| -openstackstatus- NOTICE: Jobs using devstack-gate (legacy devstack jobs) have been failing due to an ara update. We use now a newer ansible version, it's safe to recheck if you see "ImportError: No module named manager" in the logs. | 09:57 | |
| *** tonyb has quit IRC | 12:02 | |
| *** tonyb has joined #openstack-powervm | 12:19 | |
| *** efried is now known as fried_rice | 13:38 | |
| *** fried_rice is now known as fried_rolls | 16:27 | |
| *** openstackstatus has quit IRC | 18:58 | |
| *** fried_rolls is now known as fried_rice | 19:23 | |
| *** openstackstatus has joined #openstack-powervm | 19:38 | |
| *** ChanServ sets mode: +v openstackstatus | 19:38 | |
| *** openstackstatus has quit IRC | 19:55 | |
| *** openstackstatus has joined #openstack-powervm | 19:58 | |
| *** ChanServ sets mode: +v openstackstatus | 19:58 | |
| *** openstackstatus has quit IRC | 20:11 | |
| *** openstackstatus has joined #openstack-powervm | 20:12 | |
| *** ChanServ sets mode: +v openstackstatus | 20:12 | |
| fried_rice | edmondsw: That log message is a red herring. If you look right after, it recovers. | 20:13 |
|---|---|---|
| edmondsw | ok, good | 20:14 |
| edmondsw | so something else going on. You still looking? | 20:14 |
| edmondsw | I haven't had a chance to even pull the logs up | 20:14 |
| fried_rice | edmondsw: I'm... multitasking. Which I patently suck at. But yeah, it's somewhere in my swap space. | 20:19 |
| edmondsw | fried_rice thanks | 20:20 |
| fried_rice | edmondsw: Um, there's something very wrong here, though. The compute log is way too short to have handled the passing tests. | 20:21 |
| fried_rice | unless none of the 700-odd of the passing tests actually touch the compute service. | 20:22 |
| fried_rice | Which would beg the question: why do we need them in our CI? | 20:22 |
| edmondsw | yeah... | 20:22 |
| edmondsw | we really need to do a scrub of what we test and don't test in CI | 20:23 |
| fried_rice | well that's easy. | 20:23 |
| edmondsw | I don't mean just determining what we test and don't test, but rather figuring out where we test things we shouldn't or don't test things we should and changing accordingly | 20:24 |
| *** openstackgerrit has joined #openstack-powervm | 20:30 | |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: DNM: See what in the CI doesn't need our code https://review.openstack.org/599066 | 20:30 |
| fried_rice | edmondsw: Assuming the compute service will still start, we ought to be able to remove any tests that still pass this ^ | 20:31 |
| fried_rice | (well, we *certainly* ought to be able to remove tests that pass if the compute service *doesn't* start...) | 20:31 |
| fried_rice | I guess technically I ought to do the same thing to networking-powervm and ceilometer-powervm. | 20:32 |
| fried_rice | But I don't seriously expect us to take action on this; just playing around. | 20:32 |
| *** openstackstatus has quit IRC | 20:36 | |
| fried_rice | edmondsw: All the allocation candidate requests are coming back empty. No way to tell at this point without going in and looking at an actual system, but this may be because the allocation ratios are getting zeroed out. | 20:37 |
| fried_rice | edmondsw: How frequent is the failure? | 20:37 |
| *** openstackstatus has joined #openstack-powervm | 20:38 | |
| fried_rice | I'm suspecting https://bugs.launchpad.net/nova/+bug/1789654 btw | 20:38 |
| openstack | Launchpad bug 1789654 in OpenStack Compute (nova) "placement allocation_ratio initialized with 0.0" [High,In progress] - Assigned to Matt Riedemann (mriedem) | 20:38 |
| *** ChanServ sets mode: +v openstackstatus | 20:38 | |
| edmondsw | fried_rice I think esberglu said 90% | 20:38 |
| fried_rice | edmondsw: Well, https://review.openstack.org/#/c/598365/ is the fix. And the fact that it passed our CI is a good sign. I'll recheck that bad boy a time or two and see. | 20:39 |
| edmondsw | fried_rice thanks | 20:45 |
| fried_rice | edmondsw: I may have a way to confirm that that's the/a problem. Matt put up a couple of patches with additional debug logs. If we failed CI on those, the logs will tell us if this is the problem. I'm looking now... | 20:46 |
| fried_rice | edmondsw: Bingo: http://184.172.12.213/60/597560/6/check/nova-in-tree-pvm/9b7c798/logs/n-cpu.txt.gz#_Aug_30_16_43_09_412748 | 20:48 |
| fried_rice | oo, that highlight is distracting. | 20:48 |
| fried_rice | ute[9212]: New: ComputeNode(cpu_allocation_ratio=>>>>>>>>this>>>>>>>>0.0<<<<<<<<<<<<,cpu_info='{"a | 20:48 |
| edmondsw | yep, good catch | 20:49 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: DNM: See what in the CI doesn't need our code https://review.openstack.org/599066 | 21:15 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: DNM: See what in the CI doesn't need our code https://review.openstack.org/599066 | 21:16 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: DNM: See what in the CI doesn't need our code https://review.openstack.org/599066 | 21:16 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: DNM: See what in the CI doesn't need our code https://review.openstack.org/599066 | 22:01 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!