU metric for instance | 14:37 |
rafaelweingartne | or maybe, the aggregation is using a different time frame | 14:37 |
zigo | rafaelweingartne: How would I do that? | 14:38 |
rafaelweingartne | gnocchi resource show <resource_id> | 14:38 |
rafaelweingartne | get the metric ID there | 14:38 |
rafaelweingartne | then, gnocchi metric show | 14:38 |
zigo | rafaelweingartne: We wont solve it during this meeting, but ok, we can investigate together if you have time for it. | 14:39 |
rafaelweingartne | we can interact in the Story board then | 14:39 |
zigo | One solution could indeed be cloudkitty nicely and gracefully printing the advise you're giving. | 14:39 |
rafaelweingartne | you could post the data there, and we move on from that point | 14:39 |
zigo | I'm not sure how to get the <resource_id> you're mentionning above ... | 14:40 |
rafaelweingartne | I see, it will be easier for us to interact there in the story board them | 14:40 |
rafaelweingartne | can you post your doubts there, and we move on via Story board then | 14:40 |
zigo | Well, ok, though it'd be faster if we interact on IRC, IMO. | 14:41 |
zigo | Then we can post on Storyboad when I know what to post. | 14:41 |
rafaelweingartne | I do not use IRC besides attending meeting | 14:41 |
zigo | :/ | 14:41 |
zigo | Wont be easy then. | 14:41 |
rafaelweingartne | Therefore, to me, it would be more problematic. I would prefer an Async method of interaction | 14:41 |
zigo | rafaelweingartne: Can we just take 10 minutes on IRC just after this meeting ? | 14:42 |
zigo | After that, I'm ok to use Storyboard. | 14:43 |
rafaelweingartne | zigo it will be lunch time here. I am already late for it :) | 14:44 |
zigo | :) | 14:44 |
zigo | Nv mind then... | 14:44 |
rafaelweingartne | Moving on... | 14:45 |
rafaelweingartne | #topic AOB | 14:45 |
*** openstack changes topic to "AOB (Meeting topic: cloudkitty)" | 14:45 | |
rafaelweingartne | Now, I open for general questions and topics that people might have. | 14:45 |
mkarpiarz | Nothing on my side. | 14:45 |
mkarpiarz | I'll see if I can reproduce zigo's problem in my test environment. | 14:47 |
priteau | I've tried to reproduce the metadata issues I saw with Monasca, but it works in DevStack | 14:48 |
priteau | Maybe it was just a wrong configuration | 14:48 |
rafaelweingartne | mkarpiarz: I am suspecting a mismatch between the aggregation method or time frame used in CloudKitty and the one actually used in Gnocchi | 14:48 |
mkarpiarz | zigo: Could you paste your metrics.yaml file under the Story? | 14:48 |
rafaelweingartne | Anyways. Thank you guys for participating. Have a nice week. | 14:49 |
rafaelweingartne | #endmeeting | 14:49 |
*** openstack changes topic to "Rating component for OpenStack https://wiki.openstack.org/wiki/CloudKitty" | 14:49 | |
openstack | Meeting ended Mon Feb 22 14:49:28 2021 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 14:49 |
openstack | Minutes: http://eavesdrop.openstack.org/meetings/cloudkitty/2021/cloudkitty.2021-02-22-14.00.html | 14:49 |
openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/cloudkitty/2021/cloudkitty.2021-02-22-14.00.txt | 14:49 |
openstack | Log: http://eavesdrop.openstack.org/meetings/cloudkitty/2021/cloudkitty.2021-02-22-14.00.log.html | 14:49 |
*** rafaelweingartne has quit IRC | 14:49 | |
zigo | mkarpiarz: As I wrote, I can't find the matching <resource-id> ... | 14:49 |
zigo | mkarpiarz: Can you help me finding it out ? | 14:52 |
mkarpiarz | I don't know where this file is in Devstack but you should see a file which looks like this one: | 14:52 |
mkarpiarz | https://github.com/mkarpiarz/cloudkitty-playground/blob/gnocchi-collector/kolla/config/cloudkitty/metrics.yml | 14:52 |
priteau | zigo: if you install at the usual locations, it should be in /etc/cloudkitty | 14:53 |
zigo | I do have a /etc/cloudkitty/metrics.yml | 14:53 |
zigo | Then what ? | 14:53 |
zigo | FYI, that's the "normal" OpenStack Victoria installation that I have (the same thing which I all uploaded to Debian Sid and that we will release with the already frozen Debian 11 aka Bullseye). | 14:55 |
mkarpiarz | The one you see in that github repo works for me so if you share yours metrics.yaml file, I can try it in my test environment. | 14:55 |
zigo | http://paste.openstack.org/show/802886/ | 14:57 |
zigo | That's what I have. | 14:57 |
mkarpiarz | Better to post the content of this file as a comment in Storyboard. | 14:59 |
mkarpiarz | (IRC's not idea for messages with multiple lines) | 15:00 |
zigo | Well, the Debian package which I maintain just takes it from etc/cloudkitty/metrics.yml from the original Cloudkitty sources. | 15:01 |
mkarpiarz | Ah, OK, this simplifies things. I'll grab it from there then. | 15:01 |
zigo | What the Debian package does is (from debian/rules): | 15:02 |
zigo | for i in api_paste.ini metrics.yml ; do \ | 15:02 |
zigo | install -D -m 0664 etc/cloudkitty/$$i $(CURDIR)/debian/cloudkitty-common/usr/share/cloudkitty-common/$$i ; \ | 15:02 |
zigo | done | 15:02 |
zigo | then in the postinst, it takes files in /usr/share/cloudkitty-common and installs them in /etc/cloudkitty if none already exist there. | 15:02 |
zigo | Then chown them to root:cloudkitty with mode 640... | 15:03 |
zigo | Just as most OpenStack config files. | 15:03 |
zigo | I do that to make sure all is owned in a safe way that even the service itself can't write to. | 15:03 |
zigo | That's common security practice, I believe... | 15:03 |
mkarpiarz | Yeah, this should work. | 15:04 |
mkarpiarz | As for getting that resource_id, I'd say start with `openstack metric list`. | 15:05 |
mkarpiarz | This will show you all metrics stored in Gnocchi. | 15:06 |
MrClayPole | Hi, just getting started with CloudKitty. I've got it charging for volume sizes and all instances per project but its just showing total billing of each in horizon. We would like to break it out so we can see how much each instance is costing the customer for detailed billing reporting. Can this be done it CloudKitty? | 15:08 |
zigo | mkarpiarz: It takes forever, and I have at least hundreds of them. How do I filter ? | 15:10 |
zigo | mkarpiarz: I'm only interested to find out the metric that makes the processor fails ... | 15:10 |
zigo | This isn't devstack ... :) | 15:10 |
priteau | MrClayPole: You can extract "dataframes" via the CLI which will show per-resource and per-time period charging | 15:12 |
MrClayPole | Thanks I'll take a look | 15:15 |
openstackgerrit | Merged openstack/cloudkitty master: remove unicode from code https://review.opendev.org/c/openstack/cloudkitty/+/770236 | 15:54 |
openstackgerrit | Merged openstack/cloudkitty master: Remove six https://review.opendev.org/c/openstack/cloudkitty/+/768506 | 15:54 |
zigo | mkarpiarz: priteau: What can I do then to find out what metric is wrong? | 16:12 |
*** vishalmanchanda has quit IRC | 16:34 | |
*** sboyron has joined #cloudkitty | 18:33 | |
*** sboyron has quit IRC | 18:37 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!