| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-04-27 | |||
| 15:36:05 | mriedem | on a complete tangent, it seems like we should be recording instance actions for swap volume, since that's a thing that happens mostly out of band from a nova api pov | |
| 15:36:11 | mriedem | since cinder initiates it via retype | |
| 15:38:15 | mriedem | dansmith: btw, https://review.openstack.org/460377 turned out to be kind of a beast | |
| 15:38:34 | dansmith | sweet | |
| 15:38:39 | johnthetubaguy | dansmith: so we don't really use the fqdn in here: https://github.com/openstack/nova/blob/master/nova/conf/netconf.py#L55 although I could be reading the docs wrongly: https://docs.python.org/2/library/socket.html#socket.gethostname | |
| 15:39:06 | dansmith | johnthetubaguy: but that's for things like telling other computes how to contact us | |
| 15:39:12 | dansmith | johnthetubaguy: not related to service identity, AFAIK | |
| 15:39:28 | johnthetubaguy | dansmith: we use the conf to register the service though, I thought | |
| 15:39:32 | dansmith | oh, host | |
| 15:39:47 | dansmith | sorry, I was reading the lines wrong, and thought you were talking about the storage_ip | |
| 15:39:55 | johnthetubaguy | dansmith: ah, doh | |
| 15:40:13 | dansmith | johnthetubaguy: so, yeah I guess they could configure it to be wrong | |
| 15:40:16 | mdbooth | zioproto: Assuming that the instance in question was created (from Nova's pov) from an image, does if have hw_architecture defined? | |
| 15:40:29 | mdbooth | zioproto: Is it a 32bit arch by any chance? | |
| 15:40:33 | dansmith | johnthetubaguy: but again, if they have dupes, we'll fail consistency checks AFAIK | |
| 15:40:56 | johnthetubaguy | dansmith: well, not in cells v1, its a different DB, I think | |
| 15:41:05 | dansmith | johnthetubaguy: but it's all merged at the top | |
| 15:41:11 | dansmith | johnthetubaguy: https://github.com/openstack/nova/blob/master/nova/db/sqlalchemy/models.py#L81-L84 | |
| 15:41:28 | dansmith | johnthetubaguy: you'll still end up with all the services up there, right? else none of those APIs will work | |
| 15:41:31 | johnthetubaguy | dansmith: good point, it must be like that today | |
| 15:41:43 | zioproto | mdbooth: <type arch='x86_64' machine='pc-i440fx-xenial'>hvm</type> | |
| 15:41:51 | johnthetubaguy | dansmith: seems we are covered either way then, cool | |
| 15:42:04 | mdbooth | zioproto: I mean does the glance image have a hw_architecture image property? | |
| 15:42:18 | dansmith | johnthetubaguy: just like getting rid of these ambiguous id-using apis, if we have people with dupes here, we might as well flush them out :) | |
| 15:42:23 | zioproto | So I tried booting with only 1 volume attached, and the attach the second volume later. But I get libvirtError: Operation not supported: disk bus 'ide' cannot be hotplugged. | |
| 15:42:59 | johnthetubaguy | dansmith: seems a good opportunity, it can only be creating strange subtable bugs in other places | |
| 15:43:08 | dansmith | johnthetubaguy: if even possible, yeah | |
| 15:43:39 | zioproto | mdbooth: no, but now I see it has hw_disk_bus='ide' | |
| 15:43:44 | zioproto | so maybe hda comes from there | |
| 15:43:45 | mdbooth | boom! | |
| 15:44:25 | zioproto | but the volume created from that image does not have this property | |
| 15:44:29 | zioproto | at least I cant find it | |
| 15:44:48 | openstackgerrit | Rick Bartra proposed openstack/nova master: Add policy granularity to the Flavors API https://review.openstack.org/449288 | |
| 15:45:14 | mdbooth | zioproto: It's looking at the glance image | |
| 15:45:36 | zioproto | can I remove the property from the glance image ? | |
| 15:45:47 | mdbooth | Actually, no. It's looking at the glance image metadata, which will have been stored invisibly with the volume. | |
| 15:46:19 | zioproto | mdbooth: in the cinder database ? | |
| 15:46:28 | mdbooth | Yes | |
| 15:46:46 | mdbooth | I don't think anything exposes it, though | |
| 15:46:50 | mdbooth | But I could be wrong | |
| 15:47:50 | mriedem | yeah the glance image meta is stored all over the place... | |
| 15:48:11 | mdbooth | So that's why it's ide | |
| 15:48:17 | openstackgerrit | Balazs Gibizer proposed openstack/nova master: Remove **kwargs passing in payload __init__ https://review.openstack.org/418489 | |
| 15:48:19 | mdbooth | Why does it fail to boot? | |
| 15:48:23 | mriedem | and when you snapshot a server with volumes attached, the bdm meta goes into the image snapshot meta | |
| 15:49:16 | zioproto | cant find this ide metadata doing select * from volume_glance_metadata where volume_id=uuid | |
| 15:51:06 | zioproto | I give up for today | |
| 15:51:10 | zioproto | thanks for the help | |
| 15:51:20 | zioproto | I will give feedback once this is fixed | |
| 15:51:45 | zioproto | it is probably a corner case because this VM with this cinder volume was created in a weird way | |
| 15:51:54 | zioproto | what is sure is that it booted once in the past | |
| 15:52:00 | zioproto | thank you | |
| 15:52:02 | mriedem | zioproto: i also wouldn't be surprised if swap volume or something was involved | |
| 15:52:17 | mriedem | maybe cinder has a record of volume retype/migrations you can find? | |
| 15:52:26 | zioproto | I have no idea | |
| 15:52:34 | zioproto | in the action log of nova I can find only the start and stop | |
| 15:53:02 | mriedem | right we don't record instance actions for swap volume apparently | |
| 15:53:05 | mriedem | which seems like a gap to me | |
| 15:57:16 | johnthetubaguy | mriedem: there was a spec up at one point to add instance actions into all the missing APIs, I think that had a list of them missing ones | |
| 15:57:24 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Plumbing for tagged nic attachment https://review.openstack.org/415622 | |
| 15:57:26 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Delete nic metadata when detaching interface https://review.openstack.org/435387 | |
| 15:57:27 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Libvirt support for tagged volume attachment https://review.openstack.org/418100 | |
| 15:57:31 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Delete disk metadata when detaching volume https://review.openstack.org/435082 | |
| 15:57:37 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Plumbing for tagged volume attachment https://review.openstack.org/391941 | |
| 15:57:37 | mriedem | johnthetubaguy: ok so just an oversight and not intentional | |
| 15:57:37 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: API support for tagged device attachment https://review.openstack.org/376021 | |
| 15:57:40 | openstackgerrit | Artom Lifshitz proposed openstack/nova master: Libvirt support for tagged nic attachment https://review.openstack.org/415520 | |
| 15:57:51 | johnthetubaguy | mriedem: yeah, its just a not finished thing, I believe | |
| 15:58:56 | johnthetubaguy | mriedem: as volume operations don't change the server state, generally, they are in a slight bit of limbo in our error handling, honestly | |
| 16:00:15 | mriedem | the task state changes | |
| 16:00:26 | imacdonn | mdbooth: still around ? | |
| 16:00:37 | prometheanfire | mriedem: you know about the novaclient issues (8.0.0 not in UC because it looks like it breaks things) https://review.openstack.org/454164 | |
| 16:00:42 | mriedem | nova/cinder meeting in #openstack-meeting-cp | |
| 16:00:56 | mriedem | prometheanfire: i know osc and horizon had issues | |
| 16:00:58 | mriedem | b/c of nova-network | |
| 16:01:11 | prometheanfire | k, just making sure you were aware | |
| 16:01:20 | mriedem | prometheanfire: it came up in the mailing list about a month ago i think | |
| 16:01:27 | mriedem | but yeah | |
| 16:01:27 | mriedem | thanks | |
| 16:01:35 | prometheanfire | ya, it's a older review now | |
| 16:02:10 | sdague | prometheanfire: horizon said they were dropping it | |
| 16:02:16 | mdbooth | zioproto: That metadata will only be set on the root volume, btw | |
| 16:02:28 | mdbooth | imacdonn: Yep | |
| 16:02:44 | imacdonn | mdbooth: hi ... just checking if you're still following https://bugs.launchpad.net/nova/+bug/1683972 ? | |
| 16:02:44 | openstack | Launchpad bug 1683972 in os-brick "Overlapping iSCSI volume detach/attach can leave behind broken SCSI devices and multipath maps." [Undecided,New] | |
| 16:02:58 | mriedem | johnthetubaguy: also related to cinder telling nova to do things https://specs.openstack.org/openstack/nova-specs/specs/pike/approved/nova-support-attached-volume-extend.html | |
| 16:03:07 | mriedem | ^ i brought up the error handling and instance actions stuff in that too | |
| 16:03:14 | mriedem | b/c if/when it fails we need a way to report it | |
| 16:03:42 | imacdonn | mdbooth: I added os-brick and geguileo to it, but haven't heard anything from that side either | |
| 16:03:54 | mdbooth | imacdonn: Just reading your comments | |
| 16:04:03 | imacdonn | mdbooth: k, tnx | |
| 16:04:54 | mdbooth | imacdonn: Did you get the chance to try with the very latest os-brick? | |
| 16:05:24 | mdbooth | Ignore me, I see you did | |
| 16:05:40 | imacdonn | mdbooth: I applied the patch for that specific fix... I suppose it's possible that something else changed in another commit, but I doubt it | |
| 16:10:36 | sfinucan | johnthetubaguy: This patch hit a merge conflict. Fancy reviewing it again to get it in? https://review.openstack.org/#/c/418489 | |
| 16:10:50 | sfinucan | (previously +w'd, but very prone to merge conflict) | |
| 16:11:15 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Replace reserved_host_disk_mb with reserved_host_disk_gb https://review.openstack.org/460661 | |
| 16:11:15 | openstackgerrit | Stephen Finucane proposed openstack/nova master: trivial: Remove dead code https://review.openstack.org/460662 | |
| 16:11:24 | mdbooth | imacdonn: So you're referring to _disconnect_volume_iscsi in os-brick, right? | |
| 16:11:34 | sfinucan | ^ edleafe, jaypipes: You might like the first of those | |