opendevreview | Miguel Lavalle proposed openstack/neutron stable/wallaby: Add rate-limiting to metadata agents https://review.opendev.org/c/openstack/neutron/+/892754 | 02:01 |
---|---|---|
opendevreview | Miguel Lavalle proposed openstack/neutron stable/wallaby: Add rate-limiting to metadata agents https://review.opendev.org/c/openstack/neutron/+/892754 | 05:06 |
opendevreview | morice proposed openstack/neutron master: [ovn] AZs distribution in L3 port scheduler https://review.opendev.org/c/openstack/neutron/+/892604 | 06:40 |
slaweq | ykarel lajoskatona ralonsoh haleyb mlavalle hi, I think that patch https://review.opendev.org/c/openstack/neutron/+/891580 is ready to review. I checked logs from it and it makes big difference in the sql queries we are doing in Neutron. Please check it when You will have few minutes. | 07:52 |
ralonsoh | slaweq, sure, let me check | 07:53 |
ralonsoh | ah, this is to reduce the number of queries, right | 07:53 |
ralonsoh | slaweq, qq, why not using PS1 + the additional change in the api_collections? | 07:56 |
slaweq | ralonsoh I don't remember exactly now but there were some AttributeError exceptions raised there | 08:07 |
slaweq | and it wasn't working at all | 08:07 |
ralonsoh | ok, we can just change the loading method, that's the minimum change needed | 08:08 |
slaweq | yes | 08:08 |
lajoskatona | slaweq: checking | 08:10 |
ralonsoh | frickler, the issue in ndr will be solved once we have a new Neutron release, that will have https://review.opendev.org/c/openstack/neutron/+/886213 | 08:15 |
opendevreview | Lajos Katona proposed openstack/neutron-lib master: WIP: Tap mirror API def and API ref https://review.opendev.org/c/openstack/neutron-lib/+/885354 | 08:19 |
frickler | ralonsoh: hmm, but that should affect only the job running with sqla 2.0/master, not the normal unit test jobs? also I don't understand why it would start failing only now | 08:22 |
frickler | also I assumed that ndr master unit tests would also test neutron master | 08:22 |
ralonsoh | frickler, I'm going to limit temporarily the n-lib version | 08:22 |
ralonsoh | until we have a new neutron release | 08:23 |
ralonsoh | then I'll remove the limitation and bump the n-lib version | 08:23 |
ralonsoh | and bump the neutron version* | 08:23 |
ralonsoh | --> https://bugs.launchpad.net/neutron/+bug/2033493 | 08:23 |
frickler | ah, release mismatch, ok | 08:34 |
ykarel | slaweq, ack | 08:34 |
ykarel | slaweq, lgtm, just curios how much queries got reduced with that patch | 08:36 |
ralonsoh | that was the most recurrent query. In the logs provided by QA, from 750K queries, that was around 150K (if I recall correctly) | 08:37 |
ralonsoh | 750K queries only from Neutron | 08:37 |
ralonsoh | frickler, I can't force n-lib 3.7.0 because tox is claiming that the required version is 3.8.0 | 08:39 |
ralonsoh | but this is not true, this is the upper version | 08:39 |
lajoskatona | frickler, ralonsoh: the same is true for other stadiums as we discussed yesterday during the CI meeting, And we can actually make these jobs to have Neutron also in the required-project list | 08:40 |
ralonsoh | lajoskatona, the point is that we need a new Neutron version | 08:40 |
ralonsoh | we'll have it during the next week | 08:40 |
lajoskatona | ralonsoh: yes, that solves the issue (at least what I checked, and hope there will be no more :P) | 08:41 |
ralonsoh | ok, I'm going to send a new beta version today | 08:46 |
ralonsoh | I'll send you the releases patch | 08:46 |
lajoskatona | ralonsoh:thanks | 08:48 |
slaweq | ykarel I wrote it in the comment in gerrit: "from about 130k to less than 1k." | 08:54 |
slaweq | of that one type of SELECT query | 08:54 |
ralonsoh | lajoskatona, frickler slaweq https://review.opendev.org/c/openstack/releases/+/893138 | 08:54 |
ykarel | Thanks slaweq i see it now after expanding the comments | 08:56 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Populate the "router.distributed" flag in ML2/OVN https://review.opendev.org/c/openstack/neutron/+/886992 | 09:53 |
ralonsoh | slaweq, https://review.opendev.org/c/openstack/releases/+/893138/comments/1e0e8e3b_ca774a84 | 09:54 |
ralonsoh | please consider this comment | 09:55 |
ralonsoh | slaweq, I had some comments in https://review.opendev.org/c/openstack/neutron/+/884474/14 not commented/considered | 09:57 |
slaweq | ralonsoh done, sorry for the noise | 10:02 |
ralonsoh | thanks! | 10:02 |
slaweq | and I'm just now addressing Your comments in https://review.opendev.org/c/openstack/neutron/+/884474/14 | 10:02 |
slaweq | please give me 10 minutes and new PS should be send | 10:02 |
opendevreview | Slawek Kaplonski proposed openstack/neutron master: Default SG rules - use new rules templates to create rules for SGs https://review.opendev.org/c/openstack/neutron/+/884474 | 10:06 |
opendevreview | Elvira GarcĂa Ruiz proposed openstack/neutron master: Add ML2/OVN security group logging documentation https://review.opendev.org/c/openstack/neutron/+/890151 | 10:08 |
slaweq | lajoskatona ralonsoh ykarel I have question regarding https://review.opendev.org/c/openstack/neutron/+/893097/ and comment made there by lajoskatona | 10:10 |
slaweq | I know that it works much slower with db migration before every test | 10:10 |
slaweq | but without that it behaves wrong with patch https://review.opendev.org/c/openstack/neutron/+/884474 and fullstack job is broken, even if I try to just run db migration for some of the tests, not for all of them | 10:11 |
slaweq | it's causing failures in some other tests then | 10:11 |
slaweq | so I see 2 possible solutions: | 10:11 |
slaweq | 1. increase timeout of the fullstack job to 3h (it's 2.5h now) and accept that job will be much slower | 10:12 |
slaweq | or | 10:12 |
slaweq | 2. monkeypatch get_default_security_group_rules method in the fullstack tests, before neutron server will start to return always hardcoded SG rules there. That way we can simply keep fullstack tests working as it works now | 10:13 |
slaweq | Wdyt? Which one would You prefer? | 10:13 |
ralonsoh | I prefer using a new DB per test, that would skip potential errors | 10:14 |
ralonsoh | but eventually we'll need to remove/refactor tests to lower the time spent in this job | 10:14 |
ralonsoh | 3 hours is too much | 10:14 |
slaweq | also, looking at https://zuul.opendev.org/t/openstack/builds?job_name=neutron-fullstack-with-uwsgi&branch=master&skip=0 it seems that this job is already running for about 2h in most cases | 10:14 |
slaweq | so my change don't increase it a lot | 10:14 |
slaweq | it seems that we are almost on the 2.5 limit already in many cases | 10:15 |
ralonsoh | yes, 2:10, 2:20, more or less | 10:15 |
slaweq | with that (I just checked it) I would prefer to change timeout to 3h and go with currently proposed patch - is that ok for You? | 10:15 |
ralonsoh | I would prefer, as commented, to have a fresh new DB per test and increase, for now, the job time | 10:16 |
ralonsoh | yes | 10:16 |
slaweq | ++ | 10:16 |
slaweq | thx | 10:16 |
opendevreview | Slawek Kaplonski proposed openstack/neutron master: [Fullstack] Use new DB for each running test https://review.opendev.org/c/openstack/neutron/+/893097 | 10:18 |
opendevreview | Slawek Kaplonski proposed openstack/neutron master: Default SG rules - use new rules templates to create rules for SGs https://review.opendev.org/c/openstack/neutron/+/884474 | 10:18 |
opendevreview | Slawek Kaplonski proposed openstack/neutron master: Default SG rules template - Update related docs and add release note https://review.opendev.org/c/openstack/neutron/+/887664 | 10:18 |
slaweq | done ^^ | 10:19 |
slaweq | I also addressed Your comments ralonsoh | 10:19 |
ralonsoh | thanks! I'll check i tnow | 10:19 |
slaweq | please review that once again when You will have few minutes | 10:19 |
opendevreview | Slawek Kaplonski proposed openstack/neutron master: DNM just test for the neutron-fullstack tests https://review.opendev.org/c/openstack/neutron/+/890189 | 10:21 |
opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/2023.1: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893140 | 10:33 |
opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/zed: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893141 | 10:33 |
opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/yoga: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893142 | 10:34 |
opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/xena: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893143 | 10:34 |
opendevreview | Lucas Alvares Gomes proposed openstack/neutron stable/wallaby: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893144 | 10:34 |
ykarel | slaweq, ack will check | 10:54 |
ykarel | hmm so if the stability of job do not get worse should be fine to go this route | 10:55 |
opendevreview | Merged openstack/neutron master: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893030 | 11:12 |
opendevreview | Lajos Katona proposed openstack/neutron-tempest-plugin master: WIP: Tap Mirror API tests https://review.opendev.org/c/openstack/neutron-tempest-plugin/+/886004 | 11:45 |
opendevreview | Lajos Katona proposed openstack/tap-as-a-service master: WIP: CLI for Tap Mirrors https://review.opendev.org/c/openstack/tap-as-a-service/+/886085 | 11:45 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/2023.1: [FT] Make explicit the "publish" call check in "test_port_forwarding" https://review.opendev.org/c/openstack/neutron/+/893168 | 12:45 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/zed: [FT] Make explicit the "publish" call check in "test_port_forwarding" https://review.opendev.org/c/openstack/neutron/+/893169 | 12:45 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/yoga: [FT] Make explicit the "publish" call check in "test_port_forwarding" https://review.opendev.org/c/openstack/neutron/+/893170 | 12:45 |
opendevreview | Rodolfo Alonso proposed openstack/neutron stable/xena: [FT] Make explicit the "publish" call check in "test_port_forwarding" https://review.opendev.org/c/openstack/neutron/+/893171 | 12:46 |
ralonsoh | folks, please check the bobcat cycle highlights: https://review.opendev.org/c/openstack/releases/+/893174 | 12:54 |
ralonsoh | ^^^ ykarel lajoskatona slaweq mlavalle obondarev_ amotoki frickler | 12:55 |
ralonsoh | I'll add you to the review | 12:55 |
ralonsoh | thanks in advance | 12:55 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: [OVN] Populate the "router.distributed" flag in ML2/OVN https://review.opendev.org/c/openstack/neutron/+/886992 | 12:58 |
lajoskatona | ralonsoh: thanks, I read it | 12:59 |
fnordahl | lajoskatona: hello, about your comment on 893023, it was my impression that feature freeze is at the end of this week? It's part of the greater effort of active active l3 gw which is now basically missing docs to be complete | 13:03 |
ralonsoh | fnordahl, only docs? there are code patches still pending, if I'm not wrong | 13:07 |
ralonsoh | https://review.opendev.org/q/topic:2023-aa-l3-gw-multihoming+status:open | 13:07 |
fnordahl | We've been awaiting reviews for a while, yes. Are you saying that you'd rather revert the whole thing than help with reviewing it? | 13:08 |
ralonsoh | no, what I'm saying is that the current code that is merged is not active | 13:08 |
ralonsoh | and we should not continue merging these patches at least until C release | 13:08 |
ralonsoh | this feature is extensive in code and complex | 13:09 |
ralonsoh | and has not been tested | 13:09 |
ralonsoh | what we should revert is the two release notes referring to this LP | 13:09 |
fnordahl | Ok :( I understand your decision, it would be riskful to rush this now. Pleased to hear that the code will be intact allow continuing the rest of the work. Caracal it is then, onward and upward! | 13:13 |
fnordahl | I guess I can raise he reverts for the release note pieces *goes to look where they are* | 13:14 |
ralonsoh | only the release notes, nothing else | 13:17 |
ralonsoh | fnordahl, https://031729a193aba78a0197-13270c601ca2148a93462eeebb4d313b.ssl.cf1.rackcdn.com/886992/7/check/build-openstack-releasenotes/070e7cb/docs/unreleased.html | 13:17 |
ralonsoh | 1) Add enable_default_route_bfd.... | 13:17 |
ralonsoh | 2) Added support for the external-gateway-multihoming.... | 13:18 |
lajoskatona | fnordahl, ralonsoh: sorry for -1 on the patch, but from yesterdays meeting (https://meetings.opendev.org/meetings/networking/2023/networking.2023-08-29-14.00.log.html#l-19 and see the on demand at the end of meeting) that we wait with new features now | 13:19 |
opendevreview | Merged openstack/neutron master: Use HasStandardAttributes as parent class for Tags DB model https://review.opendev.org/c/openstack/neutron/+/891580 | 13:34 |
opendevreview | Frode Nordahl proposed openstack/neutron-lib master: Drop release note for l3-ext-gw-multihoming and adjacent features https://review.opendev.org/c/openstack/neutron-lib/+/893181 | 13:34 |
opendevreview | Frode Nordahl proposed openstack/neutron-lib master: Add release note for l3-ext-gw-multihoming and adjacent features https://review.opendev.org/c/openstack/neutron-lib/+/893182 | 13:34 |
opendevreview | Frode Nordahl proposed openstack/neutron master: Drop release notes for l3-ext-gw-multihoming and adjacent features https://review.opendev.org/c/openstack/neutron/+/893184 | 13:35 |
opendevreview | Frode Nordahl proposed openstack/neutron master: Add release notes for l3-ext-gw-multihoming and adjacent features https://review.opendev.org/c/openstack/neutron/+/893185 | 13:35 |
fnordahl | ralonsoh, lajoskatona: https://review.opendev.org/c/openstack/neutron-lib/+/893181 and https://review.opendev.org/c/openstack/neutron/+/893184 | 13:38 |
ralonsoh | fnordahl, thanks | 13:41 |
ralonsoh | fnordahl, actually now I think about it, the n-lib won't be needed anymore | 13:48 |
ralonsoh | we have already released the last B version | 13:48 |
fnordahl | ok, I'll abandon 893181 | 13:49 |
fnordahl | ah you did, thx | 13:50 |
opendevreview | Rodolfo Alonso proposed openstack/neutron-dynamic-routing master: Bump Neutron to 23.0.0.0b3 https://review.opendev.org/c/openstack/neutron-dynamic-routing/+/893189 | 13:55 |
opendevreview | Joel Capitao proposed openstack/neutron-dynamic-routing master: DNM Test with latest neutron https://review.opendev.org/c/openstack/neutron-dynamic-routing/+/893192 | 14:06 |
opendevreview | Rodolfo Alonso proposed openstack/neutron master: [UT] Reduce the binding retry loop in ``L3HATestFramework`` https://review.opendev.org/c/openstack/neutron/+/889071 | 15:14 |
ihrachys | haleyb do we close https://bugs.launchpad.net/neutron/+bug/2031087 now? | 15:43 |
ralonsoh | slaweq, we have a problem with the fullstack job and the DB cleanup. The job is taking more than 3 hours, regardless of the host | 16:15 |
ralonsoh | we'll need to check that tomorrow morning | 16:15 |
*** ralonsoh is now known as ralonsoh_ooo | 16:20 | |
opendevreview | Merged openstack/neutron master: [Fullstack] Use new DB for each running test https://review.opendev.org/c/openstack/neutron/+/893097 | 17:28 |
opendevreview | Merged openstack/neutron master: Drop release notes for l3-ext-gw-multihoming and adjacent features https://review.opendev.org/c/openstack/neutron/+/893184 | 17:28 |
haleyb | ihrachys: why don't we keep it open until your patch merges (thanks for that!) and we can list OVN versions that have it | 17:50 |
ihrachys | haleyb sure. the lp is yours, just checking | 19:24 |
opendevreview | Alban PRATS proposed openstack/neutron master: Enabling routing of routed subnets through snat. https://review.opendev.org/c/openstack/neutron/+/890459 | 19:50 |
opendevreview | Merged openstack/neutron stable/2023.1: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893140 | 22:04 |
opendevreview | Merged openstack/neutron stable/zed: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893141 | 22:05 |
opendevreview | Merged openstack/neutron stable/yoga: [OVN] Cleanup old Hash Ring node entries https://review.opendev.org/c/openstack/neutron/+/893142 | 23:51 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!