| *** jwcroppe has quit IRC | 00:51 | |
| *** jwcroppe has joined #openstack-powervm | 00:52 | |
| *** thorst has quit IRC | 01:09 | |
| *** adreznec has quit IRC | 02:37 | |
| *** adreznec has joined #openstack-powervm | 02:39 | |
| *** esberglu has quit IRC | 02:43 | |
| *** thorst has joined #openstack-powervm | 03:10 | |
| *** thorst has quit IRC | 03:15 | |
| *** esberglu has joined #openstack-powervm | 03:26 | |
| *** tjakobs has joined #openstack-powervm | 03:28 | |
| *** esberglu has quit IRC | 03:30 | |
| *** tjakobs has quit IRC | 03:33 | |
| *** tjakobs has joined #openstack-powervm | 03:34 | |
| *** tjakobs has quit IRC | 03:38 | |
| *** edmondsw has joined #openstack-powervm | 03:53 | |
| *** edmondsw has quit IRC | 03:57 | |
| *** tjakobs has joined #openstack-powervm | 04:42 | |
| *** tjakobs has quit IRC | 05:08 | |
| *** thorst has joined #openstack-powervm | 05:12 | |
| *** thorst has quit IRC | 05:16 | |
| *** edmondsw has joined #openstack-powervm | 05:41 | |
| *** edmondsw has quit IRC | 05:45 | |
| *** AndyWojo has quit IRC | 06:24 | |
| *** AndyWojo has joined #openstack-powervm | 06:26 | |
| *** openstackgerrit has quit IRC | 06:28 | |
| *** jwcroppe has quit IRC | 06:36 | |
| *** jwcroppe has joined #openstack-powervm | 06:36 | |
| *** esberglu has joined #openstack-powervm | 06:38 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 06:39 | |
| *** esberglu has quit IRC | 06:43 | |
| *** AlexeyAbashkin has quit IRC | 06:46 | |
| *** edmondsw has joined #openstack-powervm | 07:29 | |
| *** efried has quit IRC | 07:32 | |
| *** edmondsw has quit IRC | 07:33 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 07:38 | |
| *** efried has joined #openstack-powervm | 07:44 | |
| *** esberglu has joined #openstack-powervm | 08:26 | |
| *** esberglu has quit IRC | 08:30 | |
| *** esberglu has joined #openstack-powervm | 09:20 | |
| *** esberglu has quit IRC | 09:20 | |
| *** esberglu has joined #openstack-powervm | 09:20 | |
| *** esberglu has quit IRC | 09:21 | |
| *** AlexeyAbashkin has quit IRC | 09:53 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 09:53 | |
| *** esberglu has joined #openstack-powervm | 10:14 | |
| *** esberglu has quit IRC | 10:18 | |
| *** AlexeyAbashkin has quit IRC | 11:04 | |
| *** edmondsw has joined #openstack-powervm | 11:05 | |
| *** edmondsw has quit IRC | 11:09 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 11:20 | |
| *** thorst has joined #openstack-powervm | 11:54 | |
| *** thorst has quit IRC | 11:54 | |
| *** thorst has joined #openstack-powervm | 11:56 | |
| *** edmondsw has joined #openstack-powervm | 12:09 | |
| *** esberglu has joined #openstack-powervm | 12:38 | |
| *** kylek3h has quit IRC | 14:03 | |
| *** esberglu has quit IRC | 14:46 | |
| *** tjakobs has joined #openstack-powervm | 14:48 | |
| *** tjakobs has quit IRC | 14:49 | |
| *** tjakobs has joined #openstack-powervm | 14:49 | |
| *** tjakobs_ has joined #openstack-powervm | 14:56 | |
| *** tjakobs has quit IRC | 14:56 | |
| *** thorst has quit IRC | 15:37 | |
| mdrabe | efried could you shlap https://review.openstack.org/#/c/511343 on your review list pretty please | 15:48 |
|---|---|---|
| efried | mdrabe Ack. Ready to go, is it? | 15:58 |
| mdrabe | Yep and live tested | 15:58 |
| efried | cool | 16:04 |
| *** AlexeyAbashkin has quit IRC | 16:36 | |
| *** esberglu has joined #openstack-powervm | 17:45 | |
| *** tjakobs_ has quit IRC | 17:57 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 17:59 | |
| *** AlexeyAbashkin has quit IRC | 18:03 | |
| *** tjakobs_ has joined #openstack-powervm | 18:04 | |
| efried | mdrabe See review, me confused. | 18:19 |
| *** edmondsw has quit IRC | 19:04 | |
| *** edmondsw has joined #openstack-powervm | 19:04 | |
| *** edmondsw has quit IRC | 19:09 | |
| *** openstackgerrit has joined #openstack-powervm | 19:23 | |
| openstackgerrit | Matt Rabe proposed openstack/nova-powervm master: Persist existing LPAR wrapper attributes in DefaultStandardize on resize https://review.openstack.org/511343 | 19:23 |
| *** kylek3h has joined #openstack-powervm | 19:23 | |
| *** kylek3h has quit IRC | 19:26 | |
| *** edmondsw has joined #openstack-powervm | 19:54 | |
| mdrabe | efried: history quiz, remember why we changed the raised ConsoleTypeUnavailable to InteralError here https://github.com/openstack/nova-powervm/blob/master/nova_powervm/virt/powervm/driver.py#L1744? | 20:34 |
| mdrabe | It's not evident in the review https://review.openstack.org/#/c/435114/6/nova_powervm/virt/powervm/driver.py | 20:35 |
| efried | mdrabe Finding the in-tree review... | 20:46 |
| efried | mdrabe https://review.openstack.org/#/c/409402/14/nova/tests/unit/virt/powervm/test_driver.py@269 | 20:49 |
| efried | blaayum | 20:49 |
| mdrabe | Ah thnx couldn't find that | 20:51 |
| mdrabe | efried: Can we change it back to ConsoleTypeUnavailable? | 20:51 |
| mdrabe | InternalError isn't a handled exception by the API, but ConsoleTypeUnavailable is | 20:51 |
| mdrabe | So all InternalErrors thrown from get_vnc_console are treated as unexpected exceptions from the API, masking the original error | 20:53 |
| *** edmondsw has quit IRC | 20:56 | |
| mdrabe | efried: Here's the handled exceptions by the API https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/remote_consoles.py#L54-L62 | 20:58 |
| efried | mdrabe "masking the original error"? | 21:08 |
| efried | Is there some kind of bug for this that you're trying to resolve? | 21:09 |
| mdrabe | Because get_vnc_console (the API) doesn't handle InternalError exceptions, it gets blown up as an unhandled exception | 21:09 |
| mdrabe | Sorta, side effect of a bug | 21:09 |
| efried | I think that's what we want, in the case of an Exception. | 21:10 |
| mdrabe | This is just a small improvement thing | 21:10 |
| efried | If there's a more specific thing that's happening, we should be detecting it and raising something more specific. | 21:10 |
| efried | But if it's anything else, it's truly unexpected/unhandled, and it should be blowing up hard. | 21:10 |
| mdrabe | Yea but this is a handled exception by the driver | 21:11 |
| efried | But "masking" worries me. That's a code flow problem, if it's truly happening. Not necessarily an improper use of exceptions by us. | 21:11 |
| mdrabe | Which means it should be a handled exception by the API | 21:11 |
| efried | What is it? | 21:11 |
| mdrabe | Here's what it looks like from the API when InternalError is raised: | 21:11 |
| mdrabe | HTTP exception thrown: Unexpected API Error. Please report this at http://bugs.launchpad.net/nova/ and attach the Nova API log if possible. | 21:11 |
| efried | Uhm, right, and now you look in the compute log and it says what? | 21:11 |
| efried | My point is, whatever happened there is NOT necessarily a 400. | 21:12 |
| mdrabe | Instead if it's a ConsoleTypeUnavailable exception then the API would catch it and relay the actual exception message | 21:12 |
| efried | In fact, it's most likely NOT a 400. | 21:12 |
| efried | Yeah, I don't think we want it to. | 21:12 |
| efried | You must have an example of the actual exception that prompted this discussion. | 21:12 |
| efried | What is it? | 21:12 |
| mdrabe | Grabbin it just a sec | 21:13 |
| mdrabe | ultimately it was the firewall | 21:13 |
| mdrabe | Not sure what compute actually gets though | 21:13 |
| mdrabe | efried: here it is from compute: http://paste.openstack.org/show/623763/ | 21:17 |
| mdrabe | More generally I think that any exception _explicitly_ raised by the driver should be handled in the API | 21:18 |
| efried | What does the InstanceNotFound code path do? | 21:19 |
| mdrabe | That's handled in the API | 21:19 |
| efried | Where? | 21:19 |
| efried | oh, I see it. | 21:19 |
| mdrabe | https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/remote_consoles.py#L57 | 21:19 |
| efried | Thing is, ConsoleTypeUnavailable is not appropriate. | 21:20 |
| mdrabe | wainot? existing vterm? Based off the name I'd think that'd fit | 21:20 |
| efried | "Unavailable console type %(console_type)s." | 21:21 |
| efried | As in, "you asked for console type X, but I only support Y and Z". | 21:21 |
| mdrabe | InstanceNotReady then perhaps? | 21:21 |
| efried | Meh. That's slightly better. But unless we override the message it doesn't really help the user. | 21:22 |
| efried | And overriding the message is kind of ew. | 21:22 |
| efried | But | 21:22 |
| efried | I could get behind it | 21:23 |
| efried | If we catch VNCBasedTerminalFailedToOpen specifically. | 21:23 |
| efried | Not for the general Exception case. | 21:23 |
| mdrabe | subclass one of those maybe? | 21:23 |
| efried | could do. | 21:23 |
| mdrabe | Ight I'll open a bug and whip it up | 21:25 |
| *** edmondsw has joined #openstack-powervm | 22:25 | |
| *** edmondsw has quit IRC | 22:29 | |
| *** esberglu has quit IRC | 22:36 | |
| *** esberglu has joined #openstack-powervm | 23:11 | |
| *** esberglu has quit IRC | 23:15 | |
| *** tjakobs_ has quit IRC | 23:17 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!