| *** hamalq has quit IRC | 00:06 | |
| *** sapd1 has quit IRC | 00:19 | |
| *** artom has quit IRC | 00:28 | |
| *** hemanth_n has joined #openstack-nova | 00:41 | |
| *** hemanth_n has quit IRC | 00:50 | |
| openstackgerrit | likui proposed openstack/nova master: Replace getargspec with getfullargspec https://review.opendev.org/c/openstack/nova/+/790405 | 00:54 |
|---|---|---|
| *** gyee has quit IRC | 01:06 | |
| *** hoonetorg has quit IRC | 01:06 | |
| *** __ministry has joined #openstack-nova | 01:12 | |
| openstackgerrit | likui proposed openstack/nova master: Replace getargspec with getfullargspec https://review.opendev.org/c/openstack/nova/+/790405 | 01:28 |
| *** hemanth_n has joined #openstack-nova | 02:09 | |
| *** whoami-rajat has quit IRC | 02:18 | |
| openstackgerrit | Qiu Fossen proposed openstack/nova-specs master: Allow migrating PMEM's data https://review.opendev.org/c/openstack/nova-specs/+/785563 | 02:52 |
| *** swp20 has joined #openstack-nova | 03:29 | |
| *** viks____ has joined #openstack-nova | 05:08 | |
| *** ralonsoh has joined #openstack-nova | 05:10 | |
| *** pmannidi has joined #openstack-nova | 05:12 | |
| *** pmannidi has quit IRC | 05:13 | |
| *** k_mouza has joined #openstack-nova | 05:14 | |
| *** k_mouza has quit IRC | 05:18 | |
| *** cgoncalves has quit IRC | 05:27 | |
| *** cgoncalves has joined #openstack-nova | 05:27 | |
| *** bnemec has quit IRC | 05:27 | |
| *** bnemec has joined #openstack-nova | 05:29 | |
| *** ratailor has joined #openstack-nova | 05:42 | |
| *** elzim has joined #openstack-nova | 05:51 | |
| *** vishalmanchanda has joined #openstack-nova | 06:22 | |
| *** elzim has quit IRC | 06:23 | |
| *** dklyle has quit IRC | 06:39 | |
| *** links has joined #openstack-nova | 06:51 | |
| *** macz_ has joined #openstack-nova | 07:02 | |
| gibi | sean-k-mooney: thank you | 07:06 |
| *** macz_ has quit IRC | 07:07 | |
| *** andrewbonney has joined #openstack-nova | 07:23 | |
| *** aarents has quit IRC | 07:25 | |
| *** LinPeiWen80 has quit IRC | 07:27 | |
| *** aarents has joined #openstack-nova | 07:32 | |
| *** ociuhandu has joined #openstack-nova | 07:35 | |
| *** rpittau|afk is now known as rpittau | 07:38 | |
| *** tosky has joined #openstack-nova | 07:39 | |
| *** mgoddard has quit IRC | 07:45 | |
| *** ociuhandu has quit IRC | 07:45 | |
| *** ociuhandu has joined #openstack-nova | 07:46 | |
| *** mgoddard has joined #openstack-nova | 07:50 | |
| *** ociuhandu has quit IRC | 07:50 | |
| openstackgerrit | likui proposed openstack/nova master: Replace getargspec with getfullargspec https://review.opendev.org/c/openstack/nova/+/790405 | 08:06 |
| *** lucasagomes has joined #openstack-nova | 08:11 | |
| *** alexe9191 has joined #openstack-nova | 08:11 | |
| *** ociuhandu has joined #openstack-nova | 08:16 | |
| *** ociuhandu has quit IRC | 08:17 | |
| *** ociuhandu has joined #openstack-nova | 08:17 | |
| sean-k-mooney | gibi: bauzas: im not going to make a commitment now but im going to explore if i can provide a limited replacement for upstream elk stack hosted on my home infrastucture | 08:22 |
| bauzas | well | 08:22 |
| *** lpetrut has joined #openstack-nova | 08:22 | |
| bauzas | I can't say how I'm a sad panda | 08:22 |
| sean-k-mooney | my basic plan is when i redeploy my home infrastucure in a week or two im going to redeply my ci and add a job that will trigger only on upstrem zuul -1/-2 comments and try and download the logs form those jobs and ingest them into elk | 08:23 |
| sean-k-mooney | ill start with just nova and see if i can get that to work | 08:24 |
| sean-k-mooney | only looking at failed jobs should keep the bandwith mangeable, the full ci generate maybe 4TB of logs a month which i cans store but my isp would not be happy with me downloadign that much | 08:27 |
| *** martinkennelly has joined #openstack-nova | 08:31 | |
| *** rcernin has quit IRC | 08:32 | |
| swp20 | stephenfin: i use admin context to show a project's vm by id and it's success, but by name it says no name or id exist. is this a bug? | 08:33 |
| stephenfin | swp20: no, API resources are organized by ID. You need to list and manually search or use a specific search API. Look at the 'Resource.find' method in openstacksdk for an example | 08:40 |
| *** macz_ has joined #openstack-nova | 08:45 | |
| *** macz_ has quit IRC | 08:50 | |
| swp20 | stephenfin: yeah, i find the 'Resource.find' has no 'all_tenants' param, if i add the {'all_tenants':1} to kwargs, then i can get the vm by name. | 08:50 |
| swp20 | so need we claim this when we exec 'nova help show' command, it's really a bit confused. | 08:54 |
| *** LinPeiWen93 has joined #openstack-nova | 09:01 | |
| *** rcernin has joined #openstack-nova | 09:09 | |
| *** avolkov has joined #openstack-nova | 09:21 | |
| *** rcernin has quit IRC | 09:27 | |
| openstackgerrit | likui proposed openstack/osc-placement stable/ussuri: Update command help information https://review.opendev.org/c/openstack/osc-placement/+/790618 | 09:34 |
| *** rcernin has joined #openstack-nova | 09:34 | |
| *** alexe9191 has quit IRC | 09:35 | |
| openstackgerrit | likui proposed openstack/osc-placement master: setup.cfg: Replace dashes with underscores https://review.opendev.org/c/openstack/osc-placement/+/790630 | 09:41 |
| *** ociuhandu has quit IRC | 09:52 | |
| *** rcernin has quit IRC | 09:53 | |
| *** macz_ has joined #openstack-nova | 09:55 | |
| openstackgerrit | David Vallee Delisle proposed openstack/os-vif master: Creating oslo.config.opts entry_points for plugins https://review.opendev.org/c/openstack/os-vif/+/789645 | 09:57 |
| *** macz_ has quit IRC | 10:00 | |
| *** Luzi has joined #openstack-nova | 10:03 | |
| *** k_mouza has joined #openstack-nova | 10:05 | |
| *** k_mouza has quit IRC | 10:05 | |
| *** k_mouza has joined #openstack-nova | 10:07 | |
| *** ociuhandu has joined #openstack-nova | 10:10 | |
| *** dtantsur|afk is now known as dtantsur | 10:15 | |
| stephenfin | lyarwood: Could you stick this on your review queue for the week, if you've time? https://review.opendev.org/c/openstack/nova/+/676209 | 10:21 |
| stephenfin | bauzas: Any chance you could keep going through the instance_type -> flavor series? If any of them are too difficult to review, I can try break them up more https://review.opendev.org/q/topic:%2522compute_rpc_6.0%2522+status:open | 10:22 |
| bauzas | sean-k-mooney: any reason why you try to plug vifs *after* we try to reboot the instance or restart it ? https://github.com/openstack/nova/blob/7953c0197d1a4466cb5b78070d47626c92f9db6e/nova/compute/manager.py#L1137 | 10:23 |
| bauzas | stephenfin: yup, I started to look at those | 10:23 |
| bauzas | stephenfin: today, I'll do some reviews, specs first but I can try to look at your series too | 10:23 |
| stephenfin | good enough for me, ty :) | 10:24 |
| bauzas | stephenfin: btw. I'm off tonight until next week | 10:24 |
| stephenfin | okay, gtk | 10:24 |
| bauzas | (perpetual PTOs as mriedem said, that's it) | 10:24 |
| bauzas | or... just May | 10:24 |
| *** k_mouza_ has joined #openstack-nova | 10:25 | |
| *** k_mouza has quit IRC | 10:28 | |
| *** lpetrut has quit IRC | 10:30 | |
| sean-k-mooney | bauzas: we plug before we start the instance | 10:30 |
| sean-k-mooney | althopugh i guess its after in some cases | 10:31 |
| sean-k-mooney | hum | 10:31 |
| bauzas | sean-k-mooney: look at the above | 10:31 |
| sean-k-mooney | ya i am | 10:31 |
| bauzas | sean-k-mooney: we try to reboot the instance first | 10:31 |
| bauzas | and then we also try to restart it | 10:32 |
| bauzas | *before we plug the vifs* | 10:32 |
| sean-k-mooney | we might be trying to plug wtice but i dont see any reason not to plug first | 10:32 |
| sean-k-mooney | that would be more correct i think | 10:32 |
| bauzas | ok, I'm asking this because I'd prefer to pass the allocations to the new driver method I created *before* the two above calls | 10:32 |
| sean-k-mooney | ya i also think that proably makes sense | 10:33 |
| bauzas | if you reboot the instance with not the mdevs, then you got a qemu exception | 10:33 |
| sean-k-mooney | yep so it should be create mdevs then plug vifs then reboot | 10:33 |
| * bauzas also needs to find a vGPU env, btw. :) | 10:33 | |
| sean-k-mooney | maybe see if you can borrow one form james if they are not using it for qe tests | 10:35 |
| sean-k-mooney | bauzas: so ya just checked the code | 10:36 |
| sean-k-mooney | that is litgimatly a bug that plug vifs is after the vm start | 10:36 |
| sean-k-mooney | the reboot code path is assumign its done before | 10:36 |
| bauzas | kk | 10:37 |
| sean-k-mooney | it wont break things entirly in that we will fix it after but we should move it earlier | 10:37 |
| bauzas | ok, so I'll write the code | 10:37 |
| bauzas | sean-k-mooney: do you want me to write the bug report for the vifs one ? | 10:37 |
| bauzas | or you ? | 10:37 |
| sean-k-mooney | hehe i mean if you feel like it go for it | 10:38 |
| sean-k-mooney | you can point to https://github.com/openstack/nova/blob/7953c0197d1a4466cb5b78070d47626c92f9db6e/nova/virt/libvirt/driver.py#L3773-L3775 | 10:38 |
| sean-k-mooney | as evidnce that we expect ti to be alredy plugged | 10:38 |
| bauzas | ok, gotta go lunching but I'll write it later | 10:38 |
| * bauzas already has a shit ton of open tabs | 10:38 | |
| bauzas | hopefully, my new laptop would arrive... | 10:39 |
| sean-k-mooney | did you get a ship date or just hoping its on its way | 10:39 |
| bauzas | I asked | 10:39 |
| bauzas | "I just checked with our vendor and the laptop is currently in the status 'Allocated', which means that it should be sent soon (in a matter of days)." | 10:39 |
| bauzas | (I know we are in an upstream channel, but that's not a secret phrase ;) ) | 10:40 |
| bauzas | this was 7 days ago, so I'm not against a delivery surprise | 10:40 |
| bauzas | context : CPU shortages on vendors, leaving my refreshment order open for 6 months, craziness of the post-covid economy | 10:41 |
| bauzas | (also the fact that most of the vendors are fabless...) | 10:42 |
| *** rcernin has joined #openstack-nova | 10:44 | |
| sean-k-mooney | well hopefully you will get it sooner then your broadband upgrade | 10:45 |
| *** rcernin has quit IRC | 10:48 | |
| *** lpetrut has joined #openstack-nova | 10:56 | |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: Add regression test for bug #1928063 https://review.opendev.org/c/openstack/nova/+/790658 | 11:06 |
| openstack | bug 1928063 in OpenStack Compute (nova) "SEV enabled instance unable to hard reboot" [Undecided,New] https://launchpad.net/bugs/1928063 - Assigned to Lee Yarwood (lyarwood) | 11:06 |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: WIP compute: Persist image id within the system_metadata of an instance https://review.opendev.org/c/openstack/nova/+/790659 | 11:06 |
| openstackgerrit | Lee Yarwood proposed openstack/nova master: WIP hardware: Use image_meta.id within get_mem_encryption_constraint https://review.opendev.org/c/openstack/nova/+/790660 | 11:06 |
| lyarwood | stephenfin: ack I'll review the mypy stuff this week, apologies for not getting to it before PTO | 11:07 |
| stephenfin | ta | 11:14 |
| *** k_mouza_ has quit IRC | 11:17 | |
| *** ociuhandu has quit IRC | 11:21 | |
| *** ociuhandu has joined #openstack-nova | 11:22 | |
| *** ociuhandu has quit IRC | 11:26 | |
| *** lpetrut has quit IRC | 11:27 | |
| gibi | sean-k-mooney: thanks for trying to cover some part of the ELK stack. | 11:38 |
| sean-k-mooney | gibi: ill be reinstalling my home system later in the week so ill try and see if i can get something working at the weekend | 11:44 |
| sean-k-mooney | im hoping i can just use kolla's supprot for deploying elk or efk technically to maintian the stack | 11:45 |
| sean-k-mooney | what i need to figure out is what is the best way to upload the logs to it form the upstream jobs | 11:46 |
| sean-k-mooney | im thinking of trying the http plugin for fluentd to submit the logs with basically curl. | 11:49 |
| sean-k-mooney | https://docs.fluentd.org/input/http | 11:49 |
| *** k_mouza has joined #openstack-nova | 11:53 | |
| *** mkrai has joined #openstack-nova | 11:56 | |
| *** ociuhandu has joined #openstack-nova | 11:58 | |
| *** ociuhandu has quit IRC | 12:02 | |
| *** dpawlik has quit IRC | 12:12 | |
| *** admin0 has joined #openstack-nova | 12:18 | |
| *** ociuhandu has joined #openstack-nova | 12:19 | |
| admin0 | hi guys .. is it possible to add bus and slot on pci_passthrough alias whitelist ? | 12:19 |
| admin0 | alias = { "vendor_id":"8086", "product_id":"154d", "device_type":"type-PF", "name":"a1", "numa_policy":"preferred" } | 12:19 |
| admin0 | if there is multiple network-cards or multiple pci cards or the same vendor and product-id , how do I differencitate it | 12:19 |
| *** dpawlik3 has joined #openstack-nova | 12:20 | |
| *** ratailor has quit IRC | 12:22 | |
| admin0 | in whitelist we can specify the bus slot and function .. but how to do it in alias ? | 12:22 |
| admin0 | https://docs.openstack.org/nova/latest/configuration/config.html#pci.alias -- does not show this feature | 12:23 |
| *** dpawlik3 is now known as dpawlik | 12:26 | |
| admin0 | the use case is this ... say i have 4 devices, and i want to pick device 1 and 4 for vm1 and 2 and 3 for vm2 .. what happens is since i cannot pass the bus/slot on alias, it picks up device 1 and 2 for first vm and 3 and 4 for 2nd vm .. if its possible to pass the exact in alias somehow, it might solve that issue | 12:27 |
| sean-k-mooney | admin0: no | 12:30 |
| sean-k-mooney | alias are not specifc to any host | 12:30 |
| sean-k-mooney | or device | 12:30 |
| admin0 | you mean there is no feature for me to alias 4 specifc device of the same kind ? | 12:31 |
| sean-k-mooney | we do not allow alias to select a specifc devcie by design | 12:32 |
| sean-k-mooney | you can have muplite device of the same type on the host but an alais can only be used to select a device of a given type not a specific one | 12:32 |
| admin0 | sean-k-mooney, thanks ... so what possbility exits if I have multiple devices of the same type and I want to target specific ones ? | 12:33 |
| sean-k-mooney | you cannot do that with nova | 12:33 |
| sean-k-mooney | what is your usecase for doing that | 12:33 |
| sean-k-mooney | oh sorry you said it above just reading it again | 12:34 |
| sean-k-mooney | am not we cant do the 1,4 2,3 split you want | 12:34 |
| sean-k-mooney | are you doing that for numa reasons or soemthign else? | 12:35 |
| admin0 | numa reasons | 12:35 |
| sean-k-mooney | are tehy multi numa guest or singel numa | 12:35 |
| sean-k-mooney | i assume the vm has 2 numa nodes and you want one form each numa node? | 12:36 |
| sean-k-mooney | so you want to spread the allcoation across host numa nodes? | 12:36 |
| admin0 | right | 12:37 |
| *** ociuhandu has quit IRC | 12:43 | |
| sean-k-mooney | admin0: we currentl do not have a feature to enable that and if you did it today it woudl reduce your performance | 13:02 |
| sean-k-mooney | we have a downstream backlog item https://bugzilla.redhat.com/show_bug.cgi?id=1762119 which would allow use to create a pcie root complex per virtual guest numa node | 13:03 |
| openstack | bugzilla.redhat.com bug 1762119 in openstack-nova "[RFE] Create a virtual PCI root complex per guest NUMA node when using q35" [High,New] - Assigned to nova-maint | 13:03 |
| sean-k-mooney | but without that today all pci device are effectivly assocated with the guest virtual numa node 0 | 13:04 |
| sean-k-mooney | so you would have corss numa traffic if you enabeld it today | 13:04 |
| sean-k-mooney | admin0: if we create a virtual PCI root complex per guest NUMA node that will allow use to optimise the virtual toplogy but its a large change | 13:05 |
| sean-k-mooney | nova would have to fully generate the virtual pci topology and assign pci addres to each device | 13:06 |
| *** hemanth_n has quit IRC | 13:06 | |
| sean-k-mooney | its not imposiable to do and we hope to do it eventually but its of similar complexity to cpu pinning | 13:06 |
| admin0 | sean-k-mooney, thank you for the info | 13:08 |
| *** ociuhandu has joined #openstack-nova | 13:09 | |
| *** rcernin has joined #openstack-nova | 13:13 | |
| *** ociuhandu has quit IRC | 13:14 | |
| sean-k-mooney | gibi: by the way for the vdpa move ops are you oke with a specless blueprint or would you like a mini spec? | 13:18 |
| sean-k-mooney | im working on the pci in placement spec draft first in either case but cant recall if what we said in the ptg | 13:19 |
| sean-k-mooney | ah we just said "do the life cycle operations except suspend and live migrate" but did not say how we would track | 13:20 |
| *** ociuhandu has joined #openstack-nova | 13:23 | |
| *** rcernin has quit IRC | 13:26 | |
| *** ignaziocassano has joined #openstack-nova | 13:29 | |
| ignaziocassano | Hello Everyone, I got issues on GARP during live migrations on Rocky and Stein with centos 7. I applyed some patches to force legacy port binding. Now I am igrating on train and I got same issues. Any help please ? | 13:31 |
| openstackgerrit | Tobias Urdin proposed openstack/nova master: Stop leaking ceph df cmd in RBD utils https://review.opendev.org/c/openstack/nova/+/789374 | 13:36 |
| sean-k-mooney | there is not much more we can do really. | 13:37 |
| ignaziocassano | Tobias, did you solve your problems on train for live migration ? | 13:37 |
| *** Roamer` has quit IRC | 13:38 | |
| sean-k-mooney | ignaziocassano: remind me again are you using ml2/ovs with iptables | 13:38 |
| sean-k-mooney | or are you suing the ovs firewall dirver or ovn? | 13:39 |
| ignaziocassano | my configuration is firewall_driver = iptables_hybrid and I am not usinn ovn | 13:40 |
| *** artom has joined #openstack-nova | 13:41 | |
| ignaziocassano | I applyed neutron patch for evacuate and nova patches but it does not solve | 13:42 |
| ignaziocassano | https://review.opendev.org/c/openstack/neutron/+/640258/ | 13:43 |
| ignaziocassano | https://review.opendev.org/c/openstack/neutron/+/753314/ | 13:43 |
| ignaziocassano | https://review.opendev.org/c/openstack/neutron/+/766277/ | 13:43 |
| ignaziocassano | https://review.opendev.org/c/openstack/nova/+/742180/12 | 13:43 |
| ignaziocassano | https://review.opendev.org/c/openstack/nova/+/747454/4 | 13:43 |
| sean-k-mooney | so for ovs with iptables we were already pluging the port into ovs in prelive migrate | 13:44 |
| sean-k-mooney | and by that i mean creating the ovs port in the ovs db | 13:45 |
| sean-k-mooney | so with https://review.opendev.org/c/openstack/neutron/+/640258/ and https://review.opendev.org/c/openstack/neutron/+/753314/ | 13:45 |
| sean-k-mooney | actully you dont need either of those pathces | 13:46 |
| sean-k-mooney | the l2 agent woudl have wired it up because with iptabels we are createing a linux bridge and a veth pair | 13:46 |
| ignaziocassano | I am facing same problems I got with stein | 13:46 |
| sean-k-mooney | so the only neutorn patch you needed was https://review.opendev.org/c/openstack/neutron/+/766277/ to avoid the dhcp server race | 13:47 |
| ignaziocassano | On stein I solved with your workaround force legacy port binding | 13:47 |
| sean-k-mooney | well that wont actuly solve it | 13:48 |
| sean-k-mooney | when you use legacy port bindign neutron has the port boudn to the souce node until much much later in the live migrtion | 13:48 |
| sean-k-mooney | we still create the port on the destitaiton at the same time in both flows | 13:49 |
| sean-k-mooney | but we use stale info form nuetorn to do it | 13:49 |
| sean-k-mooney | ignaziocassano: are you still forcing legacy mode in train | 13:50 |
| sean-k-mooney | if you backport https://review.opendev.org/c/openstack/neutron/+/766277 it will only help if you are using multiple port bindings | 13:51 |
| ignaziocassano | Sean, I desabled the workaround for using legacy port binding. I think It is automatically disabled when I upgraded from stein to train because workaround.py is covered be new version. Right ? | 13:51 |
| sean-k-mooney | well you are expecting the new install to overright the exsiting files | 13:52 |
| sean-k-mooney | it should but that depens on how you isntalled | 13:52 |
| ignaziocassano | yum update | 13:52 |
| sean-k-mooney | i think that will yes the rpm should unpack over the modifed one | 13:53 |
| sean-k-mooney | so on the neutron side you have backported https://review.opendev.org/c/openstack/neutron/+/766277 to train | 13:53 |
| sean-k-mooney | and enabled it by seeing [nova]/live_migration_events=true | 13:54 |
| sean-k-mooney | https://review.opendev.org/c/openstack/neutron/+/766277/10/neutron/conf/common.py#177 | 13:54 |
| ignaziocassano | So | 13:54 |
| ignaziocassano | let me to verify if I understodd well | 13:54 |
| ignaziocassano | 766277 patch is alrady in centos 7 packages | 13:55 |
| *** stephenfin has quit IRC | 13:56 | |
| ignaziocassano | and I must enable live_migration_events=true | 13:57 |
| sean-k-mooney | am if 766277 patch is alrady in centos 7 packages its because we have backported that downstream and it got pulled into rdo that may be the case but i dont know why ralonsoh woudl not have also done an upstream backport | 13:58 |
| sean-k-mooney | but yes enabling live_migration_events=true will prevent the race between the dhcp agent and the l2 agent | 13:58 |
| *** mlavalle has joined #openstack-nova | 13:58 | |
| sean-k-mooney | that race exists for legacy and mutltiple prot binding workflows | 13:59 |
| *** Luzi has quit IRC | 13:59 | |
| ignaziocassano | So I must check if 766277 is in centos7 rpm packages ? | 13:59 |
| ignaziocassano | Sorry for my english | 14:00 |
| sean-k-mooney | yes i am not sure if it will be | 14:00 |
| ralonsoh | sean-k-mooney, what patch? | 14:01 |
| ignaziocassano | OK. Let me to recap: 1) check if patch is includes. If not I patch code with patch command. 2) enable live_migration_events=true | 14:01 |
| ignaziocassano | the patch is https://review.opendev.org/c/openstack/neutron/+/766277/10/neutron/conf/common.py#177 | 14:02 |
| ignaziocassano | or all 766277 patches ? | 14:03 |
| ralonsoh | I didn't backported this patch because was not required and not supported in older versions | 14:03 |
| sean-k-mooney | ralonsoh: well the dhcp race has always exsited so that technially affecte every neutron release | 14:04 |
| *** tbachman has joined #openstack-nova | 14:05 | |
| ralonsoh | only for hybrid plugin | 14:05 |
| ralonsoh | now nova has this code, I can backport it in U/S | 14:05 |
| ralonsoh | I'll do it today | 14:05 |
| sean-k-mooney | well hybrid plug was the default deployment mode until very very recently | 14:06 |
| gibi | sean-k-mooney: does the vdpa lifecycle operations handled as bugfix from API perspective or we bump a microversion for it? | 14:07 |
| sean-k-mooney | we do not bump the microversion no | 14:07 |
| sean-k-mooney | we are removing the 400 error and allowign the operations to proceed | 14:08 |
| gibi | then I think a specless bp is OK | 14:08 |
| gibi | except if you need RPC or DB changes | 14:08 |
| ignaziocassano | Ralonsoh and Sean, so what I must to do for solving the issue on centos 7 train ? | 14:08 |
| sean-k-mooney | ack ya if i need any rpc/object/db change ill file a spec | 14:08 |
| ralonsoh | ignaziocassano, wait until neutron and nova code is merged in stable/train | 14:09 |
| sean-k-mooney | ignaziocassano: we need to backport the pathces. honestly though im not planning to spend to much more time workign on this. ill start some of the backport but i started trying to get this fixed in 2018 so i dont really have the energy to keep working on this after 3 years | 14:10 |
| ignaziocassano | ralonsoh Can I tary to patch the code with patch command or by hand meanwhile the code is merged ? | 14:14 |
| ignaziocassano | sorry : tary = try | 14:14 |
| ralonsoh | for sure | 14:14 |
| ralonsoh | I'm pushing now neutron stable patches | 14:14 |
| ignaziocassano | ok | 14:14 |
| ignaziocassano | then I must enable under [nova] section in neutron.conf live_migration_events=True ? | 14:15 |
| *** macz_ has joined #openstack-nova | 14:15 | |
| gibi | sean-k-mooney: ack | 14:16 |
| ignaziocassano | ralonsoh then I must enable under [nova] section in neutron.conf live_migration_events=True ? | 14:16 |
| ralonsoh | yes | 14:17 |
| ignaziocassano | ralonsoh Ok, so I wait your push. OK ? Many thanks | 14:18 |
| ignaziocassano | sean-k-mooney Many thks | 14:19 |
| ignaziocassano | ralonsoh: when I can try yum update to see new packages ? | 14:21 |
| *** iurygregory has quit IRC | 14:22 | |
| *** iurygregory has joined #openstack-nova | 14:22 | |
| *** sapd1 has joined #openstack-nova | 14:23 | |
| *** happyhemant has joined #openstack-nova | 14:24 | |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Add image metadata property reference guide https://review.opendev.org/c/openstack/nova/+/756867 | 14:26 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Add 'nova:image-meta:' directive, role https://review.opendev.org/c/openstack/nova/+/781530 | 14:26 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: docs: Add cross-references to image metadata properties https://review.opendev.org/c/openstack/nova/+/781531 | 14:26 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Remove dead code https://review.opendev.org/c/openstack/nova/+/786291 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Remove 'nova.db.sqlalchemy.utils' https://review.opendev.org/c/openstack/nova/+/786292 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Remove unused DB methods https://review.opendev.org/c/openstack/nova/+/786293 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Use module-level imports for sqlalchemy https://review.opendev.org/c/openstack/nova/+/786295 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Fold in indexes https://review.opendev.org/c/openstack/nova/+/786296 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Fold in ForeignKey constraints https://review.opendev.org/c/openstack/nova/+/786297 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Remove 'nova.db.base' module https://review.opendev.org/c/openstack/nova/+/786298 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Copy docs from 'nova.db.*' to 'nova.db.sqlalchemy.*' https://review.opendev.org/c/openstack/nova/+/786299 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Synchronize function signatures https://review.opendev.org/c/openstack/nova/+/786300 | 14:30 |
| openstackgerrit | Stephen Finucane proposed openstack/nova master: db: Clean up migration code https://review.opendev.org/c/openstack/nova/+/786301 | 14:30 |
| *** stephenfin has joined #openstack-nova | 14:34 | |
| openstackgerrit | Artom Lifshitz proposed openstack/nova stable/wallaby: Test SRIOV port move operations with PCI conflicts https://review.opendev.org/c/openstack/nova/+/790710 | 14:38 |
| openstackgerrit | Artom Lifshitz proposed openstack/nova stable/wallaby: Update SRIOV port pci_slot when unshelving https://review.opendev.org/c/openstack/nova/+/790711 | 14:38 |
| *** ociuhandu has quit IRC | 15:09 | |
| *** dave-mccowan has quit IRC | 15:10 | |
| openstackgerrit | Archit Modi proposed openstack/nova stable/wallaby: Fix typo in test_utils https://review.opendev.org/c/openstack/nova/+/790724 | 15:10 |
| *** ignaziocassano has quit IRC | 15:11 | |
| *** dklyle has joined #openstack-nova | 15:14 | |
| *** dave-mccowan has joined #openstack-nova | 15:15 | |
| *** ociuhandu has joined #openstack-nova | 15:17 | |
| *** ociuhandu has quit IRC | 15:22 | |
| *** ociuhandu has joined #openstack-nova | 15:22 | |
| openstackgerrit | Merged openstack/nova master: Fix typo in test_utils https://review.opendev.org/c/openstack/nova/+/790510 | 15:25 |
| openstackgerrit | Tobias Urdin proposed openstack/nova master: Stop leaking ceph df cmd in RBD utils https://review.opendev.org/c/openstack/nova/+/789374 | 15:29 |
| *** ignaziocassano has joined #openstack-nova | 15:30 | |
| ignaziocassano | ralonsoh I tried to apply this patch https://review.opendev.org/c/openstack/neutron/+/766277/10/neutron/conf/common.py#177 and I configured thhe section nova in neutron.conf with live_migration_events=True but when I migrate a vm on a provider network it stops to respond to ping requests. It starts to respond only when it initiate some traffic for | 15:31 |
| ignaziocassano | example when it search the ntp server | 15:31 |
| ignaziocassano | If I logon using the console (becuse network is unreacheable) and I ping the router, it starts to respond | 15:32 |
| ralonsoh | ignaziocassano, please open a LP bug for this | 15:33 |
| ralonsoh | maybe we need to force a GARP to refresh the APR tables | 15:33 |
| ignaziocassano | ralonsoh, How can open a bug ? | 15:33 |
| ralonsoh | https://bugs.launchpad.net/neutron | 15:34 |
| ralonsoh | --> report a bug | 15:34 |
| ignaziocassano | thanks | 15:36 |
| sean-k-mooney | ralonsoh: qemu should be sendign GARP packet or thecnically RARP packets | 15:36 |
| ralonsoh | right | 15:37 |
| ralonsoh | so the arp table should be updated when the VM is migrated | 15:37 |
| sean-k-mooney | and for hybrid plug nova added a veth pair to ovs so the l2 agent shoudl have wired that up before we migrated | 15:37 |
| ralonsoh | arp tables, in all hosts | 15:37 |
| ralonsoh | right | 15:37 |
| sean-k-mooney | yep | 15:37 |
| sean-k-mooney | when qemu start on the dest it will send the RARP packets | 15:38 |
| sean-k-mooney | so i dont really know how ignaziocassano is having issue unless that happend after the dhcp server responded but before the l2 agent finsihed wiring up the ports | 15:38 |
| sean-k-mooney | on a system with a lot of port that could happen but with your patch it should not and it shoudl be rare without it | 15:39 |
| sean-k-mooney | unless the l2 agent is consitntly slow for some reason on the host | 15:39 |
| ignaziocassano | I have few virtual machines on this two kvm nodes | 15:41 |
| ralonsoh | kvm? | 15:42 |
| sean-k-mooney | ralonsoh: as in libvirt/kvm | 15:42 |
| sean-k-mooney | them mean compute nodes | 15:42 |
| ignaziocassano | yes | 15:42 |
| ignaziocassano | kvm nodes | 15:42 |
| ignaziocassano | libvirt kvm | 15:42 |
| ignaziocassano | centos 7 | 15:42 |
| ignaziocassano | I have just tryed again now. Startded I new vm. It responds to pings. Migrated it on new node. It stops to respond. I logged into vm using the console and I pinged the default router. It started immediately to respond to ping requests | 15:47 |
| sean-k-mooney | ignaziocassano: can you do that again but this time use tcpdump or tshark to dump the traffic | 15:49 |
| ignaziocassano | on the vm ? | 15:50 |
| *** elzim has joined #openstack-nova | 15:50 | |
| sean-k-mooney | on the destination host | 15:50 |
| ignaziocassano | ummh.... How can I do that ? On which interface ? | 15:51 |
| sean-k-mooney | if you log into the souce node and check the name of the tap device you can set up a watch command that will try do start dumping the packets | 15:52 |
| ignaziocassano | ok | 15:52 |
| sean-k-mooney | e.g. watch -n1 " tcpudump -i tap..." | 15:52 |
| sean-k-mooney | basicaly what im wondering is will you see the RARP packets form qemu and will they be recived before ovs is configured | 15:53 |
| sean-k-mooney | you coudl similarly check on teh br-ex/phsyical interface connected to ovs | 15:54 |
| sean-k-mooney | to see if they are sent onto the physical network | 15:54 |
| ignaziocassano | on the source node the interfcace is tap11fdfb84-4a | 15:57 |
| *** lucasagomes has quit IRC | 15:58 | |
| ignaziocassano | so I must run watch -n1 " tcpudump -i tap11fdfb84-4a" on the destination node ? | 15:58 |
| ignaziocassano | tcpdump: verbose output suppressed, use -v or -vv for full protocol decode | 16:01 |
| ignaziocassano | listening on tap11fdfb84-4a, link-type EN10MB (Ethernet), capture size 262144 bytes | 16:01 |
| ignaziocassano | 18:00:49.552314 ARP, Request who-has 10.138.209.47 (Broadcast) tell 10.138.209.1, length 46 | 16:01 |
| ignaziocassano | 18:00:49.552325 ARP, Request who-has 10.138.209.47 (Broadcast) tell 10.138.209.1, length 46 | 16:01 |
| ignaziocassano | 18:00:49.556221 ARP, Request who-has 10.138.209.48 (Broadcast) tell 10.138.209.1, length 46 | 16:01 |
| ignaziocassano | 18:00:49.556230 ARP, Request who-has 10.138.209.48 (Broadcast) tell 10.138.209.1, length 46 | 16:01 |
| ignaziocassano | 18:00:49.855902 IP6 :: > ff02::16: HBH ICMP6, multicast listener report v2, 1 group record(s), length 28 | 16:01 |
| ignaziocassano | 18:00:50.009075 IP6 fe80::fc16:3eff:fe6b:8bee > ff02::16: HBH ICMP6, multicast listener report v2, 1 group record(s), length 28 | 16:01 |
| ignaziocassano | 18:00:50.009117 IP6 fe80::fc16:3eff:fe6b:8bee > ff02::2: ICMP6, router solicitation, length 16 | 16:01 |
| ignaziocassano | 18:00:50.265259 ARP, Reverse Request who-is fa:16:3e:6b:8b:ee (oui Unknown) tell fa:16:3e:6b:8b:ee (oui Unknown), length 46 | 16:01 |
| ignaziocassano | 18:00:50.285607 ARP, Reverse Request who-is fa:16:3e:6b:8b:ee (oui Unknown) tell fa:16:3e:6b:8b:ee (oui Unknown), length 46 | 16:01 |
| ignaziocassano | 18:00:50.293932 ARP, Request who-has 10.138.248.68 tell 10.138.248.68, length 28 | 16:01 |
| ignaziocassano | 18:00:50.294153 IP6 fe80::f816:3eff:fe6b:8bee > ff02::1: ICMP6, neighbor advertisement, tgt is fe80::f816:3eff:fe6b:8bee, length 32 | 16:01 |
| ignaziocassano | 18:00:50.344367 ARP, Request who-has 10.138.248.68 tell 10.138.248.68, length 28 | 16:01 |
| ignaziocassano | 18:00:50.344397 IP6 fe80::f816:3eff:fe6b:8bee > ff02::1: ICMP6, neighbor advertisement, tgt is fe80::f816:3eff:fe6b:8bee, length 32 | 16:01 |
| ignaziocassano | 18:00:50.435854 ARP, Reverse Request who-is fa:16:3e:6b:8b:ee (oui Unknown) tell fa:16:3e:6b:8b:ee (oui Unknown), length 46 | 16:01 |
| ignaziocassano | 18:00:50.494425 ARP, Request who-has 10.138.248.68 tell 10.138.248.68, length 28 | 16:01 |
| ignaziocassano | 18:00:50.494456 IP6 fe80::f816:3eff:fe6b:8bee > ff02::1: ICMP6, neighbor advertisement, tgt is fe80::f816:3eff:fe6b:8bee, length 32 | 16:01 |
| ignaziocassano | the above is on destination node | 16:02 |
| sean-k-mooney | so the Reverse Request who-is fa:16:3e:6b:8b:ee (oui Unknown) tell fa:16:3e:6b:8b:ee | 16:02 |
| sean-k-mooney | i think was the RARP packet | 16:03 |
| sean-k-mooney | so the question is has the ovs agent finished installing the openflow rules wehn that was sent and did tha tpacket also get sent on the physical network correnctly encapsulated | 16:04 |
| *** lbragstad has quit IRC | 16:04 | |
| sean-k-mooney | ignaziocassano: can you confirm that fa:16:3e:6b:8b:ee is the vms mac address | 16:05 |
| ignaziocassano | fa:16:3e:6b:8b:ee ues | 16:06 |
| ignaziocassano | fa:16:3e:6b:8b:ee ues | 16:06 |
| sean-k-mooney | ignaziocassano: by the way longer message like logs are best shared using http://paste.openstack.org/ | 16:06 |
| ignaziocassano | fa:16:3e:6b:8b:ee yes it is | 16:06 |
| sean-k-mooney | :) | 16:06 |
| sean-k-mooney | cool so we can see that the RARP packets are sent but we need to look at the timestamp and compore that to the neutron agent logs | 16:06 |
| ignaziocassano | do you want I send on paste.openstack.org the openvswitch agent log related to destination node ? | 16:07 |
| *** LinPeiWen93 has quit IRC | 16:08 | |
| ignaziocassano | sean-k-mooney which logs do you need ? | 16:11 |
| ignaziocassano | I have 3 controllers and 2 compute nodes | 16:11 |
| *** ociuhandu_ has joined #openstack-nova | 16:13 | |
| sean-k-mooney | ignaziocassano: if you have a subset of the log for the time perod whne it booted say a 2-5 minute span around 18:00:49 | 16:14 |
| sean-k-mooney | the we coudl take a look | 16:15 |
| *** jamesden_ has joined #openstack-nova | 16:16 | |
| *** lbragstad has joined #openstack-nova | 16:16 | |
| ignaziocassano | Paste #52iiZxCu2PRtXAhf318N | 16:17 |
| *** ociuhandu has quit IRC | 16:17 | |
| ignaziocassano | is the dhacpagent log | 16:17 |
| *** jamesdenton has quit IRC | 16:17 | |
| *** ociuhandu_ has quit IRC | 16:18 | |
| ignaziocassano | sorry | 16:18 |
| ignaziocassano | http://paste.openstack.org/show/52iiZxCu2PRtXAhf318N/ | 16:18 |
| sean-k-mooney | can you do the same thing for the ovs l2 agent log, actully it might be better to grep by the port uuid or tap name | 16:19 |
| sean-k-mooney | e.g. grep -i -E "tap11fdfb84-4a|fa:16:3e:6b:8b:ee|<port uuid>" | 16:20 |
| sean-k-mooney | what im looking for specificly is the l2 agent in debug mode logs when ports are added to ovs and when the prot is "treated" | 16:21 |
| sean-k-mooney | which means it has configured ovs to handel data form the port | 16:21 |
| ignaziocassano | http://paste.openstack.org/show/805250/ | 16:22 |
| sean-k-mooney | hum i guess its only logged in dbug mode i assume you dont have this in debug | 16:23 |
| ignaziocassano | unfortunately it is not in debug mode | 16:24 |
| *** ociuhandu has joined #openstack-nova | 16:25 | |
| sean-k-mooney | ok | 16:25 |
| *** ralonsoh has quit IRC | 16:25 | |
| sean-k-mooney | ignaziocassano: i was trying to see if any of the logs form https://github.com/openstack/neutron/blob/master/neutron/plugins/ml2/drivers/openvswitch/agent/ovs_neutron_agent.py#L1793 would be printed but the ones i wanted i think are debug only | 16:26 |
| sean-k-mooney | ignaziocassano: i belive the log we were seeing before was https://github.com/openstack/neutron/blob/a12d9e41fdaf16dfefc7fe30e6198984a7588036/neutron/plugins/ml2/drivers/openvswitch/agent/ovs_neutron_agent.py#L1936 | 16:26 |
| sean-k-mooney | both of those logs are at 18:00:52.542 and 18:00:54.541 | 16:28 |
| sean-k-mooney | which is after teh RARP packets were sent | 16:28 |
| ignaziocassano | So, what can I do to help ? | 16:29 |
| sean-k-mooney | that is implying you are currntly hitting the race | 16:29 |
| sean-k-mooney | the last log in the dhcp agent was at 18:00:48.400 | 16:29 |
| sean-k-mooney | right before the vm started on the dest | 16:29 |
| sean-k-mooney | ignaziocassano: so it does look like if the dhcp race is fixed then it would fix your issue | 16:30 |
| *** ociuhandu has quit IRC | 16:30 | |
| sean-k-mooney | ignaziocassano: so i think we just need to wait for rodolfo to finish backporting that patch to train | 16:30 |
| sean-k-mooney | this is inline whit what i would expect without https://review.opendev.org/c/openstack/neutron/+/766277 | 16:31 |
| ignaziocassano | Ok, I did not understand if when Rodolfo will finish I can apply new code with yum update command .... | 16:31 |
| sean-k-mooney | well eventually yes. after teh backport is done a new rpm will have to be build by RDO | 16:32 |
| sean-k-mooney | so it will be included in the next stable releas after its backported | 16:32 |
| *** dtantsur is now known as dtantsur|afk | 16:32 | |
| ignaziocassano | So I will wait. I am not running for upgrading to train. The first step will be update from queens to stein in the next month. | 16:33 |
| * bauzas is done for the week, wrote some reviews | 16:33 | |
| sean-k-mooney | bauzas: ack enjoy your pto | 16:34 |
| ignaziocassano | Then I will upgrade to train. | 16:34 |
| bauzas | see you on Monday folks, but I'll errand on my laptop tomorrow morning | 16:34 |
| bauzas | (I have a few specs I'd continue to look) | 16:34 |
| bauzas | gibi: I was terrible with both of your specs, don't judge me. | 16:34 |
| bauzas | :p | 16:34 |
| bauzas | sean-k-mooney: fwiw, I'm still blocked with the mdev recreate bugfix | 16:35 |
| gibi | bauzas: thanks for the review | 16:35 |
| sean-k-mooney | bauzas: did you push anything yet | 16:35 |
| gibi | bauzas: so I guess you will be back next week | 16:35 |
| ignaziocassano | sean-k-mooney I must arrive to train and then I must decide which operating system I will use on newer openstack releases | 16:35 |
| bauzas | sean-k-mooney: I can pass the allocations, but as we only have the rp uuid, I also need to get the provider tree | 16:36 |
| sean-k-mooney | i can take a look while your off otherwise ping me on monday and i can take a look with you or talke things thorugh | 16:36 |
| bauzas | gibi: indeed, Monday that's it | 16:36 |
| *** rpittau is now known as rpittau|afk | 16:36 | |
| gibi | OK, then we I will prepare response for Monday | 16:37 |
| sean-k-mooney | ignaziocassano: ok well once an upstream backport is doen it shoudl flow into the distro versions | 16:37 |
| bauzas | sean-k-mooney: basically, I'm torn between #A : look up the allocations in the compute first and pass the inventories | 16:37 |
| gibi | bauzas: have a nice PTO | 16:37 |
| bauzas | #B : somehow get a provider tree *before* we call the RT | 16:37 |
| bauzas | gibi: well, I'll stay at home but sure | 16:37 |
| ignaziocassano | sean-k-mooney many thanks for your help. Bye | 16:37 |
| gibi | bauzas: I hope you have beers at least | 16:38 |
| sean-k-mooney | b seam more complex ok lets talk about it when your back | 16:38 |
| bauzas | gibi: I have coronas | 16:38 |
| gibi | lol | 16:38 |
| gibi | I hope you refer to the beer not the virus | 16:38 |
| bauzas | sean-k-mooney: I'll upload something tomorrow | 16:38 |
| sean-k-mooney | bauzas: gibi said beer :P | 16:38 |
| bauzas | gibi: I'm vaccinated against both :p | 16:38 |
| *** ignaziocassano has quit IRC | 16:38 | |
| gibi | :) | 16:39 |
| sean-k-mooney | corona is the beer peopel recommend to people that like cider like i do but i cant stand it | 16:39 |
| bauzas | I really have coronas in my fridge but I don't want to drink them :) | 16:39 |
| bauzas | I don't like them | 16:39 |
| bauzas | and I'm also vaccinated (I'm an election official for some June regional elections, so I got a priority) | 16:40 |
| bauzas | first jab | 16:40 |
| bauzas | next jab on June 7th | 16:40 |
| sean-k-mooney | i see good way to get peopel to volenterr to do that i guess | 16:40 |
| bauzas | sean-k-mooney: yup, indeed, a good way for a civic volunteering | 16:41 |
| bauzas | (but I wasn't knowing I could get vaccinated when someone asked me to be an official :) ) | 16:41 |
| bauzas | sean-k-mooney: fwiw, I'll upload my bugfix as a WIP tomorrow morning | 16:42 |
| sean-k-mooney | gibi: ill follow up with infra and see if the review priority lable is sticky or not | 16:42 |
| sean-k-mooney | if it is or can be made sticky across patchset ill enable that | 16:43 |
| bauzas | sean-k-mooney: gibi: I also reviewed this (the label thingie) | 16:43 |
| gibi | sean-k-mooney: thanks | 16:43 |
| bauzas | and my only concern is that I wonder why we would need to have a label if only cores use it | 16:43 |
| gibi | bauzas: I responded yesterday to your comments | 16:43 |
| gibi | ohh, then you responded to me | 16:43 |
| bauzas | gibi: and I replied today | 16:43 |
| gibi | cool | 16:43 |
| bauzas | yup | 16:43 |
| gibi | we as cores need to know what are the reviews we should looking at | 16:44 |
| bauzas | my take is that we know we have contributors that don't use IRC for pinging us to review | 16:44 |
| bauzas | and in between us, we already ping us for knowing what to review | 16:44 |
| gibi | bauzas: true, you are doing it nicely and asking time to time that is there anything you should review | 16:45 |
| gibi | which is cool | 16:45 |
| sean-k-mooney | gibi: ah i think its configurable in the labe definiton | 16:45 |
| gibi | but if anybody can set priority label then every patch will have it as everybody wants to land their patch | 16:46 |
| stephenfin | yeah, this ^ | 16:46 |
| bauzas | actually the question I have is "what do we want to do by using labels ?" | 16:46 |
| stephenfin | if everything is a priority, nothing is | 16:46 |
| bauzas | that's why I proposed a second level | 16:46 |
| bauzas | or this could be a second label | 16:46 |
| gibi | and that is why I propose to have only the second level :) | 16:46 |
| bauzas | I guess I need to see how this works | 16:47 |
| gibi | we can do an experiment | 16:47 |
| stephenfin | I don't think that's necessary. I occasionally review the list of open bugs looking for things that are important. I can set the label for anything I find | 16:47 |
| stephenfin | or anyone else | 16:47 |
| bauzas | gibi: yep, hence my concerns are nitty | 16:47 |
| bauzas | gibi: but again, I need to understand what we would do | 16:47 |
| bauzas | like, I'm seeing some changes and think those are nice for other cores | 16:48 |
| bauzas | then I'll set the label to +1 | 16:48 |
| bauzas | magically, other cores will see those changes in their dashboards | 16:48 |
| bauzas | and will consider signing off for reviews | 16:48 |
| bauzas | so, the label is just a signal to the other cores ? | 16:49 |
| bauzas | or, | 16:49 |
| bauzas | the other way | 16:49 |
| bauzas | someone pings me on IRC begging for reviews | 16:49 |
| bauzas | i don't have time now but I want to review such | 16:49 |
| gibi | it is signal for other cores that this patch seems to be ready and important. And it is signal for the author to expect incoming reviews | 16:49 |
| bauzas | so I'd tag the change for later | 16:49 |
| gibi | bauzas: your second use case already covered by starring a review | 16:49 |
| bauzas | gibi: right, I know | 16:50 |
| bauzas | gibi: okay, then the first usecase I mentioned, which I honestly think won't help | 16:50 |
| bauzas | because we assume the signal isn't lost | 16:50 |
| gibi | even if we look at the marked review on the weekly | 16:50 |
| gibi | ? | 16:50 |
| bauzas | and we assume the cores will commit to it, whatever their respective bandwidthes are | 16:51 |
| bauzas | gibi: the only benefit I see honestly is the weekly meeting | 16:51 |
| bauzas | gibi: because we already signal important changes we care by other means, like pings | 16:52 |
| gibi | right. I consider adding +1 priority only to a patch when I commit to review it within a week or even less | 16:52 |
| bauzas | ok, so the signal is different | 16:52 |
| bauzas | you're not sending a signal that's "cores, please review" | 16:53 |
| bauzas | but rather a signal saying "I have free time to this change and I commit to it" | 16:53 |
| bauzas | that's not an actional item for others then, it's an actional item for the core | 16:53 |
| bauzas | (which is better than now, agreed= | 16:54 |
| gibi | my runway experience was that thing in the slot did not get review bandwidht because we missed the commitment part | 16:54 |
| bauzas | agreed | 16:54 |
| bauzas | 100% to it | 16:54 |
| bauzas | but I think I wasn't clear about the use of this new label | 16:54 |
| bauzas | and i also think the change isn't explaining how we'll use it | 16:55 |
| bauzas | if it's a label for me committing to a change, then OK | 16:55 |
| bauzas | that being said, ones could tell we can just use gerrit CCs | 16:56 |
| sean-k-mooney | gibi: bauzas so just asked fungi , the review priorty lable is not defiend globlly so we can define it as we want for nova and we can make it sticky yes | 16:56 |
| fungi | yeah, look at cinder's acl for an example | 16:56 |
| bauzas | I'm not debating over the label itself | 16:56 |
| sean-k-mooney | https://opendev.org/openstack/project-config/src/branch/master/gerrit/acls/openstack/cinder.config#L25-L35 | 16:56 |
| bauzas | I'm more concerned by the solely usage we, as nova contributors, will make of it | 16:56 |
| gibi | bauzas: point taken. I need to write the procee up in a patch to the intree contributor doc | 16:56 |
| sean-k-mooney | bauzas: sure just confirm ing we can customise it to have the behavior we desire | 16:57 |
| bauzas | gibi: one last thing before I stop | 16:57 |
| gibi | bauzas: I wanted to start the discussion before I do that. Now I think I got enough input from you to do it | 16:57 |
| bauzas | gibi: I'm not sure we need to restrict such label to the cores | 16:57 |
| bauzas | if it's a signal about committing to review | 16:57 |
| bauzas | this could also be a way for attracting people to the vows of review | 16:58 |
| gibi | good point | 16:58 |
| sean-k-mooney | well if you look at cinders usage its quite interesting | 16:58 |
| sean-k-mooney | value = -1 Branch Freeze | 16:58 |
| sean-k-mooney | value = 0 No Priority | 16:58 |
| sean-k-mooney | value = +1 Important Change | 16:58 |
| sean-k-mooney | value = +2 Gate Blocker Fix / Urgent Change | 16:58 |
| bauzas | sean-k-mooney: that's very different from what gibi was suggesting | 16:59 |
| sean-k-mooney | they have it resticted to the core group | 16:59 |
| fungi | i have to agree, restricting access because of a fear of abuse is a premature optimization which can stifle involvement from newcomers. i recommend only restricting it if actual abuse is observed and is otherwise uncontrollable | 16:59 |
| sean-k-mooney | bauzas: it is which is why i was highigly a different usage patteren | 16:59 |
| bauzas | sounds like we have a debate :) | 17:00 |
| bauzas | sean-k-mooney: just before you came back, I was exposing to gibi the fact that I don't think the signal is useful if it's just a ask for cores to review | 17:00 |
| sean-k-mooney | fungi: wel part of the idea was to use it to replace oru old runway proces where anyone could add something to the queue but only core move thing into the runway slots | 17:00 |
| bauzas | sean-k-mooney: and gibi told me the crux of the idea was "commitment" | 17:01 |
| sean-k-mooney | bauzas: yes that is why it was going to be restrited to cores to set | 17:01 |
| fungi | maybe look at it as being similar to bug triage. do you restrict who can set the severity on bug reports? | 17:01 |
| bauzas | sean-k-mooney: the problem is that we had a process with runways which didn't really work because of the lack of commitment | 17:01 |
| bauzas | fungi: nope | 17:01 |
| sean-k-mooney | fungi: its an open team but ithink the motivates are different | 17:01 |
| sean-k-mooney | bauzas: well not entirly | 17:02 |
| bauzas | sean-k-mooney: if we're creating labels for marking changes for reviews, we'll just recreate the runway process but gerrit driven this time | 17:02 |
| bauzas | which honestly didn't greatly work | 17:02 |
| sean-k-mooney | bauzas: partly | 17:02 |
| bauzas | gibi: sean-k-mooney: here is what I'm proposing | 17:02 |
| * gibi listens but pretty out of steam already today | 17:03 | |
| bauzas | we should Depend-On the ACL change to some new change in the contribs docs | 17:03 |
| bauzas | that would explain the process | 17:03 |
| bauzas | b/c the acl change is just implementation | 17:03 |
| bauzas | but we miss the design phase | 17:04 |
| sean-k-mooney | sure no issue with that | 17:04 |
| fungi | but yes, if the goal is to only ever have x number of "priority" flagged reviews, and you expect to have a limited number of people with the necessary perspective to decide which those are, then restricting the label might be a way of avoiding problems with coordination. it all depends on how you expect to use that signal | 17:04 |
| gibi | bauzas: I agree | 17:04 |
| bauzas | gibi: take a break and just commit yourself about describing your ideal view on what should be the process for those labels in a separate change | 17:04 |
| gibi | bauzas: and I sign up to write the proposal | 17:04 |
| bauzas | cool | 17:04 |
| bauzas | I'm way more interested in debating over this change rather than the gerrit one :p | 17:05 |
| gibi | thank you all for the input I really appreciate it | 17:05 |
| bauzas | np, and it's late and I'm thristy typing | 17:05 |
| * bauzas will avoid the coronas and just use his keg again | 17:06 | |
| sean-k-mooney | bauzas: i will happily try and translate what ever ye decied into gerrit | 17:06 |
| bauzas | sean-k-mooney: thanks | 17:06 |
| bauzas | folks, should be around unofficially tomorrow morning | 17:06 |
| bauzas | ++ | 17:06 |
| *** elzim has quit IRC | 17:07 | |
| *** andrewbonney has quit IRC | 17:07 | |
| gibi | bauzas: o/. | 17:08 |
| *** elzim has joined #openstack-nova | 17:15 | |
| *** sapd1 has quit IRC | 17:17 | |
| *** dklyle has quit IRC | 17:26 | |
| *** dklyle has joined #openstack-nova | 17:27 | |
| *** lpetrut has joined #openstack-nova | 17:55 | |
| *** lpetrut has quit IRC | 17:57 | |
| *** happyhemant has quit IRC | 18:23 | |
| *** whoami-rajat_ has joined #openstack-nova | 18:33 | |
| *** vishalmanchanda has quit IRC | 18:34 | |
| *** gyee has joined #openstack-nova | 18:45 | |
| *** hamalq has joined #openstack-nova | 18:50 | |
| *** yoctozepto has quit IRC | 19:13 | |
| *** yoctozepto6 has joined #openstack-nova | 19:14 | |
| *** links has quit IRC | 19:29 | |
| *** elzim has quit IRC | 20:07 | |
| *** hamalq has quit IRC | 20:09 | |
| *** macz_ has quit IRC | 20:09 | |
| *** rpioso has quit IRC | 20:09 | |
| *** erbarr has quit IRC | 20:11 | |
| *** hack-char has quit IRC | 20:12 | |
| *** hack-char has joined #openstack-nova | 20:12 | |
| *** hamalq has joined #openstack-nova | 20:13 | |
| *** macz_ has joined #openstack-nova | 20:13 | |
| *** rpioso has joined #openstack-nova | 20:13 | |
| *** erbarr has joined #openstack-nova | 20:13 | |
| *** damiandabrowski has quit IRC | 20:16 | |
| melwitt | dansmith: I noticed you reviewed the rbd error sanitize patch earlier today, it depends on this one below it that re-enables the rbd unit tests, if you wanted to hit that too https://review.opendev.org/c/openstack/nova/+/790511 | 20:28 |
| dansmith | oh sorry | 20:40 |
| dansmith | I actually had that one open too and then just didn't look at it | 20:40 |
| *** avolkov has quit IRC | 20:41 | |
| melwitt | thanks :) | 20:42 |
| *** avolkov has joined #openstack-nova | 20:47 | |
| lyarwood | *facepalm* sorry about that melwitt | 21:00 |
| melwitt | np lyarwood | 21:01 |
| amodi | Stable branch requires either cherry-pick -x headers or [stable-only] tag! | 21:02 |
| amodi | https://review.opendev.org/c/openstack/nova/+/790724 can i not cherry=pick from gerrit ui? | 21:02 |
| amodi | what did i miss | 21:02 |
| openstackgerrit | Merged openstack/nova master: rbd: Get rbd_utils unit tests running again https://review.opendev.org/c/openstack/nova/+/790511 | 21:03 |
| melwitt | amodi: the gerrit ui only does the needed "cherry-pick -x" after the change you want to backport has merged. if you do the cherry pick in the ui before it merges, it will miss the -x option to cherry-pick | 21:04 |
| amodi | melwitt: ohh, ok so i guess i need to abandon this and re-cherry-pick it? now that the change has merged | 21:05 |
| melwitt | amodi: I'm not sure if there's a way to re-kick it from the ui perspective. you could try it but I'm not sure if that will work. usually we just do an edit to the commit message to add the line manually if this happens | 21:06 |
| amodi | i want to try with ui itself, one sec | 21:06 |
| openstackgerrit | melanie witt proposed openstack/nova stable/wallaby: rbd: Get rbd_utils unit tests running again https://review.opendev.org/c/openstack/nova/+/790836 | 21:07 |
| openstackgerrit | Archit Modi proposed openstack/nova stable/wallaby: Fix typo in test_utils https://review.opendev.org/c/openstack/nova/+/790724 | 21:07 |
| amodi | melwitt: yeah looks like it re-kicked it correctly, https://review.opendev.org/c/openstack/nova/+/790724 | 21:08 |
| amodi | thanks!! | 21:08 |
| melwitt | kewl | 21:08 |
| *** slaweq has quit IRC | 21:19 | |
| *** hack-char has quit IRC | 21:40 | |
| *** hack-char has joined #openstack-nova | 21:40 | |
| *** rcernin has joined #openstack-nova | 22:05 | |
| *** mkrai has quit IRC | 22:24 | |
| *** rcernin has quit IRC | 22:27 | |
| *** rcernin has joined #openstack-nova | 23:02 | |
| *** dave-mccowan has quit IRC | 23:02 | |
| *** rcernin has quit IRC | 23:02 | |
| *** rcernin has joined #openstack-nova | 23:02 | |
| *** whoami-rajat_ is now known as whoami-rajat | 23:05 | |
| *** macz_ has quit IRC | 23:17 | |
| *** tosky has quit IRC | 23:24 | |
| *** ebbex has quit IRC | 23:31 | |
| *** zzzeek has quit IRC | 23:34 | |
| *** zzzeek has joined #openstack-nova | 23:35 | |
| *** mlavalle has quit IRC | 23:44 | |
| *** ebbex has joined #openstack-nova | 23:51 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!