*** sapd1_x has joined #openstack-self-healing | 03:15 | |
*** sapd1_x has quit IRC | 04:08 | |
*** tojuvone has quit IRC | 04:31 | |
*** tojuvone has joined #openstack-self-healing | 04:31 | |
*** zhangchi has joined #openstack-self-healing | 08:47 | |
ricolin | o/ | 08:51 |
---|---|---|
aspiers | hi ricolin | 08:54 |
*** witek has joined #openstack-self-healing | 08:57 | |
ricolin | aspiers, howdy | 09:01 |
aspiers | hey ricolin :) how are you doing? | 09:01 |
aspiers | going to Shanghai? | 09:01 |
aspiers | I got my place confirmed | 09:02 |
ricolin | aspiers, I'm for sure:) | 09:02 |
aspiers | cool | 09:02 |
ricolin | aspiers, good to have you there | 09:02 |
aspiers | I can't remember where you live? | 09:02 |
ricolin | Taiwan/Taipei | 09:02 |
aspiers | ah ok | 09:03 |
aspiers | a lot closer to Shanghai than I am ;-) | 09:03 |
ricolin | aspiers, it's next door for me:) | 09:03 |
aspiers | yep :) | 09:03 |
aspiers | that must be a nice change for your OpenStack summits! | 09:03 |
aspiers | normally you have a huge trip | 09:03 |
ricolin | yeah, I think I will do a lot of haha to all the jetlag you guys have:) | 09:04 |
aspiers | hahah | 09:04 |
aspiers | for sure! | 09:04 |
aspiers | XD | 09:04 |
aspiers | like I do to the Americans when we have it in Europe XD | 09:04 |
ricolin | aspiers, totally | 09:05 |
aspiers | anyone else around for a quick self-healing meeting? witek / tojuvone maybe? | 09:05 |
ricolin | FYI, there's a wired situation for ceilometer metering I think not just for auto-scaling case, might also related to self-healing | 09:06 |
ricolin | http://lists.openstack.org/pipermail/openstack-discuss/2019-August/008216.html | 09:06 |
aspiers | looking | 09:06 |
aspiers | it seems like the issue is resolved? https://github.com/gnocchixyz/gnocchi/issues/1044 | 09:07 |
ricolin | I been told the real issue is not in gnocchi but aodh, hope that's also not a problem too | 09:09 |
* ricolin is searching for the irc log | 09:10 | |
ricolin | aspiers, http://eavesdrop.openstack.org/irclogs/%23heat/%23heat.2019-08-07.log.html#t2019-08-07T07:20:37 | 09:12 |
witek | hi aspiers and ricolin | 09:12 |
aspiers | hi witek! | 09:12 |
aspiers | did you guys see https://logs.opendev.org/38/675138/4/check/openstack-tox-docs/344de18/html/testing/tools-list.html ? | 09:12 |
ricolin | I did not have time to look into for detail here, so maybe not a problem at all | 09:12 |
* ricolin was referring the metering stuff | 09:13 | |
witek | aspiers: no, thanks for reference | 09:15 |
aspiers | witek: https://review.opendev.org/#/c/675138/ | 09:16 |
aspiers | witek: sorry I forgot to add you before | 09:16 |
witek | added myself now | 09:17 |
ricolin | aspiers, what exactly is that table referring? tests each tool can perform? | 09:19 |
aspiers | not just tests | 09:20 |
aspiers | functionality | 09:20 |
ricolin | aspiers, right! functionality! also just leave some comment in line | 09:25 |
aspiers | thanks | 09:25 |
aspiers | ricolin: yes, looks like you agree with my last comment | 09:26 |
ricolin | great works! | 09:26 |
ricolin | d | 09:26 |
ricolin | d'oh! | 09:26 |
aspiers | I only just posted it a few minutes ago :) | 09:27 |
ricolin | WILL FASTER THAN YOU NEXT TIME...(murmuring) | 09:29 |
aspiers | hahaha :) | 09:30 |
ricolin | witek, do you happen to know about the answer about http://lists.openstack.org/pipermail/openstack-discuss/2019-August/008216.html :) | 09:32 |
witek | no, I don't, I just followed the discussion | 09:37 |
* ricolin also post this ML in #openstack-auto-scaling | 09:40 | |
*** tojuvone has quit IRC | 10:01 | |
*** tojuvone has joined #openstack-self-healing | 10:01 | |
*** zhangchi has quit IRC | 10:10 | |
*** ricolin has quit IRC | 16:29 | |
*** ihti has joined #openstack-self-healing | 16:32 | |
*** ekcs has joined #openstack-self-healing | 16:57 | |
*** witek has quit IRC | 17:15 | |
aspiers | oops, lost track of time | 17:17 |
aspiers | hi ekcs | 17:17 |
*** ekcs has quit IRC | 17:18 | |
*** ekcs has joined #openstack-self-healing | 17:19 | |
aspiers | hi ekcs | 17:21 |
ekcs | hi aspiers ! | 17:25 |
ekcs | how’s it been? | 17:25 |
aspiers | busy :) | 17:25 |
aspiers | you? | 17:26 |
aspiers | I got approved for Shanghai btw | 17:26 |
aspiers | Did you see https://review.opendev.org/#/c/675138/ ? | 17:26 |
aspiers | I think we need to resurrect the plan to try and make something happen with testing | 17:27 |
ekcs | oh very good! | 17:27 |
ekcs | ah yes I skimmed it but havevn’t done a review. | 17:27 |
ihti | hi | 17:28 |
aspiers | hi ihti | 17:29 |
aspiers | ekcs: ihti is the author of that review | 17:29 |
ekcs | hi ihti ! | 17:29 |
aspiers | ihti: ekcs is co-chair of the SIG | 17:29 |
ekcs | the patch looks like a very helpful reference. | 17:30 |
ihti | aspiers, ah okay | 17:30 |
aspiers | 3 counts as critical mass, so ... :) | 17:30 |
aspiers | #startmeeting self-healing | 17:30 |
openstack | Meeting started Wed Aug 14 17:30:35 2019 UTC and is due to finish in 60 minutes. The chair is aspiers. Information about MeetBot at http://wiki.debian.org/MeetBot. | 17:30 |
openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 17:30 |
*** openstack changes topic to " (Meeting topic: self-healing)" | 17:30 | |
openstack | The meeting name has been set to 'self_healing' | 17:30 |
aspiers | #topic automated testing | 17:30 |
*** openstack changes topic to "automated testing (Meeting topic: self-healing)" | 17:30 | |
aspiers | ihti: anything you want to bring up? | 17:30 |
aspiers | for the minutes ... | 17:31 |
aspiers | #link https://review.opendev.org/#/c/675138/ adds a list of testing frameworks to the SIG repo | 17:31 |
ihti | not much from my side yet, playing around with Rally on a dev setup | 17:31 |
aspiers | cool | 17:31 |
aspiers | the matrix clearly shows an awful lot of overlap | 17:31 |
aspiers | it occurred to me that we should try to get all the maintainers of the various tools to review that matrix | 17:32 |
ihti | yes, it have, rally covering most of the features | 17:32 |
aspiers | as they may be able to offer extra info | 17:32 |
ihti | yes, that would be very helpful with and would make the whole list more accurate | 17:33 |
aspiers | #link https://logs.opendev.org/38/675138/4/check/openstack-tox-docs/344de18/html/testing/tools-list.html Zuul build of the new doc | 17:33 |
aspiers | Yardstick + Doctor also seems to cover everything | 17:33 |
ihti | +NFV bench | 17:34 |
aspiers | ihti: Oh I just noticed, Eris is not in there yet | 17:34 |
aspiers | I guess that's intentional? | 17:34 |
ihti | yes | 17:34 |
ihti | as it will be worked on so i thought not include it yet? | 17:34 |
ihti | not sure | 17:35 |
aspiers | I think it would be better to include it, but add a new column "maturity" | 17:35 |
aspiers | e.g. unfinished/alpha/beta/immature/mature | 17:35 |
ihti | that's good idea. | 17:36 |
aspiers | also maybe another column for "OpenStack-specific" | 17:36 |
aspiers | since tools like Jepsen are agnostic | 17:36 |
aspiers | and there is no column for performance testing, unless that is under "Scalability testing"? | 17:37 |
aspiers | but I think they are different things | 17:37 |
aspiers | so separate columns would be useful | 17:37 |
ihti | 'Openstack-specific' one I added but then removed it afterwards, as I jepsen was the only one falling in that category | 17:37 |
aspiers | Hmm | 17:37 |
aspiers | OK, so maybe change the Jepsen cell to read "Jepsen (not specific to OpenStack)" | 17:38 |
aspiers | Another suggestion: add a separate column for reliability testing | 17:39 |
ihti | For performance I thought it overlaps with 'data plane testing'? I could be off | 17:39 |
aspiers | since the ability to inject disruptions / failures is separate from the ability to test resilience against them | 17:39 |
aspiers | I think we need performance testing for control plane operations oto | 17:39 |
aspiers | too | 17:39 |
ihti | okay I will mention that for jepsen | 17:40 |
aspiers | so no, I would consider performance testing somewhat separate to data plane | 17:40 |
aspiers | control plane operations can be slow | 17:40 |
ihti | alright | 17:40 |
aspiers | especially things like failover | 17:40 |
aspiers | Sorry, I am just thinking of this new feedback now :) | 17:41 |
ihti | no worries, I appreciate it :) | 17:41 |
aspiers | Another column we need is around the level of community activity | 17:41 |
aspiers | Some of these tools are dead | 17:41 |
aspiers | That is an important datapoint | 17:41 |
ihti | alright, yes that's true | 17:41 |
aspiers | Other tools are not dead, but only have a few contributors | 17:42 |
ihti | not update for a couple of years | 17:42 |
aspiers | so a column "size of active community" would help | 17:42 |
aspiers | but maybe it's not just size, but also level of activity | 17:42 |
aspiers | e.g. 2 contributors could do total of 500 commits in a month, or 20 contributors could do total of 10 commits in a month ... | 17:43 |
aspiers | These are important considerations when designing a full solution | 17:43 |
ihti | that makes sense and can be easily gathered by having a look at repos | 17:45 |
ekcs | yea maybe instead of putting up a static “grade” we can just link to the repo activities page. | 17:45 |
aspiers | well it's useful to be able to compare them all at a glance | 17:47 |
aspiers | that's kind of the point of this matrix | 17:47 |
aspiers | any cell could change, but that's OK :) | 17:47 |
ekcs | yea that makes sense too. | 17:47 |
ihti | I can add both a link and value | 17:48 |
aspiers | perfect! | 17:48 |
ihti | to each sell | 17:48 |
ihti | *cell | 17:48 |
aspiers | that would be great | 17:48 |
aspiers | sorry, so it seems I asked for like 59 new columns :) | 17:48 |
ihti | I think its good and makes the decision making process much easier | 17:49 |
aspiers | yes | 17:49 |
aspiers | maybe once you've updated the matrix we can try to contact the communities of the various tools for feedback | 17:49 |
aspiers | some of them are already cc'd on the review | 17:50 |
aspiers | but we'll need to find others probably | 17:50 |
aspiers | ihti, ekcs: OK anything else on this topic? | 17:50 |
ihti | cool | 17:50 |
ekcs | great | 17:50 |
ihti | no | 17:50 |
ihti | no topic from my side | 17:51 |
aspiers | so we can move on, or end the meeting | 17:51 |
aspiers | ekcs: anything to discuss about Shanghai at this point, or anything else? | 17:51 |
aspiers | ihti: thanks a LOT for all your great work so far! | 17:51 |
aspiers | it's hugely appreciated | 17:51 |
ekcs | not really. my focus now is just to find time to complete this task finally: https://storyboard.openstack.org/#!/story/2005641 | 17:51 |
ihti | thanks for the reviews and feedback :) | 17:52 |
ekcs | thank you ihti ! | 17:52 |
aspiers | ekcs: oh nice, that would be great!! | 17:52 |
aspiers | ihti: BTW soon we should email openstack-discuss announcing this new doc | 17:52 |
aspiers | maybe we can announce after merging | 17:52 |
aspiers | OK, sounds like we are done for today | 17:53 |
ihti | yes, after reviews by the tool devs is done, it would be good idea to spread it out | 17:53 |
aspiers | yup | 17:53 |
aspiers | it will help restart the discussion around how to fix this | 17:53 |
aspiers | ihti: don't hesitate to ask me if you have any questions or need support on anything | 17:54 |
aspiers | see you both online, bye for now! o/ | 17:55 |
ihti | aspiers, sure i will, thanks very much till now :) | 17:55 |
aspiers | :) | 17:55 |
ekcs | bye! | 17:55 |
aspiers | #endmeeting | 17:55 |
*** openstack changes topic to "https://wiki.openstack.org/wiki/Self_healing_SIG | https://storyboard.openstack.org/#!/project/openstack/self-healing-sig" | 17:55 | |
ihti | bye! :) | 17:55 |
openstack | Meeting ended Wed Aug 14 17:55:31 2019 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 17:55 |
openstack | Minutes: http://eavesdrop.openstack.org/meetings/self_healing/2019/self_healing.2019-08-14-17.30.html | 17:55 |
openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/self_healing/2019/self_healing.2019-08-14-17.30.txt | 17:55 |
openstack | Log: http://eavesdrop.openstack.org/meetings/self_healing/2019/self_healing.2019-08-14-17.30.log.html | 17:55 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!