Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2020-06-17
18:58:35 melwitt make sense
18:58:36 sean-k-mooney but the downside is we loose contol over that
18:58:45 sean-k-mooney so our guest abi will alter with theres
18:59:03 sean-k-mooney that is ok i guess since you can use the image property if you care
18:59:16 melwitt and yeah that's the code I was looking and the helper function being called there
19:00:21 sean-k-mooney ya so this https://github.com/openstack/nova/blob/6ed556f4adfd8a98fae2e91204f1ce0484f5dd9d/nova/virt/libvirt/driver.py#L5351-L5367 is basically working around libvit/qemu not doing the right thing by default
19:01:43 sean-k-mooney well and the fact we set a default in the xml generation class
19:01:51 sean-k-mooney so if we remove that default
19:02:02 sean-k-mooney then we might not need all those specal cases
19:02:37 melwitt ok, well one step at a time I guess :P
19:02:52 melwitt I do some local tests to see what happens when I remove the default
19:02:57 melwitt *I'll
19:03:45 sean-k-mooney ok let me know an i can review. https://www.kraxel.org/blog/2019/09/display-devices-in-qemu/ is a review of some to the option and recomendations
19:04:21 melwitt thanks
19:04:21 sean-k-mooney i need to remove the gop device i added by mistake last cycle and add https://www.kraxel.org/blog/2019/09/display-devices-in-qemu/#bochs-display
19:04:45 sean-k-mooney apparently gop is only for the bhve hypervior but the libvirt doc did not mention that
19:06:40 sean-k-mooney bochs implements a uefi graphic output protocol (gop) device for kvm
19:07:03 sean-k-mooney so that is the one i should have added
19:10:06 sean-k-mooney melwitt: during the ptg we disucssed breifly the idea of a resize for something other then a flavor change
19:10:58 sean-k-mooney maybe this could be another usecause for that
19:12:19 sean-k-mooney i think this was in the context of changing machine_types where i was suggeign a new retype server action but perhaps something more generic that would update the embed flavor and image with the latest would make more sense
19:12:40 sean-k-mooney e.g. a recreate action but that preserved all ports, volume and disks
19:14:13 melwitt how was that suggestion received, I'm curious heh
19:14:57 melwitt like, I know lots of people would like to do that. in my case, I'm trying to recover from what I consider a bug, the default being set to a deprecated thing and glitching out the vnc console
19:15:20 sean-k-mooney reusing resize was dansmith idea i was thinking it shoudl be its onw server action. not sure if dansmith actully was in favor or just saying if it was to be done i would start there
19:15:45 melwitt ah, cool
19:16:08 sean-k-mooney melwitt: the main think would be that recreate would have to be a move operation like resize so we can validate the new requirement
19:16:19 sean-k-mooney since we cant assume the current host is valid anymore
19:16:36 melwitt yeah. well, we validate the new stuff on a rebuild too since a few releases ago
19:16:51 sean-k-mooney but i think it would fix a lot of the thing that hard reboot fix but which require the flaovr or image properties to be modified
19:17:06 sean-k-mooney melwitt: only if the image changes
19:17:13 melwitt yeah, it would make life a lot easier for a lot of people
19:17:46 sean-k-mooney yeah so i think it has merit but not sure where it sits in the priortiy queue
19:17:49 melwitt yeah, I'm just saying that we validate even during rebuild so I think it makes sense to validate on a recreate/resize if we're consuming new image props etc
19:18:03 melwitt we're already doing it
19:18:10 melwitt *doing similar
19:18:21 sean-k-mooney for me i think it would be a good way to opt into changing the machine type for example
19:19:59 sean-k-mooney anyway im going to go buy icecream and i guess other food before the stores start to close
19:22:53 sean-k-mooney hum not sure if good pr or pr blunder https://www.benjerry.ie/flavours/netflix-and-chillld-ice-cream
19:28:03 melwitt thanks for the video model chat o/
19:36:30 mnaser is there reasoning behind what has ended up being validated in the api layer and what doesn't get validated in the api layer?
19:37:03 mnaser i haven't replicated this yet (but in theory i can imagine it's possible): end user tried to create a vm where image size > disk size when doing bfv
19:37:16 mnaser and they don't really get much more of a helpful error (and their request gets refused too obviously)
19:37:35 mnaser but only after it fails, in what i assume nova-compute trying to ask cinder to create a disk that's impossible to create
19:53:43 melwitt mnaser: generally we validate whatever we can in the api layer to fail fast and avoid the behavior you're describing (fail later on nova-compute). within reason though, as some checks would involve rpc call to the compute host and we avoid that to some degree
19:54:24 melwitt mnaser: I found this part where we validate the image_min_disk against the volume disk size, but that might be different than just the image size as in your example? https://github.com/openstack/nova/blob/master/nova/compute/api.py#L681
19:56:15 mnaser melwitt: oh interesting, you're right, i wonder if this is a case where the user did not have min_disk specified..
19:56:58 mnaser melwitt: oh but we don't do image_size > dest_size
19:57:15 mnaser unlike https://github.com/openstack/nova/blob/master/nova/compute/api.py#L693-L695
19:58:01 mnaser so i _think_ the scenario is no min_disk but no checks on image_size / dst_size
19:58:24 melwitt right thats what I was saying I wonder if thats different
19:59:05 melwitt actually don't know why it's different but I'm not a bfv expert
19:59:28 mnaser melwitt: i think it is, min_size is taken from a glance property which is user-provided (unless a snapshot is taken)
19:59:34 mnaser so i assume user here uploaded an image without min_size
20:01:06 melwitt yeah I mean I don't know why we wouldn't do the same checks for the volume case
20:01:15 melwitt like is that intentional or an oversight
20:02:55 mnaser melwitt: yeah -- in my opinion its an oversight but indeed, maybe someone knows better :)
20:02:59 mnaser anyhow
20:03:04 mnaser ill have a patch up shortly
20:05:39 mnaser melwitt: i swear i saw you mention a while back a key to click before copying from github which gave you a commit hash'd url instead of master
20:05:56 melwitt it's 'y'
20:09:09 mnaser oh awesome, thanks
20:09:49 mnaser https://bugs.launchpad.net/nova/+bug/1883969
20:09:49 openstack Launchpad bug 1883969 in OpenStack Compute (nova) "Nova doesn't fail at API layer when image_size > volume_size with BFV" [Undecided,New]
20:13:08 melwitt sweet mnaser++
20:14:25 openstackgerrit Mohammed Naser proposed openstack/nova master: bfv: check if image_size > volume_size in API https://review.opendev.org/736370
20:14:40 mnaser melwitt: ^ voila, the unit test i wrote passed locally but i didnt run the whole suite
20:14:44 mnaser it makes sense to me™
20:14:44 mnaser it makes sense to meâ„¢
20:20:08 openstackgerrit Mohammed Naser proposed openstack/nova master: bfv: check if image_size > volume_size in API https://review.opendev.org/736370
20:36:35 melwitt mnaser: cool I'll check it
22:09:24 gmann sean-k-mooney: mnaser and to make listing flavor better (instead of need of --private or --all explicitly ) we should add the admin as default tenant access in private flavor. currently what all admin can list the flavors by default and what all it can delete is not matching.
22:09:55 gmann also --private and --all is all admin things and for non-admin --private is just ignored - https://github.com/openstack/nova/blob/6ed556f4adfd8a98fae2e91204f1ce0484f5dd9d/nova/api/openstack/compute/flavors.py#L106
22:13:42 gmann if we add admin as default tenant access in private flavor then i think we can solve it in backward compatible way but it will be interop issue so that still need microversion.
#openstack-nova - 2020-06-18
03:37:13 openstackgerrit Jinsheng Zhang proposed openstack/nova-specs master: Add nova-support-multiple-boot-volume-with-boot-order-selection spec https://review.opendev.org/736422
04:05:12 openstackgerrit Jinsheng Zhang proposed openstack/nova-specs master: fix "Line limited" error https://review.opendev.org/736425
05:07:20 openstackgerrit Jinsheng Zhang proposed openstack/nova-specs master: Add nova-support-multiple-boot-volume-with-boot-order-selection spec https://review.opendev.org/736422
05:43:46 openstackgerrit Jinsheng Zhang proposed openstack/nova-specs master: Add nova-support-multiple-boot-volume-with-boot-order-selection spec fix document error https://review.opendev.org/736422
07:50:29 gibi sean-k-mooney, melwitt: re: resize without changing the flavor, the context on the PTG was that changing a resource request of a port might also needs a move operation without changing the flavor
07:51:28 gibi sean-k-mooney, melwitt: I'll try to propose a spec for discussion during this cycle so your input about machine_type change and video mode change being similar use case helps.
07:52:45 gibi sean-k-mooney, melwitt: however I think this spec will be targeted to W not V due to time constraints and discussion needs
08:00:25 sean-k-mooney gibi: ah yes i could not recall what it was in relation too
08:01:48 sean-k-mooney gibi: i think there are other usecases too such as recreatign the vm with the same image and flavor but just updating the embeded copy to pick up change to extra_specs and image metadata without actually imaging the disk
08:03:24 gibi but still cold migrating the VM if the current host is not good for the updted extra_spec / image metadata?
08:03:35 sean-k-mooney gibi: yes
08:03:48 sean-k-mooney i would guess in generall it would be a move opertation
08:03:49 gibi OK, that seems like the same operation that I would need for the qos update
08:04:13 sean-k-mooney it might be vaild to use the same host but likely the schduler would not select the same host
08:04:30 sean-k-mooney it might but basically the same as same host resize
08:05:11 sean-k-mooney i guess we could always add a weigher to make it prefer the same host
08:05:44 sean-k-mooney but thats not really important
08:08:54 gibi yeah, I can imagine that the end user needs an to update these data if it does not cause downtime (cold migration) of the VM
08:09:13 gibi * needs an option to update
08:09:28 gibi but same host resize is still downtime today
08:09:57 sean-k-mooney gibi: currently you need to do a db edit and a hard reboot
08:10:42 sean-k-mooney the downstream issue that melwitt was looking at is related to the fact that in rhel 8.2 the default we use in nova "cirrus" was deprecated
08:10:44 sean-k-mooney https://bugzilla.redhat.com/show_bug.cgi?id=1651994
08:10:44 openstack sean-k-mooney: Error: Error getting bugzilla.redhat.com bug #1651994: NotPermitted
08:11:01 sean-k-mooney oh thats private well that is what it was tracking
08:11:43 sean-k-mooney anyway 16.0 which is based on train was released on 8.1 16.1 will be on 8.2

Earlier   Later