Earlier  
Later
New Search
Posted Nick Remark
#openstack-nova - 2019-07-04
10:01:05 kashyap [It starts with: "Overall, turned the spec a bit "upside down" to reorganize some content:"]
10:01:48 kashyap johnthetubaguy: Hmm, didn't hear that before -- got a full error somewhere?
10:02:33 johnthetubaguy kashyap: not handily, I will let you know and get that into a bug report
10:02:37 kashyap johnthetubaguy: By 'os-info', you mean `osinfo-query`?
10:02:40 johnthetubaguy yeah
10:03:19 johnthetubaguy mostly os-distro flags just have a warning message with no match found... historially bad at setting them correctly, but everything blew up when I tried to "fix" that
10:03:33 johnthetubaguy its an unknown virtio device type I think, but I will get you the details
10:03:35 openstackgerrit Lee Yarwood proposed openstack/nova master: libvirt: Remove unreachable native QEMU iSCSI initiator config code https://review.opendev.org/668750
10:04:01 lyarwood stephenfin: ^ thanks for the review, I've removed disconnect_volume and addressed the nits in the commit
10:04:33 kashyap johnthetubaguy: Sure, happy to look and check w/ the virtio folks if I don't know
10:05:15 openstackgerrit Brin Zhang proposed openstack/nova master: Specify availability_zone to unshelve https://review.opendev.org/663851
10:05:22 openstackgerrit Dirk Mueller proposed openstack/nova stable/queens: Changing scheduler sync event from INFO to DEBUG https://review.opendev.org/669126
10:05:34 openstackgerrit Dirk Mueller proposed openstack/nova stable/pike: Changing scheduler sync event from INFO to DEBUG https://review.opendev.org/669127
10:06:41 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: Remove MIN_{QEMU,LIBVIRT}_LUKS_VERSION https://review.opendev.org/668924
10:06:42 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: Remove MIN_LIBVIRT_PERF_VERSION https://review.opendev.org/669128
10:06:42 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: Remove support for Intel CMT `perf` events https://review.opendev.org/669129
10:14:50 openstackgerrit Merged openstack/nova-specs master: Add "Secure Boot support for KVM & QEMU guests" spec https://review.opendev.org/506720
10:15:23 kashyap johnthetubaguy: Thanks!
10:27:44 openstackgerrit Stephen Finucane proposed openstack/nova master: libvirt: Remove native LUKS compat code https://review.opendev.org/669121
10:36:10 johnthetubaguy kashyap: "Instance failed to spawn: UnsupportedHardware: Requested hardware 'virtio1.0-net' is not supported by the 'kvm' virt driver"
10:36:41 johnthetubaguy its with os_distro="centos7.0"
10:38:49 kashyap johnthetubaguy: And only with CentOS 7?
10:39:12 johnthetubaguy haven't tried many of the others right now
10:41:05 johnthetubaguy kashyap: I have written it up here for now: https://bugs.launchpad.net/nova/+bug/1835400
10:41:05 openstack Launchpad bug 1835400 in OpenStack Compute (nova) "Issues booting with os_distro=centos7.0" [Undecided,New]
10:41:26 openstackgerrit Brin Zhang proposed openstack/nova-specs master: Support for changing deleted_on_termination after boot https://review.opendev.org/580336
10:41:38 openstackgerrit Brin Zhang proposed openstack/nova-specs master: Support delete_on_termination in volume attach api https://review.opendev.org/612949
10:42:03 openstackgerrit Lee Yarwood proposed openstack/nova master: Get rid of args to RBDDriver.__init__() https://review.opendev.org/668564
10:42:21 kashyap johnthetubaguy: I don't see what you mean by "isn't known by 'os-info'". Do you mean this:
10:42:31 kashyap Whe nyou run `osinfo-query os | grep -i centos`, you don't see CentOS listed there at all?
10:42:42 johnthetubaguy correct, only centos7.0
10:43:00 kashyap Ah, it's the exact string match
10:43:59 johnthetubaguy yeah, although we get the "correct" match in the end, thats OK
10:44:05 johnthetubaguy just when it matches it all seems to fail
10:44:26 kashyap Right
10:44:58 kashyap johnthetubaguy: I'll check and report back on the bug
10:46:04 johnthetubaguy kashyap: thanks, I suspect we are meant to just put "virtio" in there? I am not sure though
10:47:21 kashyap johnthetubaguy: So ... was talking to DanPB on #virt, he does indeed say it's a Nova bug:
10:47:48 johnthetubaguy kashyap: I saw this patch: https://github.com/virt-manager/virt-manager/commit/617b92710f50015c5df5f9db15d25de18867957d
10:47:53 johnthetubaguy I guess we could do the same
10:48:00 kashyap 12:45 < danpb1> kashyap: nova is checking is_vif_model_valid_for_virt()
10:48:00 kashyap 12:45 < danpb1> kashyap: and that method has an incomplete list of hardware models
10:48:02 johnthetubaguy I don't mind patching that, if I know what we should do :)
10:48:07 kashyap ^ See above
10:48:11 johnthetubaguy ah...
10:49:48 johnthetubaguy kashyap: do you know where the full list lives? unsure if libvirt does this check
10:51:51 kashyap johnthetubaguy: /me is unsure, too; will check post lunch and get back
10:52:05 johnthetubaguy kashyap: sweet, thank you
11:04:23 openstackgerrit Balazs Gibizer proposed openstack/nova stable/rocky: Stabilize unshelve notification sample tests https://review.opendev.org/669118
11:14:24 kashyap johnthetubaguy: Still here?
11:14:27 johnthetubaguy yeah
11:15:35 kashyap johnthetubaguy: So Dan tells me there are two issues here.
11:16:28 kashyap 1. Nova's validation of n/w models is incomplete
11:16:51 kashyap 2. The "model names returned by libosinfo are *NOT* required to be valid libvirt network model names".
11:17:21 kashyap On point 2, there is no such "virtio-1.0-net" model name in libvirt, if we see here: https://libvirt.org/git/?p=libvirt.git;a=blob;f=src/conf/domain_conf.c#l515
11:17:29 johnthetubaguy yeah, it was (2) I was suspecting, given this: https://github.com/virt-manager/virt-manager/commit/617b92710f50015c5df5f9db15d25de18867957d
11:17:42 kashyap Apparently libosinfo is just telling us that the guest supports 'virio-1.0-net'
11:18:04 kashyap And there are two ways to provide that: https://libvirt.org/formatdomain.html#elementsVirtioTransitional
11:18:18 kashyap The "virtio-transitional" vs. "virtio-non-transitional" devices.
11:19:01 kashyap Where "virtio-transitional" provides a device that will auto-negotiate virtio 1.0 support; while "virtio-non-transitional" provoides a device that is fixed to always use virtio 1.0 support.
11:19:38 johnthetubaguy and "virtio" what does that do?
11:20:35 johnthetubaguy thinking ln520 in libvirt... I am wondering what we should do here
11:20:55 johnthetubaguy seems like (1) doesn't matter, the issue is our os-info logic seems totally bogus
11:21:03 kashyap johnthetubaguy: On 'virtio':
11:21:04 kashyap virtio
11:21:05 kashyap This device will work like a virtio-non-transitional device when plugged into a PCI Express slot, and like a virtio-transitional device otherwise; libvirt will pick one or the other based on the machine type. This is the best choice when compatibility with libvirt versions older than 5.2.0 is necessary, but it's otherwise not recommended to use it.
11:21:10 kashyap ---
11:21:29 johnthetubaguy ah, that's libvirt docs somewhere?
11:22:07 kashyap Yeah, I'm quoting from here: https://libvirt.org/formatdomain.html#elementsVirtioTransitional
11:22:15 kashyap (The link I posted earlier)
11:22:28 johnthetubaguy ah, cool, I only got this far: https://libvirt.org/formatdomain.html#elementsNICSModel
11:22:46 johnthetubaguy ah... so this text makes sense now
11:25:17 johnthetubaguy kashyap: maybe a better question... what is the optimum settings for all Cento7.0 devices, and how does os-info help me do the right thing
11:25:29 kashyap I think the optimium should be 'virtio', isn't it?
11:25:41 kashyap Given that it behaves correctly with versions older than 5.2.0?
11:27:23 johnthetubaguy just wondering if we set the machine type "correctly"
11:27:59 kashyap johnthetubaguy: Nova doesn't pick a machine type automatically for x86 (but it does for AArch64 and s390x)
11:28:10 kashyap johnthetubaguy: https://bugs.launchpad.net/nova/+bug/1780138
11:28:10 openstack Launchpad bug 1780138 in OpenStack Compute (nova) "Don't assume the guest machine type to be of 'pc'" [Medium,Confirmed] - Assigned to Kashyap Chamarthy (kashyapc)
11:31:11 johnthetubaguy just seems like we need to ignore libosinfo for now, and just default everything to virtio
11:31:27 kashyap Right
11:32:08 johnthetubaguy hmm, OK
11:34:10 kashyap johnthetubaguy: Added a comment here based on the chat here: https://bugs.launchpad.net/nova/+bug/1835400
11:34:10 openstack Launchpad bug 1835400 in OpenStack Compute (nova) "Issues booting with os_distro=centos7.0" [Undecided,New]
12:32:27 openstackgerrit Merged openstack/nova master: libvirt: Add a rbd_connect_timeout configurable https://review.opendev.org/667421
12:38:00 openstackgerrit Artom Lifshitz proposed openstack/nova master: [WIP-until-series-is-ready] Introduce live_migration_claim() https://review.opendev.org/635669
12:38:01 openstackgerrit Artom Lifshitz proposed openstack/nova master: New objects for NUMA live migration https://review.opendev.org/634827
12:38:01 openstackgerrit Artom Lifshitz proposed openstack/nova master: LM: add support for sending NUMAMigrateData to the source https://review.opendev.org/634828
12:38:02 openstackgerrit Artom Lifshitz proposed openstack/nova master: LM: add support for updating NUMA-related XML on the source https://review.opendev.org/635229
12:38:02 openstackgerrit Artom Lifshitz proposed openstack/nova master: RPC changes to prepare for NUMA live migration https://review.opendev.org/634605
12:38:03 openstackgerrit Artom Lifshitz proposed openstack/nova master: NUMA live migration support https://review.opendev.org/634606
12:38:03 openstackgerrit Artom Lifshitz proposed openstack/nova master: CONF.workarounds.enable_numa_live_migration is not needed >= Train https://review.opendev.org/640021
12:41:29 openstackgerrit Lee Yarwood proposed openstack/nova stable/stein: libvirt: Add a rbd_connect_timeout configurable https://review.opendev.org/669167
12:41:51 openstackgerrit Josephine Seifert proposed openstack/nova-specs master: Spec for the Nova part of Image Encryption https://review.opendev.org/608696
12:41:51 openstackgerrit Lee Yarwood proposed openstack/nova stable/rocky: libvirt: Add a rbd_connect_timeout configurable https://review.opendev.org/669168
12:42:28 openstackgerrit Lee Yarwood proposed openstack/nova stable/queens: libvirt: Add a rbd_connect_timeout configurable https://review.opendev.org/669169
12:59:37 openstackgerrit Kashyap Chamarthy proposed openstack/nova master: libvirt: vif: Remove MIN_{LIBVIRT,QEMU}_RX_QUEUE_SIZE https://review.opendev.org/669171
13:04:04 openstackgerrit Artom Lifshitz proposed openstack/nova master: Revert resize: wait for events according to hybrid plug https://review.opendev.org/667177
13:04:05 openstackgerrit Artom Lifshitz proposed openstack/nova master: Pass migration to finish_revert_migration() https://review.opendev.org/668631
13:04:05 openstackgerrit Artom Lifshitz proposed openstack/nova master: [DNM] testing bug/1813789 revert resize events https://review.opendev.org/664442

Earlier   Later