| *** AlexeyAbashkin has joined #openstack-manila | 00:19 | |
| *** AlexeyAbashkin has quit IRC | 00:23 | |
| *** threestrands has joined #openstack-manila | 00:44 | |
| *** caowei has joined #openstack-manila | 01:42 | |
| *** kaisers_ has joined #openstack-manila | 02:50 | |
| *** kaisers has quit IRC | 02:54 | |
| *** tinwood has quit IRC | 03:05 | |
| *** tinwood has joined #openstack-manila | 03:06 | |
| *** cfey has quit IRC | 03:21 | |
| *** cfey has joined #openstack-manila | 03:21 | |
| *** vivsoni__ has joined #openstack-manila | 03:55 | |
| *** vivsoni_ has quit IRC | 03:58 | |
| *** AlexeyAbashkin has joined #openstack-manila | 04:18 | |
| *** AlexeyAbashkin has quit IRC | 04:23 | |
| *** kaisers_ has quit IRC | 05:48 | |
| *** kaisers has joined #openstack-manila | 05:49 | |
| *** threestrands has quit IRC | 06:13 | |
| *** junboli has joined #openstack-manila | 06:20 | |
| *** openstackgerrit has joined #openstack-manila | 06:49 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/manila-ui master: Imported Translations from Zanata https://review.openstack.org/546873 | 06:49 |
|---|---|---|
| *** pcaruana has joined #openstack-manila | 06:56 | |
| *** pcaruana has quit IRC | 07:02 | |
| *** pcaruana has joined #openstack-manila | 07:02 | |
| *** rishabh has joined #openstack-manila | 07:09 | |
| *** rishabh is now known as Guest67743 | 07:10 | |
| *** rishabh_ has joined #openstack-manila | 07:11 | |
| *** Guest67743 has quit IRC | 07:13 | |
| *** rishabh__ has joined #openstack-manila | 07:20 | |
| *** rishabh_ has quit IRC | 07:20 | |
| *** rishabh_ has joined #openstack-manila | 07:23 | |
| *** rishabh__ has quit IRC | 07:24 | |
| *** rishabh_ has quit IRC | 07:26 | |
| *** rishabh_ has joined #openstack-manila | 07:27 | |
| *** kaisers has quit IRC | 07:30 | |
| *** kaisers has joined #openstack-manila | 07:57 | |
| *** aspiers[m] has quit IRC | 08:00 | |
| *** vivsoni__ has quit IRC | 08:05 | |
| *** vivsoni has joined #openstack-manila | 08:08 | |
| *** kaisers has quit IRC | 08:11 | |
| *** rishabh__ has joined #openstack-manila | 08:18 | |
| *** rishabh_ has quit IRC | 08:19 | |
| *** rishabh_ has joined #openstack-manila | 08:20 | |
| *** rishabh_ has quit IRC | 08:22 | |
| *** rishabh__ has quit IRC | 08:23 | |
| *** a-pugachev has joined #openstack-manila | 08:24 | |
| *** kaisers has joined #openstack-manila | 08:33 | |
| *** aspiers[m] has joined #openstack-manila | 09:07 | |
| *** junboli has quit IRC | 09:22 | |
| *** vivsoni has quit IRC | 09:31 | |
| *** vivsoni has joined #openstack-manila | 09:39 | |
| *** belmoreira has joined #openstack-manila | 09:42 | |
| *** a-pugachev has quit IRC | 09:45 | |
| arnewiebalck | The docs at https://docs.openstack.org/manila/pike/admin/shared-file-systems-share-replication.html say to check replication support for drivers “in the manila-scheduler service reports”. Could someone point me to what is meant by “service reports”? | 09:45 |
| *** a-pugachev has joined #openstack-manila | 09:46 | |
| arnewiebalck | Is it what is sent to syslog when sending a USR1 signal? | 09:47 |
| *** masber has quit IRC | 10:01 | |
| *** vivsoni has quit IRC | 10:19 | |
| *** vivsoni has joined #openstack-manila | 10:19 | |
| *** rishabh has joined #openstack-manila | 10:33 | |
| *** rishabh is now known as Guest91486 | 10:33 | |
| *** a-pugachev has quit IRC | 10:43 | |
| *** a-pugachev has joined #openstack-manila | 10:47 | |
| *** ganso has joined #openstack-manila | 10:49 | |
| *** a-pugachev has quit IRC | 11:00 | |
| *** a-pugachev has joined #openstack-manila | 11:03 | |
| *** Guest91486 has quit IRC | 11:04 | |
| *** gouthamr has joined #openstack-manila | 11:08 | |
| *** ubijtsa is now known as assassin | 11:56 | |
| *** a-pugachev has quit IRC | 12:02 | |
| *** rishabh has joined #openstack-manila | 12:22 | |
| *** rishabh is now known as rishabh_ | 12:22 | |
| openstackgerrit | Rodrigo Barbieri proposed openstack/manila master: TESTING CI https://review.openstack.org/543524 | 12:35 |
| *** a-pugachev has joined #openstack-manila | 12:35 | |
| *** masber has joined #openstack-manila | 12:38 | |
| *** rishabh_ has quit IRC | 12:49 | |
| openstackgerrit | Merged openstack/manila-ui master: Imported Translations from Zanata https://review.openstack.org/546873 | 12:49 |
| *** gouthamr has quit IRC | 13:04 | |
| *** gouthamr has joined #openstack-manila | 13:04 | |
| tbarron | arnewiebalck: I suspect what is meant are the periodic updates to the scheduler from the backends. | 13:29 |
| tbarron | arnewiebalck: these include reports of capabilities in addition to usage,etc. | 13:29 |
| *** rishabh has joined #openstack-manila | 13:37 | |
| *** rishabh is now known as Guest57466 | 13:38 | |
| *** dustins has joined #openstack-manila | 13:46 | |
| *** gouthamr has quit IRC | 13:50 | |
| *** dustins has quit IRC | 13:52 | |
| arnewiebalck | tbarron: Thanks! I checked these logs, but didn’t see the ‘replication_type’ capability being reported by the driver we use. | 13:55 |
| arnewiebalck | tbarron: The docs imply that every driver would report this capability. | 13:55 |
| tbarron | arnewiebalck: well ceph drivers do not support replication and | 13:56 |
| tbarron | arnewiebalck: I think the the default value is False | 13:57 |
| arnewiebalck | tbarron: Right. | 13:57 |
| tbarron | arnewiebalck: perhaps the docs need revision, or if there was an expectation that every driver would explicitly report then I don't think that was | 13:57 |
| tbarron | communicated to driver back ends owners, etc. | 13:58 |
| arnewiebalck | tbarron: The docs sayThe replication_type capability reported can have one of the following values: | 13:58 |
| tbarron | I could be wrong though, ask again when gouthamr is back from his honeymoon :) | 13:58 |
| arnewiebalck | tbarron: … None | 13:58 |
| arnewiebalck | The driver does not support Share Replication. | 13:58 |
| arnewiebalck | tbarron: Just me making assumptions ;) | 13:59 |
| tbarron | arnewiebalck: does the doc explicitly say that every driver will report? | 13:59 |
| arnewiebalck | tbarron: Ok, ta! | 13:59 |
| arnewiebalck | tbarron: Well ... | 13:59 |
| arnewiebalck | The docs say to check the service reports | 13:59 |
| tbarron | Fair enough. | 13:59 |
| arnewiebalck | There should be a replication_type. | 14:00 |
| arnewiebalck | And if the driver does not support replication, its value should be ‘None’. | 14:00 |
| arnewiebalck | So, I read the docs that every driver is reporting. | 14:00 |
| tbarron | I think it will be a better use of time to modify the doc a little bit than to require every back end to report None for every new capability that isn't supported. | 14:01 |
| arnewiebalck | :-D | 14:01 |
| tbarron | arnewiebalck: on a slight tangent, this kind of replication makes sense for | 14:01 |
| tbarron | back ends where file systems are built on top of block devices | 14:02 |
| tbarron | arnewiebalck: if we build cephfs on rbd then rbd replication could be exploited. | 14:02 |
| tbarron | arnewiebalck: but cephfs isn't like that | 14:03 |
| arnewiebalck | tbarron: no | 14:03 |
| tbarron | arnewiebalck: of course the underlying objects are being replicated anyways | 14:03 |
| arnewiebalck | tbarron: I was looking into this to accomodate a use case where the old service (based on NFS) offers replication to a remote DC | 14:04 |
| tbarron | arnewiebalck: not to say that there may not be a place for some kind of data protection that involves some form of replication to | 14:04 |
| tbarron | another ceph cluster | 14:04 |
| arnewiebalck | exactly | 14:04 |
| arnewiebalck | disaster recovery | 14:04 |
| arnewiebalck | the current solution relies on snasphots,backups, and replication | 14:05 |
| arnewiebalck | none of these we offer for CephFS atm | 14:05 |
| arnewiebalck | snapshots will come with Mimic I understand | 14:05 |
| tbarron | arnewiebalck: snapshot support is there but off by default still | 14:05 |
| arnewiebalck | tbarron: yes, too risky for prod (for now) | 14:06 |
| tbarron | I've asked whether snapshots would be a good basis for async replication. | 14:06 |
| tbarron | greg farnum replied at the time "maybe, but probably we'd do it at a lower level than that" | 14:07 |
| arnewiebalck | ok | 14:07 |
| arnewiebalck | do you happen to know if there are plans for geo-replication for CephFS? | 14:07 |
| tbarron | arnewiebalck: there are plans for everything :) | 14:08 |
| arnewiebalck | tbarron: ha ha ha, stupid question :-D | 14:08 |
| tbarron | arnewiebalck: I'm not really the person to ask :) | 14:08 |
| arnewiebalck | ok | 14:08 |
| tbarron | just a manila guy | 14:08 |
| arnewiebalck | :) | 14:08 |
| *** eharney has joined #openstack-manila | 14:08 | |
| arnewiebalck | thanks for your help! | 14:08 |
| tbarron | batrick is on the west coast US and may read the backlog later | 14:09 |
| tbarron | and sage___ ^^^ is lurking | 14:09 |
| tbarron | oh, and gregsfortytwo1 ^^^ himself | 14:10 |
| *** assassin has quit IRC | 14:38 | |
| *** assassin has joined #openstack-manila | 14:41 | |
| batrick | arnewiebalck: tbarron: async replication baked into CephFS is a possibility but more likely we'd offer primitives that makes existing POSIX FS replication more efficient | 14:58 |
| batrick | i.e. rsync | 14:58 |
| batrick | perhaps provide recursive versioning so rsync knows where to look | 14:58 |
| arnewiebalck | batrick: Hmm, of course, I’d prefer something built into CephFS, rather than having (efficiency) support to build my own machinery on top. | 15:03 |
| *** masber has quit IRC | 15:05 | |
| *** dustins has joined #openstack-manila | 15:14 | |
| *** kaisers has quit IRC | 15:15 | |
| *** kaisers has joined #openstack-manila | 15:15 | |
| *** arnewiebalck has quit IRC | 16:01 | |
| *** a-pugachev has quit IRC | 16:30 | |
| *** openstackgerrit has quit IRC | 16:33 | |
| *** a-pugachev has joined #openstack-manila | 16:33 | |
| *** a-pugachev has quit IRC | 16:33 | |
| *** a-pugachev has joined #openstack-manila | 16:35 | |
| *** assassin has quit IRC | 16:51 | |
| *** belmoreira has quit IRC | 17:25 | |
| *** a-pugachev has quit IRC | 17:45 | |
| *** lpetrut has joined #openstack-manila | 17:48 | |
| *** kaisers has quit IRC | 18:09 | |
| gregsfortytwo1 | tbarron: batrick: (guess Arne is gone) geo-rep is cool and all but my recollection is that manila replication just meant "this share is replicated into another NFS box [so it is resilient against power surges etc]" | 18:14 |
| gregsfortytwo1 | if that works for geo-rep I guess that's cool? but we discussed what it meant in pretty explicit terms when that came up as a reporting thing and it was one of the ways Manila did not seem to fit well with a scale-out FS solution | 18:14 |
| *** openstackgerrit has joined #openstack-manila | 18:24 | |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila-ui master: Update reno for stable/newton https://review.openstack.org/547608 | 18:24 |
| openstackgerrit | OpenStack Release Bot proposed openstack/manila master: Update reno for stable/newton https://review.openstack.org/547611 | 18:24 |
| *** markstur has joined #openstack-manila | 18:32 | |
| *** pcaruana has quit IRC | 18:33 | |
| *** xyang_ has joined #openstack-manila | 18:37 | |
| *** logan_ has joined #openstack-manila | 18:43 | |
| *** logan- has quit IRC | 18:44 | |
| *** xyang has quit IRC | 18:44 | |
| *** xyang_ is now known as xyang | 18:44 | |
| *** cfey has quit IRC | 18:47 | |
| *** logan_ is now known as logan- | 18:47 | |
| *** cfey has joined #openstack-manila | 18:47 | |
| *** Guest57466 has quit IRC | 19:02 | |
| *** markstur has quit IRC | 19:42 | |
| *** a-pugachev has joined #openstack-manila | 20:03 | |
| *** erlon has quit IRC | 20:04 | |
| *** lpetrut has quit IRC | 20:12 | |
| *** FL1SK has quit IRC | 22:17 | |
| *** ganso has quit IRC | 22:17 | |
| *** masber has joined #openstack-manila | 22:27 | |
| *** masber has quit IRC | 22:30 | |
| *** masber has joined #openstack-manila | 22:30 | |
| *** dustins has quit IRC | 22:46 | |
| *** FL1SK has joined #openstack-manila | 22:52 | |
| *** markstur has joined #openstack-manila | 23:06 | |
| *** masuberu has joined #openstack-manila | 23:07 | |
| *** masber has quit IRC | 23:09 | |
| *** FL1SK has quit IRC | 23:20 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!