| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-10-19 | |||
| 22:04:28 | mikal | So, the commit which added any instance accessible metadata functionality | |
| 22:04:46 | cburgess | mriedem Sure | |
| 22:04:49 | mriedem | "copy what ec2 / aws does"? | |
| 22:04:54 | cfriesen | mriedem: for what it's worth, the largest EC2 instances have 48TB of disk | |
| 22:05:02 | cburgess | mikal Um... probably somewhere in in the first few commits if not the initial import. | |
| 22:05:39 | cfriesen | mriedem: for the "abort cold migration" thing, what about the case where the dest node crashes mid-transfer? | |
| 22:05:54 | mriedem | cfriesen: wouldn't that auto-kill the cold migrate? | |
| 22:05:56 | cburgess | cfriesen I think thats covered already isn't it? Don't we detect that? | |
| 22:06:15 | cburgess | mriedem cfriesen Oh maybe not now that we can't monitor disk transfer progress. | |
| 22:06:17 | cfriesen | I'm thinking the connection could stay open until the TCP connection times out | |
| 22:06:31 | cfriesen | in the scp case, at least | |
| 22:06:32 | cburgess | Oh right... libvirt will report the error back. | |
| 22:06:39 | cburgess | At some point | |
| 22:06:40 | cburgess | In theory | |
| 22:06:54 | cburgess | mikal So whats the answer? | |
| 22:07:10 | mikal | cburgess: commit bf6e6e718cdc7488e2da87b21e258ccc065fe499 | |
| 22:07:17 | mikal | cburgess: which has the commit message "initial import" | |
| 22:07:32 | mikal | cburgess: at that point, nova was 5,083 lines of code | |
| 22:07:52 | cburgess | mikal Called it! | |
| 22:08:45 | mikal | In other words, its always been weird | |
| 22:08:54 | cburgess | wow 5K lines of code. I have scripts to fix horked up volumes that are longer then that... | |
| 22:09:40 | cburgess | OK thats a lie.. its only 1349 lines of code. | |
| 22:19:20 | cburgess | mriedem Is the existing abort admin only? Or is that policy controlable? | |
| 22:19:46 | mriedem | cburgess: anytihng related to live migration is admin-only by default | |
| 22:19:55 | mriedem | and cold migration | |
| 22:20:16 | cburgess | mriedem OK | |
| 22:20:48 | mikal | cburgess: I'm not sure volumes existed in the first commit, I'd have to check... | |
| 22:21:13 | cburgess | mikal What do volumes have to do with anything? | |
| 22:21:28 | mikal | cburgess: well, your volume fix script... | |
| 22:21:36 | mikal | cburgess: probably wasn't required back then | |
| 22:21:55 | cburgess | mikal My volume fix script wasn't required until the hot mess that is/was os-brick circa liberty. | |
| 22:28:50 | openstackgerrit | Matt Riedemann proposed openstack/nova-specs master: Use neutron's new port binding API for live migration https://review.openstack.org/375580 | |
| 22:34:57 | mriedem | dansmith: i updated ^ based on some review i had after sean updated it from the call this morning | |
| 22:35:09 | mriedem | dansmith: one sticky thing in there yet which i know you won't like, | |
| 22:35:14 | mriedem | but i think we can hammer in code review | |
| 22:36:23 | dansmith | yeah I don't really think the argument of "avoid changing the object due to stuff in ml2" is sound | |
| 22:36:40 | dansmith | but as long as we can argue in the code I guess it's fine | |
| 22:45:11 | openstackgerrit | Merged openstack/nova-specs master: Use neutron's new port binding API for live migration https://review.openstack.org/375580 | |
| 23:26:05 | openstackgerrit | Merged openstack/nova-specs master: Add pagination and changes since filter support for os-instance-action API https://review.openstack.org/507762 | |
| 23:51:18 | gmann | bauzas: i replied on your comment on this spec - https://review.openstack.org/#/c/508101/7 | |
| 23:52:02 | gmann | bauzas: let me know if i need to make it clear about those 'Main policy' in spec. i can do that quickly | |
| #openstack-nova - 2017-10-20 | |||
| 00:30:12 | openstackgerrit | Ghanshyam Mann proposed openstack/nova-specs master: Spec for API extensions policy removal https://review.openstack.org/508101 | |
| 00:30:29 | gmann | bauzas: ^^ | |
| 01:33:42 | yikun_jiang | liusheng, test | |
| 01:48:36 | openstackgerrit | garyk proposed openstack/nova master: Add debug information to metadata requests https://review.openstack.org/511895 | |
| 01:51:49 | openstackgerrit | Ghanshyam Mann proposed openstack/nova-specs master: Spec for API extensions policy removal https://review.openstack.org/508101 | |
| 06:36:08 | openstackgerrit | Zhenyu Zheng proposed openstack/nova master: nova-manage db archive_deleted_rows is not multi-cell aware https://review.openstack.org/507486 | |
| 06:39:55 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Implement query param schema for simple_tenant_usage https://review.openstack.org/507480 | |
| 07:40:51 | SamYaple | will nova have python3 support by Queens? | |
| 07:41:35 | SamYaple | looks for the latest info on python3 for nova. is this up to date? PI extensions policy removal https://review.openstack.org/508101 | |
| 07:41:46 | SamYaple | oops wrong link https://wiki.openstack.org/wiki/Python3 | |
| 08:13:03 | gmann | SamYaple: this is list remaining seems - https://github.com/openstack/nova/blob/master/tests-py3.txt | |
| 08:14:10 | SamYaple | gmann: oh perfect! so it looks like just missing support for xen | |
| 08:14:11 | gmann | SamYaple: this was whole effort for py3 - https://review.openstack.org/#/q/topic:bp/goal-python35+(status:open+OR+status:merged)+project:+openstack/nova | |
| 08:14:17 | gmann | yea | |
| 08:14:24 | SamYaple | awesome. i dont use xen :) | |
| 08:14:29 | SamYaple | wonderful news | |
| 08:14:33 | SamYaple | thanks for that | |
| 08:18:06 | kashyap | stephenfin: Hi, do you know top off your head if we can define a custom flavour that will set a specific image metadata property? | |
| 08:31:19 | stephenfin | Not that I'm aware of, no. It had to be part of the image itself | |
| 08:31:25 | stephenfin | kashyap: ^ | |
| 08:31:57 | kashyap | stephenfin: Thought so, thanks for confirming. | |
| 08:40:56 | openstackgerrit | jichenjc proposed openstack/nova master: Refactor placement version check https://review.openstack.org/512497 | |
| 08:52:20 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Implement query param schema for delete assisted vol https://review.openstack.org/513668 | |
| 08:57:52 | openstackgerrit | David Rabel proposed openstack/nova master: VMware: add support for graceful shutdown of instances https://review.openstack.org/494169 | |
| 09:04:49 | openstackgerrit | David Rabel proposed openstack/nova master: VMware: add support for graceful shutdown of instances https://review.openstack.org/494169 | |
| 09:20:15 | openstackgerrit | Zhenyu Zheng proposed openstack/nova master: nova-manage db archive_deleted_rows is not multi-cell aware https://review.openstack.org/507486 | |
| 09:22:30 | kashyap | Hi bauzas, when you get a moment: | |
| 09:22:40 | bauzas | sure ? | |
| 09:22:52 | kashyap | Is such a rule possible: If the instance has "hw_machine_type=x86_64=pc-q35-2.9" property, then when migrating it, please choose a Compute node that has that machine type configured (because: there's also the same nova.conf attribute that can be set for the whole Compute) | |
| 09:25:47 | openstackgerrit | Dmitry Tantsur proposed openstack/nova stable/pike: Pick ironic nodes without VCPU set https://review.openstack.org/513678 | |
| 10:03:52 | openstackgerrit | Ghanshyam Mann proposed openstack/nova master: Implement query param schema for host index https://review.openstack.org/513699 | |
| 10:05:20 | openstackgerrit | Chris Dent proposed openstack/nova master: WIP: [placement] Enable limiting GET /allocation_candidates https://review.openstack.org/513526 | |
| 11:20:29 | openstackgerrit | Huan Xie proposed openstack/nova master: VGPU: Define vgpu resource class https://review.openstack.org/511180 | |
| 11:21:50 | artom_ | stephenfin, since https://review.openstack.org/#/c/345396/22/nova/console/websocketproxy.py is being held up by CI, do you think https://review.openstack.org/#/c/345399/29/nova/console/websocketproxy.py@82 could be addressed? | |
| 11:22:01 | artom_ | My OCD would feel much better :) | |
| 11:22:10 | artom_ | (But I understand if you don't feel like chasing the +2 again) | |
| 11:22:36 | openstackgerrit | Huan Xie proposed openstack/nova master: VGPU: Define vgpu resource class https://review.openstack.org/511180 | |
| 12:18:18 | openstackgerrit | Takashi NATSUME proposed openstack/nova stable/pike: Add 'delete_host' command in 'nova-manage cell_v2' https://review.openstack.org/513717 | |
| 12:30:10 | openstackgerrit | Takashi NATSUME proposed openstack/nova stable/pike: Add 'delete_host' command in 'nova-manage cell_v2' https://review.openstack.org/513717 | |
| 12:32:15 | openstackgerrit | Takashi NATSUME proposed openstack/nova stable/ocata: Add 'delete_host' command in 'nova-manage cell_v2' https://review.openstack.org/513721 | |
| 12:34:25 | openstackgerrit | Takashi NATSUME proposed openstack/nova master: Fix 500 error while passing 4-byte unicode data https://review.openstack.org/407514 | |
| 12:38:10 | stephenfin | artom: I can split it out into a separate patch, but I won't mess with the first one once it's got +W | |
| 12:38:16 | stephenfin | *I will | |
| 12:38:44 | stephenfin | For reference, that got added because the original patch was written without Python 3 being a consideration | |
| 12:41:58 | artom | stephenfin, ah, gotcha | |
| 12:42:07 | artom | I'm obviously not super hung up about it | |
| 13:27:39 | openstackgerrit | sahid proposed openstack/nova-specs master: update: introducing isolate emulthreads on host https://review.openstack.org/511188 | |
| 13:43:39 | efried | Gah, Friday! | |
| 13:44:18 | figleaf | fried_rice: usually people have that reaction to Mondays | |
| 13:45:51 | gibi | this is the time (and the same time on Monday) when I like to be in an EU timezone. For me it is almos over. ;) | |
| 13:54:46 | fried_rice | leakypipes Now a good time to talk about same RCs on main & shared RPs? | |
| 13:55:49 | leakypipes | fried_rice: yuppers. | |
| 13:56:09 | fried_rice | Okay, so I opened a couple bugs and put up a couple test cases that demonstrate the issue: | |
| 13:56:20 | fried_rice | https://review.openstack.org/#/c/513149/ | |
| 13:56:26 | fried_rice | (bugs in the commit msg) | |
| 13:57:00 | fried_rice | The crux of the issue happens when the main RP (the compute node for now) has a resource class that's also present in the shared RP(s) | |
| 13:58:05 | fried_rice | The end game with shared RPs is post-Q, but it's going to come into play with the code alex_xu is writing to handle traits for allocation candidates. | |
| 13:58:55 | fried_rice | Using the example where the RC in question is DISK_GB, the problem is that you can say you want traits "RAID" and "SSD", but placement doesn't know that those are disk-y traits. | |
| 13:59:47 | fried_rice | So if your compute node has localdisk that's SSD, and your shared storage RP has disk that's RAID, you'll get a candidate for (compute node + shared RP) because the combination satisfies both traits. | |