| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2019-09-15 | |||
| 16:05:08 | MarkMielke | do you containerize those too? or you just mean in general | |
| 16:05:33 | sean-k-mooney | no ovs is containerised | |
| 16:05:53 | sean-k-mooney | it runs with net=host | |
| 16:05:57 | MarkMielke | interesting - I guess it's fine since the data plane is in kernel | |
| 16:06:06 | sean-k-mooney | not always | |
| 16:06:14 | sean-k-mooney | that is what i was trying to fix | |
| 16:06:29 | sean-k-mooney | if you are using ovs-dpdk where the dataplane is in userspace | |
| 16:06:34 | sean-k-mooney | the ovsdb need to be started first | |
| 16:06:36 | MarkMielke | ah | |
| 16:06:50 | sean-k-mooney | so that the vswtihcd can detect that dpdk is enabled when it starts up | |
| 16:07:31 | MarkMielke | ovs-dpdk... that's an interesting one for a related subject... | |
| 16:07:56 | MarkMielke | it's presumedly faster than kernel, but I also discovered that the Red Hat builds of Docker use seccomp | |
| 16:08:05 | MarkMielke | and seccomp normally triggers some of the more expensive Spectre mitigations | |
| 16:08:21 | MarkMielke | i think i have that right | |
| 16:08:27 | MarkMielke | do you disable that? | |
| 16:09:58 | sean-k-mooney | by default redhat does not tend to disabel security feature | |
| 16:10:10 | sean-k-mooney | im not sure it would impact dpdk performace it might | |
| 16:10:17 | sean-k-mooney | when we mesure something like that as an imparct | |
| 16:10:34 | sean-k-mooney | we usually docmet how to disable it and let it up to the custoemr | |
| 16:10:39 | MarkMielke | it disables indirect branch spectulation for the threads :-( | |
| 16:11:07 | MarkMielke | i started changing them to prctl instead of seccomp | |
| 16:11:15 | sean-k-mooney | if there threat profile allwos it then there is no reas for them to not do it but they need to make that threat assement themselves | |
| 16:11:20 | MarkMielke | but then I found that Red Hat build of OpenJDK also uses prctl :-) | |
| 16:11:59 | MarkMielke | so now we often disable it entirely (shocker!) | |
| 16:12:00 | MarkMielke | hehe | |
| 16:13:07 | sean-k-mooney | my dev machines uses to have almost all the mitgation disabled at one point | |
| 16:13:45 | MarkMielke | i settled on a config which ends up being similar to RHEL 7.7 ended up setting by default | |
| 16:13:52 | MarkMielke | spectre_v2=retpoline | |
| 16:13:54 | sean-k-mooney | it got kind fo annoying to have to keep updating my grub config | |
| 16:15:24 | MarkMielke | ok - well i should get a few other things done. Thanks for the chat. I think I'll open that bug with regard to oslo_utils unit tests being left in nova. | |
| 16:15:42 | sean-k-mooney | o/ | |
| 16:15:58 | MarkMielke | and if it's easy to add test i'll add to oslo_utils, but if too hard, i may add another test or two for nova as separate submission | |
| 16:17:02 | sean-k-mooney | ralonsoh: its a sunday what are you doing on irc | |
| 16:17:15 | ralonsoh | hahahaah and you?? | |
| 16:17:23 | ralonsoh | sean-k-mooney, I'm trying to fix an small problem | |
| 16:17:30 | MarkMielke | it's always a small problem | |
| 16:17:38 | sean-k-mooney | im playing eve online and have it open on another monitor | |
| 16:17:43 | ralonsoh | hahahahhaah | |
| 16:17:51 | MarkMielke | then before you know it | |
| 16:17:58 | MarkMielke | your whole weekend is gone on that small problem :-) | |
| 16:18:18 | ralonsoh | no no, it's going to be just a simple commit | |
| 16:18:23 | MarkMielke | so was mine | |
| 16:18:24 | MarkMielke | haha | |
| 16:18:26 | MarkMielke | two characters | |
| 16:18:37 | sean-k-mooney | ralonsoh: im also watching the gate to make sure everythin merges | |
| 17:49:45 | MarkMielke | Correction for above conversation... oslo_utils does have tests for units. I missed them because I was looking for qemu-info info sample output and "virtual size". This does exist in Nova. However, there is unit tests for parsing the value portion of the qemu-img info output in oslo_utils, and I can easily add tests here for QEMU 4.1. | |
| 20:13:55 | openstackgerrit | Merged openstack/nova master: Claim resources in resource tracker https://review.opendev.org/678452 | |
| 21:22:51 | openstackgerrit | Merged openstack/nova master: libvirt: Enable driver discovering PMEM namespaces https://review.opendev.org/678453 | |
| 21:36:47 | openstackgerrit | Merged openstack/nova master: libvirt: report VPMEM resources by provider tree https://review.opendev.org/678454 | |
| 21:36:55 | openstackgerrit | Merged openstack/nova master: libvirt: Support VM creation with vpmems and vpmems cleanup https://review.opendev.org/678455 | |
| #openstack-nova - 2019-09-16 | |||
| 02:17:53 | openstackgerrit | Merged openstack/nova master: Deprecate CONF.workarounds.enable_numa_live_migration https://review.opendev.org/640021 | |
| 03:06:45 | openstackgerrit | ya.wang proposed openstack/nova master: Fix typor of cpu model when check CPU compatibility https://review.opendev.org/682267 | |
| 03:25:06 | openstackgerrit | Luyao Zhong proposed openstack/nova master: objects: use all_things_equal from objects.base https://review.opendev.org/681397 | |
| 03:56:05 | openstackgerrit | Boxiang Zhu proposed openstack/nova master: Make evacuation respects anti-affinity rule https://review.opendev.org/649963 | |
| 05:46:44 | nnsingh | HI all, i have one doubt, why this file name is .yaml.txt "https://github.com/openstack/placement/blob/master/etc/placement/README-policy.yaml.txt"? what the reason behind this. | |
| 05:56:32 | openstackgerrit | garyk proposed openstack/nova master: Deconstruct the mother of all locks https://review.opendev.org/682242 | |
| 05:59:44 | openstackgerrit | Boxiang Zhu proposed openstack/nova master: Fix live migration break group policy simultaneously https://review.opendev.org/651969 | |
| 07:27:50 | openstackgerrit | Bhagyashri Shewale proposed openstack/nova master: Ignore root_gb for BFV in simple tenant usage API https://review.opendev.org/612626 | |
| 09:12:13 | openstackgerrit | Brin Zhang proposed openstack/nova-specs master: Allow specify user to reset password https://review.opendev.org/682302 | |
| 09:39:09 | openstackgerrit | Arthur Dayne proposed openstack/nova-specs master: Proposal for a safer noVNC console with password authentication https://review.opendev.org/623120 | |
| 10:22:09 | donnyd | sean-k-mooney: that was exactly what I was looking for. Thanks a bunch | |
| 10:34:16 | sean-k-mooney | donnyd: you should be aware that it does not always work. on some distors there is a scrip that shuts down any running vms on host reboot. if the compute agent is not stopped first it might notice this and mark the vm as shutdown. the livbrt vm shutdown service file is there to prevent filesystem curroption by gracefully shuting down the vms instead of sig killing them | |
| 10:35:18 | sean-k-mooney | so if it does cause you issue then you have to consider if you should disable the serivce file or not. | |
| 10:37:15 | sean-k-mooney | there is also another config option in nova to disabel reporting of the vm state in the db | |
| 10:38:07 | sean-k-mooney | i think disabling that will also prevent this issue butthen if the guest does a poweroff it wont be reflected in nova status | |
| 10:38:21 | sean-k-mooney | un less they do it via that api. | |
| 10:40:31 | openstackgerrit | Merged openstack/nova master: Parse vpmem related flavor extra spec https://review.opendev.org/678456 | |
| 10:43:32 | sean-k-mooney | pmem is almost done... the PCUP can merge and all the pending feature should be landed | |
| 11:14:01 | osmanlicilegi | greetings. i'm having problem with nova-conductor, it gives "errno 111 econnrefused" while trying to connecto to rabbitmq. i know it's not a network/firewall issue because all other rabbitmq related services work like a charm. i had the same problem with nova-api and the root cause was monkey patching but this time it's not because nova-conductor does not use monkey patching. anybody had a similar | |
| 11:14:03 | osmanlicilegi | issue before? | |
| 11:19:02 | aspiers | in case anyone didn't get a chance to review my draft SEV blog post last week, here is a link which will last 24 hours (I think) https://blog.adamspiers.org/?p=1871&preview=1&_ppp=dbc2fbd3ce | |
| 11:19:20 | sean-k-mooney | the conductor shuold be useing monkey patching | |
| 11:19:21 | artom | osmanlicilegi, you'll have bette luck in #openstack, operators hang out there, this channel is for development | |
| 11:19:28 | artom | osmanlicilegi, see /topic :) | |
| 11:19:34 | sean-k-mooney | osmanlicilegi: e.g. it should be using eventlet | |
| 11:19:52 | sean-k-mooney | but i have not seen that so i dont know how to help | |
| 11:20:29 | lyarwood | artom: https://review.opendev.org/#/c/672595/ - do you think you'll have time today to work on this and potentially break out the per n-cpu service connection setting into another change? | |
| 11:20:49 | lyarwood | artom: if not I should be able to get to it this afternoon | |
| 11:21:26 | artom | lyarwood, so... Yes, but! | |
| 11:21:46 | lyarwood | But! | |
| 11:22:18 | lyarwood | Nothing good ever follows but ;) | |
| 11:22:18 | artom | But! https://review.opendev.org/#/c/681060/ is in the gate | |
| 11:22:20 | sean-k-mooney | we might want to hold of doing any upstream work until https://review.opendev.org/#/q/topic:bp/cpu-resources+(status:open+OR+status:merged) lands | |
| 11:22:40 | artom | lyarwood, which does the same thing, right? At least for the hostname part | |
| 11:22:47 | artom | For the connection I need to play around with it some more | |
| 11:24:01 | artom | lyarwood, wait, I misunderstood you, didn't I? | |
| 11:24:12 | lyarwood | artom: it does, I'd just rather do this in a helper method in the fixture instead of in a loop per test as in that series | |
| 11:24:42 | artom | You just need the heterogeneous computes helper | |
| 11:24:43 | lyarwood | artom: and again I really need something I can backport for a few fixes | |
| 11:24:52 | lyarwood | artom: yeah pretty much | |
| 11:25:13 | artom | lyarwood, well, I have to stack on top of https://review.opendev.org/#/c/681060/11 regardless | |
| 11:25:18 | sean-k-mooney | aspiers: the assertion you cant snopp if you have physical acess is a little strong. sev and mktme both do not encyrpt cache content so if you have phyical acess to the server and snoop cache you can see the unencypted state of what ever the vm is doing | |
| 11:25:19 | artom | Otherwise we'll conflict | |
| 11:25:21 | artom | But yeah | |
| 11:26:39 | sean-k-mooney | aspiers: you cant read its ram but you could constuct a view of it by recodering the reads/writes to cache | |
| 11:27:05 | lyarwood | sean-k-mooney: sorry re you're earlier comment are you also suggesting avoding POSTing new stuff at all still? I've been holding off for the last week anyway but thought we were passed the bulk of it now. | |
| 11:27:10 | artom | I think what sean-k-mooney's saying is that the SEV work was *completely* pointless >;) | |
| 11:27:10 | lyarwood | your* | |
| 11:27:52 | sean-k-mooney | lyarwood: i personally am holding off posting new stuff until stephens changes land | |
| 11:28:04 | sean-k-mooney | or we decide to punt it | |