*** lixiaoy1 has joined #openstack-cinder | 00:32 | |
*** LiangFang has joined #openstack-cinder | 00:34 | |
*** yikun has joined #openstack-cinder | 00:52 | |
openstackgerrit | Merged openstack/cinder master: api-ref: add docs for os-migrate_volume API https://review.openstack.org/595379 | 00:58 |
---|---|---|
openstackgerrit | Merged openstack/cinder master: Avoid using dict.get() in assertions https://review.openstack.org/597377 | 00:58 |
*** Dinesh_Bhor has joined #openstack-cinder | 01:23 | |
openstackgerrit | Merged openstack/cinder master: Drop legacy backup service support https://review.openstack.org/595372 | 01:54 |
openstackgerrit | Merged openstack/cinder master: Fix api-ref title levels and index https://review.openstack.org/592026 | 01:54 |
openstackgerrit | Merged openstack/cinder master: Fix deprecation warnings https://review.openstack.org/595833 | 01:54 |
*** erlon has quit IRC | 02:01 | |
*** Dinesh_Bhor has quit IRC | 02:02 | |
*** zhaochao has joined #openstack-cinder | 02:07 | |
*** Dinesh_Bhor has joined #openstack-cinder | 02:11 | |
*** psachin has joined #openstack-cinder | 02:39 | |
*** Bhujay has joined #openstack-cinder | 02:49 | |
*** imacdonn has quit IRC | 02:50 | |
*** imacdonn has joined #openstack-cinder | 02:50 | |
openstackgerrit | LisaLi proposed openstack/cinder-specs master: Driver reinitialization after failure https://review.openstack.org/599505 | 02:51 |
openstackgerrit | Yikun Jiang proposed openstack/cinder-specs master: Add volume re-image api https://review.openstack.org/605317 | 03:04 |
*** Bhujay has quit IRC | 03:07 | |
*** LiangFang has quit IRC | 03:11 | |
openstackgerrit | Merged openstack/cinder master: Fix some inconsistencies in messages api ref https://review.openstack.org/603293 | 03:11 |
openstackgerrit | Merged openstack/cinder master: Fix some inconsistencies in qos-specs api ref https://review.openstack.org/604020 | 03:11 |
openstackgerrit | Merged openstack/cinder master: ZFSSA iSCSI implement get_manageable_volumes() https://review.openstack.org/604230 | 03:11 |
*** gnufied has quit IRC | 03:23 | |
*** gnufied has joined #openstack-cinder | 03:40 | |
*** gnufied has quit IRC | 03:40 | |
*** dave-mccowan has quit IRC | 03:44 | |
*** dave-mccowan has joined #openstack-cinder | 03:46 | |
*** itlinux has joined #openstack-cinder | 03:55 | |
*** pcaruana has joined #openstack-cinder | 04:14 | |
*** pcaruana has quit IRC | 04:38 | |
*** Dinesh_Bhor has quit IRC | 04:47 | |
*** Dinesh_Bhor has joined #openstack-cinder | 04:54 | |
*** e0ne has joined #openstack-cinder | 05:08 | |
*** abishop_ has quit IRC | 05:08 | |
*** abishop has joined #openstack-cinder | 05:12 | |
*** Bhujay has joined #openstack-cinder | 05:26 | |
*** Bhujay has quit IRC | 05:32 | |
*** dave-mccowan has quit IRC | 05:39 | |
*** pcaruana has joined #openstack-cinder | 05:43 | |
*** e0ne has quit IRC | 05:49 | |
*** Bhujay has joined #openstack-cinder | 05:49 | |
*** jistr has quit IRC | 05:55 | |
*** jistr has joined #openstack-cinder | 05:56 | |
*** dpawlik has joined #openstack-cinder | 06:23 | |
whoami-rajat | Hi tommylikehu , need some help regarding my patch related to multiattach, https://review.openstack.org/#/c/604040/4 (don't refer latest patchset), Zuul passes but the CI seems to be failing where the tempest test fail with timeout and can't reach volume-type-2 . https://github.com/openstack/tempest/blob/master/tempest/api/volume/admin/test_volume_retype.py#L81 here. | 06:51 |
tommylikehu | whoami-rajat: will take a look a little later today:) | 06:52 |
whoami-rajat | tommylikehu: ok, thanks. | 06:53 |
*** LiangFang has joined #openstack-cinder | 07:00 | |
*** rcernin has quit IRC | 07:02 | |
*** alexchadin has joined #openstack-cinder | 07:03 | |
*** amoralej|off is now known as amoralej | 07:06 | |
*** Dinesh_Bhor has quit IRC | 07:15 | |
*** psachin has quit IRC | 07:21 | |
*** alexchadin has quit IRC | 07:25 | |
*** psachin has joined #openstack-cinder | 07:26 | |
tobias-urdin | jungleboyj: no worries, didn't read anything in the release notes about resource_filters.json though, was it enforced silently because that file did not exist in queens and there was no issues then? | 07:45 |
tobias-urdin | a release note just saying that that file is used would have helped when searching through the notes for "query" since there was issue with query options for api calls | 07:45 |
*** alexchadin has joined #openstack-cinder | 07:51 | |
*** rcernin has joined #openstack-cinder | 07:56 | |
*** alexchadin has quit IRC | 07:57 | |
*** Dinesh_Bhor has joined #openstack-cinder | 07:58 | |
*** alexchadin has joined #openstack-cinder | 07:59 | |
*** e0ne has joined #openstack-cinder | 08:10 | |
*** moshele has joined #openstack-cinder | 08:12 | |
*** gkadam has joined #openstack-cinder | 08:16 | |
*** Emine has joined #openstack-cinder | 08:17 | |
*** e0ne has quit IRC | 08:20 | |
*** jistr has quit IRC | 08:30 | |
*** jistr has joined #openstack-cinder | 08:31 | |
*** LiangFang has quit IRC | 08:49 | |
*** lixiaoy1 has quit IRC | 08:53 | |
tobias-urdin | jungleboyj: is it a regression that cinderclient volume_snapshots.list() doesn't output the user_id for the snapshot? | 08:55 |
tobias-urdin | that's breaking ceilometers volume.snapshots.size metric | 08:55 |
*** Dinesh_Bhor has quit IRC | 09:03 | |
*** e0ne has joined #openstack-cinder | 09:04 | |
*** alexchadin has quit IRC | 09:06 | |
tobias-urdin | sorry for noice again, probably ceilometer not using api_version >= 3.41 | 09:08 |
*** moshele has quit IRC | 09:10 | |
*** rcernin has quit IRC | 09:16 | |
*** alexchadin has joined #openstack-cinder | 09:20 | |
openstackgerrit | Neha Alhat proposed openstack/cinder master: Enable split logging for cinder-keystoneclient interaction https://review.openstack.org/579765 | 09:21 |
*** alexchadin has quit IRC | 09:25 | |
*** Dinesh_Bhor has joined #openstack-cinder | 09:28 | |
*** alexchadin has joined #openstack-cinder | 09:42 | |
openstackgerrit | Merged openstack/cinder master: Add comment to online data migration command https://review.openstack.org/601684 | 09:42 |
openstackgerrit | yenai proposed openstack/os-brick master: Remove meanless debug log https://review.openstack.org/605380 | 09:50 |
openstackgerrit | Yikun Jiang proposed openstack/cinder-specs master: Add volume re-image api https://review.openstack.org/605317 | 09:52 |
*** Dinesh_Bhor has quit IRC | 10:10 | |
*** Dinesh_Bhor has joined #openstack-cinder | 10:16 | |
*** Dinesh_Bhor has quit IRC | 10:18 | |
*** e0ne has quit IRC | 10:27 | |
*** Luzi has joined #openstack-cinder | 10:42 | |
*** alexchadin has quit IRC | 10:49 | |
*** ganso has joined #openstack-cinder | 10:50 | |
*** alexchadin has joined #openstack-cinder | 10:58 | |
*** alexchadin has quit IRC | 11:03 | |
*** erlon has joined #openstack-cinder | 11:07 | |
*** luizbag has joined #openstack-cinder | 11:11 | |
*** psachin has quit IRC | 11:13 | |
*** pcaruana has quit IRC | 11:15 | |
*** psachin has joined #openstack-cinder | 11:27 | |
*** pooja-jadhav is now known as pooja_jadhav | 11:33 | |
*** lixiaoy1 has joined #openstack-cinder | 11:34 | |
*** brinzhang has joined #openstack-cinder | 11:42 | |
*** e0ne has joined #openstack-cinder | 11:53 | |
*** savvas has joined #openstack-cinder | 11:55 | |
*** alexchadin has joined #openstack-cinder | 11:57 | |
*** dpawlik has quit IRC | 11:58 | |
*** alexchadin has quit IRC | 12:17 | |
*** alexchadin has joined #openstack-cinder | 12:25 | |
*** lixiaoy1 has quit IRC | 12:32 | |
*** amoralej is now known as amoralej|lunch | 12:33 | |
*** psachin has quit IRC | 12:41 | |
*** zhaochao has quit IRC | 12:49 | |
*** alexchadin has quit IRC | 13:02 | |
*** mriedem has joined #openstack-cinder | 13:03 | |
*** alexchadin has joined #openstack-cinder | 13:17 | |
*** dustins has joined #openstack-cinder | 13:24 | |
*** _hemna has joined #openstack-cinder | 13:35 | |
*** abishop has quit IRC | 13:40 | |
*** amoralej|lunch is now known as amoralej | 13:41 | |
*** alexchadin has quit IRC | 13:44 | |
*** gnufied has joined #openstack-cinder | 13:45 | |
*** alexchadin has joined #openstack-cinder | 13:46 | |
*** brinzhang has quit IRC | 13:49 | |
*** alexchadin has quit IRC | 13:50 | |
_hemna | sup peeps | 13:54 |
*** abishop has joined #openstack-cinder | 14:09 | |
jungleboyj | _hemna: Morning sir. | 14:16 |
jungleboyj | tobias-urdin: O dpn | 14:19 |
jungleboyj | Wow, guess I am not ready to type this morning. | 14:19 |
openstackgerrit | Rajat Dhasmana proposed openstack/cinder master: Fix multiattach set to false after retype https://review.openstack.org/604040 | 14:24 |
_hemna | not awake yet? | 14:25 |
_hemna | heh | 14:25 |
jungleboyj | _hemna: No. | 14:25 |
_hemna | I'm not sure I want to adult today | 14:25 |
jungleboyj | _hemna: OMG, I hear you. | 14:25 |
*** Luzi has quit IRC | 14:27 | |
jungleboyj | tobias-urdin: So, it looks like the resource_filters were added almost a year ago. The only change to that file that went in for this release was supporting availability-zone type. With that said, when it merged it didn't really have a good explanation as to the possible impact on end users. So that definitely needs improvement. | 14:29 |
jungleboyj | tobias-urdin: As for the volume_snapshots.list() it sounds like you nailed that one. We have had a few people confused by things due to not specifying the microversion. That was a problem that we discussed at the PTG. We have updated the help text so that it will show all commands regardless of the requested microversion. Hoping that will reduce some confusion. | 14:32 |
openstackgerrit | Merged openstack/cinder master: Fix bug of renaming volume with same name https://review.openstack.org/602275 | 14:39 |
*** Bhujay has quit IRC | 14:44 | |
*** LiangFang has joined #openstack-cinder | 14:49 | |
*** pcaruana has joined #openstack-cinder | 15:01 | |
tobias-urdin | jungleboyj: ack, thanks | 15:01 |
jungleboyj | tobias-urdin: Welcome. Were you upgrading from Queens to Rocky? | 15:12 |
tobias-urdin | yeah https://bugs.launchpad.net/ceilometer/+bug/1794475 | 15:13 |
openstack | Launchpad bug 1794475 in Ceilometer "volume.snapshot.size broken due to user_id attributeerror" [Undecided,New] | 15:13 |
tobias-urdin | manually fixing that for now, can't we do some cooler version discovery here https://github.com/openstack/ceilometer/blob/stable/rocky/ceilometer/volume/discovery.py#L32 | 15:13 |
*** finix has joined #openstack-cinder | 15:17 | |
tobias-urdin | i still dont understand why it would stop working, iirc it worked in queens without any issue | 15:17 |
tobias-urdin | that's why i wondered if it might be a issue somewhere or a regression introduced | 15:17 |
*** dpawlik has joined #openstack-cinder | 15:32 | |
*** dpawlik has quit IRC | 15:36 | |
*** savvas has quit IRC | 15:37 | |
*** finixlei has joined #openstack-cinder | 15:38 | |
*** finix has quit IRC | 15:41 | |
*** dklyle has joined #openstack-cinder | 15:45 | |
*** Bhujay has joined #openstack-cinder | 15:46 | |
*** xyang has joined #openstack-cinder | 15:47 | |
whoami-rajat | jungleboyj: https://review.openstack.org/#/c/604067/2 i think this patch needs to add a unit test for the case where it will fail with 3 default filters but will pass with CapabilitiesFilter | 15:48 |
jungleboyj | whoami-rajat: Wow, surprised I missed that. | 15:49 |
jungleboyj | whoami-rajat: You are right. | 15:49 |
whoami-rajat | jungleboyj: maybe its a small fix that's why. | 15:52 |
jungleboyj | whoami-rajat: But you are right. | 15:54 |
whoami-rajat | jungleboyj: 🙂 | 15:56 |
*** dave-mccowan has joined #openstack-cinder | 16:01 | |
*** LiangFang_ has joined #openstack-cinder | 16:07 | |
*** LiangFang has quit IRC | 16:08 | |
*** Emine has quit IRC | 16:16 | |
*** tpsilva has joined #openstack-cinder | 16:19 | |
*** gkadam has quit IRC | 16:25 | |
*** Bhujay has quit IRC | 16:29 | |
*** LiangFang_ has quit IRC | 16:30 | |
_hemna | geguileo, jungleboyj so, I'm sorry I wasn't able to attend the PTG this time around to talk about the cinderlib and extracting drivers from tree | 16:39 |
_hemna | but I was just wondering how that all went | 16:39 |
geguileo | _hemna: we didn't discuss that topic, it took us the whole hour to decide we were going to merge it into the cinder repo | 16:42 |
_hemna | ah ok | 16:42 |
_hemna | at some point I'd like to bring up a few issues that I have with the current driver mechanism in cinder | 16:42 |
*** e0ne has quit IRC | 16:43 | |
*** dave-mccowan has quit IRC | 16:49 | |
tpsilva | jungleboyj: ping | 16:57 |
jungleboyj | tpsilva: Hold on. Wrapping up meeting. | 16:58 |
tpsilva | alright | 16:58 |
openstackgerrit | Merged openstack/cinder master: Remove unecessary pass https://review.openstack.org/605173 | 16:58 |
tpsilva | jungleboyj: sorry, thought it was finished :) | 16:58 |
tpsilva | sorry, messed up the schedule for the meetings | 17:00 |
jungleboyj | tpsilva: What's up? | 17:03 |
tpsilva | jungleboyj: hey | 17:03 |
tpsilva | I got a -2 on this: https://review.openstack.org/#/c/597193/ | 17:04 |
tpsilva | but its parent was merged | 17:04 |
tpsilva | https://review.openstack.org/#/c/597192/ | 17:04 |
tpsilva | I don't remember who exactly talked about this in the past, but it was agreed to be merged | 17:04 |
jungleboyj | tpsilva: Looking. | 17:06 |
jungleboyj | tpsilva: And I merged it. *laughing* | 17:08 |
jungleboyj | Ok. Looking at it again I think I can re-address. | 17:08 |
tpsilva | alright | 17:08 |
tpsilva | thank you! | 17:08 |
jungleboyj | tpsilva: Sorry about that. | 17:09 |
tpsilva | jungleboyj: no problem | 17:09 |
jungleboyj | _hemna: Topic for the mid-cycle? | 17:14 |
_hemna | sure | 17:15 |
*** dpawlik has joined #openstack-cinder | 17:16 | |
*** dpawlik has quit IRC | 17:16 | |
*** dpawlik has joined #openstack-cinder | 17:17 | |
openstackgerrit | Chuck Short proposed openstack/cinder master: inspur: Fix raise https://review.openstack.org/605495 | 17:36 |
*** dpawlik has quit IRC | 17:47 | |
imacdonn | _hemna: still around ? | 17:49 |
_hemna | yah | 17:50 |
imacdonn | pls review https://review.openstack.org/593955 if you have a moment ... ta | 17:50 |
_hemna | ok | 17:51 |
imacdonn | tnx | 17:51 |
_hemna | probably should have unit tests | 17:53 |
imacdonn | yeah, that seems fair .. thanks | 17:56 |
_hemna | do you remember where the dummer driver is? | 18:00 |
_hemna | found it | 18:00 |
_hemna | fake_driver | 18:00 |
openstackgerrit | Merged openstack/cinder stable/queens: VMAX driver - Block revert to snapshot for replicated volumes https://review.openstack.org/597049 | 18:05 |
*** amoralej is now known as amoralej|off | 18:20 | |
*** dpawlik has joined #openstack-cinder | 18:21 | |
*** pcaruana has quit IRC | 18:24 | |
*** dpawlik has quit IRC | 18:25 | |
jungleboyj | I have submitted our forum topics and added links from the Wiki to the submissions. | 18:34 |
jungleboyj | mriedem: Ping. | 18:49 |
mriedem | PONG | 18:50 |
*** abishop_ has joined #openstack-cinder | 18:58 | |
*** jdillaman has quit IRC | 18:58 | |
*** jdillaman has joined #openstack-cinder | 18:59 | |
*** abishop has quit IRC | 19:00 | |
jungleboyj | mriedem: Hey. | 19:01 |
jungleboyj | mriedem: Have a couple questions about scheduling VMs in a heterogeneous environment. | 19:01 |
*** dave-mccowan has joined #openstack-cinder | 19:05 | |
*** rtjure has quit IRC | 19:05 | |
*** rtjure has joined #openstack-cinder | 19:08 | |
*** asbishop has joined #openstack-cinder | 19:12 | |
*** asbishop is now known as abishop | 19:13 | |
*** abishop_ has quit IRC | 19:15 | |
*** e0ne has joined #openstack-cinder | 19:15 | |
mriedem | jungleboyj: ok | 19:24 |
mriedem | i might have a half assed answer | 19:24 |
jungleboyj | mriedem: I will take half assed. | 19:24 |
jungleboyj | mriedem: So, say I have two different kinds of compute nodes. | 19:25 |
jungleboyj | Some are HCI with Ceph and some only have Ephemeral storage. | 19:25 |
*** Emine has joined #openstack-cinder | 19:26 | |
jungleboyj | Is there some way with flavors or something to say I want this default VM config to go to they systems that are HCI (using Ceph for the volume backing) and the other config I want to go to the compute nodes just using ephemeral storage? | 19:26 |
mriedem | so like 3 compute servers on the same ceph pool in the same rack? | 19:26 |
mriedem | host aggregates is how you can tie flavors to certain sets of compute hosts | 19:27 |
jungleboyj | Right and then three compute nodes without Ceph that we want to just use ephemeral local storage. | 19:27 |
mriedem | so when you say ephemeral you mean local to the compute host right? | 19:28 |
jungleboyj | mriedem: Right. | 19:28 |
mriedem | b/c compute hosts connected to a ceph cluster can still have "ephemeral" root disks | 19:28 |
mriedem | i.e. non-volume-backed vms | 19:28 |
mriedem | yeah so if you're using libvirt, there is an backend config to say you want hosts using qcow2 or raw local disk images and the other compute hosts you configure to use the ceph cluster | 19:29 |
mriedem | and then put those different compute hosts in aggregates, | 19:29 |
jungleboyj | mriedem: Right, but if we have a flavor that uses a host aggregate and then say that we want to use Ceph as the volume-backing for the VM, it will go to the right place. | 19:29 |
mriedem | and then associate your 2 flavors to each host aggregate | 19:29 |
jungleboyj | mriedem: Sweet. That was what I was hoping you would say. | 19:30 |
*** e0ne has quit IRC | 19:30 | |
mriedem | https://docs.openstack.org/nova/latest/configuration/config.html#libvirt.images_type | 19:30 |
mriedem | volume-backed is a bit different, | 19:30 |
mriedem | since it doesn't have anything to do with that images_type config | 19:30 |
mriedem | since nova does'nt control the storage for the root disk in the case of a volume-backed vm | 19:31 |
mriedem | pinning volume-backed vms to specific compute hosts gets a bit hairier with stuff like AZs | 19:32 |
jungleboyj | Ok, but if we wanted to do that we just need to specify the right aggregate and then say we want volume backed with the right flavor. | 19:32 |
mriedem | i.e. affinity wise you want the ceph volume-backed vm to be closer to the compute hosts sharing the same ceph cluster i'd imagine | 19:32 |
mriedem | well, | 19:32 |
jungleboyj | Well, it is HCI, so they are going to be close. | 19:33 |
mriedem | yeah if the user is using the ceph flavor | 19:33 |
mriedem | so boot from volume with rbd volume type and then use the flavor pinned to the ceph compute hosts | 19:33 |
jungleboyj | mriedem: Right. | 19:33 |
jungleboyj | This is a PoC so it doesn't have to be super fancy right now. | 19:33 |
mriedem | the user could use the non-ceph flavor and it would still work as long as the other computes can connect to the rbd volume backend | 19:34 |
mriedem | but that's where i said things get weird with stuff like AZs | 19:34 |
jungleboyj | Right. | 19:34 |
jungleboyj | Just wanted to make sure I wasn't talking out my ass when I said that we should be able to have two different flavors. | 19:35 |
jungleboyj | Tastes like Ceph and doesn't taste like Ceph. | 19:35 |
mriedem | https://docs.openstack.org/nova/latest/admin/configuration/schedulers.html#host-aggregates-and-availability-zones | 19:35 |
mriedem | https://docs.openstack.org/nova/latest/user/aggregates.html | 19:36 |
jungleboyj | mriedem: Thanks. I will take a look. | 19:36 |
mriedem | when you start needing to restrict a vm and volume to the same set of hosts, that's when https://docs.openstack.org/nova/latest/configuration/config.html#cinder.cross_az_attach comes in | 19:36 |
mriedem | but that configuration has it's own set of bugs... | 19:37 |
mriedem | sorrison: can tell you all about it | 19:37 |
mriedem | tell miro i said hi | 19:39 |
jungleboyj | mriedem: Will do. :-) | 20:03 |
jungleboyj | He is keeping me busy. | 20:03 |
*** luizbag has quit IRC | 20:11 | |
*** evrardjp has joined #openstack-cinder | 20:11 | |
*** jarbassaidai has joined #openstack-cinder | 20:33 | |
jungleboyj | mriedem: Ah, I just re-read the discussion and understand why we were talking past each other at one point. There RBD backing for ephemeral storage which is a different concept from volume backed images. Gotcha. | 20:48 |
* jungleboyj has a volume centric view. :-) | 20:48 | |
mriedem | yar | 20:48 |
jungleboyj | Yar | 20:48 |
mriedem | this is why we've (nova) talked about replacing those various ephemeral image backends in nova with a cinder backend that just does boot from volume for all of those other types of things, like scaleio, rbd, ploop, etc | 20:49 |
*** dustins has quit IRC | 20:49 | |
jungleboyj | mriedem: What happened to that discussion. People were excited about it for a bit. | 20:49 |
mriedem | but some people want local disk for HPC | 20:49 |
mriedem | *people that were not going to actually work on making it happen were excited about it for a bit | 20:49 |
mriedem | https://review.openstack.org/#/c/511965/ was a semi compromise after the boston summit | 20:50 |
mriedem | where i vomited some ideas into a spec | 20:50 |
mriedem | we talked about ^ again at the last ptg, and i told people to read that and weigh in on if it's something nova should do | 20:50 |
mriedem | tl;dr put a flag on flavors saying the vm will be volume-backed | 20:51 |
mriedem | and nova orchestrates the stuff for you | 20:51 |
mriedem | so you don't need to pass the ugly block_device_mapping_v2 dict-o-doom | 20:51 |
jungleboyj | He he. | 20:51 |
*** abishop has quit IRC | 21:03 | |
*** erlon has quit IRC | 21:04 | |
jungleboyj | mriedem: Maybe a topic to bring up again at some point. | 21:05 |
mriedem | lyarwood has proposed a forum session that it could tie into so i added a comment about it there | 21:13 |
openstackgerrit | Jon Bernard proposed openstack/cinder master: RBD: add support for multiattach https://review.openstack.org/595827 | 21:27 |
imacdonn | jungleboyj: working on unit test(s) for https://review.openstack.org/#/c/593955/ ... it occurs to me; should there even be an option to disable the v3 API? It's the only non-deprecated one we have........ | 22:00 |
*** dklyle has quit IRC | 22:02 | |
jungleboyj | Well right now first option doesn't make sense. In the future if we go to a V4 we may need to disable but I think the hope is we never have to. | 22:04 |
jungleboyj | *the option | 22:05 |
imacdonn | hmm, right, yeah | 22:05 |
imacdonn | still, seems that the option to not enable v3 should only come into existence when there's a v4 | 22:08 |
imacdonn | I'll just use ddt for now and apply to 2.0 and 3.0 | 22:08 |
imacdonn | side-note: I think there's a minor bug here - this test doesn't do anything version specific, so there's no need (that I can see) to use ddt: https://github.com/openstack/cinder/blob/master/cinder/tests/unit/api/test_versions.py#L62 | 22:11 |
*** ganso has quit IRC | 22:17 | |
imacdonn | jungleboyj: so I want to submit my update to smcginnis's change, with this new unit test .. how do I handle this? He's got "Signed-off-by" in hos commit message. I'm thinking I remove that, and add a Co-Authored-By instead ? | 22:19 |
*** ThomasWhite has joined #openstack-cinder | 22:20 | |
ThomasWhite | Hi, I'm currently running cinder 10.0.7 with a Ceph RBD backend, and I am trying to find a way to rate-limit the activity of snapshots, backups etc. I have found the volume_copy_bps_limit parameter and set it to 20971520 in the hope of limiting it to 200MB/s, but it seems to have had absolutely no effect. Have I misunderstood the usage of the parameter? | 22:22 |
jungleboyj | Just add the co-authored. | 22:22 |
imacdonn | jungleboyj: and leave the "Signed-off-by"? ok... | 22:23 |
imacdonn | jungleboyj: er, sorry, dumb Q .. should the co-authored-by be me or him? not sure how gerrit handles change of submitter .... | 22:24 |
* imacdonn finds and starts reading https://openstack.nimeyo.com/27153/openstack-dev-gerrit-co-authors-and-ticket-stealing | 22:25 | |
ThomasWhite | For reference, my present cinder.conf: https://pastebin.com/CKs09QEK | 22:28 |
imacdonn | ThomasWhite: haven't looked closely, but I think that option just rate-limits a single copy transaction ... doesn't span transactions, nor limit transactions per time period | 22:34 |
*** rcernin has joined #openstack-cinder | 22:35 | |
ThomasWhite | imacdonn, even on single transactions, still seeing up to 1000 MB/s for both read and write individually. | 22:40 |
imacdonn | ThomasWhite: what release are you running? and have you applied that config on the cinder-volume host(s) ? | 22:42 |
ThomasWhite | 10.0.7/Ocata | 22:42 |
ThomasWhite | No, set it to [DEFAULT], I did try the per-host too and same result of no impact | 22:43 |
*** dpawlik has joined #openstack-cinder | 22:48 | |
imacdonn | ThomasWhite: I meant; did you add it to cinder.conf on the host(s) that run the cinder-volume service, as opposed to, say, the one(s) running cinder-api, etc ? | 22:52 |
*** dpawlik has quit IRC | 22:52 | |
*** rcernin has quit IRC | 22:53 | |
*** rcernin has joined #openstack-cinder | 22:55 | |
jungleboyj | imacdonn: So, Sean originally submitted it. If you push up another patch you will show up as the Committer but Sean will be the Author. | 23:00 |
jungleboyj | I believe that is how that shows up. | 23:00 |
jungleboyj | Anyway, If you want to add yourself you would be co-author since you are helping out. | 23:00 |
imacdonn | jungleboyj: yeah, that's what I figured - will try it .. currently trying to figure out why some VMAX unit tests are puking when I do a full 'tox -e py27' | 23:01 |
jungleboyj | Strange. Good luck. I need to go get food for my offspring. :-) | 23:02 |
imacdonn | jungleboyj: tnx .. later | 23:02 |
*** tpsilva has quit IRC | 23:19 | |
*** dklyle has joined #openstack-cinder | 23:26 | |
*** mriedem is now known as mriedem_away | 23:30 | |
openstackgerrit | iain MacDonnell proposed openstack/cinder master: Exclude disabled API versions from listing https://review.openstack.org/593955 | 23:31 |
*** finixlei has quit IRC | 23:40 | |
*** rcernin_ has joined #openstack-cinder | 23:41 | |
*** rcernin has quit IRC | 23:43 | |
imacdonn | smcginnis: looks like I stole the author on that change .... not intentionally! | 23:43 |
*** openstackgerrit has quit IRC | 23:49 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!