Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2017-04-26
16:46:04 cburgess This compute node can support local, this can't.
16:46:08 cburgess But thats a long way out again.
16:46:22 johnthetubaguy cburgess: yeah, I mean that all makes sense
16:46:26 mriedem johnthetubaguy: i knew about max_local_block_devices because people thought if you set it to 0 it meant nova would automagically always do bfv for you
16:46:35 johnthetubaguy mriedem: ah... right
16:46:44 johnthetubaguy mriedem: I remember that coming up now you mention it
16:46:46 cburgess I actually don't know what that setting does.
16:46:54 mgagne https://wiki.openstack.org/wiki/BlockDeviceConfig#max_local_block_devices
16:46:59 mriedem cburgess: traits are closer than everything else at this point
16:47:00 mgagne it's only to prevent DoS
16:47:04 johnthetubaguy gives you 400 errors, I presume
16:47:22 mgagne so people don't create an infinite amount of local disks
16:47:32 cburgess Ahh...
16:47:34 cburgess OK..
16:47:35 mriedem cburgess: so that's an alternative we haven't talked about before,
16:47:40 cburgess mriedem Traits?
16:47:48 mriedem traits on the resource provider to say it's local_disk=False
16:47:58 mriedem then tag your flavor with local_disk=False
16:48:08 mriedem scheduler / placement has to find a compute that matches those
16:48:12 cburgess mriedem Why do we need to tag the flavr?
16:48:22 melwitt I thought BFV was supposed to get solved with placement. are you just talking about how to do it in placement?
16:48:35 cburgess mriedem I'm not wild about having to set something in the flavor.
16:48:54 cburgess The fact that a compute node can or can not host local disks is something the compute node is aware of.
16:48:55 mriedem cburgess: how are you going to know where to put instances?
16:49:02 cburgess You shouldn't have to also set the right thing on a flavor.
16:49:14 cburgess I flavor you should I need local disk and the compute node should be able to say "nope not me".
16:49:27 mriedem so all of your bfv goes to the local_disk=False computes, and all non-bfv goes to the local_disk=True computes?
16:49:49 cburgess mriedem Thinking through all the use cases...
16:50:01 mriedem dansmith: super fun cells v2 meeting in 10 minutes i hope
16:50:04 cburgess mriedem This is why I was suggesting a working sessions focused on this...
16:50:12 dansmith oh yep
16:50:14 mriedem cburgess: well, we have the forum session
16:50:17 dansmith mriedem: didn't realize it was that late
16:50:21 cburgess mriedem So many different use case
16:50:26 mriedem cburgess: that's about as working as it's going to get at the summit
16:50:32 cburgess mriedem I know...
16:50:55 cburgess mriedem How badly bike shedded would we get trying to do this on the ML?
16:51:06 cburgess mriedem Where this = discuss different options.
16:51:17 openstackgerrit melanie witt proposed openstack/nova master: Add get_count_by_vm_state() to Instance object https://review.openstack.org/446244
16:51:18 openstackgerrit melanie witt proposed openstack/nova master: Remove useless quota_usage_refresh from nova-manage https://review.openstack.org/446243
16:51:18 openstackgerrit melanie witt proposed openstack/nova master: Remove 'reserved' count from used limits https://review.openstack.org/446242
16:51:19 openstackgerrit melanie witt proposed openstack/nova master: Count instances to check quota https://review.openstack.org/416521
16:51:21 openstackgerrit melanie witt proposed openstack/nova master: Add online migration to move quotas to API database https://review.openstack.org/410946
16:51:21 openstackgerrit melanie witt proposed openstack/nova master: Make Quotas object favor the API database https://review.openstack.org/410945
16:51:27 mriedem cburgess: idk, it's probably worth starting to prime the pump for that session at the forum,
16:51:42 mriedem so we don't start a 40 minute session discussing all of the history and previously discussed options
16:51:56 mriedem and then find out we have 5 minutes left with nothing new to say
16:52:22 cburgess mriedem What I don't like about most of the solutions we have discussed is that if you forget to flag a flavor correctly then it could get placed on a compute node that doesn't want local storage.
16:52:36 mriedem cburgess: before starting a ML thread,
16:52:47 mriedem maybe i start an etherpad, which i'll need for the session anyway,
16:52:48 cburgess mriedem I feel like we need a method that ensures a compute node can assert that it can never ever host a local disk.
16:52:58 mriedem and we can start listing the use cases and options, and pros/cons for each option
16:53:01 mgagne maybe just list the use cases, we don't need to agree on the validity of the need, just list them, list associated workaround/limitations, etc. and start from there
16:53:07 mriedem *then* put that in the ML for discussion
16:53:08 cburgess mriedem Yeah.. lets start with an etherpad. Once we get it to an ok state we can post it to the ML.
16:53:14 cburgess mriedem I like it.
16:53:27 cburgess mgagne Right.. only rule out the most obviously insane ideas.
16:53:39 mriedem mgagne: cburgess: i'll create an etherpad after the cells meeting and report back here
16:53:47 mgagne we can list them and explain later why we can't or don't want to support them
16:53:47 cburgess mriedem Rainbows
16:54:09 mgagne so people don't come back later and ask why that one isn't listed
16:54:38 mriedem looks like the etherpads have started https://wiki.openstack.org/wiki/Forum/Boston2017
16:54:43 mriedem which was on my todo list anyway
16:54:55 cburgess mgagne I'm only advocating ruling out things like "create a new top level openstack project" or other things of that level.
17:06:01 mordred cdent: what does placement use in the id field of its version doc?
17:08:37 cdent mordred: the min version
17:08:51 cdent which is a fudge
17:09:17 cdent it might make sense to split on '.' and use [0]
17:14:28 mordred cdent: ok. so it does have a value in id
17:14:48 cdent yeah, but it puts nothing in the url that indicates version and has no plan to
17:17:04 sumedhdegaonkar Hello all, this is my first review request, so please bear with me. I'd like to get your opinion/review of https://review.openstack.org/#/c/459794/ - It's just removing a restriction to use of 'sata' disk bus type for qemu and kvm. No change to default behavior or selection and hence in my opinion, no need to update test cases
17:30:18 openstackgerrit Rafael Folco proposed openstack/nova master: placement: Specific error for inventory in use https://review.openstack.org/460231
17:31:10 cdent mriedem: way back in earlier conversation you mentioned a bug in gabbi with left hand side json path substitutions. I meant to link you to this: https://github.com/cdent/gabbi/issues/170
17:31:30 cdent but also to ask: was that germane to what were talking about at the time, or just an observation?
17:32:15 mriedem it was related
17:32:23 mriedem i was also looking through the gabbits for an example of a response
17:32:24 mriedem to see the keys
17:33:37 cdent mriedem: I could fix that bug in the case you're talking about it is often easier to make the right hand side more complete (a structure instead of a string)
17:37:35 mriedem not a priority for me, i just noticed the comments in the gabbits about the limitation
17:37:43 mriedem api-ref is the priority,
17:37:49 mriedem because then i don't need to look at gabbits
17:38:23 cdent i regret slinging that name around
17:43:57 openstackgerrit Steven Webster proposed openstack/nova master: Improve comment for PCI port binding update https://review.openstack.org/460233
17:51:02 mriedem mgagne: cburgess: johnthetubaguy: stackempty: here you go https://etherpad.openstack.org/p/BOS-forum-using-cinder-for-nova-ephemeral-storage
17:51:47 cburgess mriedem Are we open to having a 'short term' and a 'long term' set of goals and solutions?
17:52:12 mriedem i think you could tag things that way
17:52:24 mriedem i'm open to eating lunch
17:52:27 mriedem so bbiab
17:55:39 openstackgerrit Lee Yarwood proposed openstack/nova master: libvirt: Attach and detach encryptors during swap_volume https://review.openstack.org/460244
17:55:39 openstackgerrit Lee Yarwood proposed openstack/nova master: libvirt: Refactor encryptor attach and detach calls https://review.openstack.org/460243
17:57:51 cburgess mriedem Just doing that myself
18:04:37 mgagne cburgess: so there are 2 use cases? : 1) Provide a default BDM 2) Prevent/forbid source=local in BDM
18:05:19 mgagne in fact, destination=local
18:06:38 dansmith does anyone know if test_delete_server_while_in_building_state is known to be flaky at the moment? logstash doesn't seem to think it's spiking but I see plenty of failures that aren't in my patch, like even on periodic stable jobs
18:06:40 dansmith mriedem: sdague ^
18:12:02 mgagne k policy.
18:12:02 mgagne cburgess: what do you think about the idea of having a "default bdm" for flavor-root-disk-none ? If you provide no BDM, the root_gb size is used to create a volume for you for your root disk. (flavor could also provide default volume type) This means no 400 HTTP error by default as a sane behavior/default is provided. An other extra spec or flavor attribute could forbid destination=local so if BDM is provided, it still respect the no-local-dis
18:12:46 cburgess mgagne Thats a decent idea. Pull size from root_gb and image_id from request spec.
18:12:52 mgagne yes
18:13:04 mgagne that's basically what we did on our side

Earlier   Later