| *** a-pugachev has quit IRC | 01:00 | |
| *** kaisers_ has joined #openstack-manila | 02:28 | |
| *** kaisers has quit IRC | 02:31 | |
| *** caowei has joined #openstack-manila | 02:57 | |
| *** caowei has quit IRC | 03:53 | |
| *** caowei has joined #openstack-manila | 03:53 | |
| *** ianychoi__ has quit IRC | 04:30 | |
| *** kaisers_ has quit IRC | 06:04 | |
| *** kaisers has joined #openstack-manila | 06:31 | |
| *** pcaruana has quit IRC | 07:18 | |
| *** pcaruana has joined #openstack-manila | 07:19 | |
| *** kaisers has quit IRC | 08:19 | |
| *** kaisers has joined #openstack-manila | 08:19 | |
| *** ianychoi has joined #openstack-manila | 08:21 | |
| *** kaisers has quit IRC | 08:59 | |
| *** kaisers has joined #openstack-manila | 09:01 | |
| *** kaisers has quit IRC | 09:05 | |
| *** kaisers has joined #openstack-manila | 09:05 | |
| *** dsariel has joined #openstack-manila | 09:16 | |
| *** ianychoi_ has joined #openstack-manila | 09:46 | |
| *** ianychoi has quit IRC | 09:49 | |
| *** arnewiebalck_ has joined #openstack-manila | 10:36 | |
| *** cfey has quit IRC | 10:50 | |
| *** cfey has joined #openstack-manila | 10:51 | |
| *** cfey has quit IRC | 11:26 | |
| *** BV9GP2olspookish has joined #openstack-manila | 11:46 | |
| *** cfey has joined #openstack-manila | 12:46 | |
| *** cfey has quit IRC | 12:58 | |
| *** cfey has joined #openstack-manila | 13:00 | |
| *** cfey has quit IRC | 13:13 | |
| *** cfey has joined #openstack-manila | 13:14 | |
| *** openstackgerrit has joined #openstack-manila | 13:25 | |
| openstackgerrit | zhongjun proposed openstack/manila-specs master: Support metadata for access rule resource https://review.openstack.org/551106 | 13:25 |
|---|---|---|
| *** zhongjun_ has joined #openstack-manila | 13:46 | |
| zhongjun_ | bswartz tbarron : Could you take a look at the access metadata spec when you have free time? Thanks https://review.openstack.org/551106 | 14:33 |
| *** assassin has quit IRC | 14:34 | |
| tbarron | zhongjun_: Thanks, I'm going to prioritize doing spec reviews this week. | 14:40 |
| tbarron | zhongjun_: What happended to the *squash attributes proposal that you presented in PTG? I don't see a spec for that. | 14:41 |
| tbarron | s/happended/happened/ | 14:41 |
| zhongjun_ | Yes, the squash attributes did not proposal now | 14:43 |
| tbarron | Are you saying that you are not going to propose that? Is that because you are going to rely on the metadata field instead? | 14:43 |
| zhongjun_ | yes, I am not going to propose that in rocky | 14:47 |
| tbarron | bswartz: ^^^ that's what bothers me about the metadata proposal, it makes it too easy to fork off a separate implementation even when the community agrees a new capability should be public, cross-vendor, etc. | 14:47 |
| tbarron | zhongjun_: I understand that it's the shortest path for you, it's just not really what we as a community said about those *squash* attributes | 14:48 |
| zhongjun_ | It could be proposed in next version | 14:49 |
| zhongjun_ | We will not use the metadata field in squash attributes stuff in open source manila version | 14:52 |
| zhongjun_ | tbarron: the metadata proposal won’t relate to a new capability in open source version | 14:55 |
| tbarron | Understood. My point is that putting metadata fields in and encouraging their use in non-open-source forks makes it | 14:59 |
| tbarron | too easy for vendors to do their own forked non-open-source implementations of features that *should* be | 14:59 |
| tbarron | part of the open-source distro. | 14:59 |
| tbarron | The open source distro loses energy and contributions that belong there. | 15:00 |
| tbarron | I agree however that some features are going to be needed just in forks and don't belong in the open source distro. | 15:00 |
| tbarron | However I think we as a community agreed at the Rocky PTG that the *squash* features were not in this last category. | 15:01 |
| zhongjun_ | The access meatada is the same as share metadata data, it just used for tag something what user wants | 15:02 |
| tbarron | It tags something the user wants and what they want is not implemented in the open source version of manila ! | 15:03 |
| tbarron | zhongjun_: you talked the community into believing that what the user wants here is valuable, and | 15:04 |
| tbarron | not valuable only to huawei cloud users. | 15:04 |
| tbarron | zhongjun_: maybe we need this as a topic for Thurs. meeting. | 15:04 |
| zhongjun_ | sure | 15:06 |
| *** caowei has quit IRC | 15:14 | |
| zhongjun_ | The metadata use case is the same as the share metadata use case. It’s not about aquash or some other stuff. | 15:20 |
| zhongjun_ | It is not about huawei use case, it is community use case. The metadata Like: user:jdjhf role:hdbbej. Then you can easily to filter what access you want by name or role. | 15:22 |
| zhongjun_ | I will add the separate feature about squash stuff later | 15:23 |
| *** e0ne has joined #openstack-manila | 15:29 | |
| *** knylander has quit IRC | 15:59 | |
| *** knylander has joined #openstack-manila | 16:00 | |
| *** gouthamr has joined #openstack-manila | 16:05 | |
| *** dsariel has quit IRC | 16:31 | |
| *** e0ne has quit IRC | 16:40 | |
| *** gouthamr has quit IRC | 17:03 | |
| *** gouthamr has joined #openstack-manila | 17:12 | |
| *** gouthamr has quit IRC | 17:27 | |
| *** zhongjun_ has quit IRC | 17:46 | |
| *** cfey has quit IRC | 18:18 | |
| *** arnewiebalck_ has quit IRC | 18:28 | |
| *** lpetrut has joined #openstack-manila | 18:56 | |
| *** dsariel has joined #openstack-manila | 19:23 | |
| *** kaisers_ has joined #openstack-manila | 20:05 | |
| *** kaisers has quit IRC | 20:08 | |
| bswartz | tbarron: I agree it makes it easier to solve problems downstream, in a way that the solutions never go upstream | 20:56 |
| bswartz | tbarron: the first part I like, but I agree the second part is a problem | 20:57 |
| bswartz | tbarron: wrt to the specific stuff we discussed at PTG, I think the only cross-vendor compatible one we could find was the "priority" field for access rules | 20:58 |
| bswartz | tbarron: for both all_squash and per-rule-qos, we found that it's not possible to implement in the core without some new capabilities | 20:59 |
| *** a-pugachev has joined #openstack-manila | 21:00 | |
| *** kaisers_ has quit IRC | 21:01 | |
| *** lpetrut has quit IRC | 21:09 | |
| *** kaisers has joined #openstack-manila | 21:10 | |
| tbarron | bswartz: I don't think we agreed per-rule-qos was at all suitable cross-vendor -- so hard to | 21:18 |
| tbarron | bswartz: define cross-vendor that we didn't even really talk about it | 21:19 |
| tbarron | bswartz: *squash on the other hand seemed to be almost a de facto standard, all but a few | 21:19 |
| tbarron | bswartz: back ends can do it (all the *nix based back ends likely support it) so that we thought | 21:20 |
| tbarron | bswartz: it a useful feature but yes, one that would require capabities since not all back ends support it | 21:20 |
| tbarron | bswartz: so it would be in there with replication, create from snapshot, mount from snapshot, etc. | 21:21 |
| tbarron | bswartz: not all back ends support these but they are useful and can their behavior can be | 21:21 |
| tbarron | bswartz: described in a public way | 21:22 |
| tbarron | bswartz: my issue here is that zhongjun spent a good bit of time arguing that *squash* is | 21:22 |
| tbarron | bswartz: a generally useful capability, persuaded the community, but now the proposal is to | 21:23 |
| tbarron | bswartz: handle the need for their customers only with the short-cut do-it-only-in-our-fork way | 21:23 |
| tbarron | bswartz: I won't block the metadata proposal but I will say that I think that I think this kind | 21:25 |
| tbarron | bswartz: of thing weakens the core open source project. | 21:25 |
| tbarron | bswartz: if we had been purists w.r.t. never allowing any capabilities that are not supportable by | 21:26 |
| tbarron | bswartz: *every* back end it would be a different matter; manila made a decision to go the other way | 21:26 |
| tbarron | bswartz: a long time ago and these capabilities seem as legitimate as the others we support | 21:26 |
| *** gouthamr has joined #openstack-manila | 21:42 | |
| *** dsariel has quit IRC | 21:43 | |
| *** pcaruana has quit IRC | 21:54 | |
| *** kaisers has quit IRC | 22:00 | |
| *** kaisers has joined #openstack-manila | 22:10 | |
| *** kaisers has quit IRC | 22:14 | |
| *** a-pugachev has quit IRC | 22:24 | |
| *** kaisers has joined #openstack-manila | 22:30 | |
| *** arnewiebalck has quit IRC | 23:06 | |
| *** unznjlw has joined #openstack-manila | 23:12 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!