| notmyname | clayg: mattoliverau: their twitter feed confirmed it. as did kemp. and vish. and our swiftstack employees who used to work there | 00:00 |
|---|---|---|
| notmyname | clayg: mattoliverau: assuming it were fake, it would be _terrible_ for their sales pipeline. they wouldn't joke about that | 00:01 |
| mattoliverau | peluse: I assume a similar place will exist, there is a place holder for the summit. https://wiki.openstack.org/wiki/Design_Summit/Liberty usually (last time) we had our own etherpad to gather ideas, and then created the etherpads (on the kilo link) once we had a plan on what we wanted to talk about in each session | 00:02 |
| mattoliverau | notmyname: true, well there you go then.. just a bad day to annouce.. but I guess sometimes you can't control that. Looks like Matt Garrett has moved to coreOS. | 00:03 |
| torgomatic | mattoliverau: glad to hear it | 00:03 |
| clayg | torgomatic: how's this for blow your mind - the proxy diff is smaller than the diskfile diff (!?) | 00:03 |
| torgomatic | clayg: uh | 00:03 |
| clayg | i think it was all the diskfile tests | 00:04 |
| clayg | or maybe the suffix hashes | 00:04 |
| notmyname | mattoliverau: peluse: so far, the etherpad is at https://wiki.openstack.org/wiki/Design_Summit/Planning | 00:04 |
| notmyname | mattoliverau: peluse: however, this is something we need to discuss | 00:04 |
| notmyname | remember earlier at the meeting when I said I didn't want to get too distracted by it this week? ;-) | 00:04 |
| clayg | probably the tests - we reogranized the diskfile test module like *a lot* | 00:04 |
| notmyname | so we need to decide if we use etherpads like last time or if we use the old system that we used to have to propose talks | 00:05 |
| notmyname | and that's something I wanted to discuss more at the next meeting | 00:05 |
| clayg | torgomatic: yeah it was tests - like 4.5K lines just in test_diskfile | 00:06 |
| mattoliverau | notmyname: I remember that's fine :) | 00:06 |
| mattoliverau | We still have plenty of time :) | 00:06 |
| clayg | hrmm... yeah our ratio on tests in the proxy change isn't very good for all the lines we added to obj/controller :\ | 00:06 |
| clayg | i mean it's like 2:1 tests to code instead of 4:1 | 00:06 |
| clayg | so isn't very good ~= not super insanely over done right | 00:07 |
| notmyname | :-) | 00:10 |
| notmyname | IIRC storage policies were something like 3:1 tests to code | 00:10 |
| notmyname | ok, with an eye to ec_review landing, and based on what I said at the meeting about what else is in the upcoming release, I starred some more stuff in gerrit | 00:17 |
| notmyname | when clayg proposes the ec_review chain, I'll update it | 00:17 |
| *** thumpba_ has quit IRC | 00:17 | |
| openstackgerrit | Samuel Merritt proposed openstack/swift: EC: support multiple ranges for GET requests https://review.openstack.org/166576 | 00:24 |
| *** dmorita has joined #openstack-swift | 00:25 | |
| *** erlon has quit IRC | 00:31 | |
| *** rdaly2 has joined #openstack-swift | 00:33 | |
| *** annegentle has joined #openstack-swift | 00:38 | |
| *** annegentle has quit IRC | 00:43 | |
| *** haigang has joined #openstack-swift | 00:59 | |
| *** logan2 has quit IRC | 01:19 | |
| *** bkopilov has quit IRC | 01:20 | |
| *** logan2 has joined #openstack-swift | 01:21 | |
| *** kota_ has joined #openstack-swift | 01:23 | |
| kota_ | good morning, again | 01:24 |
| mattoliverau | kota_: moring...again | 01:24 |
| clayg | ok, neat - so you know how PUT with x-copy-from works between ec and repl because of that trick we do in GETorHEAD to recreate the controller instead of calling into self | 01:24 |
| clayg | we don't do that in copy - and up until now I don't think I knew this was broken | 01:24 |
| clayg | which is .... werid .... because functests caught it - the versioning middleware tests acctually | 01:25 |
| *** annegentle has joined #openstack-swift | 01:39 | |
| *** annegentle has quit IRC | 01:44 | |
| clayg | charz: i'm not seeing any of our community cluster jobs on any new-ish reviews? | 01:48 |
| charz | clayg: checking | 01:49 |
| clayg | yeah sorry i don't remember how to troubleshoot | 01:49 |
| clayg | .... err well like that system :P | 01:49 |
| clayg | is there like a wiki page some basics? I'm not sure how much of it is really public infrastructure vs maybe some stuff we have internally? | 01:50 |
| charz | clayg: There is 21 jobs in the queue and jenkin is going to consume it. | 01:51 |
| *** panbalag has joined #openstack-swift | 02:05 | |
| *** haomaiwang has joined #openstack-swift | 02:15 | |
| *** panbalag has quit IRC | 02:20 | |
| openstackgerrit | paul luse proposed openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 02:32 |
| openstackgerrit | paul luse proposed openstack/swift: EC Reconstructor Probe Tests https://review.openstack.org/164291 | 02:33 |
| openstackgerrit | paul luse proposed openstack/swift: Fix ssync sender behavior for EC reconstructor jobs https://review.openstack.org/169052 | 02:33 |
| *** jrichli has quit IRC | 02:36 | |
| *** annegentle has joined #openstack-swift | 02:40 | |
| *** annegentle has quit IRC | 02:45 | |
| *** ujjain has quit IRC | 02:46 | |
| peluse | went through the cards on trello and updated a few - a few needs updates from others as well so if you're working on EC stuff please check/update your cards - thanks!! | 03:02 |
| *** rdaly2 has quit IRC | 03:12 | |
| openstackgerrit | paul luse proposed openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 03:17 |
| openstackgerrit | paul luse proposed openstack/swift: EC Reconstructor Probe Tests https://review.openstack.org/164291 | 03:18 |
| openstackgerrit | paul luse proposed openstack/swift: Fix ssync sender behavior for EC reconstructor jobs https://review.openstack.org/169052 | 03:19 |
| openstackgerrit | Matthew Oliver proposed openstack/swift: EC: support multiple ranges for GET requests https://review.openstack.org/166576 | 03:19 |
| peluse | yeah baby! | 03:20 |
| mattoliverau | torgomatic: ^^ found a slight difference with ranged get output when comparing EC and REPL.. there is a quick fix | 03:20 |
| *** silor has joined #openstack-swift | 03:24 | |
| *** h_m has quit IRC | 03:31 | |
| clayg | sooooo.... any ideas on COPY | 03:38 |
| clayg | i'm not having any luck *not* getting failing functests on ec_review or feature/ec | 03:38 |
| clayg | for the version middleware stuff | 03:38 |
| clayg | oh weird, it only fails... sometimes? that's got to be worse. | 03:39 |
| *** annegentle has joined #openstack-swift | 03:41 | |
| *** h_m has joined #openstack-swift | 03:43 | |
| clayg | ok, i'm confused COPY /path/to/ec -H destination /path/to/repl - works | 03:43 |
| clayg | COPY /path/to/repl -H destination: /path/to/ec - 503's | 03:44 |
| clayg | the error in the end is from the object server complaining aobut not having a frag index to stick on the data file | 03:44 |
| clayg | ... but it's making all the connections - maybe it *is* in the right controller? | 03:45 |
| *** gvernik has joined #openstack-swift | 03:46 | |
| clayg | oh, that's probably just cause the replication controller is still picking the right ring :\ | 03:49 |
| clayg | yeah that's all that is | 03:49 |
| *** annegentle has quit IRC | 03:50 | |
| clayg | and so why does a put to the repl policy work if we're in the ec controller | 03:50 |
| clayg | oh it's all that putter if policy.requires footers and junk | 03:51 |
| *** jamielennox is now known as jamielennox|away | 03:51 | |
| *** gvernik has quit IRC | 03:54 | |
| *** annegentle has joined #openstack-swift | 04:02 | |
| *** ho has quit IRC | 04:03 | |
| clayg | well that was easy -> https://gist.github.com/clayg/17e50bccc378633c53e0 | 04:03 |
| clayg | ^ someone should care about applything that to feature/ec maybe? | 04:05 |
| *** Bsony has joined #openstack-swift | 04:13 | |
| clayg | works all the way up through the proxy -> https://review.openstack.org/#/c/169985/ | 04:21 |
| clayg | I think I did it right | 04:21 |
| *** bkopilov has joined #openstack-swift | 04:30 | |
| *** gvernik has joined #openstack-swift | 04:38 | |
| *** km has joined #openstack-swift | 04:45 | |
| *** gvernik has quit IRC | 04:48 | |
| *** Bsony has quit IRC | 05:11 | |
| *** zaitcev has quit IRC | 05:12 | |
| *** zhill has joined #openstack-swift | 05:19 | |
| *** thumpba has joined #openstack-swift | 05:23 | |
| *** thumpba has quit IRC | 05:26 | |
| mattoliverau | clayg: did you want me to raise your copy fix as a patch? | 05:38 |
| *** annegentle has quit IRC | 05:39 | |
| mattoliverau | I'm about to disappear (in 30 mins or so) to go camping for the long weekend, but am happy to raise it.. just patched locally and running tests now | 05:39 |
| *** annegentle has joined #openstack-swift | 05:40 | |
| *** annegentle has quit IRC | 05:45 | |
| *** mmcardle has quit IRC | 05:46 | |
| clayg | mattoliverau: nah fuck it | 05:51 |
| clayg | i'm over feature/ec - ec_review is where it's at | 05:51 |
| clayg | acoles_away: peluse: when you guys get up just merge *all* that reconstructor/ssync stuff - it'll be up on ec_review tomorrow evening and we can keep beating on it there. | 05:52 |
| clayg | ok - well *jenkins* likes ec_review anyway - ya'll get on it - i'm out! | 05:54 |
| mattoliverau | clayg: night man, happy easter | 05:54 |
| openstackgerrit | Matthew Oliver proposed openstack/swift: Make sure COPY requests are routed correctly https://review.openstack.org/170003 | 05:56 |
| mattoliverau | Too bad there it is ^^ adandon it if you want to, but at least now it's in the system :P | 05:57 |
| *** haigang has quit IRC | 06:02 | |
| *** ppai has joined #openstack-swift | 06:07 | |
| *** haigang has joined #openstack-swift | 06:12 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:23 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:23 | |
| *** annegentle has joined #openstack-swift | 06:41 | |
| cschwede | Good morning! | 06:41 |
| *** annegentle has quit IRC | 06:46 | |
| mattoliverau | morning cschwede | 06:54 |
| *** chlong has quit IRC | 06:55 | |
| *** geaaru has joined #openstack-swift | 07:03 | |
| *** Bsony has joined #openstack-swift | 07:10 | |
| *** jistr has joined #openstack-swift | 07:13 | |
| *** jordanP has joined #openstack-swift | 07:14 | |
| *** rdaly2 has joined #openstack-swift | 07:18 | |
| *** rdaly2 has quit IRC | 07:23 | |
| *** haigang has quit IRC | 07:30 | |
| *** haigang has joined #openstack-swift | 07:35 | |
| *** annegentle has joined #openstack-swift | 07:42 | |
| *** annegentle has quit IRC | 07:47 | |
| *** zhill has quit IRC | 07:48 | |
| *** acoles_away is now known as acoles | 07:49 | |
| acoles | good morning | 07:49 |
| acoles | mattoliverau: have fun camping | 07:55 |
| *** haigang has quit IRC | 07:56 | |
| *** Bsony has quit IRC | 07:58 | |
| *** jordanP has quit IRC | 07:59 | |
| *** vinsh has quit IRC | 08:01 | |
| *** haigang has joined #openstack-swift | 08:02 | |
| *** silor has quit IRC | 08:04 | |
| *** krykowski has joined #openstack-swift | 08:11 | |
| *** haigang has quit IRC | 08:24 | |
| *** vinsh has joined #openstack-swift | 08:24 | |
| *** haigang has joined #openstack-swift | 08:28 | |
| *** krykowski has quit IRC | 08:36 | |
| *** jordanP has joined #openstack-swift | 08:37 | |
| *** krykowski has joined #openstack-swift | 08:39 | |
| *** annegentle has joined #openstack-swift | 08:43 | |
| *** annegentle has quit IRC | 08:48 | |
| *** zhill has joined #openstack-swift | 08:48 | |
| *** zhill has quit IRC | 08:53 | |
| *** Bsony has joined #openstack-swift | 08:58 | |
| *** joeljwright has joined #openstack-swift | 09:00 | |
| *** Bsony has quit IRC | 09:03 | |
| *** krykowski has quit IRC | 09:12 | |
| *** krykowski has joined #openstack-swift | 09:13 | |
| *** rdaly2 has joined #openstack-swift | 09:20 | |
| *** haigang has quit IRC | 09:24 | |
| *** rdaly2 has quit IRC | 09:24 | |
| *** jistr has quit IRC | 09:26 | |
| *** dmorita has quit IRC | 09:31 | |
| *** haigang has joined #openstack-swift | 09:40 | |
| *** annegentle has joined #openstack-swift | 09:43 | |
| *** jistr has joined #openstack-swift | 09:44 | |
| *** annegentle has quit IRC | 09:48 | |
| *** zhill has joined #openstack-swift | 09:49 | |
| *** zhill has quit IRC | 09:54 | |
| *** bobby2 has quit IRC | 09:57 | |
| *** bobby2 has joined #openstack-swift | 09:58 | |
| *** mlanner has quit IRC | 10:05 | |
| *** remix_tj has quit IRC | 10:06 | |
| *** remix_tj has joined #openstack-swift | 10:06 | |
| *** silor has joined #openstack-swift | 10:10 | |
| *** mlanner has joined #openstack-swift | 10:11 | |
| *** haomaiwang has quit IRC | 10:37 | |
| *** annegentle has joined #openstack-swift | 10:44 | |
| *** annegentle has quit IRC | 10:49 | |
| *** zhill has joined #openstack-swift | 10:50 | |
| *** haigang has quit IRC | 10:55 | |
| *** zhill has quit IRC | 10:55 | |
| *** rdaly2 has joined #openstack-swift | 11:21 | |
| *** haigang has joined #openstack-swift | 11:23 | |
| *** rdaly2 has quit IRC | 11:26 | |
| *** joeljwright has quit IRC | 11:27 | |
| *** km has quit IRC | 11:30 | |
| *** links has joined #openstack-swift | 11:33 | |
| *** annegentle has joined #openstack-swift | 11:45 | |
| *** annegentle has quit IRC | 11:50 | |
| *** zhill has joined #openstack-swift | 11:51 | |
| *** tongli has joined #openstack-swift | 11:52 | |
| *** zhill has quit IRC | 11:56 | |
| *** panbalag has joined #openstack-swift | 11:59 | |
| *** kota_ has quit IRC | 12:00 | |
| *** kei_yama has quit IRC | 12:03 | |
| acoles | peluse: fyi i have a few changes on 169052, lets touch base before merging with ec recon. bbiab getting lunch | 12:13 |
| *** bkopilov has quit IRC | 12:15 | |
| *** haomaiwang has joined #openstack-swift | 12:26 | |
| *** ppai has quit IRC | 12:29 | |
| *** annegentle has joined #openstack-swift | 12:32 | |
| *** annegentle has quit IRC | 12:33 | |
| *** annegentle has joined #openstack-swift | 12:39 | |
| *** ujjain has joined #openstack-swift | 12:48 | |
| *** zhill has joined #openstack-swift | 12:52 | |
| *** zhill has quit IRC | 12:56 | |
| *** chlong has joined #openstack-swift | 13:01 | |
| *** SkyRocknRoll has quit IRC | 13:07 | |
| *** joeljwright has joined #openstack-swift | 13:07 | |
| *** Gues_____ has joined #openstack-swift | 13:07 | |
| *** annegentle has quit IRC | 13:09 | |
| *** annegentle has joined #openstack-swift | 13:20 | |
| *** joeljwright has quit IRC | 13:20 | |
| *** fanyaohong has joined #openstack-swift | 13:22 | |
| *** annegentle has quit IRC | 13:23 | |
| *** annegentle has joined #openstack-swift | 13:23 | |
| *** Gues_____ has quit IRC | 13:32 | |
| *** zigo has quit IRC | 13:38 | |
| *** zigo_ is now known as zigo | 13:41 | |
| *** erlon has joined #openstack-swift | 13:42 | |
| *** petertr7 has joined #openstack-swift | 13:43 | |
| *** welldannit has quit IRC | 13:45 | |
| *** chlong has quit IRC | 13:47 | |
| *** zhill has joined #openstack-swift | 13:53 | |
| *** zhill has quit IRC | 13:58 | |
| *** mahatic has joined #openstack-swift | 13:59 | |
| *** Bsony has joined #openstack-swift | 14:04 | |
| *** haigang has quit IRC | 14:08 | |
| *** Bsony has quit IRC | 14:11 | |
| *** Guest__ has joined #openstack-swift | 14:14 | |
| *** bkopilov has joined #openstack-swift | 14:17 | |
| *** aix has joined #openstack-swift | 14:18 | |
| *** aix has quit IRC | 14:19 | |
| *** aix has joined #openstack-swift | 14:20 | |
| *** jrichli has joined #openstack-swift | 14:25 | |
| *** links has quit IRC | 14:26 | |
| *** Bsony has joined #openstack-swift | 14:30 | |
| *** vinsh has quit IRC | 14:32 | |
| *** bkopilov has quit IRC | 14:34 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Fix ssync sender behavior for EC reconstructor jobs https://review.openstack.org/169052 | 14:42 |
| acoles | peluse: ^^ i'm done with this, you want me to squash into 131872 or are you going to? | 14:43 |
| *** rdaly2 has joined #openstack-swift | 14:44 | |
| *** vinsh has joined #openstack-swift | 14:48 | |
| acoles | peluse: actually, did clayg mean squash the two or just approve them to merge on feature/ec? | 14:49 |
| *** rdaly2 has quit IRC | 14:49 | |
| *** Guest__ has quit IRC | 14:50 | |
| tdasilva | acoles: hi! I was wondering if it would make sense to squash patch 167595 and its dependencies or just leave as is...any thoughts? | 14:51 |
| patchbot | tdasilva: https://review.openstack.org/#/c/167595/ | 14:51 |
| *** zhill has joined #openstack-swift | 14:54 | |
| *** Bsony has quit IRC | 14:55 | |
| *** lpabon has joined #openstack-swift | 14:56 | |
| *** Bsony has joined #openstack-swift | 14:57 | |
| *** zhill has quit IRC | 14:58 | |
| acoles | tdasilva: hi! could be, let me go take a look | 14:59 |
| *** Guest__ has joined #openstack-swift | 15:01 | |
| *** Bsony has quit IRC | 15:13 | |
| peluse | acoles, I don't think he cared which way we did it | 15:16 |
| peluse | just that we did it :) | 15:16 |
| *** Guest__ has quit IRC | 15:18 | |
| *** swiftops has quit IRC | 15:18 | |
| tdasilva | peluse: happy birthday! :-) | 15:18 |
| peluse | heh, thanks! | 15:18 |
| peluse | and I'm stuck in an all day meeting today - just fantastic :( | 15:19 |
| peluse | but I managed to get 2 workouts in already this morning (started at 5:15) so I'm feeling good about that! | 15:19 |
| tdasilva | well...party all night ;) | 15:19 |
| peluse | acoles, hey I dig the job['purge'] addition... | 15:21 |
| acoles | peluse: thx. i tried to make it a callback to reconstruct_fa but the remembered time was of the essence ;) | 15:24 |
| acoles | peluse: happy b'day! | 15:25 |
| peluse | sweet, I just +2'd it. Can you go review the probe test patch real quick, its small | 15:25 |
| jrichli | peluse: happy birthday! | 15:25 |
| peluse | acoles, https://review.openstack.org/#/c/164291/ | 15:25 |
| peluse | thanks!!! | 15:25 |
| acoles | peluse: k | 15:27 |
| peluse | acoles, and clay +2'd patch 131872 if you have time to go over the latest changes there as well. There's still work to do on but its 'good enough' to get us started | 15:27 |
| patchbot | peluse: https://review.openstack.org/#/c/131872/ | 15:27 |
| peluse | oh man, did I just say that :( | 15:27 |
| acoles | peluse: k, i'm in a meeting for a while but will get there | 15:28 |
| peluse | gracias | 15:30 |
| *** h_m has quit IRC | 15:34 | |
| *** me has joined #openstack-swift | 15:34 | |
| *** me is now known as Guest8530 | 15:35 | |
| Guest8530 | Hi guys! i have a qt ; If I add a drive to a storage node, and after rebalancing the ring on the proxy-server, do i have to copy the *.ring.gz to /etc/swift of my storage nodes anew? | 15:35 |
| mandarine | Guest8530: yes | 15:36 |
| mandarine | (as far as i know) | 15:36 |
| *** silor has quit IRC | 15:37 | |
| peluse | anyone see this bp on versioned objects for Swift? https://blueprints.launchpad.net/swift/+spec/versioned-objects Doesn't seem like we need this to address the issues its helping solve for other proejcts, no? | 15:37 |
| *** joeljwright has joined #openstack-swift | 15:39 | |
| *** dspinner has joined #openstack-swift | 15:39 | |
| *** dspinner has quit IRC | 15:41 | |
| peluse | Guest8530, yes mandarine is correct | 15:46 |
| tdasilva | peluse: the name of the BP is confusing at first in the context of swift | 15:46 |
| peluse | indeed, its targeted at the other proejcts I think and tossed up at us to see if its relevant, I'm thinking its not but wanted to get others to chime in | 15:46 |
| peluse | an intel guy submitted it so, ya know :) | 15:47 |
| tdasilva | peluse: got it :-) | 15:52 |
| *** zhill has joined #openstack-swift | 15:55 | |
| *** krykowski has quit IRC | 15:55 | |
| *** zhill has quit IRC | 15:59 | |
| *** Nadeem has joined #openstack-swift | 16:00 | |
| *** lpabon has quit IRC | 16:01 | |
| *** aix has quit IRC | 16:07 | |
| acoles | peluse: since when was +457/-9 a 'small' review? :P | 16:09 |
| openstackgerrit | Thiago da Silva proposed openstack/swift: Select policy when running functional test https://review.openstack.org/167595 | 16:14 |
| tdasilva | acoles, jrichli: squashed ^^^ | 16:14 |
| acoles | tdasilva: good call | 16:15 |
| * acoles is slow today | 16:15 | |
| tdasilva | acoles: talked to jrichli and we thought there could some confusing between the first patch and kota's patch | 16:15 |
| tdasilva | s/confusing/confusion | 16:15 |
| *** Bsony has joined #openstack-swift | 16:19 | |
| *** silor has joined #openstack-swift | 16:20 | |
| acoles | tdasilva: so clayg made some good comments a few days back about not overriding the cluster's default with specified which i think had merit | 16:21 |
| acoles | tdasilva: but makes sense to review it all in one patch | 16:21 |
| peluse | acoles, heh, anymore it seems that way! | 16:21 |
| peluse | OK, I gotta head into work and will be in an al day meeting. Will be on/off IRC | 16:21 |
| acoles | peluse: well you enjoy that now :P | 16:22 |
| jrichli | tdasilva: thanks! | 16:22 |
| *** jistr has quit IRC | 16:23 | |
| *** Bsony has quit IRC | 16:24 | |
| *** lcurtis has joined #openstack-swift | 16:29 | |
| *** silor has quit IRC | 16:29 | |
| *** zhill has joined #openstack-swift | 16:35 | |
| tdasilva | acoles: mmm...not sure about that, honestly I thought the purpose of this was to exactly override the cluster's default. | 16:35 |
| tdasilva | I prefer to be able to tell that all tests are running against the same policy | 16:36 |
| tdasilva | instead of just some of them | 16:37 |
| notmyname | good morning | 16:37 |
| clayg | tdasilva: that's fine - but the test that was written to validate when you create a container with-out specifiying the policy it gets created with the *default* policy - should just be skipped then? | 16:38 |
| *** zhill has quit IRC | 16:40 | |
| tdasilva | clayg: well, it can be run that way too right? so if you don't set the env. variable it will run again the default policy. it is only when you set the default policy that you are (in my mind) saying: don't use the default, but use this one instead | 16:40 |
| tdasilva | s/when you set the default policy/when you set the environment variable | 16:40 |
| *** Bsony has joined #openstack-swift | 16:41 | |
| clayg | charz: community links are 404'ing -> e.g. why'd this py26 job fail? https://review.openstack.org/#/c/169986/1 | 16:43 |
| acoles | tdasilva: clayg: i guess the point of that test though is to send a request with *no* policy header and check that the *cluster* assigns its default. So we need to not have the func test hook insert a policy header for that case. somehow! And we need the policy collection to have retained a notion of what the cluster default is so that that test can verify its result. | 16:44 |
| *** bkopilov has joined #openstack-swift | 16:45 | |
| clayg | acoles: +1 it's a behavior the cluster has, i see no reason you wouldn't functionally test it - tdasilva points out you could test it by *not* seeing the env var - but then the func func tests "for policy in policies + ''" | 16:47 |
| acoles | so maybe we need a way for that test to inhibit the func test 'specified policy ' being applied | 16:48 |
| clayg | tdasilva: maybe the test could evolve - i'm not sure about the the idea that functests with a "specified" policy don't use other policies - some of the tests are cross policy - you're just picking one to start with and the selecting a random other, right? | 16:48 |
| *** geaaru has quit IRC | 16:49 | |
| tdasilva | clayg: yeah | 16:49 |
| clayg | acoles: how do we allow the test to create non-specified policies? | 16:50 |
| clayg | setdefault or something? | 16:50 |
| acoles | idk, stick something in the req.environ 'swift.do.not.override.policy' so that jrichli hooks don't get applied? | 16:51 |
| acoles | or is it using swift_test_client... | 16:51 |
| clayg | acoles: can we just use 'x-storage-policy': '' to indicate to the setdefault part it should pop('x-storage-policy') - if we want to let tests do that (instead of just skipping them since a "use whatever policy the cluster likes" isn't really a thing in specified storage policy land?) | 16:51 |
| clayg | i can see both sides | 16:51 |
| clayg | the only thing I don't like is doing something other than what the test says on the tin ;) | 16:51 |
| clayg | ust raise skiptest if policy.specified | 16:52 |
| acoles | i can never remember if swob request keeps 'x-storage-policy': '' in the header dict or treats it as a delete | 16:52 |
| acoles | if it does then that would work | 16:53 |
| acoles | since jrichli hook checks for 'x-storage-policy' already in headers before overriding *iirc* | 16:53 |
| *** joeljwright has quit IRC | 16:55 | |
| *** jordanP has quit IRC | 16:56 | |
| acoles | clayg: jrichli : i think clayg's idea might work i.e. have that test set 'x-storage-policy': '' to prevent the specified_policy being applied | 16:57 |
| *** thumpba has joined #openstack-swift | 16:58 | |
| acoles | (its 'x-storage-policy'= None that the HEaderEnvironProxy treats as a delete, '' is stored) | 16:58 |
| *** rdaly2 has joined #openstack-swift | 16:58 | |
| acoles | tdasilva: ^^ | 16:58 |
| acoles | ah, but does that result in a 400 from the proxy? we could strip it out before sending | 17:00 |
| *** EmilienM is now known as EmilienM|afk | 17:01 | |
| *** thumpba_ has joined #openstack-swift | 17:01 | |
| acoles | clayg: did you see the final bullet point on https://review.openstack.org/#/c/169052/ commit msg? | 17:02 |
| *** annegentle has quit IRC | 17:03 | |
| *** thumpba has quit IRC | 17:04 | |
| tdasilva | sorry, i'm back...reading log now | 17:04 |
| *** lcurtis has quit IRC | 17:07 | |
| tdasilva | acoles, clayg: the idea of 'x-storage-policy': '' sounds interesting...let me give it a try | 17:08 |
| acoles | tdasilva: thx! | 17:09 |
| peluse | clayg, not sure if you saw the scrollback - I'm stuck in an all day meeting and basically as worthless as tits on a bull today | 17:24 |
| peluse | but wanted to see where the thinking is on freezing feature/ec - land multi range and maybe mattoliverau copy fix first? | 17:24 |
| acoles | clayg: peluse i am midway reviewing ec probe tests | 17:25 |
| peluse | and then next thing you'd like to see me on - reviews, docs or start some ECrecon fixup, or other? I'm open for whatever (starting in the am other than monitoring IRC today) | 17:25 |
| peluse | acoles, thanks | 17:25 |
| acoles | i had similar question - there's comments on EC Recon, particularly ssync pieces, that i can work on fixing? if so, do we paste diffs on feature/ec-review comments now or what? | 17:27 |
| acoles | clayg: ^^ | 17:27 |
| peluse | yup, lets nail the details on how/where we work for the next week or so and make sure we're all on the same page | 17:28 |
| acoles | peluse: +1 btw i am out of the loop sunday-tuesday on family break so use/abuse me while i am here! | 17:29 |
| peluse | did notmyname approve that break? | 17:31 |
| peluse | :) | 17:31 |
| acoles | peluse: heh | 17:32 |
| * acoles has peluse-style 'all day meetings' sunday-tuesday | 17:32 | |
| peluse | ahhh, so beer is involved. nice | 17:33 |
| *** zhill has joined #openstack-swift | 17:41 | |
| openstackgerrit | Merged openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 17:43 |
| openstackgerrit | Merged openstack/swift: Fix ssync sender behavior for EC reconstructor jobs https://review.openstack.org/169052 | 17:45 |
| openstackgerrit | Merged openstack/swift: EC Reconstructor Probe Tests https://review.openstack.org/164291 | 17:45 |
| openstackgerrit | Merged openstack/swift: Make sure COPY requests are routed correctly https://review.openstack.org/170003 | 17:45 |
| jrichli | acoles clayg tdasilva: just got back from lunch. reading scrollback. | 17:45 |
| notmyname | ok, am meetings done. now to check on patch statuses | 17:53 |
| notmyname | well look at that. the reconstructor just landed on feature/ec. that's great! | 17:53 |
| *** annegentle has joined #openstack-swift | 17:58 | |
| *** welldannit has joined #openstack-swift | 18:03 | |
| acoles | peluse: so as a start how about i pick up any outstanding comments wrt ssync and recon interaction with ssync tomorrow? | 18:08 |
| acoles | peluse: i prep a diff and leave it on the review comments? | 18:11 |
| acoles | notmyname: clayg: is that how we're making fixes now, paste diffs on ec-review's for clayg to apply? | 18:12 |
| *** annegentle has quit IRC | 18:20 | |
| acoles | clayg: nice probe tests | 18:23 |
| *** EmilienM|afk is now known as EmilienM | 18:30 | |
| *** bkopilov has quit IRC | 18:41 | |
| *** bkopilov has joined #openstack-swift | 18:50 | |
| *** annegentle has joined #openstack-swift | 18:53 | |
| *** acoles is now known as acoles_away | 18:57 | |
| *** annegentle has quit IRC | 18:59 | |
| *** annegentle has joined #openstack-swift | 18:59 | |
| *** ChanServ changes topic to "Topic: Soft freeze of master in effect | EC Merge plan: https://etherpad.openstack.org/p/ec_merge_plan | Review Dashboard: http://goo.gl/vysJqI | Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/" | 19:01 | |
| notmyname | review dashboard link updated. it now includes the feature/ec_review branch in the EC work. I also starred those patches | 19:01 |
| notmyname | (if you want a CLI for pulling starred patches in gerrit, this is what I use https://github.com/notmyname/git-stats/blob/master/get_stars.py) | 19:06 |
| notmyname | in combination with GeekTool (http://projects.tynsoe.org/en/geektool/) I now have all of the starred patches on my desktop | 19:06 |
| openstackgerrit | Richard Hawkins proposed openstack/swift: Add additional func tests for TempURLs https://review.openstack.org/155985 | 19:07 |
| *** zaitcev has joined #openstack-swift | 19:21 | |
| *** ChanServ sets mode: +v zaitcev | 19:21 | |
| *** Bsony has quit IRC | 19:31 | |
| *** mahatic has quit IRC | 19:33 | |
| *** rdaly2 has quit IRC | 19:39 | |
| *** lcurtis has joined #openstack-swift | 19:41 | |
| peluse | acoles_away, sounds good. yeah so we dont duplicate effort comments for what you're working on are great. I'll do the same but I'llhave zero time today anyways L( | 19:42 |
| *** mahatic has joined #openstack-swift | 19:46 | |
| *** panbalag has quit IRC | 19:57 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift: Select policy when running functional test https://review.openstack.org/167595 | 20:09 |
| *** panbalag has joined #openstack-swift | 20:12 | |
| *** panbalag has quit IRC | 20:20 | |
| clayg | hey guys! | 20:21 |
| clayg | sorry i've been in meetings! doh! | 20:21 |
| clayg | i'm going to start merging the reconstructor into ec_review - i'll have something up tonight - i'll probably refactor a few of things I pointed out in the review while I go | 20:22 |
| clayg | acoles_away: no i didn't see that - so you got the protcol fix in there!? can't wait to see it! | 20:23 |
| clayg | peluse: acoles_away: the best way to contribute now is on feature ec_review - take a break from the reconstructor - help me validate everything we have on ec_review and you guys can pick up on the reconstructor and ssync it comes back online in ec_review | 20:25 |
| clayg | peluse: acoles_away: I *am* done with feature/ec - multi-range GET was the only thing I saw - if that lands while we're on ec_review we can look at brining it over in the 11th hour - or we can ship without it? Same thing for some of the reconstrutor clean up - everything we get in beyond "it works" is just lagniappe | 20:26 |
| clayg | notmyname: torgomatic: I need help with the commit messages on ec_review | 20:27 |
| notmyname | clayg: at 3? | 20:27 |
| clayg | you want to pair on it at 3 PST? | 20:28 |
| clayg | sorry - not clear what your question was? | 20:28 |
| notmyname | ya. lunch now, then meeting for me | 20:28 |
| clayg | peluse: mattoliverau: I went ahead and merged copy fix to feature/ec even though i already implemented it on ec_review - just to make it easier to see the what left in the diff between the branches | 20:29 |
| clayg | all - EC-REVIEW IS *READY* | 20:30 |
| clayg | everything from the storage policy object controller diskfile - it all works - the reconstructor is coming, but I have no doubt it will be done before you make your way through all of the reviews that are there (even docs!) | 20:30 |
| clayg | I'd love to get some feedback fixes rolling in - nit away! | 20:31 |
| clayg | er... you give nits and gripes - i'll make cleanup and fixes | 20:31 |
| clayg | - unless it's docs - then just push push push - peluse wrote most of them so the spelling is only slightly better than if I would have done it | 20:31 |
| *** lcurtis has quit IRC | 20:32 | |
| annegentle | clayg: I like your style. | 20:35 |
| clayg | annegentle: do you really have an irc flag on *docs*? | 20:36 |
| *** Bsony has joined #openstack-swift | 20:37 | |
| clayg | does anyone know why acoles_away added the 'fragments' key to ECDiskFileManager's _gather_on_disk_files? | 20:38 |
| annegentle | clayg: I like interruptions super infrequently :) | 20:40 |
| *** mahatic has quit IRC | 20:43 | |
| clayg | annegentle: awwwwww (that's funny) | 20:46 |
| *** lcurtis has joined #openstack-swift | 20:48 | |
| *** annegentle has quit IRC | 20:51 | |
| *** annegentle has joined #openstack-swift | 20:54 | |
| *** annegentle has quit IRC | 20:55 | |
| *** sgza has joined #openstack-swift | 20:58 | |
| *** annegentle has joined #openstack-swift | 21:01 | |
| *** lpabon has joined #openstack-swift | 21:02 | |
| *** lcurtis has quit IRC | 21:03 | |
| *** sgza has quit IRC | 21:09 | |
| *** haomaiwang has quit IRC | 21:16 | |
| *** panbalag has joined #openstack-swift | 21:16 | |
| *** panbalag has quit IRC | 21:17 | |
| *** Bsony has quit IRC | 21:21 | |
| tdasilva | clayg: nits coming :P | 21:34 |
| *** fanyaohong_ has joined #openstack-swift | 21:34 | |
| *** zacksh has quit IRC | 21:35 | |
| *** wer has quit IRC | 21:37 | |
| *** morganfainberg has quit IRC | 21:37 | |
| *** morganfainberg has joined #openstack-swift | 21:37 | |
| *** haomaiwang has joined #openstack-swift | 21:37 | |
| *** fanyaohong has quit IRC | 21:38 | |
| *** acorwin has quit IRC | 21:38 | |
| *** jasondotstar has quit IRC | 21:38 | |
| *** fanyaohong_ is now known as fanyaohong | 21:38 | |
| *** zacksh has joined #openstack-swift | 21:39 | |
| *** jasondotstar has joined #openstack-swift | 21:39 | |
| *** acorwin has joined #openstack-swift | 21:39 | |
| *** lpabon has quit IRC | 21:41 | |
| *** annegentle has quit IRC | 21:44 | |
| clayg | tdasilva: thanks bro! | 21:45 |
| peluse | clayg, not sure if read all the way back but I'm stuck in an all day meeting - just have a few minutes free and then will be back in meetingville but I saw your note above and 10-4 that! | 21:51 |
| *** annegentle has joined #openstack-swift | 22:02 | |
| *** jrichli has quit IRC | 22:06 | |
| *** annegentle has quit IRC | 22:11 | |
| *** annegentle has joined #openstack-swift | 22:12 | |
| *** tongli has quit IRC | 22:23 | |
| *** annegentle has quit IRC | 22:27 | |
| *** km has joined #openstack-swift | 22:38 | |
| *** zhill has quit IRC | 22:46 | |
| *** welldannit has quit IRC | 22:51 | |
| *** zhill has joined #openstack-swift | 22:53 | |
| *** Nadeem has quit IRC | 22:57 | |
| clayg | i've got a doc comment - so i'm going to play with the workflow of pulling in upstream changes to the doc patch and including them in my rebase (which is about to include the reconstructer!) | 23:04 |
| vinsh | Hey. Curious... have a cluster up.. have swift-recon-cron running.. but instead of a 0 for async_pending.. all I get is "NULL" | 23:05 |
| vinsh | What am I missing? | 23:05 |
| vinsh | The googlesphere comes up empty so far. | 23:05 |
| vinsh | Ah. Maybe it just took some time for the recon cache to catch up. NO sooner did I ask and I see a couple nodes with non NULL replies. | 23:07 |
| vinsh | All set. | 23:07 |
| notmyname | vinsh: great! | 23:11 |
| vinsh | :) | 23:11 |
| *** occupant has joined #openstack-swift | 23:16 | |
| clayg | awww you guys are revewing ec-review - thanks thanks thanks!!! | 23:16 |
| peluse | so for doc - do you still want push over changes on that patch? | 23:21 |
| *** vinsh has quit IRC | 23:21 | |
| notmyname | peluse: yes | 23:27 |
| clayg | peluse: yes | 23:28 |
| peluse | cool, will add some more typos tomorrow morning :) | 23:29 |
| notmyname | lol | 23:30 |
| *** petertr7 has quit IRC | 23:31 | |
| *** kei_yama has joined #openstack-swift | 23:32 | |
| clayg | peluse: I think the big risk is when I push up a rebase while people are hacking - basically around this time or later each day I'm going to push up all the fixes I've collected and reset the state of the world (including anything that's landed in docs throughout the day) | 23:32 |
| clayg | but if you've got open non-published changes when I rebase the change you'll be stuck having to rebase your fixes on the change upstream before you push | 23:32 |
| peluse | no sweat | 23:33 |
| *** thumpba_ has quit IRC | 23:37 | |
| *** annegentle has joined #openstack-swift | 23:41 | |
| clayg | nice cinder is working on "storage policies" | 23:44 |
| clayg | yay thingee! | 23:45 |
| *** haomaiwang has quit IRC | 23:45 | |
| *** haomaiwang has joined #openstack-swift | 23:46 | |
| notmyname | that's cool | 23:48 |
| *** erlon has quit IRC | 23:51 | |
| *** dmorita has joined #openstack-swift | 23:53 | |
| *** annegentle has quit IRC | 23:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!