| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2018-05-17 | |||
| 14:44:42 | dansmith | yeah, okay | |
| 14:45:04 | dansmith | well, still, on an empty cell you'd get back the whole cell, but fair enough | |
| 14:45:10 | mriedem | are the compute node resource providers in placement within aggregates? | |
| 14:45:24 | mriedem | in other words, are you getting back 200 allocation candidates or 10K? | |
| 14:45:24 | dansmith | mriedem: I think that's the implication :) | |
| 14:45:37 | mriedem | sure but have they asserted that's all linked up properly | |
| 14:45:45 | dansmith | of they didn't, they would get back zero | |
| 14:45:51 | mriedem | sorry, i mean 1K | |
| 14:45:56 | dansmith | unless they set required=false | |
| 14:46:20 | mriedem | because we'd say give me allocation candidates that are a member of this aggregate, and if the rp isn't in that aggregate we'd get back 0, right | |
| 14:46:20 | belmorei_ | depends in the project... but should get in average less than 200 | |
| 14:46:37 | dansmith | mriedem: yeah | |
| 14:57:47 | openstackgerrit | Merged openstack/osc-placement master: Fix error message asserts in functional test https://review.openstack.org/569048 | |
| 15:01:22 | melwitt | lyarwood: thanks for jumping on https://bugs.launchpad.net/nova/+bug/1771700 | |
| 15:01:23 | openstack | Launchpad bug 1771700 in OpenStack Compute (nova) "nova-lvm tempest job failing with InvalidDiskInfo" [High,In progress] - Assigned to Lee Yarwood (lyarwood) | |
| 15:01:38 | lyarwood | melwitt: well, sorry for breaking nova-lvm ^_^ | |
| 15:02:09 | melwitt | mriedem: lyarwood had an idea of running the nova-lvm job non-voting on nova/virt/libvirt/* changes. what do you think? | |
| 15:02:26 | melwitt | no worries | |
| 15:03:42 | mriedem | why non-voting? | |
| 15:03:46 | mriedem | i think it is stable otherwise | |
| 15:04:02 | mriedem | we just don't gate on it because it's not a highly deployed image backend | |
| 15:04:16 | melwitt | I dunno. voting then? | |
| 15:04:21 | mriedem | we also had a raw imagebackend job at one point | |
| 15:04:34 | mriedem | or maybe that was just a dnm patch i had... | |
| 15:05:24 | mriedem | let's see how it looks on lee's fix first | |
| 15:05:40 | mriedem | 1 hour and 20 minutes... | |
| 15:05:43 | mriedem | but it passes | |
| 15:05:50 | mriedem | maybe we should just run the nova-lvm job on compute api tests | |
| 15:05:53 | mriedem | non-voting to start | |
| 15:05:57 | mriedem | and only on libvirt changes | |
| 15:06:23 | melwitt | I think that would be helpful | |
| 15:07:38 | openstackgerrit | Merged openstack/osc-placement master: Resource class set (v1.7) https://review.openstack.org/514644 | |
| 15:09:27 | openstackgerrit | Matt Riedemann proposed openstack/nova master: DNM: test raw image live migration with vfat config drive https://review.openstack.org/567860 | |
| 15:09:40 | mriedem | lyarwood: let's make sure we don't regress the raw + vfat config drive + live migration case :) ^ | |
| 15:11:14 | lyarwood | mriedem: thanks, I'd like to get off of this regression merry-go-round eventually :) | |
| 15:12:33 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Deduplicate config/policy reference docs from main index https://review.openstack.org/564846 | |
| 15:12:34 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Implement granular policy rules for placement https://review.openstack.org/524425 | |
| 15:12:35 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Add granular policy rules for /resource_classes* https://review.openstack.org/565578 | |
| 15:12:36 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Add granular policy rules for resource providers inventories https://review.openstack.org/568666 | |
| 15:12:37 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Add granular policy rules for usages https://review.openstack.org/568706 | |
| 15:12:38 | mriedem | lyarwood: job security! | |
| 15:13:39 | mriedem | per my comments in https://review.openstack.org/#/c/567860/ i'd also like to diversify the live migration tests we run | |
| 15:13:43 | mriedem | to mix up the image format | |
| 15:13:53 | mriedem | and to use config drive | |
| 15:14:39 | mriedem | we already run the nova-live-migration job and the neutron-tempest-multinode-full job runs live migration, | |
| 15:14:49 | mriedem | so we can make nova-live-migration run with force_config_drive=True, | |
| 15:15:12 | mriedem | and nova-live-migration runs with 2 image formats today, first is default (qcow2) and 2nd is rbd, we could change that first run to be raw | |
| 15:15:23 | mriedem | and then neutron-tempest-multinode-full would test qcow2 (default) with no config drive (default) | |
| 15:15:50 | mriedem | we still wouldn't have vfat config drive format testing, but i suspect we can remove the vfat config_drive_format option/workaround | |
| 15:16:06 | mriedem | i think that was for old versions of libvirt and we now require a minimum that has that fixed | |
| 15:25:04 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Make nova-lvm run in check on libvirt changes and compute API tests https://review.openstack.org/569149 | |
| 15:25:06 | mriedem | melwitt: lyarwood: ^ | |
| 15:25:13 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Make nova-lvm run in check on libvirt changes and compute API tests https://review.openstack.org/569149 | |
| 15:25:39 | lyarwood | mriedem: ack thanks! | |
| 15:34:39 | openstackgerrit | Matt Riedemann proposed openstack/nova master: libvirt: Skip fetching the virtual size of block devices https://review.openstack.org/569062 | |
| 15:34:40 | openstackgerrit | Matt Riedemann proposed openstack/nova master: Make nova-lvm run in check on libvirt changes and compute API tests https://review.openstack.org/569149 | |
| 15:35:44 | openstackgerrit | Matt Riedemann proposed openstack/nova master: DNM: test raw image live migration with vfat config drive https://review.openstack.org/567860 | |
| 15:42:04 | openstackgerrit | Merged openstack/nova master: Fix auth_url example in hypervisor-hyper-v.rst https://review.openstack.org/568983 | |
| 15:47:18 | mriedem | tssurya: do you need to rebase https://review.openstack.org/#/c/567878/ on top of https://review.openstack.org/#/c/568271/ or just wait for https://review.openstack.org/569112/ ? | |
| 15:47:45 | openstackgerrit | Chris Dent proposed openstack/nova master: Optional separate database for placement API https://review.openstack.org/362766 | |
| 15:56:06 | openstackgerrit | Chris Dent proposed openstack/nova master: Isolate placement database config https://review.openstack.org/541435 | |
| 15:56:26 | openstackgerrit | Chris Dent proposed openstack/nova master: WIP: Ensure that os-traits sync is attempted only at start of process https://review.openstack.org/553857 | |
| 15:59:06 | openstackgerrit | Chris Dent proposed openstack/nova master: WIP: Add PLACEMENT_DB_ENABLED=True to the nova-next job https://review.openstack.org/564067 | |
| 15:59:26 | openstackgerrit | Merged openstack/nova master: Deprecate running API services under eventlet https://review.openstack.org/549510 | |
| 16:01:02 | openstackgerrit | Merged openstack/nova master: Simplify BDM boot index checking https://review.openstack.org/564480 | |
| 16:13:35 | melwitt | mriedem: do you see any issue with us increasing the max number of allowed attached volumes per instance from 26 to 1024? I was thinking to approve this specless bp https://blueprints.launchpad.net/nova/+spec/nova-improvement-of-maximum-attach-volumes-more-than-26-vols | |
| 16:14:36 | mriedem | i am not the person to ask about that... | |
| 16:14:45 | mriedem | should ask kvm people | |
| 16:14:50 | mriedem | like kashyap re danpb | |
| 16:15:14 | melwitt | I already talked to danpb about it and he suggested 1024 as the new max | |
| 16:15:30 | mriedem | because of the default fd? | |
| 16:15:37 | melwitt | was just thinking to ask you since you've worked on the multi-attach stuff | |
| 16:15:53 | mriedem | er ulimit used to default to 1024 | |
| 16:16:16 | mriedem | multiattach volumes are >1 instance to the same volume, | |
| 16:16:20 | mriedem | not >1 volumes to the same instance | |
| 16:16:28 | mriedem | default quota on volumes is 10 anyway | |
| 16:16:40 | mriedem | so a normal tenant is going to have at most 10 volumes attached to a server | |
| 16:16:50 | mriedem | if you're using powervm, i think you get up to like 256 or something | |
| 16:16:52 | melwitt | oh, right. okay | |
| 16:16:52 | mriedem | efried: ^ | |
| 16:17:05 | mriedem | because powervm just f'ing loves volumes | |
| 16:17:11 | mriedem | also, | |
| 16:17:16 | mriedem | this reminds me, | |
| 16:17:51 | melwitt | picked 1024 arbitrarily based on the field tests described here https://rwmj.wordpress.com/2017/04/25/how-many-disks-can-you-add-to-a-virtual-linux-machine/ | |
| 16:18:00 | mriedem | Ia34d06429c1f8f0a8259616bcba0c349c4c9aa33 | |
| 16:18:07 | melwitt | danpb suggested 1000 | |
| 16:18:12 | mriedem | https://review.openstack.org/#/c/366955/ | |
| 16:18:30 | mriedem | depending on how many volumes you try to attach in a single request, you could blow out the build requests table | |
| 16:18:45 | efried | I'm asking Gerald whether 1024 is enough for us. | |
| 16:18:48 | mriedem | so would 1024 BDMs in a single bfv request get stored in MediumText? | |
| 16:18:55 | melwitt | hm | |
| 16:19:01 | mriedem | MediumText is pretty damn big so i assume yes | |
| 16:19:37 | mriedem | (11:44:41 AM) burgerk: mriedem: broke at ~ 80 volumes, want to go up to 128 | |
| 16:19:45 | mriedem | so powervm was using 128 in that case | |
| 16:20:02 | efried | okay. | |
| 16:21:04 | melwitt | okay, so increasing the max is cool, it might be that 1024 is too high | |
| 16:21:14 | mriedem | MEDIUMTEXT [CHARACTER SET charset_name] [COLLATE collation_name] A TEXT column with a maximum length of 16,777,215 (224 − 1) characters. The effective maximum length is less if the value contains multibyte characters. Each MEDIUMTEXT value is stored using a 3-byte length prefix that indicates the number of | |
| 16:21:15 | mriedem | bytes in the value. | |
| 16:21:29 | mriedem | idk | |
| 16:21:34 | mriedem | someone should test it | |