| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-06-06 | |||
| 13:16:21 | sahid | yes, that is why i wanted ping you | |
| 13:19:16 | sahid | melwitt: is this | |
| 13:19:51 | sahid | https://review.openstack.org/#/c/567472/ cannot have to be considered part of: bug/1686116 | |
| 13:20:12 | sahid | s/cannot have/cannot be | |
| 13:23:01 | mriedem | sahid: there was a blueprint for that | |
| 13:23:03 | mriedem | not sure why it's a bug now | |
| 13:24:00 | mriedem | sahid: i'll fix my nits in https://review.openstack.org/#/c/553072/ | |
| 13:31:12 | openstackgerrit | Chen proposed openstack/python-novaclient master: Fix typos https://review.openstack.org/572770 | |
| 13:32:33 | mriedem | dansmith: not sure who else should take a look at this network_data.json metadata API spec, but you're probably closest https://review.openstack.org/#/c/312626/ | |
| 13:33:29 | dansmith | I thought we had already done that actually | |
| 13:35:43 | jangutter | sahid: Thanks. It doesn't particularly bother me to use naked VIFHostDevice aside from breaking the pattern. | |
| 13:39:31 | sahid | jangutter: which pattern? for VIFHostDevice that looks redondant information | |
| 13:39:55 | mriedem | dansmith: i think that was a slightly different thing that was a bug fix | |
| 13:39:58 | openstackgerrit | Matt Riedemann proposed openstack/nova master: add mtu to libvirt xml for ethernet and bridge types https://review.openstack.org/553072 | |
| 13:40:32 | mriedem | dansmith: this https://github.com/openstack/nova/commit/4b333b989dfc778a8b61db4a1b8552e988a10471#diff-083754c4012f16fa0f9dd9cd703ee0c1 | |
| 13:41:14 | dansmith | ah yep | |
| 13:41:17 | jangutter | sahid: yep, but in https://github.com/openstack/nova/blob/stable/queens/nova/network/os_vif_util.py#L287-L297 it's used, and in https://github.com/openstack/nova/blob/stable/queens/nova/network/os_vif_util.py#L329-L347 the apparent redundancy is explained (poorly, by me) | |
| 13:41:39 | openstackgerrit | Merged openstack/nova master: Add traits check in nested provider candidates https://review.openstack.org/567150 | |
| 13:41:42 | mriedem | dansmith: and looking back, i'm still not sure why we didn't have a new metadata api version for that since it was adding a new key to the response | |
| 13:41:47 | mriedem | we'd never let that fly in the compute rest api | |
| 13:42:02 | mriedem | but metadata api is the wild west pew pew | |
| 13:42:26 | dansmith | well, probably because it's a fairly small iteration | |
| 13:42:33 | dansmith | and because of how metadata is | |
| 13:42:33 | dansmith | but yeah | |
| 13:42:43 | mriedem | sahid: +2 on https://review.openstack.org/#/c/553072/ now | |
| 13:42:56 | dansmith | mriedem: so I haven't looked at this spec in the past, but has someone gone through and looked at what the delta looks like from the existing? | |
| 13:43:15 | sahid | mriedem: ok cool | |
| 13:43:22 | dansmith | or is it just that we only ever populate networks[0] now? | |
| 13:43:39 | mriedem | dansmith: it adds a new ip_addresses field to the response which is a list of fixed ips rather than the first one from the list | |
| 13:43:47 | mriedem | we only take the first fixed ip today | |
| 13:43:53 | mriedem | which is the ip_address field | |
| 13:43:58 | sahid | jangutter: hum... i don't understand the note | |
| 13:43:59 | mriedem | this exposes the list of fixed ips | |
| 13:44:14 | mriedem | and still maintains the single ip_address field - which arguably we don't need to do since it's versioned | |
| 13:44:41 | mriedem | i don't know how sensitive apps are to doing version discovery with metadata | |
| 13:44:53 | dansmith | mriedem: this certainly looks like we're asserting multiple interfaces and ips in the test no? https://github.com/openstack/nova/blob/4b333b989dfc778a8b61db4a1b8552e988a10471/nova/tests/unit/network/test_network_info.py | |
| 13:44:59 | dansmith | er hang on | |
| 13:45:04 | dansmith | https://github.com/openstack/nova/blob/4b333b989dfc778a8b61db4a1b8552e988a10471/nova/tests/unit/network/test_network_info.py#L1118 | |
| 13:45:30 | mriedem | dansmith: how so? | |
| 13:45:30 | mriedem | https://github.com/openstack/nova/blob/4b333b989dfc778a8b61db4a1b8552e988a10471/nova/tests/unit/network/test_network_info.py#L1121 | |
| 13:45:35 | mriedem | that's a single ipv4 fixed ip | |
| 13:45:43 | mriedem | but you can have multiple fixed ips per port in neutron | |
| 13:45:53 | mriedem | which is what this spec is trying to expose via network_data.json | |
| 13:46:04 | dansmith | oh, | |
| 13:46:10 | dansmith | I thought the spec was saying it was the subsequent networks and the ips there that it needed to add | |
| 13:46:12 | jangutter | sahid: Back in the day, when I wrote that, my rationale was something like: the VIF is used to generate the xml for libvirt, the port_profile is used to find the representor and plug it into the bridge. | |
| 13:46:16 | mriedem | there is also something in the spec about multiple subnet routes which i didn't fully follow | |
| 13:46:51 | mriedem | i believe the networks list is per-vif (port) | |
| 13:46:53 | dansmith | I see, multiple fixed per port | |
| 13:47:00 | dansmith | is what they're enabling | |
| 13:47:01 | mriedem | similar to how the server "addresses" field works in the compute api | |
| 13:47:08 | mriedem | dansmith: yeah | |
| 13:47:11 | dansmith | sorry, I just skimmed when you sent and haven't really read it | |
| 13:47:32 | mriedem | so you didn't see the dan smith easter egg in there yet | |
| 13:48:17 | jangutter | sahid: Also, the symmetry will break either way, there's only so far you can take the abstraction, I'm agnostic on this, but if the pattern is continued, then at least the os-vif plugins look similar. | |
| 13:49:09 | dansmith | mriedem: well now I can't approve until I find something I think qualifies | |
| 13:52:15 | mriedem | melwitt: ack, and thanks for hitting those placement policy changes | |
| 13:53:52 | kashyap | lyarwood: Seems like your recheck has resulted in a success for the 'legacy-tempest-dsvm-neutron-full'? | |
| 13:54:02 | kashyap | (For this: https://review.openstack.org/#/c/571433/) | |
| 13:54:20 | kashyap | But there's other failures, though | |
| 13:57:01 | kashyap | (Like this one 'test_volume_extend_when_volume_has_snapshot') | |
| 14:01:26 | openstackgerrit | Merged openstack/nova-specs master: Multiple Fixed-IPs support in network information https://review.openstack.org/312626 | |
| 14:04:11 | mriedem | mdbooth: artom: replied in the ML, but if we're going to block retype/live migration of multiattach volumes with >1 read/write attachment, that should happen in the cinder API | |
| 14:04:34 | mriedem | failing in the swap volume api in nova is about 3/4 of the way through the operatoin which is a real shitty place to be failing | |
| 14:05:36 | mdbooth | mriedem: Reading. Was just responding to jaypipes understandable horror at this messy interface. | |
| 14:06:12 | artom | mriedem, unless it's called directly, no? Which apparently some people are doing | |
| 14:07:05 | mriedem | mgagne: https://blueprints.launchpad.net/nova/+spec/multiple-fixed-ips-network-information is now approved so you can post code | |
| 14:07:15 | mriedem | artom: who is doing that? | |
| 14:07:29 | artom | mriedem, see mdbooth's reply :) | |
| 14:07:31 | mriedem | swap volume in nova is an admin-only api | |
| 14:07:42 | artom | We also got a downstream but about it | |
| 14:07:42 | mdbooth | mriedem: We really do have customers who call swap volume directly. | |
| 14:07:48 | mdbooth | More than one. | |
| 14:08:03 | mriedem | why? | |
| 14:08:29 | kashyap | Presumably to move storage from one type (local) to another (NFS)? | |
| 14:08:33 | mdbooth | I still haven't nailed it down, but I *think* it's a workaround related to BFV. | |
| 14:08:41 | artom | mriedem, I mean, I would agree, it's a weird thing to do, but we can't honestly put an API out there and get defensive when people use it in unexpected ways | |
| 14:08:42 | mriedem | kashyap: that would be a retype | |
| 14:08:47 | mriedem | which you should initiate through cinder | |
| 14:08:54 | kashyap | Ah, damned if I ever remember the names | |
| 14:09:13 | mdbooth | mriedem: But regardless, even if we think it's dumb, it's a public api and they're using it. | |
| 14:10:04 | mriedem | awesome | |
| 14:10:16 | mdbooth | Indeed | |
| 14:10:18 | mriedem | i made it 70 minutes today before hating my life | |
| 14:10:25 | mdbooth | \o/ | |
| 14:10:26 | kashyap | This morning we chatted here about mdbooth's e-mail (which I haven't pulled down yet). At least in context of what libvirt APi does by default | |
| 14:10:38 | jaypipes | mdbooth: my annoyance is that this is a libvirt/QEMU-specific behaviour/implementation that is leaking out of the Compute API. | |
| 14:10:59 | kashyap | jaypipes: By _default_ libvirt doesn't copy shared disks | |
| 14:11:02 | mdbooth | jaypipes: That was the thrust of my reply. It's not libvirt/qemu specific at all. | |
| 14:11:09 | jaypipes | mdbooth: BTW, where exactly does "swap volume" imply or denote that there is a copy operation going on? :( | |
| 14:11:09 | kashyap | But if you _ask_ for it, it will migrate it | |
| 14:11:16 | mriedem | mdbooth: the only virt driver that implements this is libvirt | |
| 14:11:21 | mdbooth | jaypipes: It could be implemented by any hypervisor, and they'd have to do the copy. | |
| 14:11:22 | jaypipes | mdbooth: yes, it is. | |
| 14:11:32 | mriedem | same with guest-assisted volume snapshot | |
| 14:11:42 | mdbooth | jaypipes: I replied in depth on the ML, anyway. | |
| 14:11:50 | jaypipes | yes, I read your reply. | |
| 14:11:57 | mriedem | i have no idea how non-libvirt drivers handle volume retype/live migration either | |
| 14:12:07 | jaypipes | mriedem: they don't. wisely. | |
| 14:12:10 | mdbooth | But Cinder can't do this, regardless of which hypervisor is attached. | |