| *** matsuhashi has joined #openstack-swift | 00:01 | |
| *** mwstorer has quit IRC | 00:05 | |
| *** shakamunyi has quit IRC | 00:18 | |
| *** mlipchuk has quit IRC | 00:26 | |
| *** zhiyan_ is now known as zhiyan | 00:34 | |
| *** dmorita has joined #openstack-swift | 00:36 | |
| *** zhiyan is now known as zhiyan_ | 00:36 | |
| *** sungju has joined #openstack-swift | 00:38 | |
| *** zhiyan_ is now known as zhiyan | 00:38 | |
| *** kenhui has joined #openstack-swift | 00:40 | |
| *** zhiyan is now known as zhiyan_ | 00:41 | |
| *** shri has quit IRC | 00:44 | |
| *** Edward-Zhang has quit IRC | 00:58 | |
| *** mlipchuk has joined #openstack-swift | 01:10 | |
| openstackgerrit | gholt proposed a change to openstack/swift: Add debug logging to container sync requests https://review.openstack.org/95911 | 01:19 |
|---|---|---|
| *** saschpe has quit IRC | 01:23 | |
| *** saschpe has joined #openstack-swift | 01:24 | |
| *** kenhui has quit IRC | 01:25 | |
| *** zhiyan_ is now known as zhiyan | 01:26 | |
| *** kenhui has joined #openstack-swift | 01:29 | |
| *** kenhui has quit IRC | 01:29 | |
| *** kenhui has joined #openstack-swift | 01:29 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fixes probe tests with non-zero default storage policy https://review.openstack.org/87492 | 01:34 |
| openstackgerrit | A change was merged to openstack/swift: Add ability to remove subsections from /info https://review.openstack.org/94458 | 01:38 |
| openstackgerrit | A change was merged to openstack/swift: Plumb insecure flag to functests https://review.openstack.org/88438 | 01:38 |
| *** kenhui1 has joined #openstack-swift | 01:39 | |
| *** kenhui has quit IRC | 01:41 | |
| *** RockKuo_Office has joined #openstack-swift | 01:45 | |
| *** nosnos has joined #openstack-swift | 01:48 | |
| *** kenhui1 has quit IRC | 01:50 | |
| *** diegows has quit IRC | 02:09 | |
| *** judd7 has joined #openstack-swift | 02:30 | |
| *** ZBhatti_ has joined #openstack-swift | 02:31 | |
| *** mkerrin1 has joined #openstack-swift | 02:32 | |
| *** byeager has joined #openstack-swift | 02:32 | |
| *** judd7_ has quit IRC | 02:33 | |
| *** ZBhatti has quit IRC | 02:33 | |
| *** mkerrin has quit IRC | 02:33 | |
| *** ZBhatti_ is now known as ZBhatti | 02:33 | |
| *** byeager has quit IRC | 02:34 | |
| *** byeager has joined #openstack-swift | 02:35 | |
| *** zhiyan is now known as zhiyan_ | 02:36 | |
| *** kenhui has joined #openstack-swift | 02:45 | |
| *** gyee has quit IRC | 02:49 | |
| *** zhiyan_ is now known as zhiyan | 02:49 | |
| *** kenhui has quit IRC | 02:51 | |
| *** kenhui has joined #openstack-swift | 02:51 | |
| *** haomaiwa_ has joined #openstack-swift | 02:59 | |
| *** byeager has quit IRC | 03:03 | |
| *** madhuri has quit IRC | 03:03 | |
| *** madhuri has joined #openstack-swift | 03:04 | |
| madhuri | Hi, Can anyone please help reviewing this https://review.openstack.org/#/c/85351/ ? | 03:04 |
| *** Edward-Zhang has joined #openstack-swift | 03:09 | |
| clayg | madhuri: do you need help getting started on a test for that change? | 03:15 |
| *** kenhui1 has joined #openstack-swift | 03:15 | |
| *** zhiyan is now known as zhiyan_ | 03:16 | |
| *** kenhui has quit IRC | 03:18 | |
| *** kenhui1 has quit IRC | 03:18 | |
| *** kenhui has joined #openstack-swift | 03:19 | |
| *** shakamunyi has joined #openstack-swift | 03:19 | |
| *** kenhui has quit IRC | 03:21 | |
| madhuri | I will look for it to add | 03:22 |
| *** zhiyan_ is now known as zhiyan | 03:22 | |
| openstackgerrit | A change was merged to openstack/python-swiftclient: change assertEquals to assertEqual https://review.openstack.org/93603 | 03:23 |
| *** shakamunyi has quit IRC | 03:24 | |
| madhuri | clayg: https://review.openstack.org/#/c/84819/ this too | 03:24 |
| *** Midnightmyth has joined #openstack-swift | 03:31 | |
| *** Midnightmyth has quit IRC | 03:38 | |
| madhuri | clayg: How is the container server or object server selected for processing any request? | 03:38 |
| clayg | https://www.youtube.com/watch?v=o0YWRXJsMyM | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Documentation https://review.openstack.org/96026 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support https://review.openstack.org/96027 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update FakeRing and FakeLogger https://review.openstack.org/96028 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Containers https://review.openstack.org/96029 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Object Server https://review.openstack.org/96030 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to Object Updates https://review.openstack.org/96031 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to the Auditor https://review.openstack.org/96032 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add storage policy support for the Replicator https://review.openstack.org/96033 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to ssync https://review.openstack.org/96034 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Put X-Timestamp in object 404 responses https://review.openstack.org/96035 | 03:40 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend interface on InternalClient https://review.openstack.org/96036 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend direct_client https://review.openstack.org/96037 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add container-reconciler daemon https://review.openstack.org/96038 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add reconciler probetest outline https://review.openstack.org/96039 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Merge container storage_policy_index https://review.openstack.org/96040 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Accounts https://review.openstack.org/96041 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add functional tests for Storage Policy https://review.openstack.org/96042 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update swift-container-info to be storage policy aware https://review.openstack.org/96043 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Enqueue misplaced objects during container replication https://review.openstack.org/96044 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add LRUCache to common.utils https://review.openstack.org/96045 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to list_endpoints https://review.openstack.org/96046 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Container Sync https://review.openstack.org/96047 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy Support to Recon Middleware https://review.openstack.org/96048 | 03:41 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy support to the Account Reaper https://review.openstack.org/96049 | 03:41 |
| *** zhiyan is now known as zhiyan_ | 03:41 | |
| madhuri | clayg: What all new features are up for this release? | 03:45 |
| *** nosnos has quit IRC | 03:47 | |
| *** Edward-Zhang has quit IRC | 03:51 | |
| *** shakamunyi has joined #openstack-swift | 03:53 | |
| *** ppai has joined #openstack-swift | 03:54 | |
| *** zhiyan_ is now known as zhiyan | 04:13 | |
| *** matsuhashi has quit IRC | 04:15 | |
| openstackgerrit | A change was merged to openstack/swift: Add Ratelimit parameters to /info https://review.openstack.org/88633 | 04:20 |
| *** psharma has joined #openstack-swift | 04:22 | |
| *** Edward-Zhang has joined #openstack-swift | 04:25 | |
| *** matsuhashi has joined #openstack-swift | 04:26 | |
| *** nosnos has joined #openstack-swift | 04:27 | |
| *** shakamunyi has quit IRC | 04:27 | |
| *** bkopilov has quit IRC | 04:29 | |
| *** kashyapk has joined #openstack-swift | 04:35 | |
| *** kashyapk has quit IRC | 04:36 | |
| *** bvandenh has joined #openstack-swift | 04:50 | |
| *** shakamunyi has joined #openstack-swift | 04:52 | |
| *** zhiyan is now known as zhiyan_ | 04:53 | |
| zaitcev | daarn, I'm just getting to bed and SP reviews landed | 04:53 |
| *** shakamunyi has quit IRC | 04:54 | |
| *** shakamunyi has joined #openstack-swift | 05:11 | |
| *** bvandenh has quit IRC | 05:12 | |
| *** zhiyan_ is now known as zhiyan | 05:14 | |
| *** zaitcev has quit IRC | 05:20 | |
| *** zhiyan is now known as zhiyan_ | 05:25 | |
| *** shakamunyi has quit IRC | 05:40 | |
| *** baojg has joined #openstack-swift | 05:40 | |
| *** shakamunyi has joined #openstack-swift | 05:40 | |
| *** baojg_ has joined #openstack-swift | 05:47 | |
| *** baojg_ has quit IRC | 05:50 | |
| *** baojg_ has joined #openstack-swift | 05:50 | |
| *** baojg has quit IRC | 05:51 | |
| openstackgerrit | Yuan Zhou proposed a change to openstack/swift: Update container storage_policy_index when auditing https://review.openstack.org/83093 | 05:53 |
| *** yuanzz is now known as yuan | 05:54 | |
| *** mlipchuk has left #openstack-swift | 07:07 | |
| *** zhiyan_ is now known as zhiyan | 07:18 | |
| *** zhiyan is now known as zhiyan_ | 07:27 | |
| *** haomaiwa_ has quit IRC | 07:28 | |
| *** haomaiwang has joined #openstack-swift | 07:28 | |
| *** bkopilov has joined #openstack-swift | 07:29 | |
| *** sungju has quit IRC | 07:30 | |
| *** haomai___ has joined #openstack-swift | 07:37 | |
| *** haomaiwang has quit IRC | 07:41 | |
| *** bvandenh has joined #openstack-swift | 07:41 | |
| *** nshaikh has joined #openstack-swift | 07:44 | |
| *** baojg_ has quit IRC | 07:49 | |
| *** baojg has joined #openstack-swift | 07:51 | |
| *** haomai___ has quit IRC | 08:05 | |
| *** haomaiwang has joined #openstack-swift | 08:05 | |
| *** haomai___ has joined #openstack-swift | 08:07 | |
| *** haomaiwang has quit IRC | 08:10 | |
| *** matsuhashi has quit IRC | 08:14 | |
| *** matsuhashi has joined #openstack-swift | 08:15 | |
| *** matsuhashi has quit IRC | 08:17 | |
| *** matsuhashi has joined #openstack-swift | 08:17 | |
| openstackgerrit | Yuan Zhou proposed a change to openstack/swift: Allow to deprecate policy https://review.openstack.org/86827 | 08:19 |
| *** sungju has joined #openstack-swift | 08:21 | |
| *** haomai___ has quit IRC | 08:22 | |
| *** haomaiwa_ has joined #openstack-swift | 08:22 | |
| *** nacim has joined #openstack-swift | 08:32 | |
| *** haomai___ has joined #openstack-swift | 08:37 | |
| *** sungju has quit IRC | 08:37 | |
| *** haomaiwa_ has quit IRC | 08:40 | |
| *** shakamunyi has quit IRC | 08:43 | |
| *** haomai___ has quit IRC | 08:57 | |
| *** haomaiwang has joined #openstack-swift | 08:58 | |
| *** haomai___ has joined #openstack-swift | 09:02 | |
| *** mlipchuk has joined #openstack-swift | 09:04 | |
| *** haomaiwang has quit IRC | 09:05 | |
| *** mlipchuk has quit IRC | 09:09 | |
| *** haomai___ has quit IRC | 09:17 | |
| *** haomaiwa_ has joined #openstack-swift | 09:17 | |
| *** zhiyan_ is now known as zhiyan | 09:19 | |
| *** nacim has quit IRC | 09:21 | |
| *** zhiyan is now known as zhiyan_ | 09:28 | |
| *** haomaiwa_ has quit IRC | 09:35 | |
| *** haomaiwa_ has joined #openstack-swift | 09:35 | |
| *** zhiyan_ is now known as zhiyan | 09:37 | |
| *** shakamunyi has joined #openstack-swift | 09:40 | |
| *** shakamunyi has quit IRC | 09:45 | |
| *** zhiyan is now known as zhiyan_ | 09:46 | |
| *** nacim has joined #openstack-swift | 09:48 | |
| *** sungju has joined #openstack-swift | 09:56 | |
| *** sungju has quit IRC | 09:59 | |
| *** mkollaro has joined #openstack-swift | 10:01 | |
| *** sungju has joined #openstack-swift | 10:01 | |
| *** sungju has quit IRC | 10:03 | |
| *** sungju has joined #openstack-swift | 10:09 | |
| *** mlipchuk has joined #openstack-swift | 10:21 | |
| *** nacim has quit IRC | 10:25 | |
| *** Edward-Zhang has quit IRC | 10:29 | |
| *** erlon has quit IRC | 10:39 | |
| *** erlon has joined #openstack-swift | 10:41 | |
| *** matsuhashi has quit IRC | 10:48 | |
| *** ppai has quit IRC | 10:49 | |
| *** diegows has joined #openstack-swift | 10:53 | |
| *** sungju has quit IRC | 10:55 | |
| *** judd7_ has joined #openstack-swift | 10:56 | |
| *** judd7 has quit IRC | 10:57 | |
| *** tdasilva has joined #openstack-swift | 11:01 | |
| *** ppai has joined #openstack-swift | 11:02 | |
| *** baojg has quit IRC | 11:09 | |
| *** baojg has joined #openstack-swift | 11:10 | |
| *** dmorita has quit IRC | 11:11 | |
| *** baojg has quit IRC | 11:15 | |
| *** saschpe has left #openstack-swift | 11:18 | |
| *** RockKuo_Office has quit IRC | 11:32 | |
| *** miqui has quit IRC | 11:51 | |
| *** sagar has joined #openstack-swift | 11:51 | |
| *** sagar is now known as Guest95754 | 11:51 | |
| *** jamie_h has joined #openstack-swift | 11:52 | |
| *** pconstantine has quit IRC | 11:53 | |
| *** baojg has joined #openstack-swift | 11:54 | |
| *** Guest95754 is now known as sagar_ | 11:54 | |
| sagar_ | Hello everyone. I am having problems installing swift and integrating it with keystone. Can I get some help? | 11:58 |
| *** sagar_ has quit IRC | 12:08 | |
| *** sagar_ has joined #openstack-swift | 12:09 | |
| *** nacim has joined #openstack-swift | 12:10 | |
| sagar_ | Hello, I am trying to install the latest version of swift from icehouse. After authenticating with keystone, I am able to run successfully the command "swift stat", but the command "swift upload myfiles temp" throws the following error http://paste.openstack.org/show/81860/ | 12:11 |
| *** tdasilva has quit IRC | 12:14 | |
| *** psharma has quit IRC | 12:15 | |
| *** sagar_ has quit IRC | 12:15 | |
| *** ppai has quit IRC | 12:27 | |
| *** tdasilva has joined #openstack-swift | 12:31 | |
| *** tdasilva has left #openstack-swift | 12:44 | |
| *** tdasilva has joined #openstack-swift | 12:46 | |
| *** nosnos has quit IRC | 12:50 | |
| *** miqui has joined #openstack-swift | 12:53 | |
| *** nacim has quit IRC | 12:56 | |
| gholt | Dayum. Are my search settings wrong or does clayg actually own the whole front review page for Swift? | 12:57 |
| gholt | I guess it's storage policy review time, eh? | 12:57 |
| *** baojg has quit IRC | 12:58 | |
| gholt | I would also assume that means the "soft freeze" is in effect and folks should not hit merge on anything until we get the SP stuff sorted out. | 13:01 |
| *** fifieldt has quit IRC | 13:02 | |
| gholt | I believe this is the root change to start with: https://review.openstack.org/#/c/96026 | 13:05 |
| *** haomaiwa_ has quit IRC | 13:11 | |
| *** haomaiwang has joined #openstack-swift | 13:11 | |
| *** bkopilov has quit IRC | 13:18 | |
| *** bkopilov has joined #openstack-swift | 13:19 | |
| *** pberis has joined #openstack-swift | 13:22 | |
| *** pberis has quit IRC | 13:27 | |
| *** haomai___ has joined #openstack-swift | 13:28 | |
| *** haomaiwang has quit IRC | 13:30 | |
| gholt | clayg: When we talked about split patchsets and one patchset before I'd mentioned split would be much better because any Joe should be able to merge them all to a single branch locally if they want. However, I didn't realize there'd be SO MANY. ;) Is there a way you could also post a single patchset version of all this for testing purposes for us? | 13:49 |
| gholt | If it's just me, don't worry about it, I'll do the merge down of all the patches to a single branch. But if it's more than me, it'd save a lot of time overall if you pushed a merged patchset for us to leach off of. :D | 13:49 |
| *** pberis has joined #openstack-swift | 13:52 | |
| *** nacim has joined #openstack-swift | 13:55 | |
| *** shakamunyi has joined #openstack-swift | 13:55 | |
| *** bvandenh has quit IRC | 13:59 | |
| *** zaitcev has joined #openstack-swift | 13:59 | |
| *** ChanServ sets mode: +v zaitcev | 14:00 | |
| *** nacim has quit IRC | 14:02 | |
| *** nacim has joined #openstack-swift | 14:22 | |
| *** foexle has joined #openstack-swift | 14:26 | |
| *** mlipchuk has quit IRC | 14:30 | |
| *** Midnightmyth has joined #openstack-swift | 14:42 | |
| *** kenhui has joined #openstack-swift | 14:43 | |
| *** kenhui has quit IRC | 14:43 | |
| *** kenhui has joined #openstack-swift | 14:43 | |
| *** diegows has quit IRC | 14:55 | |
| *** kenhui1 has joined #openstack-swift | 15:03 | |
| *** kenhui1 has quit IRC | 15:03 | |
| *** haomai___ has quit IRC | 15:04 | |
| *** kenhui1 has joined #openstack-swift | 15:06 | |
| *** kenhui has quit IRC | 15:06 | |
| *** nacim has quit IRC | 15:07 | |
| *** baojg has joined #openstack-swift | 15:13 | |
| *** shakamunyi has quit IRC | 15:16 | |
| *** shakamunyi has joined #openstack-swift | 15:18 | |
| *** bkopilov has quit IRC | 15:21 | |
| *** foexle has quit IRC | 15:28 | |
| gholt | clayg: I think, maybe, this is the tail? https://review.openstack.org/#/c/96049/ If so, checking that out would get me what I want, right? A branch with everything in it? | 15:34 |
| notmyname | good morning | 15:34 |
| notmyname | gholt: and ya, that was my plan. find the end and see if that got everything | 15:34 |
| gholt | If so, there's a couple warnings with python setup.py build_sphinx. :) | 15:35 |
| notmyname | I see https://review.openstack.org/#/c/96026/ ans the first patch and https://review.openstack.org/#/c/96049/ as the last one | 15:39 |
| notmyname | gholt: so I confirm what you see | 15:39 |
| *** foexle has joined #openstack-swift | 15:41 | |
| notmyname | `git review -d 96049` will get you the full storage policy patches into one branch | 15:42 |
| notmyname | unfortunately, it looks like the "ratelimit in /info" patch landed after the storage policy stuff was proposed, so the SP branch isn't at the tip of master currently | 15:43 |
| notmyname | clayg: if there are minor issues in a particular patch, how do you want us to handle them? should we simply leave a comment and you'll do what's necessary? or should we push a patch over to fix the minor stuff? | 15:45 |
| *** gyee has joined #openstack-swift | 15:46 | |
| *** GoJkOrS has joined #openstack-swift | 15:46 | |
| gholt | notmyname: Yeah, I should probably just cool it for a bit. :) Wait for you to send an email about it all. :D | 15:50 |
| * notmyname suspects the answer is to just leave gerrit comments so that we don't have to get into the dependency chain management | 15:50 | |
| notmyname | gholt: composing it now. don't wait for me :-) | 15:50 |
| cschwede_ | notmyname: morning - and thanks for the RT! Hopefully I got everything right ;) | 15:54 |
| notmyname | cschwede_: looks great :-) | 15:54 |
| cschwede_ | notmyname: cool :) | 15:54 |
| gholt | Sweet, unit and functional tests passed here without having to change anything. Usually I have some screwed up minor detail that throws everything off. | 15:55 |
| notmyname | `git diff master | wc -l` ==> 22090 | 15:55 |
| gholt | Haha | 15:55 |
| creiht | dang | 15:56 |
| *** mkollaro has quit IRC | 15:56 | |
| notmyname | `find . -name \*py | xargs wc -l` on master is 96754. with SP is 107205 | 15:57 |
| notmyname | so about 11k added | 15:57 |
| notmyname | or about 10% | 15:57 |
| gholt | wc -l `find swift -name '*.py'` | sort -n = 33220 on master, 35548 with sp | 15:58 |
| notmyname | see, it's only 2k lines added :-) | 15:58 |
| notmyname | and 9k of tests ;-) | 15:58 |
| notmyname | (guessing) | 15:58 |
| gholt | Probe tests failed for me though. Hmm. | 15:59 |
| *** ChanServ changes topic to "Storage Policies have been proposed to master. Soft freeze in effect until these patches land | First SP patch: https://review.openstack.org/#/c/96026/ | Last SP patch: https://review.openstack.org/#/c/96049/" | 16:02 | |
| openstackgerrit | Yuan Zhou proposed a change to openstack/swift: Update swift-object-info/swift-get-nodes to be storage policy aware https://review.openstack.org/82734 | 16:02 |
| creiht | according to sloccount | 16:04 |
| creiht | +1574 lines of code | 16:04 |
| creiht | +6537 lines of tests | 16:05 |
| clayg | gholt: looks like you found the end of the patch set - but yeah I wouldn't mind posting everything in a single change too something like this -> https://github.com/clayg/swift/compare/sp-base | 16:06 |
| notmyname | creiht: when you put it that way, it looks relatively tiny! | 16:06 |
| creiht | no comment | 16:06 |
| cschwede_ | looks like something nice to read on the plane :) | 16:07 |
| clayg | notmyname: i can rebase ontop of anything at anytime - it's the emails to all of you that are going to get annoying | 16:07 |
| gholt | Heheh, filters can solve that. :) | 16:08 |
| clayg | let's get a days worth of reviews commented up - i'll fix everything at rebase in in the late evening again | 16:08 |
| notmyname | clayg: sounds good. thanks | 16:08 |
| clayg | notmyname: because of the rebase mess it's probably better to just leave comments | 16:08 |
| notmyname | clayg: I added my -2 to the start of the chain | 16:08 |
| notmyname | and agreed with that | 16:08 |
| gholt | I think maybe probe tests fail because of new info stuff or something. | 16:11 |
| *** foexle has quit IRC | 16:12 | |
| gholt | Maybe they also rely on some changes to saio I haven't yet made. | 16:12 |
| gholt | Oh, hah, if I scroll up a bit I finally see "'module' object has no attribute 'SkipTest'" so uhm, maybe it's just that. | 16:13 |
| clayg | gholt: oh heh, maybe i wrote a skip in there trying to be clever and never tested the skip :\ | 16:15 |
| gholt | Probably just needs to use nose.SkipTest instead of unittest.SkipTest -- testing that now... | 16:16 |
| clayg | oh, on py 2.6 it probably does - doh! | 16:18 |
| gholt | clayg: Yeah, that was it, just nose.SkipTest instead of unittest.SkipTest -- not sure which patchset I'd comment on that, or if this irc is good enough. :) | 16:18 |
| clayg | bah, lame | 16:19 |
| clayg | well... which file ;) | 16:19 |
| gholt | Hey, I'm quite happy that was all the tests found on my crazy set up. :) | 16:19 |
| gholt | Just two files: test/probe/test_container_merge_policy_index.py test/probe/test_container_sync.py | 16:20 |
| clayg | gholt: do you want to just try to comment here -> https://github.com/clayg/swift/pull/1/files | 16:21 |
| clayg | not for this - i got this - but in general | 16:22 |
| clayg | and thanks | 16:22 |
| notmyname | heh. I'm glad I just caught that. somehow my octogenarian grandparents were added to my "swift core" email list | 16:25 |
| *** foexle has joined #openstack-swift | 16:25 | |
| glange | nepotism | 16:26 |
| briancline | octogenarians everywhere want to know | 16:27 |
| gholt | Heheh | 16:29 |
| gholt | clayg: That's an option too. I'm open for any of the above. I'm guessing it'd be better to keep it all Gerrit for posteriors though. | 16:29 |
| *** foexle has quit IRC | 16:31 | |
| clayg | gholt: oh that'd definately a good point - i was just thinking of your 'can't find the right review to comment on' problem... | 16:32 |
| clayg | i guess a second squashed change wouldn't be that hard to maintain... | 16:32 |
| gholt | Well, there's the tail, which is fine, and git blame will tell me the rest. | 16:32 |
| gholt | So I don't want to make more work when there are already solutions I can use. :D | 16:33 |
| gholt | I'll make the comments just to track stuff. You shouldn't have to track things you're told on irc too. That way you can just hit all the reviews once a night and go go go. | 16:33 |
| clayg | gholt: ok, looks like jenkins had some review comments for me too already | 16:34 |
| gholt | Yeah, Jenkins in the cases I saw was worse than useless. :) | 16:34 |
| creiht | cool tests initially pass here | 16:36 |
| creiht | now to build docs and see if I can update my saio | 16:36 |
| *** zaitcev has quit IRC | 16:41 | |
| clayg | umm... so when I was working on some of the stuff in https://review.openstack.org/#/c/96029/ I wrote these https://gist.github.com/clayg/83a504fd5aa96c077f6a - so... well just let me know if you have questions when you get to the container schema changes | 16:41 |
| *** shri has joined #openstack-swift | 16:46 | |
| *** ozialien has quit IRC | 16:46 | |
| *** pconstantine has joined #openstack-swift | 16:48 | |
| *** baojg has quit IRC | 16:50 | |
| creiht | at least the saio setup is easy with policies | 16:59 |
| *** lpabon has joined #openstack-swift | 16:59 | |
| creiht | and skipped func tests now run | 17:00 |
| creiht | some probe test failures though | 17:01 |
| * creiht waits for them to finish | 17:01 | |
| portante | so on one of my SAIOs unit tests fail with 7 errors and 1 failure | 17:05 |
| portante | do I have to rebuild an SAIO to get the unit tests to work? | 17:06 |
| *** bvandenh has joined #openstack-swift | 17:06 | |
| gholt | portante: I didn't. | 17:06 |
| portante | TypeError: sequence index must be integer, not 'str', at line 105, translate_container_headers_to_info | 17:06 |
| portante | okay, I'll go look then | 17:07 |
| *** diegows has joined #openstack-swift | 17:07 | |
| portante | wait, maybe my deps are out of date ... | 17:07 |
| gholt | That's sounds like... fun. Heheh. | 17:07 |
| portante | ;) | 17:07 |
| *** bvandenh has quit IRC | 17:08 | |
| clayg | oh this is usefulish -> https://review.openstack.org/#/q/status:open+project:openstack/swift+branch:master+topic:bp/storage-policies,n,z | 17:08 |
| portante | clayg: thanks | 17:09 |
| gholt | Ah yeah, not bad. Also, I finally found a use for those Gerrit stars. Marked the leading and trailing patchsets. :) | 17:09 |
| *** zaitcev has joined #openstack-swift | 17:09 | |
| *** ChanServ sets mode: +v zaitcev | 17:09 | |
| clayg | gholt: nice! | 17:09 |
| gholt | Of course, that's just for me, but still I like it! | 17:09 |
| clayg | portante: which changeset are you on? there's been so much rebasing, i know all the tests will pass for each change on at least my environment, but some of the shuffling may have introduced some corners that don't appear in the final product | 17:10 |
| portante | I started with the final patchset, 96049 | 17:11 |
| clayg | oh... well hehe - that should work ;) | 17:11 |
| portante | clayg, but I might have some out-of-date deps in my SAIOs | 17:11 |
| portante | I am running them under tox now to see if that works in my environment | 17:11 |
| *** diegows has quit IRC | 17:18 | |
| GoJkOrS | Hey all, I am trying to setup a new region in swift and I am wondering how I get 1 copy of the data over to that new region and vice versa. ie. region 1 has a new object come in, that object should have 3 copies in region 1 and 1 copy in region 2. Is this possible? | 17:19 |
| clayg | lol @ superdupernit: | 17:20 |
| notmyname | GoJkOrS: swift will take your 3 replicas and spread them out over the regions you have configured. so you will have 3 replicas total, and if you have two regions, you'll have at least one replica in each region | 17:21 |
| clayg | k, i'm getting on a train for a bit - i'll work on some of the doc comments - thanks | 17:21 |
| *** nshaikh has quit IRC | 17:21 | |
| GoJkOrS | Hum. That is not the behavior that I am seeing. I see all 3 copies in 1 region and nothing in the second region. Let me do a bit more testing to see if what you describe is happening. | 17:22 |
| notmyname | GoJkOrS: did you set any write affinity? | 17:22 |
| GoJkOrS | I have not | 17:22 |
| notmyname | GoJkOrS: is this an existing cluster that you are adding a new region to? | 17:23 |
| GoJkOrS | Yes. | 17:23 |
| GoJkOrS | I have 1 zone in each region also. | 17:24 |
| notmyname | GoJkOrS: is the other region a similar size (capacity) as the first one? | 17:24 |
| GoJkOrS | 1 region is 6 nodes and the second is 5 nodes. The nodes are all the same size. | 17:25 |
| notmyname | GoJkOrS: ok. does `swift-ring-builder` say that your ring is balanced? | 17:26 |
| *** bkopilov has joined #openstack-swift | 17:26 | |
| GoJkOrS | 262144 partitions, 3.000000 replicas, 2 regions, 2 zones, 11 devices, 0.00 balance | 17:27 |
| notmyname | GoJkOrS: so far so good :-) | 17:28 |
| notmyname | GoJkOrS: so how's replication in the cluster? has it completed a full pass since you added the other region? | 17:28 |
| GoJkOrS | How do i determine that? | 17:29 |
| notmyname | GoJkOrS: if you have the recon middleware installed, you can use `swift-recon --replication` | 17:29 |
| notmyname | GoJkOrS: otherwise you need to look in the logs and find the "replication pass complete" messages | 17:30 |
| GoJkOrS | OIC. Ok. Let me look. | 17:30 |
| notmyname | GoJkOrS: just looked at the code. the message is "Object replication complete. (%.02f minutes)" and is logged at INFO level | 17:31 |
| notmyname | err...just one space there | 17:31 |
| notmyname | in case you want to grep for it :-) | 17:31 |
| GoJkOrS | Should I look for that on a new node or an existing node? | 17:32 |
| GoJkOrS | As in one from the original region or one on the new region. | 17:32 |
| notmyname | GoJkOrS: should be running on every node | 17:32 |
| notmyname | GoJkOrS: but since replication is push-based, look for it on the original region first | 17:32 |
| *** bkopilov has quit IRC | 17:35 | |
| GoJkOrS | May 28 17:35:53 swift09 object-auditor Object audit (ZBF) "forever" mode completed: 0.04s. Total quarantined: 0, Total errors: 0, Total files/sec: 72.47 , Total bytes/sec: 0.00, Auditing time: 0.04, Rate: 0.98 | 17:36 |
| GoJkOrS | is that the line? | 17:36 |
| notmyname | GoJkOrS: that's the auditor, not the replicator | 17:36 |
| GoJkOrS | May 28 17:37:14 swift09 object-replicator Object replication complete. (0.00 minutes) | 17:37 |
| GoJkOrS | IC | 17:37 |
| notmyname | that seems...optimistic ;-) | 17:38 |
| GoJkOrS | There is very little in the original cluster. | 17:38 |
| *** kenhui1 has quit IRC | 17:39 | |
| notmyname | GoJkOrS: do you have a particular object you are looking at? | 17:41 |
| GoJkOrS | I am doing a du on the new devices and it shows 0. | 17:42 |
| *** bena has joined #openstack-swift | 17:43 | |
| notmyname | GoJkOrS: if you have a particular object, you can check where it's supposed to be in the cluster with `swift-get-nodes` | 17:43 |
| notmyname | GoJkOrS: then you can check that it's actually there | 17:43 |
| bena | hi I am using s3curl to connect to swift openstack | 17:43 |
| notmyname | GoJkOrS: also, you should check that the rings are all the same on all the servers | 17:43 |
| bena | I run ./s3curl.pl --id 'testing:admin' --key 'intel123' --get -- -k -v -s http://10.1.0.10:8080 | 17:44 |
| bena | but recived an AccessDenied< | 17:44 |
| portante | zaitcev: on F19, looks like a sudo pip install -U xattr busts things up a bit | 17:45 |
| portante | cffi seems broken | 17:45 |
| zaitcev | portante: why would you even do that?! | 17:45 |
| GoJkOrS | ok, thanks for the help. I will keep working on it and will check back if I need any further help. | 17:45 |
| portante | :( | 17:45 |
| portante | following the SAIO directions | 17:45 |
| portante | why wouldn't you do that? | 17:45 |
| bena | do you know the account:user and password of openstack swift | 17:45 |
| bena | ?? | 17:45 |
| zaitcev | I never let pip anywhere near my boxes. | 17:45 |
| *** jamie_h has quit IRC | 17:45 | |
| zaitcev | Well, except under tox | 17:46 |
| zaitcev | I you want to use pip, install Ubuntu. | 17:46 |
| notmyname | bena: it's whatever you have configured in your auth system. there isn't just one. swift is a multi-tenant system | 17:46 |
| portante | zaitcev: perhaps then you can make some recommendations to the SAIO documentation for Fedora to not use pip and provide the proper things there? | 17:47 |
| zaitcev | portante: I keep patching it and then some jerk breaks it again. It was a tug of war since 2010. | 17:47 |
| zaitcev | portante: At least we got rid of that awful "Swift on RHEL" wiki page | 17:48 |
| *** bkopilov has joined #openstack-swift | 17:49 | |
| *** bena_ has joined #openstack-swift | 17:49 | |
| bena_ | what is the default account:user and password swift | 17:50 |
| bena_ | ?? | 17:50 |
| *** bena has quit IRC | 17:50 | |
| notmyname | bena_: what auth system are you using? | 17:50 |
| bena_ | the default auth system | 17:51 |
| notmyname | bena_: there isn't a default. or there is, but only from the perspective of whatever install doc you followed | 17:51 |
| notmyname | bena_: "swift" doesn't have an account or password. users (with credentials) are given access to a swift account. the 401 you are getting is because the credentials you supplied either could not be authenticated or were authenticated but not authorized to access a swift account | 17:52 |
| bena_ | ./s3curl.pl --id <account>:<user> --key <password> --get -- -k -v -s http://10.1.0.10:8080 | 17:52 |
| zaitcev | portante: I'll come back and re-fix the development_saio.rst, however we still have an issue with our xattr that makes PYTHONPATH=$(pwd) required to run ./.unittests. | 17:52 |
| bena_ | how you do to connect with s3curl? | 17:53 |
| zaitcev | portante: I was thinking about creating a fake egg around xattr. | 17:53 |
| notmyname | bena_: I've never used s3curl | 17:53 |
| zaitcev | bet he forgot to add swift3 into proxy pipeline | 17:54 |
| portante | zaitcev: okay, lemme try that | 17:55 |
| portante | that is a bummer | 17:55 |
| zaitcev | portante: Functests work fine as-as, if that helps :-) | 17:55 |
| portante | zaitcev: unfortunately in my environment that pythonpath trick is not workin | 17:55 |
| zaitcev | portante: well, dropping xattr from requirements and python setup.py build help | 17:56 |
| zaitcev | portante: although I don't know what your specific issue is, it could be something else | 17:57 |
| portante | zaitcev: looks like "pip install -U xattr" from SAIO causes a new version of cffi to install, and things break | 17:57 |
| *** lpabon has quit IRC | 17:57 | |
| zaitcev | portante: maybe you need a sticky note with "rpm -e pip" on your laptop's bezel | 17:58 |
| portante | cute ;) | 17:59 |
| zaitcev | By definition pip wreaks havoc with the system and there's no way to go back except a reinstall. That is why we always package everything and do not support pip. | 17:59 |
| zaitcev | By we I mean Red Hat OpenStack team. | 18:00 |
| zaitcev | It has its downsides, like the need to package every new thing Monty throws at us, like pbr. | 18:00 |
| zaitcev | pip is just as bad as Ruby gems | 18:01 |
| *** lpabon has joined #openstack-swift | 18:01 | |
| portante | zaitcev: okay, so this seems to work: pip uninstall cffi xattr; yum install python-cffi pyxattr | 18:01 |
| *** lpabon has quit IRC | 18:01 | |
| * notmyname wishes "use packages" was a more common viewpoint in openstack | 18:01 | |
| zaitcev | It's something a latte-sipping SOMA hippies throw at their devops servers before they cash out of their startup | 18:01 |
| notmyname | zaitcev: I'll have you know I drink mochas! not lattes | 18:01 |
| notmyname | speaking of which.... | 18:02 |
| * notmyname goes to brew another pot of locally roasted, artisinally crafted, fair-trade, shade-grown, etc, etc coffee | 18:03 | |
| portante | I spoke too soon, the unit tests fail with DistributionNotFound:xattr>=0.4 | 18:03 |
| notmyname | (did I get the words right?) | 18:03 |
| zaitcev | Yeah, needs a fake egg or dropping xattr from requirements and python setup.py build | 18:03 |
| portante | ugh | 18:03 |
| portante | really? | 18:03 |
| portante | I think we need a better answer on Fedora | 18:04 |
| *** lpabon has joined #openstack-swift | 18:04 | |
| zaitcev | >> do something about xattr.so and xattr>=4.0: create a fake egg (re 1020449) | 18:04 |
| *** tdasilva has quit IRC | 18:04 | |
| zaitcev | I agree, we do | 18:05 |
| zaitcev | brb, going to visit a knee surgeon | 18:09 |
| portante | zaitcev: best of luck on that | 18:10 |
| *** diegows has joined #openstack-swift | 18:14 | |
| notmyname | knee surgeon? what's up with people working on swift getting hurt? me, mlanner, zaitcev, anyone else? | 18:19 |
| *** mwstorer has joined #openstack-swift | 18:22 | |
| *** elambert has joined #openstack-swift | 18:34 | |
| *** lpabon has quit IRC | 18:34 | |
| clayg | back | 18:36 |
| clayg | lol @ zaitcev's rant on pip | 18:38 |
| briancline | notmyname: my knees are growing increasingly unhappy as well | 18:41 |
| briancline | not something that should be happening at age 30, though it did start a few years ago | 18:42 |
| briancline | my involvement in swift also began a few years ago | 18:42 |
| briancline | huh | 18:42 |
| *** gvernik has joined #openstack-swift | 18:42 | |
| creiht | clayg: http://paste.ronin.io/cd48e891 | 18:45 |
| creiht | I enabled policies on my saio, and I am getting those to errors | 18:45 |
| clayg | creiht: looking | 18:45 |
| notmyname | oh. thank you calendar reminder. swift team meeting in 15 minutes. full meeting: storage policies patches. choo choo! | 18:45 |
| creiht | when I run probetests | 18:45 |
| notmyname | (as in that's it. not as in we'll talk about it more) | 18:45 |
| clayg | creiht: did you run the functests? | 18:46 |
| clayg | creiht: you're on the final patchset? | 18:46 |
| clayg | creiht: did gholt ever get the probetests passing? | 18:47 |
| clayg | creiht: oh nm, I think he was just running them with the skips (no sp configured yet) | 18:47 |
| clayg | creiht: i'm guessing the saio instructions didn't include a "copy etc/swift/container-reconciler.conf-sample into /etc/swift/container-reconciler.conf" :P | 18:48 |
| creiht | lol | 18:48 |
| creiht | no | 18:48 |
| creiht | functests ran fine | 18:49 |
| notmyname | clayg: I'm looking over the docs, but my saio probe tests failed too. I think because of the lack of a reconciler config | 18:49 |
| creiht | I'll copy the reconciler config and try again | 18:49 |
| clayg | creiht: yeah they don't really hit the failure modes | 18:49 |
| creiht | right | 18:49 |
| clayg | creiht: did you mostly use policies_saio | 18:49 |
| creiht | clayg: yeah | 18:49 |
| clayg | creiht: I guess I should check both docs | 18:49 |
| clayg | creiht: k, thanks | 18:49 |
| creiht | I'll leave a note | 18:49 |
| clayg | creiht: awesome! | 18:49 |
| * clayg says "did you even run this?" for creiht | 18:50 | |
| creiht | lol | 18:50 |
| creiht | it is actually going smoother than I expected :) | 18:50 |
| clayg | creiht: oh that's just cause you have't gotten too far into it yet | 18:50 |
| creiht | haha | 18:50 |
| clayg | at least you guys will be able to slap me next week in CO | 18:51 |
| creiht | well honestly I wasn't figuring I would get this far | 18:51 |
| clayg | peluse_: ^ props for decent doc! | 18:51 |
| notmyname | clayg: peluse_: are there different tmp and quarantine directories for different policies? | 18:55 |
| clayg | notmyname: that was a great comment by gholt - i'm going to look at adding that today - should be mostly tucked in DiskFile | 18:56 |
| *** mjfork has joined #openstack-swift | 18:56 | |
| notmyname | kk | 18:56 |
| notmyname | (ya, that's where I got it) | 18:57 |
| mjfork | as a user of 2 different swift clusters, is it possible to run container sync externally? or does it require privileged access from the service provider side? | 18:58 |
| notmyname | mjfork: container-sync itself requires privileged access. but an external process to poll + GET/PUT would be functionally equivalent (or close). | 18:59 |
| gvernik | a question. Each patch related storage policies depends on another patch which depends on another patch and so on... When i do for example something like that "git fetch https://review.openstack.org/openstack/swift refs/changes/40/96040/1 && git checkout FETCH_HEAD" will it automatically fetch all dependencies? | 19:00 |
| notmyname | mjfork: and whenever we have marconi (or some other distibuted, durable queue), we can make that much more flexible and efficient :-) | 19:00 |
| cschwede_ | gvernik: yes, and you should use patch 96049 - it’s the last patch | 19:00 |
| notmyname | gvernik: if you get the last patch in the chain, you will get the whole chain | 19:00 |
| notmyname | gvernik: and you can use `git review -d <gerrit patch number>` to automatically make a branch for it and download it | 19:01 |
| gvernik | thanks | 19:01 |
| mjfork | notmyname: thanks. | 19:01 |
| *** tdasilva has joined #openstack-swift | 19:02 | |
| creiht | clayg: getting closer | 19:03 |
| creiht | now I just have this error with probe tests | 19:03 |
| creiht | http://paste.ronin.io/4bf1105f | 19:03 |
| cschwede_ | anyone saw an error in testContainerListing (test.functional.tests.TestAccount) ? | 19:04 |
| clayg | creiht: hrmm... that should be the container-replicator | 19:04 |
| creiht | hrm | 19:06 |
| creiht | for saio | 19:06 |
| creiht | do I need just one container-reconciler.conf | 19:06 |
| clayg | creiht: can you sanity check swift-init container-replicator once -nv | 19:06 |
| creiht | or do I need a separate one for each | 19:06 |
| clayg | creiht: just the one | 19:06 |
| creiht | k | 19:06 |
| clayg | but the merging of storage policy index happens in the container-replicator - the reconciler is just a queue chewer | 19:07 |
| creiht | clayg: running that seems to run without error | 19:08 |
| creiht | but of course the test data will be gone since it resets for each test | 19:08 |
| clayg | creiht: well you can run just the one test "python ./swift/test/probe/test_container_merge_storage_policy_index.py -x" and it will halt leaving the environment messed up for you to inspect | 19:09 |
| creiht | k | 19:09 |
| clayg | or "...py:TestContainerMergePolicyIndex.test_merge_storage_policy_index" if you wanna get real specific | 19:10 |
| creiht | do you mean -v instead of -x? | 19:11 |
| clayg | well, -x says "stop at the first failure" | 19:11 |
| creiht | ahh | 19:11 |
| clayg | if you just run that one test then you don't really need it | 19:11 |
| creiht | well -x is for nosetests and not python right? | 19:11 |
| clayg | creiht: oh... yeah run it with nose them :P | 19:12 |
| creiht | either way running that one test | 19:12 |
| creiht | :) | 19:12 |
| creiht | hrm running that one test by iteself passed | 19:13 |
| clayg | so looking at the timestamps in that paste those containers should have settled down... i would say maybe it was a weird timing/ordering thing... but we run the replicator twice to make sure that isn't really an issue | 19:13 |
| clayg | "we" - get_to_final_state does that | 19:14 |
| *** mkollaro has joined #openstack-swift | 19:15 | |
| clayg | creiht: oh we're supposed to be in that other channel | 19:16 |
| creiht | heh | 19:16 |
| notmyname | no, you're not :-) | 19:16 |
| notmyname | you're doing what you're supposed to be doing right here :-) | 19:16 |
| creiht | lol | 19:17 |
| creiht | ok running tests on just the .py by itself worked | 19:17 |
| creiht | running full thing again | 19:17 |
| clayg | oh, that was quick | 19:18 |
| creiht | heh | 19:21 |
| creiht | clayg: ok now all tests pass | 19:21 |
| creiht | maybe there is a rare race condition there? | 19:21 |
| clayg | creiht: well, i stick it in a for loop and see if I can get it to pop | 19:22 |
| gholt | I thought probe = "possible race obe" Okay, I'm not good at backronyms. | 19:22 |
| creiht | lol | 19:22 |
| creiht | yeah probably not a big deal | 19:22 |
| creiht | unless it happens a lot | 19:23 |
| creiht | passed again | 19:26 |
| creiht | meh | 19:26 |
| creiht | :) | 19:26 |
| clayg | creiht: i think once is too much, i say we cancel storage policies | 19:26 |
| creiht | so looks like I got saio setup with policies | 19:26 |
| creiht | haha | 19:26 |
| gholt | creiht: How many unit tests do you skip per run atm? | 19:27 |
| creiht | running again to verify | 19:27 |
| gholt | Cool, thanks! 72.002s here btw | 19:27 |
| gholt | Er, 74.002s that is.. | 19:28 |
| clayg | gholt: unittests should be able to mock out storage polices anytime they need more than one... i don't know why any unittests would skip short of something trying to be cleaver with a dependency that's tricky to build or something? | 19:28 |
| *** tdasilva has quit IRC | 19:29 | |
| creiht | gholt: 0 skipped in 85.171s | 19:29 |
| gholt | I'm just wondering if I don't have something set up right so it's not testing or something. Oh, I gotcha. Yeah, unit tests shouldn't care about my set up I guess. | 19:29 |
| gholt | I skip 3 unit tests and 3 func tests. each runs about the same, ~74s. | 19:30 |
| clayg | so yeah I think some functests should skip if the system under test isn't exposing more than one storage policy | 19:30 |
| clayg | well.. there's lots of functests that may skip | 19:30 |
| creiht | I think I only skip 1 func test | 19:30 |
| gholt | I've also got some skpping domain name something something and something, so... | 19:30 |
| clayg | gholt: if you can track down the names of those skipped unittests I'd love to know why | 19:30 |
| creiht | but there were a lot more skipped when I didn't have a policy | 19:31 |
| gholt | Yeah, I'll double check what I'm still skipping to be sure. | 19:31 |
| gholt | -v | 19:31 |
| clayg | creiht: a lot more functests skipped? | 19:31 |
| creiht | yeah | 19:31 |
| creiht | I *think* | 19:31 |
| clayg | yeah i don't know how many ended up getting in | 19:31 |
| creiht | :) | 19:31 |
| clayg | and I think it may like skip the *class* or something? so i don't know how many S's you get | 19:31 |
| creiht | which is reasonable to me | 19:31 |
| creiht | just double checked, I'm not skipping any functests currently | 19:33 |
| creiht | and 76.474s :) | 19:33 |
| gholt | Ooh, I found another type toeken. I feel like I'm contributing! | 19:34 |
| gholt | Lol, another typo (with my typo) | 19:34 |
| creiht | heh | 19:34 |
| notmyname | gholt: don't worry. someday you'll be a useful contributor to swift ;-) | 19:34 |
| gholt | :) | 19:35 |
| notmyname | (the ";-)" means I'm laughing, if that doesn't translate in text--text-mode jokes are awesome!) | 19:35 |
| gholt | It'd almost be better if I misunderstood and got all in a huff. Heheheh | 19:35 |
| clayg | lol | 19:36 |
| *** tdasilva has joined #openstack-swift | 19:37 | |
| *** elambert has quit IRC | 19:38 | |
| notmyname | hmm...probetests pass (with one skip) after I run `python ./setup.py develop`. wish I would have thought of that first | 19:39 |
| clayg | notmyname: probably the expirer or container-sync or something | 19:40 |
| notmyname | clayg: so, FYI on my SAIO, unit, func, and probe tests all pass. with one skip in probe tests | 19:40 |
| notmyname | clayg: perhaps. will the -v/-x thing work to find it? | 19:40 |
| creiht | notmyname: yeah that is the same here | 19:40 |
| clayg | yeah i'm glad tests are mostly working for folks - sans creiht weird thing, and whatever is going on with portante's setup | 19:40 |
| clayg | notmyname: umm.... yeah -v will show the names, maybe a bunch of other stuff... | 19:41 |
| gholt | clayg: Yeah, all my 3 skipped func tests are due to me not have object versioning enabled. | 19:41 |
| notmyname | *cough*RHEL*cough* | 19:41 |
| gholt | I should probably enable that or something. | 19:41 |
| clayg | gholt: it's on by default in vagrant-swift-all-in-one, also has built in storage policy support and makes it easy to toggle features in configs on a saio because of conf.d | 19:42 |
| clayg | heh | 19:42 |
| clayg | gholt: oh and the autodoc script is sorta handy-ish (i obvioulsy didn't run it it or i would have seen all those build errors) | 19:43 |
| gholt | :) I'm not sure why I don't have it enabled or how I did that or how to undo it. But i'll learn | 19:43 |
| clayg | gholt: but better late than never? | 19:44 |
| clayg | something in like the container-server configs somehow? | 19:44 |
| gholt | Set ``allow_versions`` to ``True`` in the container server config | 19:44 |
| gholt | Right there in the doc/source/overview_object_versioning.rst :) | 19:44 |
| *** diegows has quit IRC | 19:47 | |
| *** jogo has left #openstack-swift | 19:48 | |
| gholt | Cool, now 0 skipped func tests. I feel complete. Thank you all. | 19:49 |
| notmyname | ah, yes. "SKIP: Unable to find container sync realms" | 19:50 |
| notmyname | gholt: copy etc/container-sync-realms.conf-sample to /etc/swift/ ? | 19:51 |
| notmyname | and rename | 19:51 |
| gholt | Er, I don't think that'd work but it might. Mine has: | 19:52 |
| gholt | $ cat /etc/swift/container-sync-realms.conf | 19:52 |
| gholt | [realm1] | 19:52 |
| gholt | key = abc | 19:52 |
| gholt | cluster_name1 = http://127.0.0.1:8080/v1/ | 19:52 |
| *** diegows has joined #openstack-swift | 19:52 | |
| clayg | notmyname: i'd be it's all right there in doc/source/overview_container_sync.rst :P | 19:53 |
| clayg | notmyname: I'd also bet that vagrant-swift-all-in-one has built in support | 19:53 |
| notmyname | lol | 19:53 |
| clayg | yup, add container-sync-realms was the last commit ;) | 19:54 |
| gholt | Heheh | 19:54 |
| *** bkopilov has quit IRC | 19:55 | |
| creiht | clayg: also, probably should ad DocImpact to all the commits :) | 20:00 |
| clayg | creiht: #willfix | 20:00 |
| creiht | hehe | 20:00 |
| creiht | but the storage policies basics seem to be working on my saio | 20:00 |
| creiht | left some notes on policies_saio.rst | 20:01 |
| creiht | goint to take a break for a bit | 20:01 |
| clayg | no no no don't stop - it gets better! | 20:03 |
| creiht | haha | 20:04 |
| *** bkopilov has joined #openstack-swift | 20:14 | |
| peluse_ | been out all morning... looks like we've started rocking on SP | 20:18 |
| clayg | does DocImpact go above or below the change-id and the implements bp thing? | 20:18 |
| notmyname | clayg: change-id has to be in the last paragraph. I'd put it right after the Implements line | 20:20 |
| notmyname | it == "DocImpact" | 20:20 |
| clayg | do late i'm going DocImpact\nImplements\nChange-Id | 20:20 |
| clayg | s/do/too/ | 20:20 |
| notmyname | that's ok too :-) | 20:20 |
| peluse_ | clayg: have you started doing anything with the SP doc comments? I will start addressing them now if not | 20:20 |
| clayg | peluse_: yeah i think i'm all caught up | 20:21 |
| clayg | peluse_: can you review yuan's changes or see about adding get_policy_string support to the /tmp and /quarantine dirs | 20:21 |
| peluse_ | clayg: way cool. yes, I can do both - consider them done! | 20:22 |
| peluse_ | clayg: quesiton though, for the /tmp and /q dir deal, do I propose that to master and hunt down the right current patch in the proposed chain that has the base function and use that as a dependency as the best way to proceed? | 20:28 |
| *** elambert has joined #openstack-swift | 20:30 | |
| *** gvernik has quit IRC | 20:30 | |
| *** mjfork has quit IRC | 20:35 | |
| peluse_ | clayg: or maybe better yet I could just push on patch https://review.openstack.org/#/c/96027/ | 20:41 |
| creiht | so with swift-get-nodes, you have to know what ring the object is in | 20:43 |
| creiht | right? | 20:43 |
| notmyname | creiht: I think yuan just proposed a WIP patch (to master?) this morning for that | 20:43 |
| creiht | oh cool | 20:44 |
| notmyname | ah, to feature/ec https://review.openstack.org/#/c/82734/ | 20:44 |
| peluse_ | https://review.openstack.org/#/c/82734/ | 20:44 |
| peluse_ | duh, sorry | 20:44 |
| notmyname | :-) | 20:44 |
| creiht | so with that, you will still need to know what storage policy an object is in right | 20:45 |
| peluse_ | I'll check with yuan tonight and see what the WIP is | 20:45 |
| peluse_ | correct | 20:45 |
| creiht | which an operator isn't likely to know | 20:45 |
| peluse_ | heh, maybe that's the WIP :) Good point though | 20:45 |
| creiht | just based on the url | 20:46 |
| peluse_ | should be able to get it from the cont in the url, no? | 20:46 |
| creiht | so it seems to me either swift-get-nodes should be smart enough to look up what policy the container is | 20:46 |
| creiht | or have a tool to easily query policies for a given container | 20:46 |
| peluse_ | that | 20:46 |
| peluse_ | that's what I'm thinking, but I haven't looked at this patch yet. I'll get Yuan the feedback though | 20:47 |
| notmyname | ya. the nice thing about swift-get-nodes now is that it's completely offline | 20:47 |
| creiht | well I guess swift-container-info could show that | 20:47 |
| creiht | notmyname: yeah I know | 20:47 |
| creiht | that's why I'm kinda waffling on it | 20:47 |
| *** bena_ has quit IRC | 20:49 | |
| clayg | peluse_: just make the change to feature/ec and I'll pull it into the chain? | 20:50 |
| peluse_ | clayg: roger | 20:51 |
| clayg | creiht: swift-container-info has the policy in it, so on a saio you can for db in $(find /srv/node*/*/container* -name \*.db); do swift-container-info $db | grep -i policy; done | 20:52 |
| clayg | creiht: otherwise you can hack something together with direct_client | 20:52 |
| creiht | yeah | 20:52 |
| creiht | I can just see it confusing for operators | 20:52 |
| clayg | creiht: swift-get-nodes always requires you "know the ring" depending on account/container/object-N - it's sorta weak for that | 20:52 |
| creiht | yeah I know | 20:52 |
| clayg | creiht: not sure I follow, but i'm open to suggestions | 20:52 |
| creiht | yeah I don't have a good answer either | 20:53 |
| clayg | creiht: yaun and peluse_ are working on it - I think we can make the tools better | 20:53 |
| creiht | cool | 20:53 |
| creiht | I'll wait to see what they come up with | 20:53 |
| clayg | creiht: are you trying to solve for like "given some number of ring locations what do the containers there think their storage policy is at the moment"? | 20:56 |
| *** miqui has quit IRC | 20:56 | |
| creiht | clayg: no, I'm thinking that I get a url from someone that is having issues | 20:56 |
| creiht | so I run a swift-get-nodes to figure out where the object is supposed to be | 20:57 |
| clayg | ah yeah - and that has to talk to the container servers | 20:57 |
| creiht | yeah I don't know what storage policy an object is | 20:58 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 21:00 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 21:00 |
| *** tdasilva has left #openstack-swift | 21:05 | |
| *** gyee has quit IRC | 21:14 | |
| *** shri has quit IRC | 21:25 | |
| *** diegows has quit IRC | 21:38 | |
| *** Midnightmyth has quit IRC | 21:42 | |
| peluse_ | clayg: just noticed we'll have issues with > single digit policy indixies in a few places... I'll knock that out too and add a test case | 21:43 |
| * clayg is trying to imagine where/how we backed in the assumption that polices will only be one digit | 21:44 | |
| notmyname | peluse_: simple! make sure they are base-36 and you've got enough storage policies for everyone! | 21:46 |
| notmyname | peluse_: from a comment "Are there any rules about what are legal policy names?" <<-- is this "utf8 string less than a few k long" since it has to be in headers? | 21:48 |
| * notmyname is imagining a storage policy with a name 3k bytes long | 21:49 | |
| clayg | notmyname: sounds like a loaded question | 21:50 |
| notmyname | the one in the comments or mine? :-) | 21:50 |
| clayg | the comment - like the next question is going to be "but what if someone creates a name like..." | 21:51 |
| peluse_ | hehe | 21:52 |
| peluse_ | FYI wrt the policy index issue its just how we're doing some parsing of dir names to extract the index... nothing to do with names on that one | 21:52 |
| notmyname | heh. the "yomamma" policy. it just disables the 413 response | 21:53 |
| clayg | oh... hrmm... do we regex the dir listing or something? | 21:53 |
| peluse_ | base, policy_idx = dir.split('-', 1) | 21:53 |
| clayg | seems legit | 21:53 |
| clayg | the 1 is the number of times to split | 21:54 |
| clayg | so you get object, 123 == 'object-123'.split('-', 1) | 21:54 |
| clayg | so you get object, -123 == 'object--123'.split('-', 1) | 21:54 |
| peluse_ | oh cool, was just reading it wrong then. never mind :) | 21:54 |
| clayg | instead of say 'object', '', '123' == 'object--123'.split('-') or something like that | 21:55 |
| clayg | I like the 413 HTTP status on the yo momma joke - most people go for 402 | 21:57 |
| notmyname | hmm...I just learned that there is a 431 response code specifically for headers being too big | 21:57 |
| clayg | notmyname: invinted to keep keystone in check no doubt | 21:58 |
| *** ekarlso has quit IRC | 22:01 | |
| notmyname | lol | 22:02 |
| notmyname | clayg: funny coincidence on that. the rfc that defined that response code was written by mark nottingham. it was accepted in 2012, which is about the same time RAX was doing lots of keystone shenanigans (IIRC). and at the time, mark was employed by rackspace. | 22:05 |
| notmyname | so it seems perfectly logical that the specific response header is a result of keystone ;-) | 22:05 |
| * notmyname is bad at writing conspiracy theories | 22:08 | |
| gholt | Heheh, so hopefully by tomorrow morning we'll all be done with the spell/grammar checking. | 22:16 |
| *** ekarlso has joined #openstack-swift | 22:26 | |
| peluse_ | sheeeeit | 22:30 |
| peluse_ | oops, wrong window :) | 22:31 |
| briancline | the judges will still allow it | 22:31 |
| clayg | gholt: if the docs look good and the tests pass we're basically ready to merge right? | 22:33 |
| notmyname | I don't know if it's worth it at this point. POLICY and POLICY_INDEX are header names | 22:40 |
| dfg | clayg: just made a non-doc related comment- you just wait... | 22:40 |
| clayg | yay yay yay! | 22:41 |
| clayg | notmyname: i was pretty sure at somepoint that those needed to change, but then I think i realized I didn't care. | 22:41 |
| clayg | notmyname: they do end up looking funny in tests next to a bunch of other headers | 22:41 |
| notmyname | clayg: ya, that's my only nit with them. it is a nit. but it's something I've had to recheck every time I've looked at the feature/ec branch | 22:42 |
| clayg | notmyname: I think we probably could get away with just using literal strings and trying not to typo - but having flake8 check it for you is nice too | 22:42 |
| clayg | notmyname: it must go away after you spend some time with it - still no time like the present if we wanna go hog wild with a search and replace? | 22:43 |
| notmyname | I'm not convinced that's it's worth it. let me get deeper into the patch chain and see what happens | 22:45 |
| clayg | dfg: i can't wait any more - submit it now! | 22:45 |
| dfg | clayg: i think i built it up too much. I'm just glad I can sit in my beanbag to get through these reviews | 22:50 |
| clayg | peluse_: notmyname: do we get "I survived the storage policy merge" t-shirts? | 22:53 |
| * clayg somehow still thinks we're all going to survive | 22:54 | |
| *** Edward-Zhang has joined #openstack-swift | 22:55 | |
| clayg | peluse_: yuan's patches are looking good - I think i took the container-auditor-migration thing and probably the non-zero-default-probetest fix, i'm testing deprecate storage policy support change now | 22:55 |
| notmyname | clayg: http://imgur.com/NdO5Jq2 | 22:56 |
| *** bkopilov has quit IRC | 22:59 | |
| peluse_ | clayg: cool, I'm struggling with some VM issues with my SAIO so am moving slow right now (literally). Argh | 23:00 |
| clayg | peluse_: vagrant destroy -f; vagrant up | 23:00 |
| clayg | go get coffee and come back to a fully configured working saio | 23:01 |
| clayg | there's been studies... 95% of the time it works everytime | 23:01 |
| *** bkopilov has joined #openstack-swift | 23:13 | |
| peluse_ | clayg: thx. not sure if my vbox is falling down or I might have an SSD issue with my laptop... great timing | 23:13 |
| clayg | awww that sucks - sorry bro | 23:13 |
| *** ozialien has joined #openstack-swift | 23:19 | |
| clayg | pandemicsyn: i don't know what any of this means, but I *do* hate load-balancers so can I do this instead -> http://bits.shutterstock.com/2014/05/22/stop-buying-load-balancers-and-start-controlling-your-traffic-flow-with-software/ | 23:19 |
| peluse_ | clayg: best I can do on the shirt is a free generic Intel t-shirt next week but I can bring a sharpie with me :) | 23:20 |
| *** kragniz has quit IRC | 23:32 | |
| clayg | yuan: peluse_: why is it a good idea to make it so you can't deprecate policy-0? | 23:38 |
| clayg | yuan: peluse_: I guess you'd have a different way to enfoce a default... if you can't guarantee zero is available | 23:39 |
| clayg | idk, maybe it'd acctually be ok if the default policy was deprecated - it would just force people to choose a policy on purpose when creating a container | 23:40 |
| *** kragniz has joined #openstack-swift | 23:44 | |
| *** mkollaro has quit IRC | 23:45 | |
| *** gyee has joined #openstack-swift | 23:53 | |
| *** foexle has joined #openstack-swift | 23:56 | |
| *** bkopilov has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!