| *** edmondsw has joined #openstack-powervm | 00:16 | |
| *** edmondsw has quit IRC | 00:20 | |
| *** thorst has joined #openstack-powervm | 00:27 | |
| *** thorst has quit IRC | 00:27 | |
| *** edmondsw has joined #openstack-powervm | 02:04 | |
| *** edmondsw has quit IRC | 02:09 | |
| *** thorst has joined #openstack-powervm | 02:28 | |
| *** thorst has quit IRC | 02:33 | |
| *** thorst has joined #openstack-powervm | 03:58 | |
| *** thorst has quit IRC | 04:03 | |
| *** chhavi has joined #openstack-powervm | 04:04 | |
| *** chhavi has quit IRC | 04:38 | |
| *** chhavi has joined #openstack-powervm | 04:41 | |
| *** chhavi has quit IRC | 04:44 | |
| *** edmondsw has joined #openstack-powervm | 05:40 | |
| *** edmondsw has quit IRC | 05:45 | |
| *** thorst has joined #openstack-powervm | 05:59 | |
| *** thorst has quit IRC | 06:05 | |
| *** edmondsw has joined #openstack-powervm | 07:29 | |
| *** edmondsw has quit IRC | 07:33 | |
| *** thorst has joined #openstack-powervm | 08:01 | |
| *** thorst has quit IRC | 08:05 | |
| *** k0da has joined #openstack-powervm | 08:08 | |
| *** edmondsw has joined #openstack-powervm | 09:17 | |
| *** edmondsw has quit IRC | 09:21 | |
| *** thorst has joined #openstack-powervm | 10:02 | |
| *** thorst has quit IRC | 10:06 | |
| *** smatzek has joined #openstack-powervm | 10:46 | |
| *** thorst has joined #openstack-powervm | 10:47 | |
| *** thorst has quit IRC | 10:49 | |
| *** thorst has joined #openstack-powervm | 11:00 | |
| *** edmondsw has joined #openstack-powervm | 11:05 | |
| *** edmondsw has quit IRC | 11:09 | |
| *** thorst has quit IRC | 11:36 | |
| *** efried_backJuly5 is now known as efried | 11:40 | |
| *** svenkat has joined #openstack-powervm | 11:56 | |
| *** svenkat has quit IRC | 12:00 | |
| *** svenkat has joined #openstack-powervm | 12:08 | |
| *** edmondsw has joined #openstack-powervm | 12:22 | |
| *** mdrabe has joined #openstack-powervm | 12:43 | |
| *** thorst has joined #openstack-powervm | 12:44 | |
| *** thorst_ has joined #openstack-powervm | 12:45 | |
| *** thorst has quit IRC | 12:48 | |
| *** esberglu has joined #openstack-powervm | 13:01 | |
| *** jwcroppe has quit IRC | 14:01 | |
| *** jwcroppe has joined #openstack-powervm | 14:02 | |
| *** jwcroppe has quit IRC | 14:06 | |
| *** miltonm has joined #openstack-powervm | 14:15 | |
| *** jwcroppe has joined #openstack-powervm | 14:16 | |
| *** dwayne has joined #openstack-powervm | 15:20 | |
| *** smatzek has quit IRC | 16:17 | |
| *** smatzek has joined #openstack-powervm | 16:53 | |
| thorst_ | efried: 5531 | 16:59 |
|---|---|---|
| edmondsw | thorst_ 5518 | 17:21 |
| thorst_ | LGTM | 17:23 |
| edmondsw | thorst_ I commented on 5531 | 17:32 |
| thorst_ | thx | 17:32 |
| thorst_ | also 5533 | 17:38 |
| edmondsw | esberglu could we run CI with 5518 before merging? | 17:38 |
| edmondsw | what's the best way to do that? | 17:39 |
| esberglu | edmondsw: I can put a manual run through with it | 17:58 |
| edmondsw | esberglu tx | 17:58 |
| edmondsw | esberglu are the tests in 5534 new, or why are we just needing to skip them now? | 18:58 |
| efried | thorst_ Whoah | 19:43 |
| thorst_ | whoah what? | 19:43 |
| efried | thorst_ Whoah assuming my partition ID is 1 if the lparcfg file doesn't exist. | 19:44 |
| thorst_ | heh | 19:44 |
| thorst_ | I thought you'd hate on that | 19:44 |
| thorst_ | nvcastet was asking me why such a small change would take so much time | 19:44 |
| thorst_ | and I told him, I'm channeling my inner efried | 19:45 |
| efried | thorst_ Okay, so talk me through the possibilities here. | 19:45 |
| thorst_ | distro just doesn't populate it. | 19:46 |
| efried | Are they just 1) running nvl on a distro that doesn't have that file; and 2) running remote API? | 19:46 |
| thorst_ | we're remote (ex. running from my laptop) | 19:46 |
| thorst_ | those were all I could really think of | 19:46 |
| thorst_ | the alternative I could think of is returning the value from get_mgmt_partition...but I didn't really like that | 19:47 |
| efried | thorst_ So my main concern is future redundancy. | 19:48 |
| thorst_ | hmm | 19:50 |
| thorst_ | well, the redundancy there will be I/O | 19:50 |
| thorst_ | I don't think nova-powervm will ever run on a redundant LPAR | 19:50 |
| thorst_ | (to my knowledge) | 19:50 |
| thorst_ | I guess anything is possible | 19:50 |
| efried | That's an interesting discussion. | 19:51 |
| thorst_ | yeah | 19:51 |
| efried | Got a few minutes to do a brainstorm on the phone? | 19:51 |
| thorst_ | I mean, isn't the only reason we coded that the way we did is for the CI? | 19:51 |
| thorst_ | sure, though my voice is kinda going out | 19:51 |
| thorst_ | give me a ring | 19:51 |
| efried | mm, we can do it here. | 19:51 |
| efried | So the point of redundancy is what? | 19:52 |
| efried | To keep the VMs going if one of the I/O servers goes away? | 19:52 |
| efried | Or to keep the compute service responsive? | 19:52 |
| efried | What happens in other drivers if the compute service on a host stops responding, but the VMs are still okay? Can that even happen? | 19:53 |
| thorst_ | I/O | 19:53 |
| thorst_ | if the management layer gets restarted or what not...peace. We can deal with that because we have other hosts in the pool that can take new VM requests | 19:54 |
| thorst_ | but you really, ideally, don't want the I/O to the VM going out | 19:54 |
| thorst_ | then you have to start doing things like remote rebuild, LPM, possibly client VM outages, etc... | 19:54 |
| thorst_ | actual workload impacts. | 19:54 |
| thorst_ | that jumps from inconvenience to big deal territory | 19:54 |
| efried | So we really have a primary and a backup, and if the primary goes down, the backup picks up the I/O just long enough for us to get the primary back (how?). The nvl and openstack services run only on the primary. | 19:57 |
| thorst_ | so that's magic of the hypervisor | 19:58 |
| thorst_ | I think for storage its just standard MPIO | 19:58 |
| thorst_ | network is black magic that I won't pretend to understand completely | 19:58 |
| efried | thorst_ No MPIO for localdisk. For SSP it's literally vscsi connections to both I/O servers. | 20:00 |
| thorst_ | that's MPIO | 20:00 |
| efried | Heck, I don't even know if that works. | 20:00 |
| thorst_ | it does. | 20:01 |
| efried | I don't see how OVS is going to do redundancy. | 20:05 |
| thorst_ | yikes...OVS is ... a challenge | 20:06 |
| thorst_ | there will be tears | 20:06 |
| esberglu | edmondsw: 5518 ran through no issues | 20:29 |
| *** thorst_ has quit IRC | 20:55 | |
| *** esberglu has quit IRC | 20:55 | |
| *** thorst has joined #openstack-powervm | 20:57 | |
| *** svenkat has quit IRC | 20:59 | |
| *** thorst has quit IRC | 21:01 | |
| *** smatzek has quit IRC | 21:16 | |
| edmondsw | esberglu thanks! I'll go ahead and merge then | 21:26 |
| *** edmondsw has quit IRC | 21:31 | |
| *** thorst has joined #openstack-powervm | 21:31 | |
| *** edmondsw has joined #openstack-powervm | 21:33 | |
| *** thorst has quit IRC | 21:35 | |
| *** edmondsw has quit IRC | 21:38 | |
| *** edmondsw has joined #openstack-powervm | 22:11 | |
| *** edmondsw has quit IRC | 22:16 | |
| *** esberglu has joined #openstack-powervm | 22:21 | |
| *** esberglu has quit IRC | 22:26 | |
| *** jwcroppe has quit IRC | 22:43 | |
| *** jwcroppe has joined #openstack-powervm | 22:43 | |
| *** jwcroppe has quit IRC | 22:47 | |
| *** jwcroppe has joined #openstack-powervm | 23:00 | |
| *** jwcroppe has quit IRC | 23:01 | |
| *** jwcroppe has joined #openstack-powervm | 23:01 | |
| *** jwcroppe has quit IRC | 23:06 | |
| *** thorst has joined #openstack-powervm | 23:12 | |
| *** thorst has quit IRC | 23:12 | |
| *** thorst has joined #openstack-powervm | 23:19 | |
| *** jwcroppe has joined #openstack-powervm | 23:31 | |
| *** jwcroppe has quit IRC | 23:36 | |
| *** k0da has quit IRC | 23:45 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!