Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2018-03-07
14:56:56 openstackgerrit Dan Smith proposed openstack/nova master: Add --purge helper flag to archive_deleted_rows https://review.openstack.org/550182
14:56:56 openstackgerrit Dan Smith proposed openstack/nova master: Add simple db purge command https://review.openstack.org/550171
14:56:56 openstackgerrit Dan Smith proposed openstack/nova master: Run post-test archive against cell1 https://review.openstack.org/550194
14:57:13 mriedem stephenfin: ceph doesn't work on py3 if you're taking notes
14:57:20 mriedem well, librados doesn't
14:57:24 kashyap stephenfin: I see; "theoretically complete" means, in practise it is not? :D
14:57:37 stephenfin kashyap: All unit and functional tests are passing with Python 3. However, I doubt anyone is running a Python 3 deployment
14:57:38 mriedem kashyap: nova works on py3
14:57:39 kashyap mriedem: How about eventlet?
14:57:41 stephenfin mriedem: It does now?
14:57:50 stephenfin I submitted patches saying as much a few months back
14:57:53 mriedem the dependent libraries, based on your configuration, might not work on py3
14:58:12 mriedem last i checked the py3 ceph job wasn't working, melwitt was investigating that
14:58:14 kashyap mriedem: I see, is there a test job with PY3?
14:58:21 mriedem yes
14:58:21 stephenfin mriedem: See commit 1e70cb8d4dace3c3372cd6e1bb13060becf25c6a
14:58:46 stephenfin It does require ceph v11.1.0 though
14:58:56 stephenfin ...based on my notes from that
14:58:57 mriedem stephenfin: might be that we don't have a new enough version of ceph in the UCA
14:59:14 mriedem https://bugs.launchpad.net/glance-store/+bug/1706405
14:59:15 stephenfin Without knowing anymore about it, That's probably it, aye
14:59:15 openstack Launchpad bug 1706405 in glance_store "ceph jobs failing to upload images in pike due to "AttributeError: 'NoneType' object has no attribute 'Rados'" on py35" [Undecided,Confirmed]
15:01:18 mriedem legacy-tempest-dsvm-py35-full-devstack-plugin-ceph is the job to look for
15:02:30 jaypipes dansmith: needs to be manually execute()d... :(
15:02:37 dansmith ah
15:02:47 mriedem http://logs.openstack.org/80/549880/1/experimental/legacy-tempest-dsvm-py35-full-devstack-plugin-ceph/7923b02/job-output.txt.gz#_2018-03-05_23_04_06_724259
15:02:48 jaypipes but like I said, femto-optiomization
15:02:56 mriedem looks like the ceph job is busted on the tempest regex change
15:03:02 mriedem mtreinish: ^ is there a template for how to fix that?
15:03:08 mriedem tempest run: error: argument --regex/-r: expected one argument
15:03:10 dansmith jaypipes: yeah, but seems worth it.. i
15:03:16 dansmith jaypipes: I'll tack it on the end
15:05:40 mtreinish mriedem: hmm, that looks like a job not setting a regex and it was expecting one
15:06:04 mtreinish mriedem: I have no idea how/where any of the job definitions are setup now, maybe andreaf can help
15:06:05 mriedem mtreinish: so likely this http://logs.openstack.org/80/549880/1/experimental/legacy-tempest-dsvm-py35-full-devstack-plugin-ceph/7923b02/job-output.txt.gz#_2018-03-05_22_47_04_710732
15:06:13 mriedem mtreinish: it's in openstack-zuul-jobs
15:06:37 mriedem https://github.com/openstack-infra/openstack-zuul-jobs/blob/master/playbooks/legacy/tempest-dsvm-py35-full-devstack-plugin-ceph/run.yaml#L61
15:06:43 mriedem ^ doesn't exist in nova,
15:06:47 mriedem i think jbernard was working on adding it
15:07:05 mtreinish that probably is the cause then
15:07:25 mtreinish especially if that's being expected to set the regex things run with
15:07:26 andreaf mriedem I defined a new style job for ceph some time ago but the patch has been waiting for reviews for long time https://review.openstack.org/#/c/543048/
15:07:29 mriedem heh https://review.openstack.org/#/c/459774/
15:07:47 mriedem andreaf: is that py3?
15:07:55 mriedem andreaf: i can't read the new zuulv3 voodoo
15:07:56 andreaf mriedem both I think
15:08:08 andreaf yeah both
15:09:07 andreaf mriedem it defines the job and also a job template that runs py27 on check and py3 on experimental
15:09:27 andreaf the project stanzas are pretty much the same as before
15:10:11 andreaf mriedem mtreinish anyways my point was rather than fixing the legacy job perhaps we should run the new style one
15:11:19 jaypipes dansmith: coolio.
15:11:42 mriedem andreaf: yeah maybe - you should have the same regex in the py3 job in there
15:11:52 mriedem andreaf: given https://review.openstack.org/#/c/459774/4/tempest-conf.sh
15:13:40 idlemind presenting svm (nested virtualization - amd) to guests. i've set my cpu_mode to host-passthrough. when i look at the /proc/cpuinfo i see my amd cpu details but it's missing the svm property ... i'm pretty sure i had this working at one point but i can't seem to think of anything else i did
15:14:08 andreaf mriedem the regex wasn't in the legacy py3 job - I was just migrating it not enhancing it
15:14:19 andreaf mriedem but ok I can add it in there
15:15:19 Spazmotic mriedem wanted to just thank you again for your testing ideas. Was great getting more experience with Mock after wanting to pull my hair out for a couple of hours :)
15:17:14 mriedem \o/
15:17:14 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add abstract OVSDB API https://review.openstack.org/476612
15:17:40 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Save admin password to sysmeta in libvirt driver https://review.openstack.org/550498
15:18:21 mriedem andreaf: yeah i think the thing was jbernard was trying to slim down the set of tests that get run in the ceph job to stabilize it
15:18:30 mriedem andreaf: and then start expanding the test bucket
15:19:29 mriedem although the job should probably be running the scenario tests in serial
15:19:42 openstackgerrit Rodolfo Alonso Hernandez proposed openstack/os-vif master: Add native implementation OVSDB API https://review.openstack.org/482226
15:22:10 mriedem stephenfin: http://logs.openstack.org/80/549880/1/experimental/legacy-tempest-dsvm-py35-full-devstack-plugin-ceph/7923b02/logs/dpkg-l.txt.gz
15:22:11 mriedem ii ceph 12.2.1-0ubuntu0.17.10.1~cloud0
15:22:18 mriedem so we should be ok version wise
15:23:11 openstackgerrit Dan Smith proposed openstack/nova master: Add --purge helper flag to archive_deleted_rows https://review.openstack.org/550182
15:23:12 openstackgerrit Dan Smith proposed openstack/nova master: Make nova-manage db purge take --all-cells https://review.openstack.org/550502
15:24:32 stephenfin mriedem: You mean we should be OK now? Looks like we were using an older one when that bug was reported https://bugs.launchpad.net/glance-store/+bug/1706405/comments/3
15:24:33 openstack Launchpad bug 1706405 in glance_store "ceph jobs failing to upload images in pike due to "AttributeError: 'NoneType' object has no attribute 'Rados'" on py35" [Undecided,Confirmed]
15:24:49 mriedem stephenfin: that's a pretty old bug,
15:24:57 mriedem i'm watching https://review.openstack.org/#/c/543048/ to see if it passes now
15:25:03 stephenfin Right. Just making sure
15:25:55 mriedem andreaf: what is your plan to drop the legacy ceph job? are you just going to restrict that to the stable branches in openstack-zuul-jobs?
15:27:38 andreaf mriedem: that's what I'm doing at first for integration gate jobs, start with master only and then go back and replace for stable branches
15:28:42 andreaf mriedem: but that takes some time, so you're right you may want to fix the legacy job as well meanwhile
15:29:03 andreaf mriedem tbh I wasn't sure whether that job was important to anyone anymore - so I just put up my patch, asked for reviews and then I let it be
15:29:07 idlemind grr my issue might be from a hypervisor that wasn't rebooted (or kvm_amd wasn't removed/inserted) after enabling nested ... we'll see
15:29:23 mriedem andreaf: ceph ci is important about once per quarter :)
15:29:46 mriedem which reminds me that http://grafana.openstack.org/dashboard/db/ceph-failure-rate still doesn't work
15:34:09 openstackgerrit Dan Smith proposed openstack/nova master: Add --purge helper flag to archive_deleted_rows https://review.openstack.org/550182
15:34:10 openstackgerrit Dan Smith proposed openstack/nova master: Make nova-manage db purge take --all-cells https://review.openstack.org/550502
15:37:06 dansmith mriedem: I think it'd be easiest to rebase the archive --all-cells patch on top of my series instead of the other way around, just based on where the touch points are
15:37:12 dansmith which I will do if mine lands first
15:37:31 dansmith but he was first, so no worries either way
15:38:53 mriedem is there a dependency?
15:40:29 dansmith no just a conflict
15:40:53 dansmith but the conflict is way smaller if he goes on top of me, because of where in manage.py he's touching I think
15:41:27 openstack Launchpad bug 1754071 in OpenStack Compute (nova) "image not found warning in logs when instance is deleted during snapshot" [Low,Triaged]
15:41:27 mriedem hongbin: https://bugs.launchpad.net/nova/+bug/1754071 if you're interested
15:41:39 mriedem hongbin: as a follow up to https://review.openstack.org/#/c/511074/
15:42:19 hongbin mriedem: ack, will look into that
15:43:42 openstackgerrit sahid proposed openstack/nova-specs master: update: introducing isolate emulthreads on host https://review.openstack.org/511188
15:44:07 Spazmotic Sorry for the nickspam.. VPN shenanigans.
15:46:18 mriedem dansmith: artom: this is pretty nifty https://review.openstack.org/#/c/539584/ - we should consider doing that in the libvirt driver too
15:46:30 mriedem it's not obvious so i had to look at the paste that claudiub|2 put into the review comments
15:47:00 mriedem when port binding fails with the libvirt driver, i think we eventually just get the virtual interface exception or whatever, which logs a generic warning
15:47:02 dansmith ooh, blame
15:47:07 dansmith love it

Earlier   Later