Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-06-20
15:26:44 mnaser is the same
15:26:47 mnaser so qemu never boots the vm
15:27:29 mnaser i added a LOG.error(disk_info) in the libvirt tooling
15:27:30 mnaser and this is what came out of it too - https://gist.github.com/mnaser/aedd44b9df018786a144f7c719ec5a13
15:27:49 mnaser notice how 'dev': 'vda' (which is ephermal) and the 'dev': 'sda' (which is attached cinder block device)
15:28:11 mnaser i am guessing that's where the issue is stemming, the first one should be sda too (because using virtio-scsi with the image hw_ parameters)
15:28:12 mriedem mnaser: https://github.com/openstack/nova/blob/9045a98145f0834446d668e0282ddb1c6f2d136c/nova/compute/api.py#L921 is eventually what gets set in the instance.root_device_name field
15:29:22 mnaser i've confirmed with two different installs that booting an image with hw_scsi_model=virtio-scsi and hw_disk_bus=scsi will work but no cinder volumes will successfully attach: libvirtError: internal error: unable to execute QEMU command '__com.redhat_drive_add': Duplicate ID 'drive-scsi0-0-0-0' for drive
15:29:30 mdbooth mnaser: what's the symptom of the weird xml?
15:29:35 mdbooth Does it barf?
15:29:50 mnaser so, if you try to attach it to an already running vm, you get the libvirt error above (duplicate id)
15:30:19 mnaser if you turn the vm off, attach (really a no-op if its offline), start it, it'll generate that invalid xml and fail
15:30:32 mnaser so im not sure if libvirt or nova is to blame in putting the duplicate address ther
15:30:45 mdbooth We really shouldn't be looking at root_device_name
15:31:34 mnaser i've just been kinda trying to understand the code and i dont know if i can point the finger at nova or libvirt yet
15:32:03 mnaser but it's obviously just an issue when using virtio-scsi
15:37:48 mgagne mriedem: supports_volume_extend vs supports_extend_volume ? RE https://review.openstack.org/#/c/454322/18/nova/virt/driver.py
15:38:31 openstackgerrit Dan Smith proposed openstack/nova master: Fix using targeted cell context when finding services in cells https://review.openstack.org/475830
15:38:53 dansmith dtantsur: mriedem ^
15:39:01 dtantsur nice
15:40:09 openstackgerrit Kaitlin Farr proposed openstack/nova master: Remove deprecated keymgr code https://review.openstack.org/439855
15:41:04 mriedem melwitt: i think you might have gotten delete happy in the tests here https://review.openstack.org/#/c/457862/
15:41:33 mriedem mgagne: i'm ok with either - i suppose supports_extend_volume to match the driver's method name
15:41:49 mgagne that's my opinion
15:50:22 mriedem mdbooth: i think that root_device_name thing is a legacy bdm v1 thing, but not totally sure
15:50:44 mriedem some kind of ec2 carry over
15:51:36 mdbooth mriedem: I haven't followed the code all the way down, but I suspect we could ignore it
15:54:36 openstackgerrit Jay Pipes proposed openstack/nova master: placement: scheduler uses allocation candidates https://review.openstack.org/475600
15:56:50 dansmith jaypipes: quick +W bait: https://review.openstack.org/#/c/475830/
15:57:01 dansmith jaypipes: fixes the ironic people
15:57:31 jaypipes dansmith: wallaby'd
15:57:38 dansmith thanks
15:57:41 dtantsur thanks all!
15:57:51 dansmith I guess cdent still isn't around today?
15:58:06 jaypipes dansmith: he's on PTO for next two weeksd.
15:58:11 dansmith oh my
15:58:20 jaypipes dansmith: after which he will be a VMWare employee.
15:59:29 dansmith oh
15:59:32 dansmith my
15:59:35 dansmith GAWD
16:01:34 openstackgerrit Merged openstack/nova master: Count networks to check quota https://review.openstack.org/446241
16:02:29 gibi_ jaypipes: good luck with the api.fault bug
16:02:46 jaypipes gibi_: yeah, I might need to pass on that :(
16:02:56 jaypipes gibi_: that's some old-ass code in there :)
16:03:13 gibi_ jaypipes: I hope somebody can say 'delete that code!'
16:04:47 jaypipes gibi_: it's possible. some day. https://twitter.com/dims/status/542002018970529792
16:06:27 gibi_ :)
16:07:36 openstackgerrit Merged openstack/nova master: Add SecurityGroupList.get_counts() https://review.openstack.org/446245
16:08:08 openstackgerrit Merged openstack/nova master: Count security groups to check quota https://review.openstack.org/457860
16:09:12 openstackgerrit Merged openstack/nova master: Add FixedIPList.get_count_by_project() https://review.openstack.org/446246
16:09:48 openstackgerrit Merged openstack/nova master: Count fixed ips to check quota https://review.openstack.org/457861
16:10:17 openstackgerrit Merged openstack/nova master: Add FloatingIPList.get_count_by_project() https://review.openstack.org/446247
16:10:23 dansmith wow, just count those counting quotas merges
16:21:09 openstackgerrit Balazs Gibizer proposed openstack/nova master: Factor out duplicated notification sample data (3) https://review.openstack.org/452820
16:21:09 openstackgerrit Balazs Gibizer proposed openstack/nova master: Factor out duplicated notification sample data https://review.openstack.org/452818
16:21:10 openstackgerrit Balazs Gibizer proposed openstack/nova master: Factor out duplicated notification sample data (2) https://review.openstack.org/452819
16:21:10 openstackgerrit Balazs Gibizer proposed openstack/nova master: Factor out duplicated notification sample data https://review.openstack.org/475860
16:25:34 openstackgerrit Gábor Antal proposed openstack/nova master: Transform keypair.delete notification https://review.openstack.org/463946
16:32:01 openstackgerrit Alex Szarka proposed openstack/nova master: [WIP]Transform instance-live_migration_abort notification https://review.openstack.org/420453
16:36:15 openstackgerrit Alex Szarka proposed openstack/nova master: [WIP]Transform instance-live_migration_abort notification https://review.openstack.org/420453
16:37:55 openstackgerrit Alex Szarka proposed openstack/nova master: [WIP]Transform instance-live_migration_abort notification https://review.openstack.org/420453
16:40:06 openstackgerrit Alex Szarka proposed openstack/nova master: [WIP]Transform instance-live_migration_abort notification https://review.openstack.org/420453
16:47:24 openstackgerrit Gábor Antal proposed openstack/nova master: Add json style checking for sample notifications https://review.openstack.org/443677
17:22:28 openstackgerrit Merged openstack/nova master: Updated from global requirements https://review.openstack.org/475563
18:26:59 mriedem gibi_: i know this was already approved, but needed to be rebased and while it's sitting there i had a question inline https://review.openstack.org/#/c/460029/8/doc/notification_samples/instance-soft_delete-start.json
18:27:07 mriedem about transforming soft delete notifications to versioned payloads
18:36:44 melwitt mriedem: I deleted those tests on purpose because after changing floating_ips to a CountableResource, there are no more "project scoped quotas" that are refreshable or syncable. the majority of the tests in nova/tests/unit/test_quota.py rely on the global QUOTAS engine and once there are no more real examples of a certain type of quota, we would have to synthesize fake resources for the tests to still work
18:37:51 melwitt instead, I figured delete them since they're no longer applicable to anything real. the quota refresh command only refreshes ReservableResources
18:38:42 mriedem i thought the refresh command was using the _sync_*_quotas stuff still, which is in the db api, but would have to dig
18:38:57 mriedem if it's only using ReservableResources, and we don't have any of those, that's also grounds to remove it - but i need to check
18:39:27 melwitt mriedem: it does, but the _sync_* functions are only used on ReservableResources. the connection is made in nova/quota.py and each patch that removes the connection stops it from working that way
18:39:55 melwitt mriedem: we still have ReservableResources, they are just only user scoped and those unit tests are testing project scoped. it's confusing :(
18:39:57 openstackgerrit Kaitlin Farr proposed openstack/nova master: Remove deprecated keymgr code https://review.openstack.org/439855
18:40:44 mriedem i'll dig a bit
18:42:15 melwitt the patch that removes the last ReservableResource is "Count instances" and we're not there yet. k. the "user scoped quotas" are the ones in the "project_user_quotas_table" and the "project scoped quotas" are in the "quotas" table (floating_ips, fixed_ips, networks). that's what the tests refer to when it says "project key" it means one of: 'floating_ips', 'fixed_ips', or 'networks'
18:43:12 mriedem melwitt: ok so https://github.com/openstack/nova/blob/master/nova/quota.py#L900 checks https://github.com/openstack/nova/blob/master/nova/quota.py#L416
18:43:15 melwitt there's also the global var in nova/db/sqlalchemy/api.py PER_PROJECT_QUOTAS that show the project keys
18:43:38 mriedem and since per_project_resources is fixedips, floatingips and networks,
18:43:48 mriedem and none of those are ReservableResources
18:44:11 melwitt yes
18:44:16 mriedem but we could still have user-scoped resources,
18:44:36 mriedem like server groups
18:44:53 melwitt yes, and we still do. we still have 'instances' 'cores' 'ram' as of that patch I think
18:45:17 mriedem let me check the tests being removed again
18:45:23 melwitt so those will still use reserve/commit/rollback and be synced/refreshed
18:45:24 melwitt k
18:46:35 mriedem rfolco: some issues in here https://review.openstack.org/#/c/460147/
18:46:49 openstackgerrit Merged openstack/nova master: Fix using targeted cell context when finding services in cells https://review.openstack.org/475830
18:46:51 mriedem rfolco: i'd really appreciate to leave the translation marker removal to a separate patch devoted to that
18:47:29 melwitt the only one we could argue is test_usage_refresh_invalid_user_key. I could do s/floating_ips/fake_resource_key/ there and the test would still work. the point of it had been, "verify you get Invalid if you try to refresh a valid key that isn't a valid user key"
18:48:17 melwitt and since there are no more project keys, we can't choose any valid keys. the test on L3270 is sort of identical
18:48:32 rfolco mriedem, looking
18:51:17 openstackgerrit Merged openstack/nova master: api-ref: fix hypervisor_hostname description for Ironic https://review.openstack.org/474723
18:51:44 openstackgerrit Merged openstack/nova master: [doc] Updated sqlalchemy URL in migrate readme https://review.openstack.org/472549
18:52:42 mriedem melwitt: removed my -1 for that - are we ok with https://review.openstack.org/#/c/457862/22/nova/network/floating_ips.py@225 ?
18:52:48 mriedem i think i talked myself into being ok with it
18:52:53 mriedem but i need you to close the deal
18:55:07 melwitt mriedem: thanks. the test on L3270 is the same test really anyway. looking at the next question ...
19:00:29 melwitt mriedem: looks like that's similar to the fixed ip one. I didn't think to call the self.deallocate* methods bc I thought to just use the inverse object functions to undo the create. but, I think technically maybe someone could create a floating ip and a dns domain in parallel and maybe could end up with orphaned dns domains that way. so it's probably most correct to call self.deallocate*
19:01:29 melwitt I'm not sure if associating the dns domain requires anything from the floating ip create response maybe. but either way it doesn't hurt to call the self.deallocate function

Earlier   Later