| *** catintheroof has quit IRC | 00:03 | |
| *** sticker has joined #openstack-manila | 00:05 | |
| *** catintheroof has joined #openstack-manila | 00:11 | |
| *** cknight1 has quit IRC | 00:17 | |
| *** ganso has quit IRC | 00:21 | |
| *** catintheroof has quit IRC | 00:25 | |
| *** hongbin has quit IRC | 00:54 | |
| *** markstur_ has joined #openstack-manila | 00:55 | |
| *** markstur has quit IRC | 00:57 | |
| *** gouthamr has quit IRC | 01:42 | |
| *** jiaohaolin1 has left #openstack-manila | 01:45 | |
| *** tuanluong has joined #openstack-manila | 02:13 | |
| *** winston-d_ has joined #openstack-manila | 02:30 | |
| *** chlong has quit IRC | 02:42 | |
| *** gouthamr has joined #openstack-manila | 02:53 | |
| bswartz | zhongjun: sorry for late response... | 03:30 |
|---|---|---|
| zhongjun | bswartz: It doesn't matter | 03:31 |
| bswartz | zhongjun: the issue of needing multiple subnets for dual v4/v6 shares when DHSS=true is something that came up it Atlanta, but only during side conversations on friday | 03:31 |
| bswartz | tbarron gouthamr and I discussed it on a white board | 03:31 |
| bswartz | some kind of change to share networks is definitely needed | 03:31 |
| bswartz | initially there was some confusion about what is a "network" and what is a "subnet" | 03:33 |
| bswartz | eventually we agreed that a network is an L2 broadcast domain and there can be multiple subnets per network (for different address types, primarily) | 03:34 |
| zhongjun | bswartz: yes, we need some kind of change to share networks. as above, there are two ways. which one is suitable for manila or there is another way. | 03:37 |
| bswartz | zhongjun: it's something to discuss with the community | 03:37 |
| bswartz | please put it on the agenda for thursday | 03:38 |
| zhongjun | bswartz: ok, How about I separate ipv6 patch to two patch: DHSS=true and DHSS=false | 03:40 |
| bswartz | I kind of don't like that | 03:41 |
| bswartz | I'd prefer 1 patch with some acknowledged limitations | 03:41 |
| bswartz | then we can solve the limitations later | 03:41 |
| bswartz | separating the patches just creates more work | 03:42 |
| zhongjun | bswartz: We need to test lvm ipv6 (DHSS=false) first | 03:42 |
| zhongjun | bswartz: It could make our patch's goalsmore clearly | 03:43 |
| bswartz | yeah but we could also make the primary goal to make dhss=false work, and dhss=true to partially work | 03:45 |
| bswartz | then later make dhss=true fully work | 03:45 |
| zhongjun | bswartz: or add this patch https://review.openstack.org/#/c/444479/ in ipv6 patch(DHSS=false), Then we can run lvm-ipv6-job https://review.openstack.org/#/c/443737/ | 03:45 |
| zhongjun | base on one patch | 03:46 |
| zhongjun | bswartz: It could a little weired, if I just let patch support both ipv6 and ipv4 for dhss=false, and only support ipv6 or ipv4 for dhss=true | 03:54 |
| zhongjun | bswartz: in one patch | 03:55 |
| *** gcb has quit IRC | 03:58 | |
| *** arnewiebalck_ has joined #openstack-manila | 04:15 | |
| *** jprovazn has joined #openstack-manila | 04:19 | |
| *** gouthamr has quit IRC | 05:23 | |
| *** gcb has joined #openstack-manila | 05:30 | |
| *** gcb has quit IRC | 05:46 | |
| *** gcb has joined #openstack-manila | 05:47 | |
| *** markstur_ has quit IRC | 05:49 | |
| *** markstur has joined #openstack-manila | 05:50 | |
| *** markstur has quit IRC | 05:55 | |
| *** sticker_ has joined #openstack-manila | 06:03 | |
| *** sticker has quit IRC | 06:06 | |
| *** pcaruana has joined #openstack-manila | 06:06 | |
| *** sticker_ is now known as sticker | 06:16 | |
| *** markstur has joined #openstack-manila | 06:31 | |
| *** td has quit IRC | 06:34 | |
| *** markstur has quit IRC | 06:36 | |
| *** td has joined #openstack-manila | 06:36 | |
| *** markstur has joined #openstack-manila | 07:13 | |
| *** markstur has quit IRC | 07:18 | |
| *** a-pugachev has joined #openstack-manila | 07:29 | |
| *** markstur has joined #openstack-manila | 07:55 | |
| *** markstur has quit IRC | 08:00 | |
| *** makowals has quit IRC | 08:38 | |
| *** makowals has joined #openstack-manila | 08:41 | |
| *** dsariel_ has joined #openstack-manila | 08:48 | |
| *** dsariel_ has quit IRC | 08:49 | |
| *** dsariel_ has joined #openstack-manila | 08:49 | |
| *** dgonzalez_ has joined #openstack-manila | 09:21 | |
| *** david_ has joined #openstack-manila | 09:21 | |
| *** tpatzig_ has joined #openstack-manila | 09:21 | |
| *** david_ has quit IRC | 09:23 | |
| *** dgonzalez_ has quit IRC | 09:23 | |
| *** tpatzig_ has quit IRC | 09:23 | |
| *** makowals has quit IRC | 09:43 | |
| *** makowals has joined #openstack-manila | 09:49 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: [WIP] Add share group quotas https://review.openstack.org/473464 | 09:54 |
| *** rraja has joined #openstack-manila | 10:00 | |
| *** arnewiebalck__ has joined #openstack-manila | 10:01 | |
| *** lpetrut has joined #openstack-manila | 10:01 | |
| *** arnewiebalck_ has quit IRC | 10:03 | |
| *** arnewiebalck__ is now known as arnewiebalck_ | 10:03 | |
| *** rraja has quit IRC | 10:10 | |
| *** markstur has joined #openstack-manila | 10:21 | |
| *** markstur has quit IRC | 10:25 | |
| *** markstur has joined #openstack-manila | 10:42 | |
| *** arnewiebalck_ has quit IRC | 10:43 | |
| *** ganso has joined #openstack-manila | 10:46 | |
| *** markstur has quit IRC | 10:46 | |
| *** dsariel_ has quit IRC | 10:55 | |
| *** markstur has joined #openstack-manila | 11:03 | |
| *** markstur has quit IRC | 11:07 | |
| *** sticker_ has joined #openstack-manila | 11:07 | |
| *** sticker__ has joined #openstack-manila | 11:08 | |
| *** sticker has quit IRC | 11:11 | |
| *** sticker_ has quit IRC | 11:12 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: [Docs] Correct glusterfs references https://review.openstack.org/473787 | 11:21 |
| *** markstur has joined #openstack-manila | 11:23 | |
| *** sticker__ is now known as sticker | 11:26 | |
| *** markstur has quit IRC | 11:28 | |
| *** markstur has joined #openstack-manila | 11:44 | |
| *** markstur has quit IRC | 11:48 | |
| *** chlong has joined #openstack-manila | 11:57 | |
| *** arnewiebalck_ has joined #openstack-manila | 12:01 | |
| openstackgerrit | Merged openstack/puppet-manila master: Add environment variable https://review.openstack.org/473271 | 12:02 |
| *** tuanluong has quit IRC | 12:04 | |
| *** dsariel_ has joined #openstack-manila | 12:11 | |
| openstackgerrit | zhongjun proposed openstack/manila master: Add like filter https://review.openstack.org/462468 | 12:15 |
| *** openstackgerrit has quit IRC | 12:18 | |
| *** openstackgerrit has joined #openstack-manila | 12:23 | |
| openstackgerrit | Tom Barron proposed openstack/manila master: [Docs] Correct glusterfs references https://review.openstack.org/473787 | 12:23 |
| *** markstur has joined #openstack-manila | 12:25 | |
| *** markstur has quit IRC | 12:30 | |
| *** catintheroof has joined #openstack-manila | 12:33 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: [WIP] Add share group quotas https://review.openstack.org/473464 | 12:44 |
| *** rraja has joined #openstack-manila | 12:55 | |
| *** arnewiebalck_ has quit IRC | 12:56 | |
| bswartz | zhongjun: yes that's what I had in mind | 13:00 |
| *** markstur has joined #openstack-manila | 13:07 | |
| *** eharney has joined #openstack-manila | 13:10 | |
| *** xyang1 has joined #openstack-manila | 13:10 | |
| *** xyang_ has joined #openstack-manila | 13:11 | |
| *** markstur has quit IRC | 13:12 | |
| xyang_ | bswartz: ping | 13:16 |
| *** dustins has joined #openstack-manila | 13:18 | |
| xyang_ | bswartz: Several of our CI's are broken because of this bug: https://bugs.launchpad.net/manila/+bug/1697581 | 13:26 |
| openstack | Launchpad bug 1697581 in Manila "create snapshot failed due to absent of snapshot['share_id']" [Critical,Confirmed] - Assigned to EMC VNX OpenStack (emc-vnx-openstack) | 13:26 |
| xyang_ | bswartz: I wonder if you ran into the same issue with NetApp CI? | 13:26 |
| *** dustins has quit IRC | 13:29 | |
| *** dustins has joined #openstack-manila | 13:29 | |
| bswartz | xyang_: it would depend on the driver | 13:36 |
| *** gouthamr has joined #openstack-manila | 13:36 | |
| xyang_ | bswartz: NetApp driver seems to be referencing it as well. that's why I asked | 13:37 |
| bswartz | xyang_: it looks like our CI is broken but it's not clear that that bug is related | 13:37 |
| bswartz | xyang_: do you know which change introduced the breakage? | 13:38 |
| bswartz | in theory whatever change removed the field from the model should have fixed the drivers in the same commit | 13:38 |
| xyang_ | bswartz: https://review.openstack.org/#/c/433854/12/manila/db/sqlalchemy/models.py | 13:39 |
| bswartz | oh! | 13:39 |
| bswartz | the one we just merged | 13:39 |
| bswartz | so maybe this is my fault for not looking at CI results | 13:39 |
| xyang_ | bswartz: :( | 13:39 |
| xyang_ | bswartz: I didn't look either:( | 13:40 |
| xyang_ | bswartz: https://github.com/openstack/manila/blob/master/manila/share/drivers/netapp/dataontap/cluster_mode/lib_base.py#L791 | 13:40 |
| xyang_ | bswartz: two of our CI's failed on that patch and I didn't notice it:( | 13:41 |
| bswartz | okay so the question is should we change the drivers or change the model | 13:42 |
| bswartz | I'll have to take a closer look | 13:42 |
| xyang_ | ok | 13:44 |
| *** markstur has joined #openstack-manila | 13:49 | |
| *** cknight has joined #openstack-manila | 13:51 | |
| *** eharney has quit IRC | 13:52 | |
| *** markstur has quit IRC | 13:54 | |
| *** eharney has joined #openstack-manila | 13:56 | |
| bswartz | vponomaryov: was the removal of the share_id property here necessary? https://review.openstack.org/#/c/433854/12/manila/db/sqlalchemy/models.py | 14:01 |
| bswartz | IMO it should have stayed | 14:02 |
| vponomaryov | bswartz: it is not removed | 14:10 |
| bswartz | vponomaryov: what happened to it then? | 14:11 |
| xyang_ | vponomaryov: share_id is still a field but it is not populated any more | 14:11 |
| bswartz | lines 729-734 look like it was removed | 14:11 |
| vponomaryov | bswartz: 727 | 14:12 |
| vponomaryov | for right | 14:12 |
| vponomaryov | bswartz: it is not proxy to other attr anymore it is standalone attr now | 14:13 |
| bswartz | according to xyang the field is blank | 14:13 |
| xyang_ | vponomaryov: but that is not populated any more | 14:13 |
| bswartz | so maybe there's a bug? | 14:13 |
| vponomaryov | xyang_: it is populated on db migration step | 14:13 |
| bswartz | because drivers should see a property called "share_id" which is exactly the share_instance_id | 14:13 |
| xyang_ | bswartz: we are fixing our drivers to use share_instance_id instead | 14:13 |
| bswartz | xyang_: no that's not the right fix | 14:14 |
| vponomaryov | it is | 14:14 |
| xyang_ | bswartz: why not? | 14:14 |
| xyang_ | bswartz: the snapshot passed in create_snapshot in the driver is an instance | 14:14 |
| bswartz | xyang_: when we introduce share instances years ago, the goal was to prevent drivers from knowing that fact | 14:14 |
| bswartz | from the driver's perspective, it was all supposed to look like plain shares and snapshots | 14:15 |
| vponomaryov | bswartz: drivers handle more than one instance | 14:15 |
| bswartz | drivers weren't supposed to know about instances | 14:15 |
| vponomaryov | bswartz: disagree | 14:15 |
| bswartz | later, when replication and migration happened it became necessary for drivers to know about instances | 14:15 |
| xyang_ | bswartz: not sure if that's possible for driver not to know about instance | 14:15 |
| bswartz | but outside of those 2 features knowledge of instances should not be reuqired | 14:16 |
| xyang_ | vponomaryov: you are saying share_id should be populated after running db migration? Our CI's are fresh installs and share_id is not populated | 14:18 |
| vponomaryov | https://review.openstack.org/#/c/433854/12/manila/db/migrations/alembic/versions/b10fb432c042_squash_share_group_snapshot_members_and_share_snapshot_instance_models.py | 14:18 |
| xyang_ | bswartz: It's strange that only 2 features can use instance but other features cannot. | 14:19 |
| xyang_ | bswartz: once driver starts to have knowledge about it, then it knows about it | 14:20 |
| vponomaryov | xyang_: if it is None then it means it was not set in place that is used as source | 14:20 |
| xyang_ | vponomaryov: that should be set in the manager though if it is ever populated | 14:21 |
| xyang_ | vponomaryov: I mean we don't see it in the driver after a fresh install in create_snapshot | 14:21 |
| vponomaryov | then, it means it was never set in the source | 14:23 |
| vponomaryov | you were depending on other attr all the time | 14:23 |
| vponomaryov | xyang_: so, I guess, you are right, this DB migration includes only port of share group snapshot members | 14:24 |
| vponomaryov | xyang_: share manager didn't write it before and does not now | 14:25 |
| bswartz | xyang_: I'm not sure what to do about the driver interface issue | 14:25 |
| bswartz | we have some older drivers which might not update to a newer interface | 14:25 |
| bswartz | so we need to provide backward compatibility | 14:25 |
| xyang_ | bswartz: if we need to always provide backward compatibility, then the share_id property needs to be added back, and we don't need to change our driver given that share_id will always be populated | 14:27 |
| bswartz | xyang_: that was the point I was trying to make | 14:29 |
| bswartz | I think we should continue to provide backward compatibility in the snapshot model | 14:29 |
| bswartz | I was trying to find out if there was a good reason to remove it | 14:29 |
| vponomaryov | bswartz: it was providing wrong value | 14:30 |
| vponomaryov | bswartz: getting "share_id" it was not SHARE id | 14:31 |
| vponomaryov | bswartz: so, if we need to have it populated then we need to update it at the same time when we update other attrs | 14:31 |
| vponomaryov | with creation of share snapshot instances | 14:32 |
| vponomaryov | not keeping previous workarounds | 14:32 |
| bswartz | okay | 14:33 |
| bswartz | as long as drivers are able to obtain the value of the share instance id from the share_id property | 14:33 |
| vponomaryov | for example, if we add "share_id" here https://github.com/openstack/manila/blame/master/manila/share/api.py#L1067 it should be populated all the time | 14:35 |
| vponomaryov | https://github.com/openstack/manila/blob/059fae0e/manila/share/api.py#L1067 | 14:36 |
| xyang_ | vponomaryov: that one is for share with replicas only | 14:37 |
| *** hongbin has joined #openstack-manila | 14:39 | |
| openstackgerrit | Dustin Schoenbrun proposed openstack/manila master: Change example value in docs for CephFS snapshots https://review.openstack.org/473526 | 14:40 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: Set share_id to share snapshot instance all the time https://review.openstack.org/473864 | 14:45 |
| vponomaryov | xyang_, bswartz: expect it to work ^ | 14:46 |
| vponomaryov | just without unit tests yet | 14:46 |
| xyang_ | vponomaryov: thanks | 14:47 |
| *** a-pugachev has quit IRC | 14:57 | |
| *** a-pugachev has joined #openstack-manila | 14:59 | |
| *** markstur has joined #openstack-manila | 15:07 | |
| *** markstur has quit IRC | 15:07 | |
| *** markstur has joined #openstack-manila | 15:08 | |
| *** fishbone has joined #openstack-manila | 15:18 | |
| fishbone | Hello all, I have searched documentation and I can't see the method to ensure configuration of the attached block storage device for the shares | 15:20 |
| bswartz | fishbone: what driver? | 15:20 |
| fishbone | When I create the shares the share instance does not configure the disk, thus leaving it unmounted | 15:20 |
| fishbone | bswartz: generic | 15:20 |
| bswartz | you mean the service instance isn't attaching the cinder volumes? | 15:21 |
| fishbone | bswartz: It attaches the drive but it doesn't format the drive, or configure the allocation table so it can be mounted as a proper file system | 15:21 |
| bswartz | fishbone: that's strange -- it should do that automatically | 15:23 |
| bswartz | you may have found a recent bug, or it might be an environmental issue particular to you | 15:23 |
| fishbone | bswartz: hmmm, I will go over this again and verify I completed each step as required | 15:23 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: Set share_id to share snapshot instance all the time https://review.openstack.org/473864 | 15:41 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: Set share_id to share snapshot instance all the time https://review.openstack.org/473864 | 15:44 |
| vponomaryov | fishbone: in case it is not complicated correctly you should have some logs with errors | 15:45 |
| vponomaryov | s/complicated/completed/ | 15:45 |
| fishbone | vponomaryov: after some testing I was confused, the nfs shares create properly just need to restart the service when adjusting permissions, however; it does not seem as if the cifs shares are created properly | 15:46 |
| vponomaryov | fishbone: release version? which image? which error? | 15:47 |
| vponomaryov | fishbone: driver mode? manila configuration? | 15:48 |
| vponomaryov | fishbone: how deployed ... etc | 15:48 |
| fishbone | vponomaryov: thanks, I'm using version 10.0.0, ocata, I manually created the manila service image using the documentation, essentially installed samba in this case, I'm using the generic driver | 15:53 |
| fishbone | vponomaryov: I launched a test share server using the dashboard, and the nfs share works fine | 15:54 |
| *** lpetrut has quit IRC | 15:57 | |
| vponomaryov | fishbone: even if you created your own image, there should be some errors in logs | 15:57 |
| openstackgerrit | Merged openstack/manila master: Change example value in docs for CephFS snapshots https://review.openstack.org/473526 | 15:59 |
| *** a-pugachev has quit IRC | 16:00 | |
| fishbone | vponomaryov: There are no errors in the logs, it seems to assume that the server is created appropriately, my conf file: http://paste.openstack.org/show/612441/ | 16:04 |
| fishbone | vponomaryov: I feel as though the issue is my interpretation of how the cifs share is configured within the provision, does it use samba to create the share? | 16:08 |
| vponomaryov | fishbone: have you enabeld registry? | 16:08 |
| vponomaryov | fishbone: is samba 3.2+ ? | 16:08 |
| vponomaryov | absence of enabled registry is the only possible reason that can cause problems without errors in logs and comes into my mind | 16:09 |
| fishbone | vponomaryov: samba is version 4.2.14, I don't believe I enabled registry, I am also unfamiliar with that | 16:10 |
| *** arnewiebalck_ has joined #openstack-manila | 16:10 | |
| vponomaryov | fishbone: https://docs.openstack.org/developer/manila/devref/generic_driver.html#requirements-for-service-image | 16:10 |
| vponomaryov | Samba server >=3.2.0, that can be configured by data stored in registry | 16:10 |
| *** dsariel_ has quit IRC | 16:10 | |
| vponomaryov | fishbone: https://github.com/openstack/manila-image-elements/blob/b0d2c6dc/elements/ubuntu-cifs/post-install.d/50-manila-samba-patch-config | 16:12 |
| vponomaryov | fishbone: you need to execute last line from above link ^ | 16:13 |
| vponomaryov | fishbone: and restart samba | 16:13 |
| *** a-pugachev has joined #openstack-manila | 16:13 | |
| fishbone | vponomaryov: ah I see, thanks a lot for this, I will test it out | 16:13 |
| vponomaryov | https://www.samba.org/samba/docs/man/manpages-3/smb.conf.5.html#idp58012816 | 16:14 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: [WIP] Add share group quotas https://review.openstack.org/473464 | 16:24 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: [WIP] Add share group quotas https://review.openstack.org/473464 | 16:25 |
| fishbone | vponomaryov: That was it, thanks a million, vponomaryov! | 16:30 |
| vponomaryov | fishbone: you are welcome | 16:30 |
| *** ianychoi has joined #openstack-manila | 17:15 | |
| *** rraja has quit IRC | 17:20 | |
| *** a-pugachev has quit IRC | 17:24 | |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: [WIP] Add share group quotas https://review.openstack.org/473464 | 17:33 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: [WIP] Add share group quotas https://review.openstack.org/473464 | 17:36 |
| openstackgerrit | Valeriy Ponomaryov proposed openstack/manila master: Add share groups quotas https://review.openstack.org/473464 | 17:41 |
| *** fishbone has quit IRC | 18:22 | |
| *** a-pugachev has joined #openstack-manila | 19:30 | |
| *** jprovazn has quit IRC | 19:38 | |
| *** xyang_ has quit IRC | 19:39 | |
| *** xyang_ has joined #openstack-manila | 19:41 | |
| *** dsariel has joined #openstack-manila | 19:48 | |
| *** a-pugachev has quit IRC | 19:49 | |
| *** eharney has quit IRC | 19:53 | |
| *** arnewiebalck_ has quit IRC | 19:59 | |
| *** ianychoi has quit IRC | 20:46 | |
| *** gouthamr has quit IRC | 21:03 | |
| *** erlon has quit IRC | 21:28 | |
| *** gouthamr has joined #openstack-manila | 21:29 | |
| *** xyang_ has quit IRC | 21:54 | |
| *** dustins has quit IRC | 21:54 | |
| *** xyang1 has quit IRC | 21:56 | |
| *** cknight has quit IRC | 22:17 | |
| *** ganso has quit IRC | 22:22 | |
| *** catintheroof has quit IRC | 22:47 | |
| *** winston-d_ has quit IRC | 23:13 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!