Thursday, 2018-01-11

*** edmondsw has quit IRC00:00
*** edmondsw has joined #openstack-powervm01:44
*** edmondsw has quit IRC01:48
*** esberglu has quit IRC04:10
*** chhavi__ has joined #openstack-powervm04:46
*** edmondsw has joined #openstack-powervm05:20
*** edmondsw has quit IRC05:24
*** edmondsw has joined #openstack-powervm07:08
*** edmondsw has quit IRC07:13
*** AlexeyAbashkin has joined #openstack-powervm08:08
*** edmondsw has joined #openstack-powervm08:56
*** edmondsw has quit IRC09:01
*** edmondsw has joined #openstack-powervm10:44
*** edmondsw has quit IRC10:49
*** AlexeyAbashkin has quit IRC11:06
*** AlexeyAbashkin has joined #openstack-powervm11:25
*** edmondsw has joined #openstack-powervm12:33
*** edmondsw has quit IRC12:38
*** chas has joined #openstack-powervm13:21
*** apearson__ has joined #openstack-powervm13:41
*** efried has joined #openstack-powervm13:53
*** edmondsw has joined #openstack-powervm14:21
*** edmondsw_ has joined #openstack-powervm14:24
*** esberglu has joined #openstack-powervm14:28
*** edmondsw has quit IRC14:28
*** edmondsw_ is now known as edmondsw14:45
edmondswefried is there a reason not to autospec nova mocks?14:46
edmondswesberglu thought we should only autospec pypowervm and not nova14:46
edmondswI thought we should autospec pretty much everything14:46
efriedThere were some bugs in mock around autospec that made that impractical, but I think they're fixed now, at least some of them.  claudiub was working that.14:52
efriedIt would be ambitious to autospec *everything*.  But a good goal long term.14:52
edmondswI meant more in new code, like in the pypowervm patches14:52
efriedWe can.  It tends to be a fair bit of overhead, but sure helps flush out bugs.14:53
esbergluOkay, I can add autospec to the new vscsi test code. I didnt autospec everything in OVS/SEA, but I will leave that since were up to core reviews there14:58
esbergluWhy do I need to set the return value there first?14:59
edmondswesberglu we could also add autospecing later. I don't want to hold things up for that. Sounds like it's not done as widely in nova as I thought14:59
edmondswesberglu what's the point of checking the return value against what you got back if you didn't set what you were supposed to get back?15:00
*** gman-tx has joined #openstack-powervm15:02
esbergluI don't see the functional difference between15:07
esberglua) self.assertEqual(mock_add_map.return_value, add_func(mock_vios))15:07
esberglub) mock_add_map.return_value = 'add_map"15:07
esbergluself.assertEqual('add_map', add_func(mock_vios))15:07
esbergluEither way it's just checking that what is returned from the add_func method is the return value of add_map right?15:07
efriedesberglu is right.  You don't need to set a return value to compare against it.  If you don't set the return value, it returns a mock, which is unique and comparable.15:07
edmondswesberglu yeah, that's not what I meant15:07
edmondswlet me take another look15:08
edmondswesberglu ok yeah, I must have misread that. That line is fine15:11
esbergluedmondsw: It is supposed to be wtsk though. But I think that keeping the checks in test_afs that are part of the add_func method is more clear. Thoughts?15:13
edmondswesberglu keeping them in there means that if that's never called (as they weren't because you had ftsk instead of wtsk), they're skipped and nobody notices15:14
edmondswI didn't think it would be any less clear to have them outside that... but if you put some other protection in place to make sure the method is called I'll be fine15:15
esbergluedmondsw: Yeah I already added a check that its actually called15:16
edmondswbrb, have to reboot15:18
*** edmondsw has quit IRC15:18
*** edmondsw has joined #openstack-powervm15:24
*** chas has quit IRC15:39
*** gman-tx has quit IRC15:45
esbergluefried: edmondsw: Latest vSCSI patch is up15:51
esbergluMy turn to reboot15:51
edmondswesberglu ack15:52
*** esberglu has quit IRC15:52
*** csky has joined #openstack-powervm15:59
*** esberglu has joined #openstack-powervm16:26
edmondswesberglu few easy comments on vscsi and then I think we're done16:28
*** AlexeyAbashkin has quit IRC16:53
esbergluedmondsw: New patch is up17:10
edmondswesberglu ack17:16
esbergluedmondsw: Do we want to port the new test_vscsi tests to the OOT driver?17:22
*** chhavi__ has quit IRC17:26
edmondswesberglu I was wrong about the log asserts... better put them back17:32
edmondswsorry bout that17:32
esbergluedmondsw: Yeah I had them in there to test the various numbers of vio_modified. Thought that maybe it was bad to assert on logs though? Will switch back17:33
edmondswyeah, you were right to begin with, and I just misread going over it too fast17:33
edmondswor misassumed something17:33
efriedI'm not a fan of asserting log text, though I've seen it done in nova.  Asserting log args is nicer, though still not wonderful.17:38
esbergluefried: I didn't like it either, just didn't see any alternative since that's the only variation in the different cases17:42
esbergluedmondsw: Done17:47
-openstackstatus- NOTICE: Due to an unexpected issue with zuulv3.o.o, we were not able to preserve running jobs for a restart. As a result, you'll need to recheck your previous patchsets17:47
esbergluCool the vSCSI patch got to the top of the zuul queue from that lol17:49
esbergluefried: edmondsw: I'm looking into a test that is sometimes failing for IT CI.18:01
esbergluThe test in question updates the server name to non-ascii characters18:01
esbergluThen there's an error trying to delete the server because the non-ascii chars can't be decoded. Seems like an issue with the test, except for the fact that it is not hitting OOT18:01
efriedWe should be sanitizing those away, I would have thought.18:01
efriedPerhaps we didn't bring the sanitizing code over properly.18:02
esbergluefried: Looks like the sanitize is done in the update here for OOT18:07
esbergluBut we don't even have that update method IT. So now I'm confused as to how the name is being updated at all18:07
esbergluHere's the test
*** AlexeyAbashkin has joined #openstack-powervm18:20
*** AlexeyAbashkin has quit IRC18:24
*** AlexeyAbashkin has joined #openstack-powervm18:57
*** AlexeyAbashkin has quit IRC19:01
edmondswesberglu anything more to test with vSCSI before we ask cores to review?19:19
edmondswefried I gave my +1 on vscsi... give it a look when you get a chance19:19
edmondswwe'll want your +1 before asking cores19:19
efriedack (in both senses)19:19
esbergluedmondsw: Nope I'm ready for cores19:21
edmondswgreat, that's what I thought19:22
edmondswefried fyi ^19:22
efriedesberglu edmondsw Is [storage.block] right?19:35
efriedI thought that was localdisk19:35
edmondswefried pretty sure [storage.block] is correct here19:42
*** edmondsw_ has joined #openstack-powervm19:48
*** edmonds__ has joined #openstack-powervm19:48
*** edmondsw has quit IRC19:52
*** edmondsw_ has quit IRC19:52
esbergluedmondsw: efried: I was asked to provide release versions for 1.4.0 FP1. Nothing has changed since last tag in ceilometer and networking.19:54
esberglunova-powervm only has 1 change19:55
esbergluI'm gonna cut another release with that and provide those unless there's any objections19:55
*** apearson__ has quit IRC20:03
*** apearson__ has joined #openstack-powervm20:05
efriedesberglu One-character fix on vSCSI and I'll be +1.20:23
esbergluefried: Done20:28
efriedThanks, +120:28
*** apearson__ has quit IRC20:42
*** apearson__ has joined #openstack-powervm20:59
edmonds__efried good catch... I made the same comment back in PS17 and guess I didn't check that it was fixed21:07
edmonds__or snuck back in21:07
*** edmonds__ is now known as edmondsw21:07
esbergluedmondsw: Saw you poked the nova guys on vscsi. I updated that etherpad21:25
esbergluWould you mind posting on the ML? I use verse and you'll be able to send it before I figure out how to make it plaintext21:26
esbergluon Notes21:26
edmondswesberglu tx21:26
edmondswesberglu sure... it's easy in notes, impossible in verse21:27
edmondswnote sent21:34
esbergluedmondsw: efried: FYI I may or may not take some or all of tomorrow off. Might take advantage of the snowstorm and get some snowboarding in this weekend21:46
esbergluNot expecting reviews to gain traction by tomorrow anyways21:46
efriedAhh, snowboarding.  I remember my teens and 20s.21:47
efriedhaha!  Traction joke21:47
edmondswhave fun!21:48
edmondswesberglu I loved skiing from the moment I tried it, but hated snowboarding. To each their own :)21:50
esbergluThanks! And I have yet to try skiing, I guess I just like snowboarding too much21:52
edmondswdo what ya love21:52
efriedI skied for years before trying snowboarding.  I love 'em both, like you could love driving a sports car and riding a motorcycle.21:57
efriedBut, as with the analogy, not both at the same time.21:57
*** csky has quit IRC22:33
*** esberglu has quit IRC22:36
*** edmondsw has quit IRC22:54
*** edmondsw has joined #openstack-powervm22:55
*** edmondsw has quit IRC22:59
*** csky has joined #openstack-powervm23:06
*** chhavi__ has joined #openstack-powervm23:22
*** chhavi__ has quit IRC23:26
*** apearson__ has quit IRC23:29
*** csky has quit IRC23:48

Generated by 2.15.3 by Marius Gedminas - find it at!