| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-09-03 | |||
| 06:21:50 | openstackgerrit | Luyao Zhong proposed openstack/nova master: Add functional tests for virtual persistent memory https://review.opendev.org/678470 | |
| 06:21:50 | openstackgerrit | Luyao Zhong proposed openstack/nova master: libvirt: Enable driver configuring PMEM namespaces https://review.opendev.org/679640 | |
| 08:32:55 | stephenfin | alex_xu: Could you look at https://review.opendev.org/#/c/678861/ and https://review.opendev.org/#/c/678902/ ? | |
| 09:11:41 | yaawang | stephenfin: Hi, cloud you review https://review.opendev.org/#/c/670298/ and https://review.opendev.org/#/c/670299/ and https://review.opendev.org/#/c/670300/ :) | |
| 10:14:07 | luyao | stephenfin: Could you help review the rest of the patch series if you have time? :) https://review.opendev.org/#/q/topic:bp/virtual-persistent-memory+(status:open+) | |
| 10:22:57 | cdent | is 'nova-grenade-multinode' not working the latest "gate's not happy"? | |
| 10:23:03 | cdent | or is it just me | |
| 10:27:57 | gibi | cdent: could you link the failing job? | |
| 10:28:33 | cdent | gibi: https://review.opendev.org/#/c/679627/ https://zuul.opendev.org/t/openstack/build/af05b1eaef0047a6a2f2f466e93b5ac5 | |
| 10:29:06 | cdent | I've seen that few different times this morning, as I'm trying to catch up. haven't dug into it yet (still trying to catch up) | |
| 10:30:18 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Simplify 'fakelibvirt.HostInfo' object https://review.opendev.org/678861 | |
| 10:30:19 | openstackgerrit | Stephen Finucane proposed openstack/nova master: objects: Rename 'fields' import to 'obj_fields' https://review.opendev.org/674103 | |
| 10:30:19 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start checking compute usage in functional tests https://review.opendev.org/678902 | |
| 10:30:20 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start reporting PCPU inventory to placement https://review.opendev.org/671793 | |
| 10:30:20 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Fold in argument to '_update_provider_tree_for_vgpu' https://review.opendev.org/676729 | |
| 10:30:21 | openstackgerrit | Stephen Finucane proposed openstack/nova master: hardware: Differentiate between shared and dedicated CPUs https://review.opendev.org/671800 | |
| 10:30:21 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: '_get_(v|p)cpu_total' to '_get_(v|p)cpu_available' https://review.opendev.org/672693 | |
| 10:30:22 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add support for translating CPU policy extra specs, image meta https://review.opendev.org/671801 | |
| 10:30:22 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Start reporting 'HW_CPU_HYPERTHREADING' trait https://review.opendev.org/675571 | |
| 10:30:23 | openstackgerrit | Stephen Finucane proposed openstack/nova master: libvirt: Always enable the NUMATopologyFilter https://review.opendev.org/679745 | |
| 10:30:23 | openstackgerrit | Stephen Finucane proposed openstack/nova master: Add reshaper for PCPU https://review.opendev.org/674895 | |
| 10:32:02 | stephenfin | alex_xu: Added resize tests to https://review.opendev.org/671801 as promised | |
| 10:32:08 | gibi | cdent: did you found the exact failure in the logs? is it https://storage.gra1.cloud.ovh.net/v1/AUTH_dcaab5e32b234d56b626f72581e3644c/logs_27/679627/2/check/nova-grenade-multinode/af05b1e/logs/error.txt.gz ? or I | |
| 10:32:13 | gibi | or I'm on a wrong track | |
| 10:37:22 | cdent | gibi: yeah, that's the one I saw, while skimming the genade log | |
| 10:44:22 | openstackgerrit | Adam Spiers proposed openstack/nova master: Make _get_cpu_feature_traits() always return a dict https://review.opendev.org/679568 | |
| 11:13:50 | openstackgerrit | sean mooney proposed openstack/nova master: [DNM] test migration with pinning https://review.opendev.org/679754 | |
| 11:13:50 | openstackgerrit | sean mooney proposed openstack/nova master: multi numa nfv testing job https://review.opendev.org/679656 | |
| 11:15:52 | sean-k-mooney | artom: ^ the second patch should get the pinning only case and i also am testign we can spawn with explcit small pages via the alternitive flavor | |
| 11:16:33 | sean-k-mooney | we are not migrating with the alternive flavor but i guess i could add a third patch to do that | |
| 11:16:52 | sean-k-mooney | e.g. hugepage+pinning+multi-numa+realtime | |
| 11:17:20 | sean-k-mooney | im going to test that config manually now however | |
| 11:19:20 | sean-k-mooney | artom: we might also want to invert the depens on relationship in a futre version so your change depnd on the zuul job instead of the other way aroudn | |
| 11:20:29 | artom | sean-k-mooney, we're not migrating with the alt flavor? That sucks, it was kinda the whole point | |
| 11:20:45 | sean-k-mooney | we migrate with the standard flaovr | |
| 11:20:51 | sean-k-mooney | we resize to the alt flavor | |
| 11:21:09 | sean-k-mooney | the standard flavor is multi-numa-128-1-2 | |
| 11:21:18 | sean-k-mooney | e.g. hugepages and 2 numa nodes | |
| 11:21:29 | artom | Oh | |
| 11:21:32 | artom | https://review.opendev.org/#/c/679754/1/playbooks/nfv/nfv.yaml@14 is fine anyways | |
| 11:21:33 | sean-k-mooney | the alt flaovr is the one with pinning and realtime + hugepages | |
| 11:21:35 | artom | It's what I had in mind | |
| 11:22:20 | sean-k-mooney | im using the secondary pinned-128-1-2 flavor to do some other testing as a side effect of the job | |
| 11:22:30 | sean-k-mooney | its unrelated to live migration | |
| 11:22:41 | sean-k-mooney | i just want to make sure that we can trun on all the things | |
| 11:22:52 | sean-k-mooney | which is chacked by the resize tests | |
| 11:28:05 | gibi | cdent: the bug with grenade + cinder seems to be already reported an tracked http://status.openstack.org/elastic-recheck/#1840355 | |
| 11:28:31 | cdent | the error message is the best ever... | |
| 11:28:37 | cdent | yes, yes it is | |
| 11:30:29 | sean-k-mooney | i mean its not wrong | |
| 11:43:22 | yonglihe | stephenfin: Comments addressed. and thanks. https://review.opendev.org/#/c/621476/ | |
| 12:01:07 | sean-k-mooney | yaawang: the vcpu model selection series sure. im just grabing lunch but ill review it after | |
| 12:14:13 | yaawang | sean-k-mooney: thk :) | |
| 12:19:43 | brtknr | Hi all, any users with experience using fedora-coreos on openstack? | |
| 12:20:00 | brtknr | do I need to explicitly inject ssh key to the instance? | |
| 12:25:49 | alex_xu | stephenfin: cool | |
| 12:36:13 | stephenfin | bauzas: Could you look at https://review.opendev.org/#/c/679745 ? | |
| 12:36:47 | sean-k-mooney | brtknr: fedora core os does not have any password by defualt so if you dont inject ssh keys by selecting a key pair you wont be able to log in | |
| 12:37:49 | brtknr | sean-k-mooney: I selected a keypair when I created my instance but it is not letting me in as core user | |
| 12:37:50 | openstackgerrit | Silvan Kaiser proposed openstack/nova stable/stein: Exec systemd-run without --user flag in Quobyte driver https://review.opendev.org/660705 | |
| 12:40:55 | brtknr | sean-k-mooney: default user is core correct? | |
| 12:41:55 | sean-k-mooney | its would be core or fedora i suspect. i dont actully use fedora core os but i knwo that the cloud images do not have a password set | |
| 12:44:16 | openstackgerrit | Bhagyashri Shewale proposed openstack/nova master: Ignore root_gb for BFV in simple tenant usage API https://review.opendev.org/612626 | |
| 12:54:28 | bauzas | stephenfin: sure | |
| 13:22:29 | sean-k-mooney | artom: Live Migration failure: 'NoneType' object has no attribute 'set' | |
| 13:27:07 | sean-k-mooney | ill replicate this in the ci job but when the souce host only hase cpu enabled in the first numa node and the dest only in the second numa node it failes with that error | |
| 13:27:35 | artom | sean-k-mooney, yeah, that's what I expected | |
| 13:27:46 | artom | sean-k-mooney, it's fixed in a local version, need to finish unit tests and push | |
| 13:28:03 | sean-k-mooney | ok im reconfriguring my vcpu_pin_set | |
| 13:28:14 | sean-k-mooney | and ill continue testing each feature independtly | |
| 13:29:38 | donnyd | sean-k-mooney: Any issues with NUMA I can help with today? | |
| 13:30:09 | sean-k-mooney | donnyd: nope your ci is working fine. we are now testing different edgecases | |
| 13:31:10 | sean-k-mooney | donnyd: i might add some new lables in the future e.g. multi-numa and nested-virt so that they are more expclit | |
| 13:31:50 | sean-k-mooney | donnyd: if you want to replicate the flaovr to a new on i can swap it over to use that | |
| 13:31:54 | donnyd | At some point we should probably create some sensical labels | |
| 13:32:03 | sean-k-mooney | yep | |
| 13:33:13 | sean-k-mooney | if you can copy the flaovr you create and give it a new name i can add a multi-numa-ubuntu-bionic lable that uses it | |
| 13:33:50 | donnyd | Ok, so you want to handle the project config end and I can do the flavor side? correct? | |
| 13:34:17 | sean-k-mooney | well i dont mind if you do the porject config stuff but im happy to do it too | |
| 13:34:26 | stephenfin | cdent: no. | |
| 13:34:37 | donnyd | I can take care of it np | |
| 13:34:51 | sean-k-mooney | cdent: he is obvioulsy the next oslo ptl :P | |
| 13:35:01 | cdent | maybe we should just dissolve things then :) | |
| 13:35:10 | artom | Did efried desist or something? | |
| 13:35:25 | sean-k-mooney | artom: no he just has not renominated himself | |
| 13:35:39 | sean-k-mooney | he has like 9 hours left | |
| 13:35:49 | sean-k-mooney | no one has nominated for nova | |
| 13:36:03 | efried | what what? | |
| 13:36:13 | artom | We're grown ups, we can be trusted without an adult, right? | |
| 13:36:34 | donnyd | 8cpu-16GBram-80GBdisk-multi-numa-nested-ubuntu-bionic | |
| 13:36:34 | donnyd | 8cpu-8GBram-80GBdisk-multi-numa-nested-ubuntu-bionic | |
| 13:36:40 | donnyd | so like those sean-k-mooney | |
| 13:37:03 | efried | artom: If I've learned anything from doing this, it's that a PTL is not useful for "adulting" the team. | |
| 13:37:34 | efried | Run meetings, ack release patches, organize milestoney things, other paperwork. | |
| 13:38:03 | artom | efried, heh - from the outside looking in, and without wanting to sound pejorative, PTL, at least for nova, looks like a more... secretary-type? job | |
| 13:38:14 | efried | just so | |
| 13:38:14 | artom | Perhaps admin/organizer would be a better word | |
| 13:38:24 | sean-k-mooney | donnyd: they dont need to be that long be sure | |
| 13:38:45 | sean-k-mooney | donnyd: if thoes are teh flavor you created thats fine | |
| 13:38:53 | kashyap | Yeah, "secretary" undermines the actual (unthankful) work involved | |