| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-05-08 | |||
| 19:09:14 | dansmith | tssurya: but I think it's going the right direction | |
| 19:19:32 | artom | aspiers, still around? You got volunteered for compute capabilities as traits during PTG, right? | |
| 19:19:47 | artom | http://lists.openstack.org/pipermail/openstack-discuss/2019-May/005906.html doesn't make much sense to me since my knowledge of placement is pretty slim | |
| 19:20:43 | artom | I want to help out with that because it helps my baby device tagging :) | |
| 19:23:47 | mriedem | kinda sucks we lost our entire ptg etherpad(s) | |
| 19:27:15 | artom | The main ones, at any rate :( | |
| 19:27:37 | artom | The governance one is still going strong :D | |
| 19:31:09 | sean-k-mooney | mriedem: i think i have a copy of the main one on my laptop | |
| 19:31:18 | sean-k-mooney | i backed up the -3 etherpad | |
| 19:41:15 | artom | sean-k-mooney, I'm specifically not asking you any of the questions that I have because I know you're on PTO and you shouldn't be here :D | |
| 19:43:54 | efried | artom, mriedem, sean-k-mooney: I copied -3 to https://etherpad.openstack.org/p/nova-ptg-train-4 | |
| 19:45:08 | efried | artom: tell me more about device tagging | |
| 19:45:10 | sean-k-mooney | artom: im technically playing a video game while listening to an audio book so im more or less on pto | |
| 19:45:28 | artom | "playing a video game while listening to an audio book" | |
| 19:45:31 | artom | Kids these days | |
| 19:45:58 | artom | efried, basically https://bugs.launchpad.net/nova/+bug/1817927 | |
| 19:46:00 | openstack | Launchpad bug 1817927 in OpenStack Compute (nova) "device tagging support is not checked during move operations" [Undecided,New] | |
| 19:46:09 | artom | And since I wrote the original device tagging stuff, I figure it's mine to fix | |
| 19:46:14 | sean-k-mooney | what can i say i get board easily. that our im optimising my time by multi tasking | |
| 19:46:55 | artom | So if the fix is using compute capabilities traits with whatever placement pre-filter black magic (excuse the ignorance) is needed, I'm on board with that, and will help | |
| 19:47:39 | efried | artom: Cool. Yes, the effort should help. In fact, I would expect we should get all of that for free once the framework is fixed up. | |
| 19:48:26 | artom | efried, right, and so that was basically my question to aspiers. IIUC he's been volunteered to fix up said framework, so wondering if/how he needed help | |
| 19:49:28 | efried | artom: He and I spent a couple of hours Saturday afternoon going over some of the background. A careful reading of that thread - especially gibi's response - should be pretty enlightening. | |
| 19:49:53 | artom | efried, ah, I was flying home then. Err, which specific thread :) | |
| 19:50:15 | artom | Summary: Implicit trait-based filters? | |
| 19:50:26 | efried | artom: Yes, the one you linked above :P | |
| 19:51:02 | artom | *snerk* *sigh* I amaze myself sometimes :( | |
| 19:53:23 | efried | The main thrust is: RequestSpec.requested_resources is currently only used for the numbered request groups coming from port bandwidth pieces. It should be accumulating *all* of the request groups - from flavor, image, request filters, and anything else that needs to inform the GET /a_c call. But apparently putting the unnumbered group in there is hard. | |
| 19:55:07 | artom | Full disclosure: I don't know the difference between a numbered request group and an unnumbered one. Or even what a request group is. I'll read the placement api-ref. | |
| 20:13:05 | tssurya | dansmith: thanks for the review :) I'll update it | |
| 20:24:06 | eandersson | The upgrade checker is confusing when it comes to Ironic Flavor Migration. The Upgrade suggests this is something that need to be done. | |
| 20:24:49 | eandersson | But the nova-manage documentation tell you do not do it https://docs.openstack.org/nova/pike/cli/nova-manage.html | |
| 20:26:22 | eandersson | Guessing that the key is this? | |
| 20:26:23 | eandersson | > This routine is provided only for advanced users that may be skipping the 16.0.0 Pike release, never able to run services normally at the Pike level | |
| 20:26:41 | eandersson | (e.g. for fast-forward?) | |
| 20:30:38 | openstackgerrit | Matt Riedemann proposed openstack/nova-specs master: Spec to pre-filter disabled computes with placement https://review.opendev.org/657884 | |
| 20:31:04 | mriedem | tssurya: dansmith: efried: ^ tis written but with lots of todo/question/alternates inline | |
| 20:31:36 | mriedem | eandersson: "never able to run services normally at the Pike level" is for FFU yes | |
| 20:32:15 | mriedem | eandersson: when you restart the ironic compute service it should migrate the embedded instance.flavor for each node it's managing automagically | |
| 20:32:28 | aspiers | artom: I learnt the difference on Sat, which is *almost* long enough ago for me to have forgotten it ;-) | |
| 20:33:06 | mriedem | https://opendev.org/openstack/nova/src/tag/18.0.0/nova/virt/ironic/driver.py#L545 | |
| 20:33:22 | artom | aspiers, hah :) So yeah, do you think you want/need help (from admittedly another placement newbie)? | |
| 20:33:48 | aspiers | artom: probably will, but have to catch up with a bunch of other stuff before I get to this | |
| 20:34:25 | aspiers | including filing expenses and sending a report about the self-healing SIG :-o | |
| 20:34:25 | mriedem | eandersson: if you go down that route, note there are some open issues https://review.opendev.org/#/c/637217/ | |
| 20:35:02 | artom | Reason I'm pestering is we have internal tracking requirements, so if I'm committing to helping with this thing, I'll need to file trackers for it sooner rather than later | |
| 20:36:34 | aspiers | artom: Got it. TBH I probably won't know how much help I need (if any) until I understand it better | |
| 20:37:24 | artom | aspiers, totally fair | |
| 20:37:46 | artom | In any case, nothing has to be decided tonight, and I have my own share of reading to do :) | |
| 20:37:52 | aspiers | +1 | |
| 20:37:52 | artom | And kids to fetch from daycare. | |
| 20:37:55 | openstackgerrit | Surya Seetharaman proposed openstack/nova-specs master: Support server power state update through external event https://review.opendev.org/636132 | |
| 20:45:40 | openstackgerrit | Matt Riedemann proposed openstack/nova-specs master: Spec to pre-filter disabled computes with placement https://review.opendev.org/657884 | |
| 21:59:16 | sorrison | mriedem: Yeah didn't think of that but that seems ok to me and desired. Controlling access to list other projects instances seems like it should be separate to the other filters I think | |
| 22:01:00 | mriedem | alright, sounds like no change necessary then | |
| 22:01:06 | mriedem | no work works for me | |
| 22:09:41 | openstackgerrit | Eric Fried proposed openstack/os-traits master: Introduce traits for NUMA and NIC subtree affinity https://review.opendev.org/657898 | |
| 22:17:28 | openstackgerrit | Rodrigo Barbieri proposed openstack/nova master: Add functional confirm_migration_error test https://review.opendev.org/657870 | |
| #openstack-nova - 2019-05-09 | |||
| 04:13:15 | openstackgerrit | Brin Zhang proposed openstack/nova master: Change the migration warning log https://review.opendev.org/657916 | |
| 04:33:36 | openstackgerrit | Merged openstack/nova stable/stein: Use migration_status during volume migrating and retyping https://review.opendev.org/657575 | |
| 06:29:15 | openstackgerrit | zhongshengping proposed openstack/nova master: Update Python 3 test runtimes for Train https://review.opendev.org/657941 | |
| 09:37:44 | openstackgerrit | zhongshengping proposed openstack/nova master: Update Python 3 test runtimes for Train https://review.opendev.org/657941 | |
| 10:07:16 | openstackgerrit | Sylvain Bauza proposed openstack/nova master: Pass allocations to virt drivers when resizing https://review.opendev.org/589085 | |
| 11:09:49 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/ocata: [DNM] Populate key_pairs from top_cell to child cells https://review.opendev.org/657995 | |
| 11:13:12 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/ocata: [DNM] Populate key_pairs from top_cell to child cells https://review.opendev.org/657995 | |
| 11:14:17 | tssurya | sorrison: ^^ its an ugly fix, but it worked for us :) I don't think it will be merged upstream but you can use it downstream. Let me know if you have any questions | |
| 11:19:21 | openstackgerrit | Surya Seetharaman proposed openstack/nova stable/ocata: [DNM] Populate key_pairs from top_cell to child cells https://review.opendev.org/657995 | |
| 13:10:34 | shilpasd | efried: Hi | |
| 13:10:42 | efried | o/ shilpasd | |
| 13:10:45 | shilpasd | Hi need bit input, i am currently working on implementation of https://review.opendev.org/#/c/650188/, Allow compute nodes to use shared storage provider for DISK_GB resources | |
| 13:11:24 | shilpasd | Gone through RESHAPE functionality implemented and here which deals with reshaping inventory and allocations | |
| 13:11:33 | shilpasd | In Shared storage Provider, similarly doing reshape where moving allocations of DISK_GB from compute node to shared storage provider | |
| 13:12:10 | shilpasd | now i want to correct local_gb_used at compute node at nova side | |
| 13:12:23 | shilpasd | so can you help me how to proceed here | |
| 13:13:34 | shilpasd | pre_start_hook() first update_available_resource() and then calls update_provider_tree() | |
| 13:14:13 | shilpasd | and during update_provider_tree() i am doing RESHAPE for DISK_GB similarly done for VGPU | |
| 13:16:01 | efried | shilpasd: let me go look at some code... | |
| 13:16:15 | efried | IIRC local_gb_used is a bit tricky | |
| 13:17:59 | efried | shilpasd: I'm sort of thinking we may want local_gb_used to be set to zero when the host is using shared disk. | |
| 13:18:10 | efried | mriedem, gibi, dansmith: You guys have a take on this? ^ | |
| 13:24:28 | mriedem | idk, i'd think local_gb_used would at least report reserved_host_disk_mb (which defaults to 0) | |
| 13:25:19 | mriedem | otherwise i'd tend to agree that it should be 0 for instance usage in the resource tracker if you're using shared storage on that node | |
| 13:25:39 | mriedem | mdbooth might have other wrinkles in mind there | |
| 13:28:42 | dansmith | yea, what mriedem said... because in the future we could have the compute *also* reporting its own disk inventory if it gained the ability to do that thing | |
| 13:28:48 | mdbooth | Hmm, I'd need to look at code. Depends how local_gb_used is used. | |
| 13:36:43 | efried | mdbooth: The only nod to thin provisioning is allocation_ratio | |
| 13:36:50 | efried | which amounts to a wild-ass guess | |
| 13:37:19 | efried | and is why cinder has shown minimal enthusiasm for adopting placement to do their tracking | |
| 13:37:22 | mdbooth | IIRC local_gb_used is all about thin provisioning. | |
| 13:37:37 | efried | oh? | |
| 13:37:52 | mdbooth | IIRC it answers the question: can I migrate here? | |
| 13:38:56 | mdbooth | It's weird that we don't track actual usage in placement. | |
| 13:39:40 | efried | mdbooth: you would have to be sending updates all the time | |
| 13:40:00 | mdbooth | Yeah | |
| 13:40:04 | efried | the actual usage of thin-provisioned storage changes every time the guest does a write | |
| 13:40:23 | efried | also, somebody told me that the storage doesn't ever actually tell you how much space it really has anyway. | |
| 13:40:31 | efried | the more expensive the storage, the less accurate | |
| 13:40:32 | mdbooth | *some* storage | |
| 13:40:44 | efried | which seems bizarre to me | |
| 13:40:46 | mdbooth | We don't use that in Nova, though. | |