Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-08-06
15:59:29 artom So need the correct magic disk model/bus imageprops
16:00:01 artom Because TRIM is no longer just for SSDs: https://ceph.com/geen-categorie/openstack-and-ceph-rbd-discard/ <--- that bit is actually genuinely interesting
16:00:01 sean-k-mooney ok so its effectivly the same thing without the schduliing bit
16:00:40 mriedem fungi: must be something weird on my end then but nope still can't access
16:00:44 sean-k-mooney they want to add the disk_bus=scsi imemage meta and enable virtio-scisis as the scsi model
16:00:58 artom mriedem, you're cut off, forced PTO
16:01:13 sean-k-mooney but they want to apply that chagne to existing instances too
16:01:14 mriedem sweet release
16:01:31 artom sean-k-mooney, yeah, we have a KCS for it: https://access.redhat.com/solutions/2532341
16:01:45 artom ... which explicitly states "it is not possible to change existing VMs configuration."
16:01:54 mriedem artom: then tell them to snapshot their existing servers, add the image meta they want on the snapshots, and re-deploy
16:01:57 sean-k-mooney well without a rebuild
16:02:01 sean-k-mooney but ya
16:02:12 openstackgerrit Boxiang Zhu proposed openstack/nova master: Make evacuation respects anti-affinity rule https://review.opendev.org/649963
16:02:14 mriedem or rebuild if they don't want to lose keys/uuids/ips whatever
16:02:23 jangutter artom: OpenStack Icehouse got referenced in that article!
16:02:37 artom mriedem, yeah, not sure why a rebuild isn't acceptable
16:02:41 fungi mriedem: yeah, maybe a network problem somewhere then, but it's still working for me (just tried again)
16:02:50 sean-k-mooney and ya trim is used with ceph and lvm thin provisioning to tell the storage backend it can use those blocks for other oversubsried images
16:02:51 artom I guess they forgot to volume-back their valuable pet disks
16:02:53 mriedem artom: b/c it will blow away the root disk
16:03:01 mriedem artom: you can't rebuild + new image on volume-backed servers
16:03:28 mriedem not yet anyway - there is approved spec and code in cinder to support it, but the nova piece was abandoned
16:03:30 artom detach volume -> rebuild -> re-attach?
16:03:40 mriedem you can't detach the root volume
16:03:44 sean-k-mooney artom: rebuild will only nuke the root disk
16:03:45 artom Assuming not bfv, obviously
16:03:46 mriedem again, there is an approved spec for that, but it's abandoned
16:03:53 sean-k-mooney if they have data volumes those would be fine
16:05:05 mriedem fungi: it's back
16:05:11 openstackgerrit Matt Riedemann proposed openstack/nova stable/queens: Replace non-nova server fault message https://review.opendev.org/674859
16:05:18 mriedem maybe charter has a beef with canonical all of a sudden
16:08:54 jangutter mriedem: it was just a brief outage while they installed the tap on the phone lines.
16:09:26 artom Who needs taps when it's Huawei's equipment >:D
16:11:15 dansmith artom: zing :)
16:11:23 sean-k-mooney mriedem: you use vxhost vms for dev right. if its really a proably you can alwasy use "ssh -D " and forward your traffic via a dev vm to get to launchpad or whatever else is down
16:12:24 sean-k-mooney *really a problem
16:12:56 mriedem sean-k-mooney: no i use vexxhost vms for devstack, which are short-lived b/c $$$
16:12:59 sean-k-mooney or x11 forward a firefox on the vm
16:13:12 mriedem i do dev on a vm on my laptop
16:13:50 sean-k-mooney mriedem: oh well i can proably give you ssh acess to one of my local systems if it contiues to be an issue
16:15:03 artom :q
16:15:11 artom doh
16:15:40 artom *missed
16:16:04 artom But a mixed-DPI setup essentially requires GNOME, which no longer allows you to swap out your VM
16:28:16 eandersson mriedem, https://bugs.launchpad.net/nova/+bug/1817633
16:28:16 openstack Launchpad bug 1817633 in OpenStack Compute (nova) "Listing placement usages causes a circular reference error" [High,Fix committed] - Assigned to Chris Dent (cdent)
16:28:32 eandersson This wouldn't be related to the issue we discussed the other day right?
16:29:15 cdent eandersson: since that pinged me, I'm not curious, what was the issue the other day?
16:29:19 cdent s/not/now/
16:29:34 eandersson https://bugs.launchpad.net/nova/+bug/1799309
16:29:34 openstack Launchpad bug 1799309 in OpenStack Compute (nova) "Migration/Resize fails with Unexpected exception in API method: Circular reference detected" [Undecided,New]
16:29:57 eandersson I attached the chat log to that bug report as well.
16:31:00 openstackgerrit Chris Dent proposed openstack/nova master: Add a document that describes profiling eventlet services https://review.opendev.org/674865
16:33:56 cdent eandersson: the placement thing was very specifically about a value being of a bad type
16:34:20 eandersson I see - yea the file referenced no longer exists
16:34:33 cdent mriedem: there's ^ a start at the profiling doc. Probably needs a few iterations (and the usual, oh, chris wrote a thing, see the typos)
16:39:15 mriedem Sundar: you asked for it https://review.opendev.org/#/c/674726/
16:39:25 mriedem cdent: thanks
16:44:17 sean-k-mooney efried: mriedem Sundar: by the way i would be fine with not supporting some operation in addtion to not support live migration but the current patch block way more operation then im comfortable with
16:45:43 sean-k-mooney with no support for snapshots resize/cold-migrate or live migrate there is not way to do upgrades of the node other then an inplace upgrade
16:47:08 mriedem i'm also not cool with the blocker really, at least without a policy rule to control it or something
16:47:30 mriedem if you do it like vgpus, then it's just docs on what works and what doesn't and anything we don't test is considered experimental and you fix it as bugs later
16:48:46 sean-k-mooney mriedem: ya that sound resonable. i am semi ok with 409s for live migratin or something we know upfont wont work
16:49:18 dansmith dear god
16:49:49 dansmith also, are we really adding a flavor property for device profile?
16:49:50 sean-k-mooney it was really blocking interface/volume attach/detach and lock that annoyed me
16:50:01 sean-k-mooney dansmith: its an extra spec
16:50:18 dansmith oh, it's added as a property on flavor, I see
16:50:30 dansmith I don't really like that
16:50:54 sean-k-mooney right but we did that not to need to change the nova apin in the first versions
16:51:15 sean-k-mooney dansmith: but at least your consitent in not liking it every time you see it
16:52:43 sean-k-mooney i think given how long it been waiting we should merge basic integation and consider it experimental. but my version of basic is more then boot, reboot, delete and status.
16:54:30 sean-k-mooney i prefer the docs approch rather then expcit blocker check given that in thory other virt driver will support cyborg at some point so i think doc are better for then api check when they could be virt dirver specifc.
16:57:22 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Implement power_off/power_on for the FakeDriver https://review.opendev.org/674876
16:57:25 openstackgerrit Matt Riedemann proposed openstack/nova stable/pike: Replace non-nova server fault message https://review.opendev.org/674877
17:22:47 openstackgerrit sean mooney proposed openstack/nova master: support pci numa affinity policies in flavor and image https://review.opendev.org/674072
17:47:58 dansmith mriedem: bauzas: hmm, we still construct reqspec by from_components() huh?
17:49:04 sean-k-mooney dansmith: i think bauzas is on PTO for the next 2-3 weeks
17:49:22 dansmith oh right, because he's in france and today ends in "day"
17:49:25 sean-k-mooney dansmith: but yes i belive we do
17:49:41 dansmith sean-k-mooney: yeah, that was rhetorical to start a bitch-fest :D
17:50:31 sean-k-mooney hehe i thought it was the from_flavor version that everyone hated
17:51:00 sean-k-mooney https://github.com/openstack/nova/blob/master/nova/scheduler/utils.py#L389-L417
17:51:28 sean-k-mooney oh wait im thinking of somthing else
17:53:26 sean-k-mooney you mean how its built form this blob of stuff https://github.com/openstack/nova/blob/master/nova/compute/api.py#L897
17:54:46 sean-k-mooney dansmith: when are you heading on PTO by the way. are you around this week?
17:55:19 dansmith sean-k-mooney: glad to know that me being gone ALL last week was so unimpactful that you didn't even notice
17:55:25 openstackgerrit Stephen Finucane proposed openstack/nova master: Add support for translating CPU policy extra specs, image meta https://review.opendev.org/671801
17:55:25 openstackgerrit Stephen Finucane proposed openstack/nova master: trivial: Remove unused '_instance_to_allocations_dict' function https://review.opendev.org/674893
17:55:26 openstackgerrit Stephen Finucane proposed openstack/nova master: scheduler: Flatten 'ResourceRequest.from_extra_specs', 'from_image_props' https://review.opendev.org/674894
17:55:26 openstackgerrit Stephen Finucane proposed openstack/nova master: Add reshaper for PCPU https://review.opendev.org/674895
17:55:35 dansmith maybe that means I could take another week and nobody would know ;)
17:56:48 sean-k-mooney well i was off since tursday so i dont remember much form last weeks other then i need to relax. no i just noticed you comment that you were going to be off around august 24th in a review
17:57:28 sean-k-mooney maybe that was someone else
17:57:40 dansmith erm, Aug 24 is a saturday, so I do plan to not be around that day, but.. I think maybe you're confusing me with someone else. Which is also super flattering :D
17:59:25 mriedem dansmith: i think i've been working on burning out from_components the last few releases
17:59:38 mriedem like getting resize reschedule to pass the request spec back to conductor so we don't have to build a new one
17:59:49 mriedem some of that compat might be deletable by now too
17:59:52 dansmith okay

Earlier   Later