| *** m_kazuhiro has joined #swift3 | 00:50 | |
| *** bill_az has quit IRC | 01:41 | |
| *** openstackgerrit has quit IRC | 11:43 | |
| *** openstackgerrit has joined #swift3 | 11:44 | |
| *** m_kazuhiro has quit IRC | 11:52 | |
| *** siva_krishnan has joined #swift3 | 13:23 | |
| *** bill_az has joined #swift3 | 13:37 | |
| *** ajiang_ has joined #swift3 | 13:55 | |
| *** hrou_ has joined #swift3 | 13:55 | |
| *** ajiang has quit IRC | 13:56 | |
| *** hrou has quit IRC | 13:56 | |
| *** hrou_ has quit IRC | 15:19 | |
| *** hrou has joined #swift3 | 15:20 | |
| *** chsc has joined #swift3 | 17:05 | |
| *** peluse_ has joined #swift3 | 18:18 | |
| *** peluse has quit IRC | 18:22 | |
| *** hrou has quit IRC | 18:57 | |
| *** hrou has joined #swift3 | 18:57 | |
| *** m_kazuhiro has joined #swift3 | 22:58 | |
| kota_ | morning | 23:00 |
|---|---|---|
| m_kazuhiro | morning | 23:00 |
| chsc | Hi everybody. | 23:01 |
| *** bhakta has quit IRC | 23:03 | |
| bill_az | Hi | 23:04 |
| kota_ | bill_az, chsc: hi! | 23:05 |
| *** bhakta has joined #swift3 | 23:05 | |
| bill_az | Hey Kota, Happy 2016! (and m_kazuhiro too) | 23:05 |
| kota_ | bill_az: i'm happy to see you again, this year :-) | 23:06 |
| bill_az | Thanks | 23:06 |
| kota_ | ok, let get started | 23:07 |
| kota_ | agenda is here, https://wiki.openstack.org/wiki/Meetings/swift3 but i think nothing special for today. | 23:07 |
| chsc | I have one patch where i would like to get feedback on. Does not have to be during the meeting, just wanted to mention in here. | 23:08 |
| kota_ | chsc: nice, which? | 23:08 |
| chsc | https://review.openstack.org/#/c/269300/ | 23:08 |
| chsc | That would fix the owner returned in the bucket listing when running with s3_acl=True | 23:09 |
| bill_az | Fixes this one: https://bugs.launchpad.net/swift3/+bug/1535459 | 23:10 |
| openstack | Launchpad bug 1535459 in Swift3 "Bucket listing returns wrong owner" [Undecided,New] | 23:10 |
| kota_ | chsc: thanks for working this | 23:10 |
| kota_ | quick impression for me, could we use response.object_acl transparently on both (s3_acl=True/False) case? | 23:12 |
| kota_ | i forgot whether nonS3acl response has object_acl attribute or not. | 23:13 |
| chsc | req.get_acl_response only exists for s3_acl=True | 23:13 |
| chsc | So we can only have a response in that case. | 23:13 |
| kota_ | yup, i mean | 23:14 |
| kota_ | if we coud set object_acl as owner is a requet user in s3_acl=False case, it's not necessary to add special condition in controller for s3acl=True or False | 23:15 |
| kota_ | just do like as SubElement(owner, 'ID').text = resp.object_acl.id | 23:16 |
| kota_ | s/special condition/special *if* condition/ | 23:17 |
| kota_ | oh, no | 23:17 |
| chsc | The current request that the controller belongs to is for the bucket. In the s3_acl=True case we need to send a request for each object to retrieve the object ACL to get the owner. | 23:17 |
| kota_ | sorry | 23:17 |
| chsc | We need at least the if for the req.get_acl_response call. | 23:18 |
| kota_ | this is on listing objects, right? | 23:18 |
| chsc | yes | 23:18 |
| kota_ | gotcha | 23:18 |
| chsc | ok | 23:19 |
| kota_ | that's a known issue I was pending. | 23:19 |
| kota_ | (and a reason I am working on last-modified support on listing containers on upstream swift. | 23:20 |
| kota_ | historically, we don't want a tons of extra HEAD's for back ground objects because listing-objecs is one of most common api for users. | 23:21 |
| kota_ | and in general a lot of objects will be in a container. | 23:21 |
| chsc | Right. Only the owner is stored with each object and required in the bucket listing. | 23:22 |
| chsc | The other option would be storing the object owner in the container database, but i don't see whether that would be possible, since the db is in core swift. | 23:22 |
| kota_ | chsc: i think it's possible on container-sysmeta or memcached | 23:23 |
| chsc | Ok. I can take a look how the owner could ne stored there. | 23:25 |
| kota_ | chsc: thanks, is it urgent? | 23:26 |
| chsc | No, not urgent. | 23:26 |
| kota_ | if so, we could consider merge it as tentatively | 23:26 |
| kota_ | chsc: ah, ok. | 23:26 |
| chsc | Thank you for taking a look. | 23:27 |
| kota_ | sure | 23:27 |
| kota_ | i am looking forward to that it gets progress :-) | 23:28 |
| kota_ | thanks for bringing up | 23:28 |
| bill_az | I had a question on the ceph-s3 tests, and posting results of run for each new release. | 23:28 |
| bill_az | I know llyrad & alpha_ori had been working on it | 23:29 |
| bill_az | but don't see them today - kota: do you know any update on that? | 23:29 |
| kota_ | bill_az: do you mean about https://github.com/swiftstack/s3compat ? | 23:30 |
| kota_ | which is a bunch of script to run ceph-s3 and to summrize the result. | 23:31 |
| kota_ | and also, I'd like to try it and to consider merge into the gate docs like swift upstream docs but not yet. | 23:32 |
| bill_az | kota: yes, that is the core of it, summarize the result in wiki format so we can easily publish it with up to date results | 23:33 |
| kota_ | i don't mean i'd like to merge whole things of s3compat into swift3 repo, just want to run at the gate and show the docs to enable us to see which feature supported. | 23:33 |
| kota_ | in the patch. | 23:33 |
| bill_az | thats a good idea running that as part of gate so that for every update we have latest results | 23:35 |
| kota_ | i wonder if someone takes the ticket :-) | 23:35 |
| kota_ | kidding | 23:36 |
| bill_az | lol - I thought the ticket was already taken... ;-) | 23:37 |
| kota_ | i wish too :-) | 23:38 |
| kota_ | so that's current status for docs | 23:38 |
| kota_ | anything else? | 23:39 |
| bill_az | I don't have anything else for today | 23:39 |
| kota_ | all: last call another topic for today? | 23:41 |
| bill_az | kota: do you know if guys from Intel were able to get started helping with Swift3? | 23:41 |
| kota_ | looks nothing. | 23:41 |
| bill_az | are there high priority items they could help with? | 23:42 |
| kota_ | i think siva is one from openstack innovation center | 23:42 |
| kota_ | but not sure if he is rackers or intels | 23:42 |
| kota_ | wait, seeing his e-mail address | 23:43 |
| kota_ | looks like an intel guy, https://review.openstack.org/#/q/owner:siva.radhakrishnan%2540intel.com+status:open | 23:44 |
| bill_az | great - looks like he's helping on improving tests | 23:44 |
| bill_az | That's all I have | 23:45 |
| kota_ | :-) | 23:45 |
| bill_az | kota: Thanks | 23:45 |
| kota_ | ok, finish up this meeting, next will be happen in 2 weeks, 3rd Feb. | 23:46 |
| kota_ | thanks for working swift3 :-) | 23:46 |
| chsc | Sounds good. Thank you for running the meeting. | 23:46 |
| *** chsc has quit IRC | 23:47 | |
| timburke | siva_krishnan has also been doing a great job of moving some older patches forward (like patch 159130 and patch 155221) | 23:55 |
| patchbot | timburke: https://review.openstack.org/#/c/159130/ - Fix DELETE Bucket to also delete segments bucket | 23:55 |
| patchbot | timburke: https://review.openstack.org/#/c/155221/ - Fix Delete Multiple Objects process when multipart... (MERGED) | 23:55 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!