Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-09-19
16:29:53 bauzas sean-k-mooney: I guess I need to update with the existing RPs but the one I wanna remove
16:30:02 lyarwood mriedem_afk: this isn't specific to encrypted volumes, the call to connect_volume is failing because once we've connected to the target we can't find a block device locally that matches the WWN provided by c-api/c-vol.
16:30:12 lyarwood mriedem_afk: at this point the fact the volume is encrypted doesn't come into play
16:30:17 sean-k-mooney bauzas: isnt the usage caulated by subtracting the allcoation form the avilabe inventory
16:30:56 bauzas sean-k-mooney: sorry, nevermind the "usage" wording, it's confusing
16:30:58 sean-k-mooney i.e. once you remvoe teh allocation againt the RP the usage should be caulated correctly when next quired
16:31:05 bauzas sean-k-mooney: I meant, say https://docs.openstack.org/api-ref/placement/?&expanded=update-allocations-detail#id72
16:31:42 bauzas sean-k-mooney: if I'd like to remove allocations against 89873422-1373-46e5-b467-f0c5e6acf08f, then I would update the endpoint with a payload having 4e061c03-611e-4caa-bf26-999dcff4284e only ?
16:31:53 bauzas sean-k-mooney: and in this case, would it remove the DB records ?
16:31:54 sean-k-mooney yes
16:32:12 sean-k-mooney that is my understanding at least
16:32:43 bauzas ok, we can test with my functional test at least
16:33:01 sean-k-mooney or with osc-placment but sure
16:33:05 bauzas we miss cdent, tetsuro, efried_pto
16:33:06 sean-k-mooney functional test are good too
16:34:40 sean-k-mooney https://github.com/openstack/placement/blob/master/placement/tests/functional/gabbits/allocations-1.28.yaml#L174-L181
16:34:43 sean-k-mooney bauzas: ^
16:35:08 sean-k-mooney so based on there functionla test if you use 1.28+ it should work
16:36:50 sean-k-mooney well i should include the next test too so
16:36:52 sean-k-mooney https://github.com/openstack/placement/blob/master/placement/tests/functional/gabbits/allocations-1.28.yaml#L174-L187
16:37:50 sean-k-mooney bauzas: so a put of empty allcoation is accpeted implying a put of a partill set of allcoation shoudl also work
16:38:43 openstackgerrit Merged openstack/nova-specs master: Re-propose cross-cell-resize spec for Ussuri https://review.opendev.org/683002
16:52:10 lyarwood mriedem_afk: actually os-brick is doing some random lookups when encrypted volumes are used
17:01:12 bauzas sean-k-mooney: it works seamlessly
17:02:13 openstackgerrit Sylvain Bauza proposed openstack/nova master: WIP: Add a placement audit command https://review.opendev.org/670112
17:03:50 lyarwood mriedem_afk: so it looks like /dev/disk/by-id/ is just slow to populate on these CI nodes and we aren't retrying the lookup https://github.com/openstack/os-brick/blob/1c02d525bb77487d5de89c4bd58116b916d44207/os_brick/initiator/connectors/iscsi.py#L521-L540
17:04:30 lyarwood mriedem_afk: the fact we are using /dev/disk/by-id/ is due to the volume being encrypted https://github.com/openstack/os-brick/blob/1c02d525bb77487d5de89c4bd58116b916d44207/os_brick/initiator/connectors/iscsi.py#L545-L553
17:05:37 lyarwood mriedem_afk: and is due to the horrid design of the original encryptors that symlink things in a horrific way such that we can't use the actual block device path on the host otherwise we'd nuke it https://bugs.launchpad.net/os-brick/+bug/1703954
17:05:38 openstack Launchpad bug 1703954 in OpenStack Compute (nova) "Attach/Detach encrypted volume problems with real paths" [Undecided,Incomplete]
17:07:23 lyarwood mriedem_afk: I'll post something in the morning to retry _get_device_link but in the longer term I think I need to sort out the symlinking in the encryptors such that it's safe to provide them with a real block device path and they don't do anything stupid.
17:07:41 lyarwood mriedem_afk: anyway both tasks for the morning \o_
17:36:20 mriedem lyarwood: ack thanks for investigating, i'll open a bug (i found https://bugs.launchpad.net/os-brick/+bug/1820007 but looks old/different) and start tracking in e-r
17:36:21 openstack Launchpad bug 1820007 in os-brick "Failed to attach encrypted volumes after detach: volume device not found at /dev/disk/by-id" [Undecided,New]
17:36:29 mriedem it also only failed on ovh nodes so maybe those are slower here, idk
17:37:51 mriedem actually a lot of the details in that bug look the same so i'll just re-use that
18:10:15 openstackgerrit Walter A. Boring IV (hemna) proposed openstack/nova stable/pike: WIP: Avoid redundant initialize_connection on source post live migration https://review.opendev.org/683008
18:35:25 openstackgerrit Merged openstack/nova master: Trigger real BuildAbortException during migrate with bandwidth https://review.opendev.org/682876
20:15:17 openstackgerrit Merged openstack/nova master: Validate CPU config options against running instances https://review.opendev.org/680107
20:15:26 openstackgerrit Merged openstack/nova master: objects: Add 'NUMACell.pcpuset' field https://review.opendev.org/680108
20:15:36 openstackgerrit Merged openstack/nova master: hardware: Differentiate between shared and dedicated CPUs https://review.opendev.org/671800
21:56:57 openstackgerrit Matt Riedemann proposed openstack/nova master: Add evacuate vs rebuild contributor doc https://review.opendev.org/683245
21:57:03 mriedem efried_pto: you asked for it ^ dansmith your review on that would also be helpful due to your blog post
21:57:06 mriedem *the* blog post
22:00:59 dansmith mriedem: https://www.youtube.com/watch?v=gVpzrxjiAtc
22:01:29 mriedem coincidentally i'm wearing that same outfit
22:01:38 mriedem just went to the scarf store last night
22:01:47 dansmith I don't doubt it
22:01:58 dansmith coincidentally, I'm wearing nothing *but* that scarf
22:02:22 mriedem man, i can't imagine the amount of coke flowing through these guys in this video
22:03:15 dansmith yeah keith looks like he hasn't slept in two weeks
22:28:34 openstackgerrit Dustin Cowles proposed openstack/nova master: Provider Config File: YAML file loading and schema validation https://review.opendev.org/673341
22:28:35 openstackgerrit Dustin Cowles proposed openstack/nova master: Provider Config File: Function to further validate and retrieve configs https://review.opendev.org/676029
22:28:36 openstackgerrit Dustin Cowles proposed openstack/nova master: WIP: Provider Config File: Merge provider configs to provider tree https://review.opendev.org/676522
23:15:41 openstackgerrit Eric Fried proposed openstack/nova master: libvirt: Enable driver configuring PMEM namespaces https://review.opendev.org/679640
#openstack-nova - 2019-09-20
00:21:22 openstackgerrit sean mooney proposed openstack/nova-specs master: resubmit image metadata prefiltering spec for ussuri https://review.opendev.org/683258
08:59:52 luyao efried_pto, stephenfin: Hi, could you help review the vpmems doc if you get time, since you both are familiar with the vpmems series. :D I think I get the content close which is focusing on the current functionality. https://review.opendev.org/#/c/680300.
09:02:32 luyao And the patch 'objects: use all_things_equal from objects.base' is ready to merge , need +W on it, https://review.opendev.org/#/c/681397/13
09:07:35 openstackgerrit Sylvain Bauza proposed openstack/nova master: Add a prelude for the Train release https://review.opendev.org/683327
09:59:32 AdamMork Hello, friends! I have a openstack contains 2 controllers an 4 compute node's. I need to enable AES-NI for one or more vm. Openstack deployed via kolla and work on docker containers(nova container, neutron container etc. on controller). I read manual https://software.intel.com/en-us/articles/openstack-epa-feature-breakdown-and-analysis . On the
09:59:33 AdamMork compute node (in bios i enable AES-NI) . But in vm use command (cat /proc/cpuinfo | grep aes) i not found aes support. How to configure nova to enable aes instructions?
10:28:35 AdamMork ok! How i can modify Nova conf "The Nova* libvirt driver takes its configuration information from a section in the main Nova file /etc/nova/nova.conf." I have docker containers and how to find /etc/nova/nova.conf.? should i connect to nova container?
11:28:54 AdamMork ok. I see more manuals and understand : in globals.yml need uncomment :node_custom_config: "/etc/kolla/config" and overrides basic config for nova (As of now kolla only supports config overrides for ini based configs.) https://github.com/openstack/kolla-ansible/blob/master/doc/source/admin/advanced-configuration.rst THX
12:54:46 mriedem gibi: i saw your resize reschedule rpc pin bug, do you know if that's a new regression in train?
12:55:28 gibi mriedem: not yet. I'm about to push a reproduction patch, then I will look into when the fault was introduced
12:55:57 gibi it is clear that the problem is with the legacy request spec in prep_resize ending up in the conductor during a re-schedule with rpc pinned to 5.0
12:58:38 gibi mriedem: this is where it blows https://github.com/openstack/nova/blob/9b2e00e015f22b2d876cd3c239af8e139040c8c8/nova/conductor/manager.py#L327
12:58:51 mriedem i know that in compute rpc api 5.1 we send the RequestSpec to compute https://opendev.org/openstack/nova/src/branch/master/nova/compute/rpcapi.py#L833
12:58:58 mriedem but backlevel if to the dict form if we can't
12:59:13 mriedem gibi: ah yeah - that's my patch
12:59:15 mriedem remember?
12:59:34 mriedem https://review.opendev.org/#/c/680762/
13:00:03 mriedem so if you have a recreate we can lay that on top
13:00:43 gibi mriedem: ohh you have a fix for it. cool. Yeah I will push a functional repro soon and the you can rebase
13:00:50 gibi or I can rebase
13:00:55 gibi your series
13:01:00 mriedem you can rebase it,
13:01:06 mriedem also, i've duplicated your bug against 1843090
13:01:17 gibi mriedem: thanks. I will update my patch accordingly
13:01:42 gibi mriedem: I reached this bug while tried to create a func test for the bandwidth case when rpc in pinned
13:02:06 gibi mriedem: it seems there will also be extra issues
13:02:07 mriedem yeah it was this patch that made me think of it in your bw series for prep_resize where you were using the request spec
13:03:36 gibi mriedem: btw do you know about a bug regading booting a instance with --availablity-zone az:node and then migrate it. I saw that in this case nova does not try to re-schedule the migration if the first dest host fails in prep_resize
13:03:57 gibi it seem the that filter_properties are not populated for re-schedule
13:04:17 gibi if I boot the server with the new host parameter then the re-schedule work
13:05:18 mriedem gibi: that's working as designed https://opendev.org/openstack/nova/src/branch/master/nova/scheduler/utils.py#L801
13:05:44 mriedem using az::node will set force_nodes
13:06:07 mriedem do you see "Re-scheduling is disabled due to forcing a host" in the logs?
13:06:13 gibi mriedem: but forcing the host during the boot allows me to migrate it without dest host specified, but it does not allow nova to re-schedule
13:06:27 gibi during that migrate
13:07:06 gibi mriedem: I have to go back and re-create the situation as I saw this while I created the func test for the rpc pin bug.
13:07:26 mriedem gibi: which release? https://review.opendev.org/#/q/I3f488be6f3c399f23ccf2b9ee0d76cd000da0e3e
13:07:47 gibi mriedem: master :)
13:07:54 mriedem oh nvm that's ignore_hosts
13:08:46 mriedem so i think force_hosts/nodes gets persisted for some reason, i'm not sure why, but for every move operation we call this to basically unset those fields https://opendev.org/openstack/nova/src/branch/master/nova/objects/request_spec.py#L692
13:08:55 mriedem it would be simpler if we just didn't persist those values
13:09:24 mriedem gibi: so maybe we're calling ^ *after* populate_retry
13:10:03 mriedem yup https://opendev.org/openstack/nova/src/branch/master/nova/conductor/tasks/migrate.py#L293
13:11:29 gibi mriedem: so is the order of pupulate and rest wrong there?
13:11:53 mriedem it appears so
13:11:54 gibi s/pupulate/populate_retry

Earlier   Later