| *** mujahidali has joined #openstack-powervm | 05:30 | |
| *** k0da has joined #openstack-powervm | 05:45 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 06:15 | |
| *** AlexeyAbashkin has quit IRC | 06:41 | |
| *** k0da has quit IRC | 06:43 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 06:58 | |
| *** Alexey_Abashkin has joined #openstack-powervm | 07:01 | |
| *** AlexeyAbashkin has quit IRC | 07:02 | |
| *** Alexey_Abashkin is now known as AlexeyAbashkin | 07:02 | |
| *** k0da has joined #openstack-powervm | 07:50 | |
| *** mujahidali has quit IRC | 08:13 | |
| *** mujahidali has joined #openstack-powervm | 08:13 | |
| *** AlexeyAbashkin has quit IRC | 10:08 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 10:11 | |
| *** AlexeyAbashkin has quit IRC | 10:18 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 10:31 | |
| *** mujahidali has quit IRC | 11:32 | |
| *** AlexeyAbashkin has quit IRC | 12:31 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 12:35 | |
| *** edmondsw has joined #openstack-powervm | 12:42 | |
| *** AlexeyAbashkin has quit IRC | 12:55 | |
| *** AlexeyAbashkin has joined #openstack-powervm | 12:56 | |
| *** esberglu has joined #openstack-powervm | 13:21 | |
| openstackgerrit | Matthew Edmonds proposed openstack/ceilometer-powervm master: add lower-constraints job https://review.openstack.org/555358 | 13:28 |
|---|---|---|
| *** mujahidali has joined #openstack-powervm | 14:08 | |
| openstackgerrit | Matthew Edmonds proposed openstack/ceilometer-powervm master: add lower-constraints job https://review.openstack.org/555358 | 14:27 |
| *** efried is now known as fried_rice | 14:32 | |
| openstackgerrit | Eric Fried proposed openstack/networking-powervm master: Match neutron's version of hacking, flake8 ignores https://review.openstack.org/582686 | 15:51 |
| *** k0da has quit IRC | 15:58 | |
| *** AlexeyAbashkin has quit IRC | 16:13 | |
| mdrabe | efried: When I run tox in my nova-powervm workspace, can I tell it to point to a nova review? | 16:18 |
| mdrabe | This is for https://review.openstack.org/#/c/579676/ | 16:18 |
| mdrabe | fried_rice* | 16:18 |
| fried_rice | mdrabe: For local testing, yes. | 16:19 |
| mdrabe | How do I do that? | 16:19 |
| fried_rice | mdrabe: Do you have a directory .tox/py27/src/nova in your nova-powervm git repo? | 16:20 |
| mdrabe | yep | 16:20 |
| mdrabe | ok mkay I see | 16:20 |
| fried_rice | okay, cd in there and git review -d whatever change number you want. | 16:20 |
| mdrabe | thnx | 16:20 |
| fried_rice | yahyoubetcha. | 16:20 |
| *** fried_rice is now known as fried_rolls | 16:24 | |
| openstackgerrit | Merged openstack/ceilometer-powervm master: add lower-constraints job https://review.openstack.org/555358 | 16:52 |
| *** mujahidali has quit IRC | 17:22 | |
| *** esberglu has quit IRC | 18:10 | |
| *** esberglu has joined #openstack-powervm | 18:10 | |
| *** esberglu has quit IRC | 18:15 | |
| *** esberglu has joined #openstack-powervm | 18:22 | |
| *** esberglu has quit IRC | 18:24 | |
| *** esberglu has joined #openstack-powervm | 18:24 | |
| openstackgerrit | Matt Rabe proposed openstack/nova-powervm master: Add configuration option for specifying MSP IPs https://review.openstack.org/584481 | 18:30 |
| *** fried_rolls is now known as fried_rice | 18:38 | |
| openstackgerrit | Merged openstack/networking-powervm master: Match neutron's version of hacking, flake8 ignores https://review.openstack.org/582686 | 18:59 |
| fried_rice | mdrabe: Help me understand how MSP works, wouldja? | 19:14 |
| *** k0da has joined #openstack-powervm | 19:14 | |
| fried_rice | mdrabe: The part that's confusing me is, "The MSP IPs on the destination host will be saved in the | 19:14 |
| fried_rice | migration data where the source host will pass this parameter and | 19:14 |
| fried_rice | the source MSP IPs to the migration job." | 19:14 |
| fried_rice | Are the MSP IPs actually *different* addresses on the source & destination hosts? | 19:14 |
| fried_rice | for a single migration? | 19:14 |
| mdrabe | fried_rice: Yes, they could be different SEAs or VEAs on the source/destination VIOSes | 19:15 |
| fried_rice | That doesn't make sense to me. I thought the MSP was responsible for carrying data for the instance. | 19:15 |
| mdrabe | So the source host has it's MSPs in conf and so does the destination | 19:15 |
| fried_rice | butbutbut | 19:15 |
| fried_rice | the IPs would be the same?? | 19:15 |
| mdrabe | Between source and destination? No | 19:15 |
| mdrabe | The migration job has those parameters SourceMSPIPAddr and DestMSPIPAddr | 19:16 |
| fried_rice | maybe we need to back all the way up and I need a crash course on how migration works. | 19:16 |
| mdrabe | Those parameters are IP addresses belong to adapters that will host the migration | 19:16 |
| mdrabe | So if you have 10G and 1G adapters on your VIOSes, you can specify the 10G ones for migration via this parameter | 19:17 |
| fried_rice | so there's a source VIOS and a dest VIOS that have to talk to each other to effect the migration? | 19:18 |
| mdrabe | Right | 19:18 |
| fried_rice | okey. | 19:18 |
| mdrabe | The source port has to be able to talk to the dest one | 19:18 |
| *** edmondsw has quit IRC | 19:19 | |
| fried_rice | mdrabe: Did we have the dest_msp_name and source_msp_name kwargs on the migrate_lpar method in pypowervm already? | 19:19 |
| mdrabe | Yes | 19:19 |
| *** edmondsw has joined #openstack-powervm | 19:20 | |
| fried_rice | mdrabe: Thanks for the help, +2. | 19:22 |
| mdrabe | fried_rice: I -WF'd it for test, also I'm pretty sure the nova commit has to merge first | 19:23 |
| mdrabe | Also I don't know when I'll get around to getting two NL hosts and setting up devstack. If you or someone ends up doing that let me know. I'd like to piggyback | 19:24 |
| fried_rice | I have one that's stacked (though a couple months old) that you can borrow, but you'll have to find another. | 19:25 |
| fried_rice | edmondsw: ? | 19:25 |
| edmondsw | I am trying to setup a multinode devstack | 19:26 |
| fried_rice | ight | 19:26 |
| edmondsw | gonna be afk for a bit, back later | 19:26 |
| *** edmondsw has quit IRC | 19:26 | |
| *** k0da has quit IRC | 19:33 | |
| *** k0da has joined #openstack-powervm | 19:34 | |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: Spec: Device Passthrough https://review.openstack.org/579359 | 19:44 |
| *** k0da has quit IRC | 19:59 | |
| *** esberglu has quit IRC | 20:19 | |
| *** edmondsw has joined #openstack-powervm | 20:20 | |
| edmondsw | I'm back | 20:21 |
| *** k0da has joined #openstack-powervm | 20:22 | |
| *** esberglu has joined #openstack-powervm | 20:40 | |
| fried_rice | Welcome back | 20:40 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: Inventory whitelist schema and loading https://review.openstack.org/579289 | 20:43 |
| fried_rice | edmondsw: Modified spec & patch for multi-feature_codes, lemme know if you're okay with the ORed list of ANDed values. | 20:44 |
| *** esberglu has quit IRC | 20:44 | |
| edmondsw | fried_rice I like it... thanks | 20:48 |
| fried_rice | coo | 20:48 |
| edmondsw | fried_rice should I +2 or should we get that NVLink TODO resolved first? | 20:48 |
| fried_rice | edmondsw: As you wish. We can always amend later. | 20:49 |
| edmondsw | yeah, that's what I'm thinking | 20:49 |
| fried_rice | edmondsw: We certainly have enough to get started. | 20:49 |
| edmondsw | yep | 20:49 |
| edmondsw | +2 | 20:49 |
| edmondsw | mdrabe ^ dive in yourself https://review.openstack.org/#/c/579359 | 20:50 |
| mdrabe | on it | 20:51 |
| mdrabe | fried_rice: What happens in update_provider_tree if a device gets removed or replaced? | 21:25 |
| mdrabe | If the host went under maintenance for example | 21:26 |
| fried_rice | swhy we namespace the providers. We look for any provider with the namespace pattern, and if it doesn't correspond to one we discover from REST, we whack it. | 21:26 |
| fried_rice | mdrabe: Described in this section: http://logs.openstack.org/59/579359/9/check/build-openstack-sphinx-docs/b6e1cd3/html/specs/rocky/device-passthrough.html#resource-providers | 21:27 |
| mdrabe | fried_rice: Ah okay, so the purging would happen in update_provider_tree? | 21:29 |
| fried_rice | mdrabe: Well, technically update_provider_tree just removes the provider from the ProviderTree object, and then we yield control back to the compute manager, where the resource tracker (via update_from_provider_tree) does the actual work to remove it from placement. | 21:30 |
| fried_rice | It's an odd little dance. | 21:30 |
| mdrabe | makes sense, thnx | 21:32 |
| mdrabe | fried_rice: Nothing would have to be done for resize? | 21:37 |
| fried_rice | mdrabe: Like what? | 21:47 |
| mdrabe | fried_rice: The PowerVMDriver will have to update spawn to query placement, parse out the DRC index, match the corresponding slot in the MS wrapper, and update it in the LPAR wrapper | 21:57 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: Inventory whitelist schema and loading https://review.openstack.org/579289 | 21:57 |
| mdrabe | Cold migration will make use of those same changes? | 21:57 |
| fried_rice | presumably. | 21:57 |
| openstackgerrit | Eric Fried proposed openstack/nova-powervm master: Spec: Device Passthrough https://review.openstack.org/579359 | 21:59 |
| fried_rice | edmondsw, mdrabe: Updated to rip out feature codes ^^ | 21:59 |
| *** fried_rice is now known as efried | 22:09 | |
| efried | mdrabe: Is there anything to fix on that spec? | 22:10 |
| mdrabe | Nope, not that I can find | 22:11 |
| edmondsw | efried +2/+A on the spec | 22:24 |
| edmondsw | mdrabe also see https://review.openstack.org/#/c/579289/ | 22:24 |
| openstackgerrit | Merged openstack/nova-powervm master: Spec: Device Passthrough https://review.openstack.org/579359 | 22:44 |
| *** k0da has quit IRC | 23:02 | |
| *** k0da has joined #openstack-powervm | 23:39 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!