*** hoonetorg has quit IRC | 00:02 | |
*** hoonetorg has joined #openstack-trove | 00:18 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/trove master: Updated from global requirements https://review.openstack.org/517505 | 00:36 |
---|---|---|
openstackgerrit | Yao Wang proposed openstack/trove master: Stop polling if nova instances goto error https://review.openstack.org/518639 | 01:42 |
openstackgerrit | chengyang proposed openstack/trove stable/pike: Fix mysql instance create failed when enable skip-name-resolve https://review.openstack.org/518642 | 02:27 |
*** miqui has quit IRC | 02:30 | |
*** chason has joined #openstack-trove | 02:39 | |
*** Jeffrey4l has quit IRC | 02:50 | |
*** daidv has quit IRC | 02:51 | |
*** maciejjozefczyk has quit IRC | 02:51 | |
*** hoonetorg has quit IRC | 02:51 | |
*** lifeless has quit IRC | 02:51 | |
*** mmasaki has quit IRC | 02:51 | |
*** mmasaki has joined #openstack-trove | 02:54 | |
*** daidv has joined #openstack-trove | 02:57 | |
*** lifeless has joined #openstack-trove | 02:58 | |
openstackgerrit | Fan Zhang proposed openstack/trove master: Support negative number as unlimited quota restraint in Trove. https://review.openstack.org/517569 | 02:58 |
*** hoonetorg has joined #openstack-trove | 03:04 | |
*** maciejjozefczyk has joined #openstack-trove | 03:04 | |
*** Jeffrey4l has joined #openstack-trove | 03:05 | |
*** Jeffrey4l has quit IRC | 03:17 | |
*** Jeffrey4l has joined #openstack-trove | 03:24 | |
*** links has joined #openstack-trove | 03:34 | |
openstackgerrit | Merged openstack/trove master: Support insecure SSL when talking to services https://review.openstack.org/410348 | 05:49 |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/trove master: Lazy load all configuration options https://review.openstack.org/466923 | 05:52 |
*** wong has quit IRC | 06:16 | |
*** wong has joined #openstack-trove | 06:20 | |
*** links has quit IRC | 06:22 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/trove-dashboard stable/pike: Imported Translations from Zanata https://review.openstack.org/493761 | 06:25 |
openstackgerrit | OpenStack Proposal Bot proposed openstack/trove-dashboard master: Imported Translations from Zanata https://review.openstack.org/518241 | 06:28 |
*** links has joined #openstack-trove | 06:28 | |
openstackgerrit | zhanggang proposed openstack/trove master: Missing import of 'assert_equal' in tests/util/__init__.py https://review.openstack.org/455549 | 06:29 |
*** hoonetorg has quit IRC | 06:48 | |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/trove master: Lazy load all configuration options https://review.openstack.org/466923 | 06:57 |
*** eanderson_ has joined #openstack-trove | 06:58 | |
*** hoonetorg has joined #openstack-trove | 07:05 | |
*** spectr has joined #openstack-trove | 07:06 | |
*** rpittau_ has joined #openstack-trove | 07:07 | |
*** maciejjozefczyk has quit IRC | 07:10 | |
*** rpittau has quit IRC | 07:11 | |
openstackgerrit | Siyi Luo proposed openstack/python-troveclient master: Update the documentation link for doc migration https://review.openstack.org/506428 | 07:14 |
chason | Hi guys, what do you think about this patch? | 07:16 |
chason | https://review.openstack.org/#/c/511664/ | 07:16 |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/trove master: Lazy load all configuration options https://review.openstack.org/466923 | 07:21 |
*** rcernin has quit IRC | 07:45 | |
*** kumarmn has joined #openstack-trove | 07:45 | |
*** tesseract has joined #openstack-trove | 08:01 | |
openstackgerrit | Merged openstack/trove master: Open the volume_support of redis https://review.openstack.org/507428 | 08:08 |
openstackgerrit | Siyi Luo proposed openstack/python-troveclient master: Update the documentation link for doc migration https://review.openstack.org/506428 | 08:14 |
*** rcernin has joined #openstack-trove | 08:37 | |
*** eanderson_ has quit IRC | 09:00 | |
openstackgerrit | Tomasz Nowak proposed openstack/trove master: Configure guestagent on Ubuntu guest images to use CA certificates https://review.openstack.org/451311 | 09:29 |
*** maciejjozefczyk has joined #openstack-trove | 09:39 | |
*** kei_yama has quit IRC | 09:44 | |
*** tosky has joined #openstack-trove | 10:10 | |
*** links has quit IRC | 10:15 | |
openstackgerrit | Fan Zhang proposed openstack/trove master: Support -1 as unlimited quota restraint in Trove. https://review.openstack.org/517569 | 10:32 |
openstackgerrit | Yao Wang proposed openstack/trove master: Add timestamp to cluster instanceName https://review.openstack.org/518701 | 10:41 |
*** gouthamr has joined #openstack-trove | 10:52 | |
*** gouthamr has quit IRC | 10:56 | |
*** rcernin has quit IRC | 11:26 | |
*** smatzek has joined #openstack-trove | 11:51 | |
*** Guest34249 has joined #openstack-trove | 12:23 | |
*** Guest34249 has quit IRC | 12:40 | |
smatzek | chason: I know there are several open reviews for documentation changes. With the documentation moving to the repos it caused a lot of doc changes. I have not yet had time to go through the set of documentation reviews. If you are familiar with the documentation I would appreciate your reviews on the open doc changes. | 12:43 |
openstackgerrit | Merged openstack/trove-dashboard master: Switch from oslosphinx to openstackdocstheme https://review.openstack.org/498319 | 12:50 |
*** tianhui has quit IRC | 13:20 | |
openstackgerrit | Fan Zhang proposed openstack/trove master: Support -1 as unlimited quota restraint in Trove. https://review.openstack.org/517569 | 13:30 |
*** pmackinn has joined #openstack-trove | 13:33 | |
*** zhaochao has quit IRC | 13:42 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/trove master: Updated from global requirements https://review.openstack.org/517505 | 14:01 |
*** smatzek has quit IRC | 14:27 | |
openstackgerrit | Yao Wang proposed openstack/trove master: Stop polling if nova instances goto error https://review.openstack.org/518639 | 14:55 |
*** smatzek has joined #openstack-trove | 14:55 | |
*** smatzek_ has joined #openstack-trove | 14:57 | |
*** smatzek has quit IRC | 14:59 | |
openstackgerrit | Yao Wang proposed openstack/trove master: Add timestamp to cluster instance name https://review.openstack.org/518701 | 15:00 |
*** kylek3h has quit IRC | 15:18 | |
*** zhaochao has joined #openstack-trove | 15:30 | |
*** spectr has quit IRC | 15:35 | |
*** spectr has joined #openstack-trove | 15:37 | |
*** zhaochao has quit IRC | 15:37 | |
*** kylek3h has joined #openstack-trove | 15:44 | |
*** spectr has quit IRC | 16:00 | |
*** itlinux has joined #openstack-trove | 16:04 | |
*** itlinux has quit IRC | 16:23 | |
*** itlinux has joined #openstack-trove | 16:24 | |
*** tesseract has quit IRC | 16:27 | |
*** tosky has quit IRC | 18:03 | |
*** itlinux has quit IRC | 18:04 | |
*** itlinux has joined #openstack-trove | 18:05 | |
openstackgerrit | Merged openstack/trove master: Support -1 as unlimited quota restraint in Trove. https://review.openstack.org/517569 | 18:16 |
openstackgerrit | Merged openstack/trove master: Missing import of 'assert_equal' in tests/util/__init__.py https://review.openstack.org/455549 | 18:22 |
*** pmackinn has quit IRC | 18:26 | |
eandersson | smatzek_, do you want to go with context, *args, **kwargs for all of them? | 18:39 |
eandersson | or maybe just *args, **kwargs for all of them | 18:39 |
smatzek_ | eandersson: how about *args, **kwargs for all. That would protect us against future constructor changes. | 19:02 |
*** smatzek_ is now known as smatzek | 19:02 | |
eandersson | Yep | 19:03 |
eandersson | Sounds good | 19:04 |
openstackgerrit | Tomasz Nowak proposed openstack/trove master: Use existing CNF_INCLUDE_DIR to create mysql-flavor directory https://review.openstack.org/437369 | 19:04 |
*** pmackinn has joined #openstack-trove | 19:07 | |
openstackgerrit | Erik Olof Gunnar Andersson proposed openstack/trove master: Lazy load all configuration options https://review.openstack.org/466923 | 19:30 |
eandersson | smatzek, ^ | 19:33 |
smatzek | thanks | 19:33 |
smatzek | that LGTM but I'm going to wait to vote until Zuul votes | 19:35 |
eandersson | Sounds good | 19:38 |
eandersson | We could probably add some basic caching for these, so that we don't have to call import_class everytime | 19:39 |
eandersson | but reviewing the code I doubt this would have much impact, and I rather just get these fixed. | 19:39 |
eandersson | This patch should make Zuul a lot more stable. | 19:39 |
smatzek | I agree. I'd rather get it fixed and add performance fixes later. | 19:40 |
smatzek | Honestly, I haven't seen Zuul failing on timing issues related to this, at least not in the voting gates. | 19:40 |
eandersson | Well at the moment config options like agent_call_high_timeout aren't respected. | 19:40 |
smatzek | The big hitter is the KVM nested virtualizatoin failures that we're seeing on some of the hosts we get from infra. I'm very close to disabling that as running with straight 'qemu' is getting higher success rates. | 19:41 |
eandersson | I see | 19:41 |
smatzek | really? I saw a great improvement in the stability after I started setting agent_call_high_timeout higher, a commit back in October. | 19:41 |
smatzek | right now I see the most failures due to kvm virt failure (due to https://bugs.launchpad.net/ubuntu/+source/linux-lts-xenial/+bug/1682077), and secondly, from mysql stop taking more than the default 30 seconds to stop. | 19:42 |
openstack | Launchpad bug 1682077 in linux-lts-xenial (Ubuntu) "nested KVM fails - KVM: entry failed, hardware error 0x0 " [High,Confirmed] | 19:42 |
eandersson | It was months since I did the research, but https://bugs.launchpad.net/trove/+bug/1694538 | 19:43 |
openstack | Launchpad bug 1694538 in OpenStack DBaaS (Trove) "Non-lazy loaded configuration options depend on import order" [High,In progress] - Assigned to Erik Olof Gunnar Andersson (eandersson) | 19:43 |
smatzek | the 30 second stop timeout is more likely to happen when running with nested qemu virt as it runs slower. | 19:43 |
eandersson | the agent_call flags were some of the affected by the bug | 19:43 |
eandersson | but yea the nested KVM fails is obviously gonna have bigger impact! :p | 19:44 |
eandersson | https://review.openstack.org/#/c/515814/ ? | 19:45 |
smatzek | and the problem with that is it doesn't happen every time. Some hosts will due nested KVM just fine, some that don't support it use qemu, but some 'say' they can support it but then fail, likely on some kernel bug in one of the hosts in the nested chain | 19:45 |
eandersson | Interesting | 19:45 |
smatzek | yep, that's the one I've been doing rechecks on to get a feeling for success rate with just qemu. On last count it was 3 failures in 16, 18% failure rate, which isn't the greatest, but the nested KVM failure rate seems awefully high too. | 19:46 |
smatzek | I've been checking every -1 Zuul in the past week or so and most are due to nested kvm failure | 19:46 |
smatzek | agent_call_high_timeout was one of the things I had to set to even get the 're-enable integration tests' commit to merge. We were getting near 100% failure without it. https://github.com/openstack/trove/commit/d7b773d2b07f0c95909ef73c4644a63f09fd82c3 | 19:48 |
eandersson | Yea - that is the commit I saw earlier. Was gonna comment on it. | 19:48 |
eandersson | It's possible I mis-identified this as one of the affect variables of bug 1694538 | 19:48 |
openstack | bug 1694538 in OpenStack DBaaS (Trove) "Non-lazy loaded configuration options depend on import order" [High,In progress] https://launchpad.net/bugs/1694538 - Assigned to Erik Olof Gunnar Andersson (eandersson) | 19:48 |
smatzek | possible. Regardless, it will be good to get it fixed. | 19:49 |
eandersson | Yep agreed | 19:49 |
smatzek | so if we merge the change to go with only qemu nested virt I suspect the stop timeout issue will be the majority of the failures. This isn' | 19:51 |
smatzek | t currently configurable and I'm considering adding a CONF setting for it. http://logs.openstack.org/14/515814/1/check/legacy-trove-scenario-dsvm-mysql-single/ad7657a/logs/devstack-gate-post_test_hook.txt.gz#_2017-11-01_19_02_28_057 | 19:51 |
openstackgerrit | Samuel Matzek proposed openstack/trove-dashboard master: Fix unstable test case https://review.openstack.org/518786 | 19:54 |
*** itlinux has quit IRC | 20:16 | |
*** smatzek has quit IRC | 20:56 | |
*** smatzek has joined #openstack-trove | 20:58 | |
*** smatzek has quit IRC | 21:03 | |
*** jdavis has joined #openstack-trove | 21:08 | |
*** itlinux has joined #openstack-trove | 21:12 | |
*** itlinux has quit IRC | 21:29 | |
*** rcernin has joined #openstack-trove | 22:01 | |
*** itlinux has joined #openstack-trove | 22:07 | |
*** smatzek has joined #openstack-trove | 22:11 | |
*** smatzek has quit IRC | 22:11 | |
*** kylek3h has quit IRC | 22:39 | |
*** pmackinn has quit IRC | 23:09 | |
*** itlinux has quit IRC | 23:26 | |
*** jdavis has quit IRC | 23:38 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!