| openstackgerrit | Samuel Merritt proposed openstack/swift: EC: support multiple ranges for GET requests https://review.openstack.org/166576 | 00:02 |
|---|---|---|
| *** jrichli has joined #openstack-swift | 00:09 | |
| *** annegentle has quit IRC | 00:14 | |
| clayg | son of a bitch | 00:16 |
| *** ho has joined #openstack-swift | 00:19 | |
| mattoliverau | clayg: you ok? | 00:26 |
| torgomatic | man vs. code; code is winning | 00:31 |
| notmyname | mattoliverau: I'm sitting next to him. he's cursing like a sailor at the code. I think he's pretty happy, actually | 00:31 |
| mattoliverau | notmyname: aah, the 'son of a bitch' was in a moment of solving/understanding the problem.. I know that feeling :) | 00:33 |
| ho | morning! | 00:37 |
| egon | I sit across from a guy from south america, so I get the equivalent in spanish. | 00:38 |
| egon | it's pretty funny, because he didn't think anyone would understand anything he said. | 00:38 |
| *** kota_ has joined #openstack-swift | 00:38 | |
| openstackgerrit | Janie Richling proposed openstack/swift: Allow middleware to override metadata header checking https://review.openstack.org/163806 | 00:41 |
| openstackgerrit | Janie Richling proposed openstack/swift: FakeFooters middleware https://review.openstack.org/165517 | 00:46 |
| *** zhill has quit IRC | 00:46 | |
| ho | I would like to have a procedure for building a saio with ec. does anyone have this info? | 00:49 |
| mattoliverau | ho: morning | 00:49 |
| ho | mattoliverau: morning! | 00:49 |
| mattoliverau | ho: if you checkout the EC branch of the code, the SAIO documentation contained is up to date regarding setting it up for EC | 00:49 |
| jrichli | morning | 00:50 |
| mattoliverau | jrichli: evening :) | 00:50 |
| jrichli | mattoliverau: well, yes... thanks! | 00:50 |
| ho | mattoliverau: thanks! | 00:50 |
| mattoliverau | ho: sample files etc.. just make sure your on the EC branch when followining the instructions and you should be fine, ask if you need any help | 00:51 |
| ho | jrichli: hello :-) | 00:51 |
| jrichli | ho: hello! | 00:52 |
| ho | mattoliverau: thanks for your help. | 00:53 |
| clayg | a'ight tox - do your thing | 00:58 |
| *** annegentle has joined #openstack-swift | 01:14 | |
| *** Tyger has quit IRC | 01:16 | |
| *** Tyger has joined #openstack-swift | 01:17 | |
| *** annegentle has quit IRC | 01:20 | |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient: Compare each chunk of large objects when uploading https://review.openstack.org/161043 | 01:36 |
| peluse | mattoliverau, small tweak on https://review.openstack.org/#/c/166307/ | 01:39 |
| clayg | I either really don't understand patch 159205 - or I should have stuck a -1 on it - because I *really* don't understand acoles_away's reasoning for having in-process functests test a pre-configured on disk ring? | 01:39 |
| patchbot | clayg: https://review.openstack.org/#/c/159205/ | 01:39 |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: Small ec diskfile refactor https://review.openstack.org/166754 | 01:42 |
| *** annegentle has joined #openstack-swift | 02:16 | |
| *** annegentle has quit IRC | 02:21 | |
| *** haomaiwang has joined #openstack-swift | 02:28 | |
| *** panbalag has quit IRC | 02:46 | |
| *** zul has quit IRC | 02:47 | |
| *** yuan has quit IRC | 02:48 | |
| *** zul has joined #openstack-swift | 02:59 | |
| *** ppai has joined #openstack-swift | 03:01 | |
| *** gyee has quit IRC | 03:06 | |
| *** bkopilov has quit IRC | 03:29 | |
| openstackgerrit | paul luse proposed openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 03:31 |
| openstackgerrit | Janie Richling proposed openstack/swift: WIP - Provides a simple skeleton of middleware for encryption feature. depends on review 165517 for FakeFooters as a work-around until EC is merged to master, and support for footers can be used from there. https://review.openstack.org/157907 | 03:41 |
| notmyname | torgomatic: ok, looking at your patches now (no tequila, but I do have ice cream) | 03:42 |
| jrichli | oops - i guess i need a space after the title | 03:42 |
| jrichli | s/space/line/ | 03:42 |
| notmyname | jrichli: ya, a blank line | 03:42 |
| openstackgerrit | Janie Richling proposed openstack/swift: WIP - Provides a simple skeleton of middleware for encryption feature. https://review.openstack.org/157907 | 03:43 |
| notmyname | jrichli: much better :-) | 03:43 |
| jrichli | lol | 03:44 |
| notmyname | jrichli: oh, I ended up getting the CODE keyboard from wasdkeyboards | 03:44 |
| notmyname | jrichli: cherry mx clear switches | 03:44 |
| jrichli | you happy with it? | 03:44 |
| notmyname | it's different, definitely. I'm still, at times, teaching my fingers how to use it | 03:45 |
| notmyname | I got a wrist rest, and I think that's essential. so much better after that. | 03:45 |
| notmyname | overall, yes, I'm pretty happy with it | 03:45 |
| jrichli | good, good. BTW, did you all decide on a T-shirt design? | 03:46 |
| notmyname | yup. 2 actually. well, same design but different colors with different color shirts | 03:47 |
| *** yuan has joined #openstack-swift | 03:48 | |
| notmyname | peluse: nice! I was just sitting down to look at that patch (conditional responses | 03:48 |
| jrichli | cool. I will be excited to see them! I saw you have lots of talks lined up. Will you need more than 2 swift shirts? | 03:48 |
| notmyname | lol. I have enough swift and openstack shirts to last for about 3 full weeks of conferences | 03:48 |
| jrichli | great, no worries then. | 03:49 |
| notmyname | hmm... I don't like how `git review -d 1234` and `git branch -d 1234` pretty much mean the exact opposite things | 03:53 |
| notmyname | mattoliverau: what are you looking at for the rest of the day? | 03:57 |
| mattoliverau | notmyname: I'm currently playing with ranged gets, then might get back on PUT refactor and then work my way down one of the patch set chains. | 03:58 |
| notmyname | mattoliverau: awesome! I'm looking at the single range gets right now too | 03:58 |
| notmyname | thanks! | 03:59 |
| mattoliverau | yay twins :P | 03:59 |
| notmyname | I'm eating ice cream. are you? (not climbing any volcanos, though) | 03:59 |
| mattoliverau | notmyname: lol, damn volcanos.. actaully I blame the icecream :P | 04:00 |
| mattoliverau | notmyname: no ice cream for me.. but I *could* go get some for afternoon tea ;P | 04:01 |
| notmyname | mattoliverau: isn't the better question "why wouldnt' you go get ice cream?" | 04:01 |
| mattoliverau | notmyname: true, that is the better question :P | 04:02 |
| jrichli | ok, now that I am hungry for sweets, I am going to bed. night! | 04:02 |
| mattoliverau | jrichli: lol, night | 04:02 |
| notmyname | :-) | 04:03 |
| *** jrichli has quit IRC | 04:03 | |
| notmyname | mattoliverau: I've got "FAILED (SKIP=12, errors=5, failures=4)" for functests with the single-range patch on top of the ifmatch patch. looking metter! | 04:04 |
| notmyname | *better | 04:04 |
| mattoliverau | notmyname: so it looks like the Melbourne Linux User Group wants me to talk about swift.. So was thinking doing the normal intro to swift talk, any chance I can steal your slide deck? If not I'll steal what I remember and use presentty ;) | 04:04 |
| mattoliverau | notmyname: nice :) I'm manually testing it first, trying to trick it and throwing data at it. | 04:05 |
| notmyname | mattoliverau: I'd be happy to give you pointers or help, but you should make your own talk instead of use mine :-) | 04:05 |
| mattoliverau | kk, just thought I'd ask :) | 04:05 |
| notmyname | hmm...a lot of ifmatch functests errors | 04:06 |
| mattoliverau | notmyname: torgomatic has another patch to get EC working with If-[None]-Match | 04:07 |
| mattoliverau | I wonder if it'll work if this was rebased off that? | 04:07 |
| notmyname | ya, didn't you +2 it and peluse just marked it +2/+1? | 04:07 |
| mattoliverau | notmyname: yup, it worked well :) | 04:07 |
| notmyname | ah, interesting. it would be because my git-fu seems to be lacking | 04:08 |
| mattoliverau | notmyname: k, so I'll play with presentty after this EC rush is over (RE: Presentation) :) | 04:08 |
| notmyname | I thought I was still working off of that patch. turns out I'm not | 04:08 |
| *** km has quit IRC | 04:09 | |
| notmyname | got it. rerunning tests | 04:10 |
| peluse | party | 04:13 |
| *** annegentle has joined #openstack-swift | 04:18 | |
| notmyname | cool. "errors=2, failures=2" with ifmatch + single range patch on functests | 04:19 |
| openstackgerrit | Merged openstack/swift: Refactoring the PUT method https://review.openstack.org/156825 | 04:19 |
| mattoliverau | ^^ looks like I don't need to look at that next :P | 04:19 |
| *** annegentle has quit IRC | 04:23 | |
| kota_ | notmyname, mattoliverau: hi, for now I could be a volunteer to review EC patches so what is highest priority? | 04:23 |
| notmyname | kota_: great! and thanks | 04:24 |
| notmyname | kota_: let me pull up the list | 04:25 |
| kota_ | notmyname: Thanks, priority reviews on wiki hides all of them for EC, right now :P | 04:26 |
| notmyname | heh, I think I just added "go look on the gerrit dashboard" since it's changing so rapidly :-) | 04:26 |
| mattoliverau | kota_: you can start with starred by PTL list on the dashboard (should match priority reviews). | 04:26 |
| notmyname | kota_: mattoliverau and I are looking at the range patches, so I think we've got those | 04:27 |
| notmyname | kota_: the patch chain starting with https://review.openstack.org/#/c/165188/ is pretty important | 04:27 |
| notmyname | kota_: that would be a great place to start | 04:27 |
| openstackgerrit | Merged openstack/swift: EC: Support If-Match and If-None-Match on GET+HEAD https://review.openstack.org/166579 | 04:28 |
| notmyname | kota_: or https://review.openstack.org/#/c/143791/ is a stand-alone patch (no dependencies) that is good to look at | 04:28 |
| kota_ | ya, I missed the line priority reviews are starred. Ok, I will start to walk through from 165188 | 04:28 |
| kota_ | notmyname, mattoliverau: thanks! | 04:29 |
| notmyname | mattoliverau: hrm. still getting one if-match error (but no range errors) | 04:32 |
| mattoliverau | notmyname: well that's better then before :P | 04:33 |
| *** silor has joined #openstack-swift | 04:41 | |
| notmyname | git st | 04:48 |
| openstackgerrit | John Dickinson proposed openstack/swift: better functional test for ranged GETs https://review.openstack.org/167092 | 04:49 |
| notmyname | mattoliverau: I had that patch in patch set 5 for the single range GETs but looks like torgomatic dropped it. there it is independently so the patch chain doesn't get messed up | 04:50 |
| mattoliverau | notmyname: ta | 04:50 |
| * mattoliverau is waiting for functional tests to complete | 04:51 | |
| clayg | mattoliverau: I'm going to screwing with the diskfile->ec-probe test patch chain shortly | 04:57 |
| *** bkopilov has joined #openstack-swift | 04:57 | |
| clayg | mattoliverau: I'm all done moving code, got a beer cracked open and working my way change-by-change via tox | 04:58 |
| mattoliverau | clayg: living the dream :) | 04:59 |
| *** ppai_ has joined #openstack-swift | 05:00 | |
| *** ppai has quit IRC | 05:01 | |
| notmyname | ok, I've looked enough at the single ranges for now (although I'm slightly suspicious about no new unit tests in the proxy object controller) | 05:03 |
| notmyname | I'm looking at the weird ifmatch bug still on HEAD of ec right now | 05:03 |
| clayg | hrmmm... so https://xkcd.com/1205/ says you can only spend about an hour reviewing https://review.openstack.org/#/c/166104/ - but I still think it would help when you're going to be running unittests dozens of times like I am now | 05:03 |
| clayg | notmyname: ranged requests are stupid - so is copy - and so is if-match - this is crunch time - we need to get the reconstructor and the proxy refactor done | 05:04 |
| notmyname | clayg: what do you need? | 05:04 |
| clayg | we can noodle around with proxy features while people are reviewing important stuff :P | 05:04 |
| notmyname | heh. I thought you were all over that! :-) | 05:04 |
| clayg | notmyname: yeah we're trying - i should plan on syncing with paul and acoles hi-bandwidth in the am | 05:05 |
| notmyname | kk | 05:05 |
| clayg | notmyname: did the master put method extraction refactor merge? | 05:05 |
| clayg | notmyname: or is it not starred - or is my dashboard old? | 05:05 |
| notmyname | looks like it did | 05:06 |
| clayg | oh shit! | 05:06 |
| mattoliverau | clayg: yup about 10 mins ago | 05:06 |
| clayg | did it include the fix for https://bugs.launchpad.net/swift/+bug/667956 | 05:07 |
| openstack | Launchpad bug 667956 in OpenStack Object Storage (swift) "chunked transfer client disconnect causes unhandled value error in object server" [Low,Confirmed] | 05:07 |
| clayg | bha, or a test for swob.Request.is_chunked - you guys are too easy | 05:07 |
| clayg | i stop paying attention for two minutes and ya'll merge anything - I seriously don't get the in-process tests running against an on-disk ring change - just because 9/10 of acoles ideas are gold - doesn't mean he can't throw out a turd occasionally as much as the rest of us | 05:08 |
| clayg | oh it did include the finally for conn in conn close - nice - no tests for it though | 05:09 |
| clayg | well i guess a few mocks had to get updated to have a close method - so I guess that probably means someone was calling it accidently ;) | 05:10 |
| clayg | neway - i'm happy it got merged - i was worried about that change | 05:10 |
| clayg | notmyname: what's going on with the master merge to feature/ec -> https://review.openstack.org/#/c/166558/ | 05:10 |
| notmyname | clayg: bug in the infra CI system. should be fixed tomorrow | 05:11 |
| notmyname | clayg: tl;dr is that global-requirements doesn't have a branch called feature/ec and so it failed and was treated as a test failure. the fix is to gracefully fall back to using master, I think. anyway, they're on it and I'm told it's good for tomorrow | 05:13 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Cleanup ECDiskFile classes https://review.openstack.org/167097 | 05:14 |
| clayg | woooh! 1 down 5 more to go! | 05:15 |
| notmyname | clayg: this is the start of a new patch chain that should be starred, right? | 05:16 |
| *** annegentle has joined #openstack-swift | 05:19 | |
| clayg | man, sometimes I be hacking and then like I get all annoyed when my vim commands don't work in browsers and shit | 05:19 |
| clayg | zz - make this line the center of the screen damnit! | 05:19 |
| clayg | glange: hey, i are you activily looking into https://bugs.launchpad.net/swift/+bug/1435506 or just raising the issue? | 05:22 |
| openstack | Launchpad bug 1435506 in OpenStack Object Storage (swift) "Account DB policy_stat Table Needs Some More Creating" [Undecided,New] | 05:22 |
| clayg | glange: because I've been trying to take a stab at https://bugs.launchpad.net/swift/+bug/1424108 but somehow we don't seem to be experiencing the issue | 05:23 |
| openstack | Launchpad bug 1424108 in OpenStack Object Storage (swift) "KeyError: 'storage_policy_index' in _really_merge_items swift/account/backend.py" [Undecided,New] | 05:23 |
| *** annegentle has quit IRC | 05:23 | |
| *** reed has quit IRC | 05:26 | |
| *** Novtopro has joined #openstack-swift | 05:27 | |
| *** km has joined #openstack-swift | 05:28 | |
| clayg | oh man top post on HN is great -> https://bugzilla.redhat.com/show_bug.cgi?id=1202858 | 05:29 |
| openstack | bugzilla.redhat.com bug 1202858 in squid "restarting squid results in deleting all files in hard-drive (rm -rf /*)" [Urgent,On_qa] - Assigned to psimerda | 05:29 |
| *** zaitcev has quit IRC | 05:34 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Add commit method to interface of DiskFile https://review.openstack.org/165208 | 05:36 |
| clayg | kota_: don't get too far on that patch chain - i'm rebasing - you could tell me what you thinkg about https://review.openstack.org/167097 | 05:40 |
| kota_ | clayg: what? | 05:45 |
| notmyname | clayg is changing everything! | 05:46 |
| kota_ | ah, great | 05:46 |
| clayg | kota_: my bad - i thought you had just put a review comment on https://review.openstack.org/#/c/165188/ - so i was just giving you the heads up that I'm moving some stuff around | 05:46 |
| kota_ | clayg: Ah, ok. I'll look at the cleanup, | 05:48 |
| *** Novtopro has quit IRC | 05:49 | |
| kota_ | clayg: I hope it would be addressed my review comment. And if everything is ok, I'll revert my patch. So please let me take a time to look at. | 05:49 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Multiple Fragment Archive support for suffix hashes https://review.openstack.org/159637 | 05:51 |
| clayg | acoles_away: ^ I think I got all of your spelling and other fixups in there | 05:52 |
| clayg | kota_: which patch - don't revert - we need all the good code we can get! | 05:53 |
| kota_ | clayg: ah, I might be in a fog, current chain of EC staff starts from 165188? | 05:53 |
| clayg | kota_: take your time - let me know if I can help :) | 05:53 |
| clayg | kota_: only just as of about 45 mins ago ;) | 05:54 |
| clayg | kota_: that's what I was trying to give you the heads up on :) | 05:54 |
| kota_ | clayg: I could make my misunderstanding you said about https://review.openstack.org/#/c/166754/ but it seems not correct. | 05:54 |
| kota_ | 166754 based on the chain and so... | 05:55 |
| kota_ | clayg: ok, I'll start again from https://review.openstack.org/#/c/165188/ | 05:57 |
| kota_ | clayg: Thanks to make my heads up on :) | 05:58 |
| *** nshaikh has joined #openstack-swift | 05:58 | |
| *** km has quit IRC | 06:02 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Imported Translations from Transifex https://review.openstack.org/167106 | 06:06 |
| notmyname | I'm turning in for the night | 06:15 |
| *** pcaruana has quit IRC | 06:41 | |
| *** chlong has quit IRC | 06:42 | |
| *** Bsony has joined #openstack-swift | 06:57 | |
| *** chlong has joined #openstack-swift | 07:03 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Add Fragment Index filter support to ssync https://review.openstack.org/165188 | 07:10 |
| *** annegentle has joined #openstack-swift | 07:20 | |
| *** annegentle has quit IRC | 07:25 | |
| *** tsg has quit IRC | 07:28 | |
| *** ppai_ has quit IRC | 07:37 | |
| *** navid__ has joined #openstack-swift | 07:38 | |
| *** mmcardle has joined #openstack-swift | 07:49 | |
| *** ppai_ has joined #openstack-swift | 07:49 | |
| *** nshaikh has quit IRC | 07:59 | |
| *** navid__ has quit IRC | 07:59 | |
| *** nshaikh has joined #openstack-swift | 08:00 | |
| *** kota_ has quit IRC | 08:04 | |
| *** jordanP has joined #openstack-swift | 08:07 | |
| *** chlong has quit IRC | 08:12 | |
| *** kota_ has joined #openstack-swift | 08:14 | |
| *** annegentle has joined #openstack-swift | 08:21 | |
| *** annegentle has quit IRC | 08:26 | |
| *** tellesnobrega has quit IRC | 08:28 | |
| *** joeljwright has joined #openstack-swift | 08:30 | |
| *** tellesnobrega has joined #openstack-swift | 08:40 | |
| *** Bsony has quit IRC | 08:41 | |
| *** geaaru has joined #openstack-swift | 08:49 | |
| *** ppai_ has quit IRC | 08:53 | |
| *** jistr has joined #openstack-swift | 09:04 | |
| *** ppai_ has joined #openstack-swift | 09:06 | |
| *** tab___ has quit IRC | 09:07 | |
| *** acoles_away is now known as acoles | 09:14 | |
| clayg | morning acoles! | 09:16 |
| acoles | clayg: good morning, hows things? | 09:16 |
| *** h_m has quit IRC | 09:16 | |
| *** egon has quit IRC | 09:19 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 09:24 |
| *** egon has joined #openstack-swift | 09:24 | |
| clayg | acoles: just peachy! | 09:24 |
| portante | morning clayg | 09:26 |
| portante | morning acoles | 09:26 |
| portante | you guys still up? | 09:26 |
| clayg | well for acoles I think it's morning | 09:26 |
| clayg | i guess *technically* it's morning for me to :\ | 09:27 |
| portante | :( | 09:27 |
| clayg | meh | 09:27 |
| cschwede | looks like some swift devs are affected by insomnia… Good Morning folks! | 09:30 |
| acoles | portante: hi, what are you doing around at this time? ;) | 09:30 |
| *** nshaikh has quit IRC | 09:31 | |
| kota_ | cschwed: lol, anyways, good morning guys :) | 09:32 |
| kota_ | woops, s/cschwed/cschwede/ | 09:33 |
| ho | good morning! | 09:36 |
| *** Bsony has joined #openstack-swift | 09:38 | |
| kota_ | ho: good evening :) | 09:38 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: wip: ec reconstructor probe test https://review.openstack.org/164291 | 09:41 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 09:41 |
| clayg | ok, and that should be working rebuild and revert | 09:41 |
| clayg | peluse: ^ | 09:41 |
| *** km has joined #openstack-swift | 09:42 | |
| ho | kota_: :-) | 09:46 |
| *** nshaikh has joined #openstack-swift | 09:59 | |
| openstackgerrit | Martin Kletzander proposed openstack/swift: Fix common misspellings https://review.openstack.org/166843 | 10:08 |
| *** nshaikh has joined #openstack-swift | 10:10 | |
| openstackgerrit | Merged openstack/swift: EC: support single ranges for GET requests https://review.openstack.org/163620 | 10:11 |
| openstackgerrit | Merged openstack/swift: Make get_dev_path() treat mount_check literally... https://review.openstack.org/166307 | 10:11 |
| openstackgerrit | Merged openstack/swift: better functional test for ranged GETs https://review.openstack.org/167092 | 10:11 |
| ho | kota_: btw thanks for the review today :) | 10:12 |
| kota_ | ho: sure :) | 10:13 |
| portante | acoles: can't sleep thinking of how hard you folks are working ... ;) | 10:17 |
| acoles | portante: are we making too much noise? :P | 10:19 |
| *** haomaiwang has quit IRC | 10:21 | |
| *** ppai_ has quit IRC | 10:30 | |
| *** Bsony_ has joined #openstack-swift | 10:34 | |
| *** Bsony has quit IRC | 10:35 | |
| *** acoles is now known as acoles_away | 10:41 | |
| *** ppai_ has joined #openstack-swift | 10:43 | |
| openstackgerrit | Alistair Coles proposed openstack/python-swiftclient: Allow reading from object body on download https://review.openstack.org/155291 | 10:45 |
| *** Bsony has joined #openstack-swift | 10:46 | |
| *** silor has quit IRC | 10:47 | |
| *** Bsony has quit IRC | 10:47 | |
| *** Bsony_ has quit IRC | 10:48 | |
| *** acoles_away is now known as acoles | 10:52 | |
| *** Bsony has joined #openstack-swift | 10:52 | |
| *** ho has quit IRC | 10:57 | |
| *** Bsony_ has joined #openstack-swift | 10:57 | |
| openstackgerrit | Merged openstack/swift: Add swift-recon feature to track swift-drive-audit error count https://review.openstack.org/163889 | 10:59 |
| *** Bsony has quit IRC | 10:59 | |
| *** Bsony_ has quit IRC | 11:02 | |
| *** Bsony has joined #openstack-swift | 11:05 | |
| *** Bsony_ has joined #openstack-swift | 11:13 | |
| *** Bsony has quit IRC | 11:16 | |
| *** Bsony_ has quit IRC | 11:18 | |
| *** panbalag has joined #openstack-swift | 11:22 | |
| *** delattec has quit IRC | 11:28 | |
| *** cdelatte has quit IRC | 11:28 | |
| *** Bsony has joined #openstack-swift | 11:31 | |
| *** km has quit IRC | 11:32 | |
| *** Bsony has quit IRC | 11:36 | |
| *** zul has quit IRC | 11:40 | |
| *** zul has joined #openstack-swift | 11:41 | |
| *** jamielennox is now known as jamielennox|away | 11:45 | |
| peluse | wow, I just missed clay by a few hrs I guess | 11:48 |
| *** silor has joined #openstack-swift | 12:01 | |
| *** cdelatte has joined #openstack-swift | 12:03 | |
| tdasilva | good morning | 12:05 |
| *** Bsony has joined #openstack-swift | 12:09 | |
| *** chlong has joined #openstack-swift | 12:11 | |
| *** chlong has quit IRC | 12:15 | |
| *** silor has quit IRC | 12:17 | |
| *** Bsony has quit IRC | 12:20 | |
| peluse | morning | 12:22 |
| *** kota_ has quit IRC | 12:27 | |
| *** silor has joined #openstack-swift | 12:29 | |
| *** jistr is now known as jistr|class | 12:30 | |
| *** chlong has joined #openstack-swift | 12:32 | |
| *** chlong has quit IRC | 12:35 | |
| *** chlong has joined #openstack-swift | 12:36 | |
| *** octavian1001 has joined #openstack-swift | 12:38 | |
| octavian1001 | hello | 12:39 |
| *** chlong has quit IRC | 12:39 | |
| octavian1001 | how can I change the access level of a given container using the api from http://developer.openstack.org/api-ref-objectstorage-v1.html | 12:40 |
| octavian1001 | from dashboard (Horizon) I can change the access level from private to public | 12:41 |
| octavian1001 | but I don't see any option to change it from the API | 12:41 |
| tdasilva | octavian1001: you would use the headers X-Container-Write and X-Container-Read | 12:44 |
| tdasilva | octavian1001: this blog post also has some good info: https://swiftstack.com/docs/cookbooks/swift_usage/container_acl.html | 12:44 |
| *** bogdan_ has joined #openstack-swift | 12:45 | |
| octavian1001 | thanks | 12:45 |
| *** octavian1001 has quit IRC | 12:55 | |
| *** chlong has joined #openstack-swift | 12:56 | |
| mandarine | Hi everyone | 13:09 |
| mandarine | I might have a question about replication | 13:10 |
| mandarine | Do the rsync logs ever stop to run crazy ? | 13:10 |
| mandarine | I've been waiting for about 25 minutes to synchronize just 55GB and it does not seem to stop yet | 13:11 |
| ctennis | You can check the output of the replication process and see how far along it is, rsync is just a byproduct of what it's doing. | 13:12 |
| ctennis | grep for "partitions" in the swift log | 13:12 |
| mandarine | thanks | 13:15 |
| openstackgerrit | paul luse proposed openstack/swift: Cleanup ECDiskFile classes https://review.openstack.org/167097 | 13:15 |
| mandarine | I'll wait for it to appear again, then | 13:15 |
| *** ppai_ has quit IRC | 13:16 | |
| mandarine | "1h remaining" | 13:17 |
| mandarine | thanks you very much | 13:17 |
| *** bogdan_ has quit IRC | 13:21 | |
| *** Gue______ has joined #openstack-swift | 13:30 | |
| *** fthiagogv has joined #openstack-swift | 13:31 | |
| *** Gue______ has quit IRC | 13:32 | |
| openstackgerrit | paul luse proposed openstack/swift: Add commit method to interface of DiskFile https://review.openstack.org/165208 | 13:34 |
| acoles | peluse: ^^ did much change? just reviewing that one | 13:35 |
| *** Bsony has joined #openstack-swift | 13:35 | |
| peluse | acoles, rebase is all. its D changed in the test code | 13:38 |
| peluse | parent I mean, not dependency | 13:38 |
| acoles | peluse: cool. thx | 13:39 |
| peluse | just rebasing the whole chain on up right now to fix the patch the patch conflict at the head | 13:39 |
| acoles | peluse: ok i have a bunch of comments on https://review.openstack.org/167097 that are going to be on last version but I'm sure clayg can live with that | 13:39 |
| peluse | acoles, cool. I've scanned all of the new ones and wil review in depth after I get them all rebased | 13:41 |
| *** nshaikh has quit IRC | 13:48 | |
| *** jistr|class is now known as jistr | 13:53 | |
| peluse | this sucks | 13:58 |
| acoles | peluse: ? | 13:58 |
| peluse | fixing rebase shit on https://review.openstack.org/#/c/159637/ - have to step out for an hour or so. will finish when i get back. who knew making it required that your target device/path actually exist would cause so many problems in this new code | 13:59 |
| openstackgerrit | Lorcan Browne proposed openstack/swift: Add lowest option to swift-recon disk usage output https://review.openstack.org/167236 | 14:12 |
| *** jrichli has joined #openstack-swift | 14:18 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift: versioned writes middleware https://review.openstack.org/134347 | 14:37 |
| *** reed has joined #openstack-swift | 14:42 | |
| *** EmilienM|PTO is now known as EmilienM | 14:51 | |
| *** annegentle has joined #openstack-swift | 14:51 | |
| openstackgerrit | Lorcan Browne proposed openstack/swift: Add lowest option to swift-recon disk usage output https://review.openstack.org/167236 | 14:52 |
| notmyname | good morning | 15:25 |
| *** Gue______ has joined #openstack-swift | 15:28 | |
| *** bkopilov has quit IRC | 15:31 | |
| *** mahatic has joined #openstack-swift | 15:31 | |
| notmyname | looks like https://review.openstack.org/#/c/167097/2 is the starting point for today | 15:32 |
| *** annegentle has quit IRC | 15:33 | |
| openstackgerrit | paul luse proposed openstack/swift: Multiple Fragment Archive support for suffix hashes https://review.openstack.org/159637 | 15:36 |
| *** tsg has joined #openstack-swift | 15:37 | |
| *** ygbo has joined #openstack-swift | 15:37 | |
| *** gyee has joined #openstack-swift | 15:39 | |
| openstackgerrit | paul luse proposed openstack/swift: Multiple Fragment Archive support for suffix hashes https://review.openstack.org/159637 | 15:40 |
| *** annegentle has joined #openstack-swift | 15:43 | |
| peluse | notmyname, yup... it and everyone up the chain needed a little work after a late night landing - almost done getting them all up to snuff gain then I'll circle back and throw in my 2 cents FWIW | 15:45 |
| notmyname | great | 15:45 |
| *** ygbo has left #openstack-swift | 15:45 | |
| tdasilva | peluse, notmyname: what's the plan with merging master to feature/ec? I imagine the next one might be a bit messy with the PUT refactoring now in master | 15:48 |
| *** bkopilov has joined #openstack-swift | 15:49 | |
| peluse | tdasilva, we've got at least a half dozen or more to land before crossing that bridge... | 15:50 |
| *** annegentle has quit IRC | 15:50 | |
| tdasilva | peluse: oh ok, cool | 15:50 |
| notmyname | well, the master to ec is different, right? | 15:51 |
| notmyname | or, rather, we shoudl be able to make progress on https://review.openstack.org/#/c/166558/ today | 15:51 |
| notmyname | I hope | 15:51 |
| notmyname | peluse: or what do you want to do there? | 15:52 |
| *** rdaly2 has joined #openstack-swift | 15:52 | |
| peluse | notmyname, yeah, would be nice to land that one as is however if we have to create a new one it will pick up the PUT refactor in whcih case we're better off waiting until we have the current WIP chain landed I think | 15:53 |
| notmyname | peluse: maybe that one there can be rechecked and then land on ec. it's from before the PUT refactor so should be easier | 15:54 |
| openstackgerrit | Lorcan Browne proposed openstack/swift: Add lowest option to swift-recon disk usage output https://review.openstack.org/167236 | 15:59 |
| notmyname | ok, coffee cup empty. time to go to the office | 16:01 |
| *** EmilienM is now known as EmilienM|PTO | 16:02 | |
| tdasilva | mm...just noticed probe tests failing on my local master branch | 16:03 |
| *** jistr has quit IRC | 16:05 | |
| *** _ahale has quit IRC | 16:13 | |
| *** ahale has joined #openstack-swift | 16:14 | |
| *** ahale has quit IRC | 16:15 | |
| *** chlong has quit IRC | 16:16 | |
| cschwede | tdasilva: which test is failing? running probetests on master now, checking | 16:17 |
| *** Gue______ has quit IRC | 16:22 | |
| cschwede | tdasilva: all probetests passed on my VM, saw only 5 skipped tests. | 16:23 |
| *** bkopilov has quit IRC | 16:25 | |
| *** Bsony has quit IRC | 16:30 | |
| clayg | acoles: oh no, the Mixin messes with your ide?! | 16:34 |
| *** Gue______ has joined #openstack-swift | 16:35 | |
| *** annegentle has joined #openstack-swift | 16:37 | |
| clayg | peluse: so you got the whole chain rebased all the way up through the probe tests? | 16:38 |
| *** GustavoJoseRuiz1 has joined #openstack-swift | 16:38 | |
| *** annegent_ has joined #openstack-swift | 16:38 | |
| peluse | clayg, not yet | 16:39 |
| clayg | what happend in the multi-fi suffix hashes? | 16:39 |
| clayg | can i help | 16:39 |
| peluse | I'm on https://review.openstack.org/#/c/159637/24 | 16:39 |
| peluse | yeah, that one :) | 16:40 |
| peluse | I'm down to jsut a few failing ones. what happened was that simple change to get_dev_path() is messing with all of the new tests | 16:40 |
| peluse | it landed right after you pushed the last set of changes | 16:40 |
| peluse | give me a few minutes.... | 16:41 |
| *** annegent_ has quit IRC | 16:42 | |
| *** annegentle has quit IRC | 16:43 | |
| clayg | acoles: I don't think you submitted your draft comments to patch set 1 - you found a test that I dropped? | 16:43 |
| peluse | actualy that one is OK now. all fixed up. the one I'm fighting is https://review.openstack.org/#/c/165188/ | 16:43 |
| clayg | peluse: ok, let me know if you get stuck | 16:45 |
| acoles | clayg: oops, just hit submit | 16:45 |
| clayg | notmyname: yeah if the master->feature/ec is working again I'd like to get the proxy put method extraction merge fix out of the way so we can start looking at minimizing churn in the proxy next | 16:46 |
| openstackgerrit | paul luse proposed openstack/swift: Add Fragment Index filter support to ssync https://review.openstack.org/165188 | 16:47 |
| GustavoJoseRuiz1 | holaaaaaaaaaaaaaaaa | 16:47 |
| peluse | clayg, ^^ there are 2 still 2 broken tests in diskfile but I can't fix them without going back and reviewing earlier patches... | 16:47 |
| peluse | so there it is for now | 16:48 |
| clayg | GustavoJoseRuiz1: Que onda? | 16:48 |
| clayg | peluse: ok - maybe I should take a stab at them then :P | 16:49 |
| GustavoJoseRuiz1 | de donde eres? | 16:49 |
| peluse | clayg, yes, please! could easily be a rebase issue between that one and the one before it... lots of code moving around | 16:49 |
| clayg | GustavoJoseRuiz1: I'm in SF, what's up is about the only spanish i know :P | 16:50 |
| *** Gue______ has quit IRC | 16:54 | |
| *** Gue______ has joined #openstack-swift | 16:55 | |
| tdasilva | cschwede: thanks for checking...it's probably my VM, | 16:55 |
| *** GustavoJoseRuiz1 has left #openstack-swift | 16:55 | |
| * notmyname is back online | 16:58 | |
| clayg | cschwede: thanks for the fix on https://github.com/swiftstack/vagrant-swift-all-in-one/pull/15 - but I don't understand why it isn't already picked up with the rest of the depends? | 16:58 |
| *** tsg has quit IRC | 16:59 | |
| clayg | cschwede: I noticed I had "fixed" all of the other source installs to use pip install -e . instead of python setup.py develop at some point - so I'm trying that on a fresh up now | 16:59 |
| cschwede | clayg: you’re welcome! i don’t know, it failed using the vagrant-all…, but worked manually. actually the important part was to install it before swift itself | 17:00 |
| acoles | clayg: peluse: i'm trying out jrichli's policy-specific func test patch, i want to verify that it is NOT using the default - can you think of a way to disable a policy so tests would break if that policy was accessed? | 17:00 |
| clayg | cschwede: i don't know sounds fishy - it's in requirements.txt - something must be wonky | 17:01 |
| clayg | acoles: hack the object-server X-Backend-Storage-pOlicy-Index to throw up? | 17:02 |
| cschwede | clayg: yes, that’s true. i wasn’t able to reproduce the problem when i installed it manually, it happened only when i provisioned the VM from scratch | 17:02 |
| clayg | acoles: I think it's in get_name_and_placement | 17:02 |
| acoles | clayg: yeah, thats the best idea i had, i can't think of any way to do it via config | 17:02 |
| acoles | clayg: thx | 17:03 |
| cschwede | clayg: i will check if it works when i do a „pip install -r requirements.txt“ before the swift installation | 17:04 |
| cschwede | clayg: if i remembered correctly, it was an error in the pyeclib installation itself | 17:04 |
| cschwede | clayg: thus the dependency was recognized, but failed to install (some gcc errors) | 17:04 |
| clayg | oh crazy!? | 17:05 |
| cschwede | but as i said - i couldn’t reproduce this manually | 17:05 |
| clayg | peluse: which tests were failing on the ssync patch? | 17:06 |
| peluse | one sec | 17:08 |
| clayg | peluse: it doesn't seem to be in test.unit.obj? | 17:08 |
| peluse | TestECDiskFile.test_yield_hashes() | 17:09 |
| peluse | TestECDiskFile.test_yield_hashes_suffixes() | 17:09 |
| clayg | oh, heh WOMM | 17:09 |
| clayg | acoles: so what are you going to do about your IDE and the Mixin's? I thought i was being so cleaver :'( | 17:10 |
| acoles | clayg: well i could just live with the hate :/ | 17:10 |
| clayg | peluse: oh, nm - totally fails for me - I've got it! | 17:10 |
| peluse | sweet! | 17:11 |
| peluse | acoles, which IDE do you use? | 17:11 |
| clayg | acoles: well, what ide is it - aren't Mixin's like a not so uncommon python thing? | 17:11 |
| acoles | clayg: i tried making the Mixin class inherit TestCase but then the tests in that class get run (and fail of course) | 17:12 |
| acoles | pycharm | 17:12 |
| clayg | acoles: right - which methods specifically does it think are unresolved? | 17:12 |
| peluse | all this talk of Mixin is making me thirsty.... | 17:12 |
| acoles | self.assertEqual etc | 17:12 |
| acoles | so what does your ide do? it can't know that those methods will get inherited from another superclass | 17:13 |
| clayg | acoles: vim doesn't complain? | 17:14 |
| *** annegentle has joined #openstack-swift | 17:14 | |
| clayg | acoles: anyway - what if we make a TestMixinABC in test.unit.utils and add unittest.Testcase abstract methods to it as we go? | 17:14 |
| clayg | http://stackoverflow.com/questions/28010960/are-mixin-classes-abstract-base-classes | 17:14 |
| acoles | clayg: huh, but vim does complain if you type self.banana()? | 17:15 |
| clayg | nah, that shits to hard in python - you never know at runtime if you'll have a bananna or not! | 17:16 |
| clayg | or test.unit.__init__ w/e | 17:16 |
| cschwede | clayg: so, just tested. it would be possible to simply do a „pip install -r requirements.txt“ before, that works too. i’m now running it again on master, will send you the bug report afterwards | 17:18 |
| acoles | clayg: true, which makes me wonder what heuristics pycharm is applying | 17:19 |
| clayg | peluse: both of those tests were replaced and moved - not sure in which change exactly - maybe this one - anyway it looks like i'm just going to delete them - they're covered in the per-manager tests | 17:19 |
| clayg | acoles: well it's obviously trying really hard - it can't know that the runner doesn't execute the non-TestCase class and it "just works" | 17:20 |
| clayg | acoles: but I like the idea of hinting that the MixIn is an abstract class that can't be used unless you define methods xyz (or subclass from someone who does) | 17:21 |
| notmyname | peluse: I'm told that -infra is ready. I just rechecked the master->ec patch | 17:21 |
| peluse | clayg, great. I saw several copies (well >1) so my next step was going to be to back track and review if there were suposed to be or not :) | 17:21 |
| peluse | notmyname, cool | 17:21 |
| clayg | acoles: how many methods do you need to add to the MixIn to get the warnings to go away? assertTrue, assertFalse, assertEqual? | 17:21 |
| acoles | clayg: i think right now i would rather live with the noise in my ide than churning/reviewing the tests again. maybe after kilo, unless you think its trivial | 17:22 |
| * peluse out for 90 min or so... | 17:23 | |
| acoles | clayg: min, i'm on a different branch. fail() from memory | 17:25 |
| clayg | peluse: I think you messed up the rebase on the ssync change - there's a bunch of tests that moved from the diskfile to the manager test case that are duplicated now | 17:26 |
| cschwede | clayg: error on master: http://paste.openstack.org/show/196089/ | 17:27 |
| peluse | clayg, OK, about to walk out the door. I can re-do it when I get back. The only required tweaks are where I changed a lot of 'dev' to 'sda1' | 17:27 |
| cschwede | the problem is in line 191: ==> default: Ran python setup.py develop && pip install -r test-requirements.txt returned 1 | 17:27 |
| peluse | or whatever its defined to in one of those classes self.expected1 or something like that | 17:27 |
| cschwede | swapped commands are working | 17:28 |
| acoles | clayg: assertRaises and fail seem to be it. but hey, its not a priority, i can most likely turn down the heat in pycharm and stop whining | 17:28 |
| clayg | hey do you guys remember back when you used to install python code with setup.py develop - and then pbr broke everything and now we type 'pip install -e .' like that's how it supposed to work? | 17:29 |
| acoles | clayg: like, i don't 'hate' that you did it, its a useful pattern, just wish we could inherit from TestCase without causing the tests to be run | 17:30 |
| clayg | acoles: *right* - someone said that nose would respect a __test__ = False class attribute - but it didn't work for me | 17:30 |
| * cschwede afk dinner, brb | 17:31 | |
| *** zhill has joined #openstack-swift | 17:32 | |
| acoles | clayg: hmmm, well maybe i'll dig some more some time. btw thanks for the changes to write_durable/commit | 17:34 |
| acoles | ok lets see if we can verify jrichli 's cool func test patch | 17:34 |
| *** annegentle has quit IRC | 17:34 | |
| notmyname | clayg: oh is that what I'm supposed to do? I end up with `python setup.py develop` after switching between master and feature/ec | 17:34 |
| clayg | peluse: ok, looks like you're out for a min - i have the ssync patch fixed up like i want - i'm going to try and apply some of the fixes acoles and kota requested into one rebase | 17:35 |
| clayg | notmyname: I think that's fine for getting the bin scripts pointing at the right versions - but python setup.py develop just doens't work for dependency resolution anymore - pbr has ruined something and no one seems to care :'( | 17:36 |
| *** Bsony has joined #openstack-swift | 17:36 | |
| clayg | that's not fair | 17:36 |
| clayg | I think we all care - but trying to fix it such a terrible mess because the obvious conclusion is just to stop using it - which for infra is a non-starter - so we give up | 17:37 |
| acoles | clayg: notmyname : i noticed recently that i have to repeat setup.py after running tox. but i do get easily confused :/ | 17:38 |
| clayg | what?! | 17:38 |
| clayg | i hate python | 17:38 |
| openstackgerrit | Merged openstack/swift: Imported Translations from Transifex https://review.openstack.org/167106 | 17:38 |
| clayg | redbo: dfg: are you guys full on the go-go-rack-labs now or still chipping away at hummingbird? | 17:39 |
| clayg | glange: you filed a bug about the policy stat table - i was thinking I should try and help - do you need any help? | 17:39 |
| acoles | clayg: re 'i hate python' - i can't remember if i hit submit for my comment that we should make our variable names hungarian e.g. strTimestamp :D | 17:40 |
| clayg | acoles: we should stop having vairables called timestamp that are strings | 17:41 |
| torgomatic | the penalty for actually doing so is fifty goulashes | 17:41 |
| *** Bsony has quit IRC | 17:41 | |
| clayg | torgomatic: like... the soup? | 17:42 |
| notmyname | clayg: looks like dfg is adding some hummingbird-specific stuff into their deployment of swift | 17:42 |
| torgomatic | clayg: none other | 17:42 |
| acoles | torgomatic: sounds worse than the lashes :) | 17:42 |
| torgomatic | :) | 17:42 |
| *** annegentle has joined #openstack-swift | 17:43 | |
| acoles | clayg: yeah, i know, and stop using ts for tombstones and timestamps | 17:44 |
| *** mmcardle has quit IRC | 17:44 | |
| dfg | notmyname: you still get emails about pull requests to our prod stuff? | 17:44 |
| clayg | acoles: re the [Tt]est in the name - I'm pretty sure unittests's test runner needs the class to be subclassed from unittest.TestCase - but nosetest will pick up anything that has name with test in it (or even is a subclass of TestCase) | 17:44 |
| notmyname | dfg: it's a public repo | 17:44 |
| dfg | notmyname: ya i know | 17:45 |
| notmyname | dfg: so yeah, I'm subscribed, just to see what you're up to and running | 17:45 |
| clayg | dfg: you never call | 17:45 |
| clayg | dfg: we worry about you | 17:45 |
| *** tsg has joined #openstack-swift | 17:45 | |
| dfg | we're still experimenting with stuff and will have somehting to show about hummingbird fairly soon. | 17:46 |
| clayg | acoles: i don't know what to do about ts - that's a damn convienient variable name :\ | 17:46 |
| clayg | dfg: yay yay yay! | 17:46 |
| *** tsg_ has joined #openstack-swift | 17:47 | |
| *** rdaly2 has quit IRC | 17:50 | |
| acoles | clayg: til from man "functions and classes whose names match testMatch and TestCase subclasses with any name will be loaded and executed as tests" | 17:50 |
| *** tsg has quit IRC | 17:50 | |
| dfg | notmyname: if that stuff pans out i'll clean it up and submit it to upstream, once things are a little more solidified on our end. which should be soon. | 17:50 |
| notmyname | dfg: great | 17:51 |
| clayg | acoles: i'm not sure if you telling me I'm full of it - or if you agree? | 17:52 |
| clayg | acoles: i like your change to test _get_ondisk_file instead of manager.get_ondisk_files | 17:53 |
| acoles | clayg: lol. the man page agrees with you so i do too! | 18:00 |
| acoles | clayg: ok good (re test) | 18:01 |
| *** Gue______ has quit IRC | 18:02 | |
| *** jamielennox|away is now known as jamielennox | 18:04 | |
| openstackgerrit | Prashanth Pai proposed openstack/swift: Refactor server side copy as middleware https://review.openstack.org/156923 | 18:05 |
| *** jordanP has quit IRC | 18:10 | |
| InAnimaTe | hey all, in a bit of a pickle. i have a cluster that is over 90% full and really need some relief asap | 18:13 |
| InAnimaTe | some of the users have started removing some stuff...however because of (im guessing) the reclaim age, its taking a while for stuff to actually be removed. | 18:14 |
| InAnimaTe | is there any way i can speed this up to almost instantly remove items ? | 18:14 |
| *** annegentle has quit IRC | 18:16 | |
| InAnimaTe | note that im running 1.8.0 | 18:20 |
| ctennis | InAnimaTe: the bytes are freed if users are actually deleting objects | 18:24 |
| *** reed has quit IRC | 18:24 | |
| ctennis | if they are deleting the account and expecting the objects to be cleaned up, then you need to run the reaper a little more aggressively | 18:24 |
| InAnimaTe | hmm that doesn't seem to be what im seeing | 18:26 |
| ctennis | when an object is deleted, a 0 byte tombstone is written out | 18:27 |
| notmyname | InAnimaTe: the reclaim age _must_ be longer than your replication cycle time (and the time it takes you to handle failed hardware) so that data isn't resurrected | 18:27 |
| InAnimaTe | ctennis: so the next time the replicator is run, it reclaims the space utilized by whatever object existed at where that tombstone is? | 18:28 |
| InAnimaTe | notmyname: but if everything is running 100% fine, then that time should be short | 18:29 |
| InAnimaTe | at least i would think | 18:29 |
| ctennis | InAnimaTe: no, that space should be reclaimed instantly on DELETE - only if you have objects in handoff locations due to drive failures or other replication issues would oyu have extra data that replication needs to clean up | 18:30 |
| ctennis | InAnimaTe: in that case then, yes replication would take care of cleaning it up | 18:30 |
| *** annegentle has joined #openstack-swift | 18:30 | |
| ctennis | InAnimaTe: any idea how long your replication cycle time is? | 18:30 |
| notmyname | InAnimaTe: yes. it's what ctennis just said. the reclaim age is to allow replication to clean up handoffs and ensure that it isn't missed | 18:31 |
| notmyname | InAnimaTe: also, check your quarantine directory to ensure that there isn't a lot of stuff there taking up space | 18:31 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Add Fragment Index filter support to ssync https://review.openstack.org/165188 | 18:34 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Cleanup ECDiskFile classes https://review.openstack.org/167097 | 18:34 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Multiple Fragment Archive support for suffix hashes https://review.openstack.org/159637 | 18:34 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Add commit method to interface of DiskFile https://review.openstack.org/165208 | 18:34 |
| *** reed has joined #openstack-swift | 18:35 | |
| clayg | peluse: I was going to rebase the ec-recon - but I'm honestly not sure which version of the disk/dir check stuff you want? | 18:42 |
| clayg | peluse: can you take over rebasing from https://review.openstack.org/#/c/131872/37 onward? | 18:42 |
| peluse | clayg, thanks, I'll take recon as I'll rebase and work on it as well | 18:42 |
| clayg | ok, i'm heading into the office | 18:43 |
| peluse | so yes for sure, and thanks thanks thanks! | 18:43 |
| *** annegentle has quit IRC | 18:45 | |
| *** fthiagogv has quit IRC | 18:46 | |
| *** lpabon has joined #openstack-swift | 18:48 | |
| openstackgerrit | paul luse proposed openstack/swift: Erasure Code Reconstructor https://review.openstack.org/131872 | 18:50 |
| *** jkugel has joined #openstack-swift | 18:53 | |
| acoles | jrichli: i got the policy-specific func tests to run :) needed some tweaks - i hacked my proxy to return 500 if the default policy was used and then used SWIFT_TEST_POLICY=<non-default> and all but two tests pass! | 18:57 |
| acoles | jrichli: the two fails are tests that explicitly setup policy headers to best across policies and so run foul of my proxy hack | 18:58 |
| acoles | s/best/test/ | 18:58 |
| jrichli | acoles: sounds good! | 18:59 |
| tdasilva | acoles, jrichli: are the two tests that are failing copy related? | 19:00 |
| acoles | tdasilva: TestCrossPolicyObjectVersioning | 19:00 |
| acoles | tdasilva: NB they are only failing in this context because I deliberately broke the proxy to reject default policy | 19:01 |
| acoles | tdasilva: i wanted to convince myself that jrichli patch really did cause all the tests to use a specified policy | 19:01 |
| *** jistr has joined #openstack-swift | 19:02 | |
| tdasilva | acoles: ok...just wondering because I saw patch 162890 and I was wondering if that would help jrichli's patch | 19:02 |
| patchbot | tdasilva: https://review.openstack.org/#/c/162890/ | 19:02 |
| acoles | tdasilva: also note i was working on master with replication policies | 19:02 |
| tdasilva | acoles: oh ok, I was going to test on feature/ec | 19:03 |
| acoles | tdasilva: maybe you could take a look at my comment on https://review.openstack.org/#/c/166097/ and see what you think about making TestCrossPolicyObjectVersioning use the specified_policy rather than a random choice | 19:03 |
| tdasilva | acoles: sure, looking now | 19:04 |
| acoles | tdasilva: jrichli : the random choice in those tests means that even if we specify a policy, the specified policy may not be used at all in those tests | 19:04 |
| acoles | tdasilva: so i suggest SWIFT_TEST_POLICY should override the random-ness | 19:04 |
| openstackgerrit | paul luse proposed openstack/swift: wip: ec reconstructor probe test https://review.openstack.org/164291 | 19:04 |
| acoles | tdasilva: i also wonder if those tests should be duplicated with reversed policies w.r.t. the version source and target?? | 19:05 |
| *** geaaru has quit IRC | 19:06 | |
| acoles | jrichli: anyway, i feel pretty positive about that! thx! | 19:06 |
| tdasilva | acoles: ok, caught up | 19:07 |
| tdasilva | acoles: I see no reason to testing the reversed way too | 19:08 |
| tdasilva | acoles: just a question: would we need two env variables to specify two policies? | 19:08 |
| acoles | tdasilva: my thinking was that (initially at least) we just want to be sure that the specified policy is exercised by all tests | 19:09 |
| acoles | tdasilva: i guess to be thorough, maybe cross policy tests should be run using all combinations of policies?? | 19:10 |
| *** torgomatic has left #openstack-swift | 19:11 | |
| acoles | tdasilva: or at least policy types, but that imho is a separate issue from jrichli patch | 19:11 |
| *** torgomatic has joined #openstack-swift | 19:11 | |
| *** ChanServ sets mode: +v torgomatic | 19:11 | |
| tdasilva | acoles: agreed | 19:11 |
| *** silor has quit IRC | 19:11 | |
| *** silor has joined #openstack-swift | 19:12 | |
| tdasilva | acoles: trying to locate the randomness you mentioned | 19:12 |
| acoles | tdasilva: i'll get a link | 19:12 |
| tdasilva | it's that select() call right? | 19:13 |
| tdasilva | acoles: ^ | 19:13 |
| acoles | tdasilva: yes, thats it | 19:13 |
| tdasilva | return random.choice(self) | 19:13 |
| acoles | yup | 19:14 |
| acoles | tdasilva: so maybe that becomes return specified_policy or random.choice() ?? | 19:14 |
| tdasilva | acoles: sounds like a good idea, but then in the case of TestCrossPolicyObjectVersioingTest the second policy selected will just be a random one? | 19:17 |
| tdasilva | I think that would be ok, right? | 19:17 |
| acoles | tdasilva: yes, random but excluding the specified_policy. i think that is ok. we at least force the specified_policy to be used as one of the test policies. | 19:18 |
| *** silor has quit IRC | 19:19 | |
| acoles | tdasilva: a follow on might be to swap the policies in the test wrt versions, so we then know the specified_policy has been tested as a source and a target for versions | 19:19 |
| acoles | 'swap' as in repeat test with policies swapped | 19:19 |
| tdasilva | acoles: yes, sounds good | 19:20 |
| acoles | jrichli: so is this making sense to you ^^? | 19:20 |
| tdasilva | acoles: so, jrichli's patch is in feature/ec atm, are you guys planning to move it to master? | 19:21 |
| acoles | tdasilva: :D jrichli has been asking that, idk! | 19:21 |
| acoles | peluse: what do you think?^^ this https://review.openstack.org/#/c/166097 could go on master, or stay on feature/ec? | 19:22 |
| tdasilva | acoles: the other issue i'm concerned with that patch is the issue of copying between two different policy types | 19:23 |
| tdasilva | I think jrichli ran into that issue, right? | 19:23 |
| acoles | tdasilva: not sure | 19:24 |
| acoles | tdasilva: whats your concern? | 19:24 |
| jrichli | tdasilva: I ran into an issue where an ECController was being used for a non-EC policy | 19:25 |
| tdasilva | jrichli: right, that's where I think this patch 162890 might come in handy | 19:25 |
| patchbot | tdasilva: https://review.openstack.org/#/c/162890/ | 19:25 |
| *** panbalag has quit IRC | 19:27 | |
| tdasilva | acoles, jrichli: so I'm thinking that jrichli's patch might need to be dependent on that | 19:27 |
| jrichli | if the patch goes on master, the tests will pass. | 19:27 |
| jrichli | and we could merge to ec from there | 19:27 |
| *** panbalag has joined #openstack-swift | 19:28 | |
| tdasilva | well...but not until the other patch is merged | 19:28 |
| acoles | tdasilva: jrichli : thats a bug with ec though not functional testing, right? | 19:28 |
| jrichli | i thought the other patch only caused issues on the ec branch. is that not correct? | 19:28 |
| tdasilva | jrichli: yes, ec only | 19:28 |
| *** tellesnobrega has quit IRC | 19:29 | |
| acoles | bbiab | 19:29 |
| *** gyee has quit IRC | 19:30 | |
| tdasilva | acoles: sorry i'm being confusing...I guess the point I'm trying to make is that motivation of jrichli patch is to allow testing with EC policies, but we won't really be able to test that until this other problem of copying between two policy types is fixed | 19:30 |
| *** jistr has quit IRC | 19:31 | |
| jrichli | tdasilva: oic. | 19:31 |
| *** tellesnobrega has joined #openstack-swift | 19:33 | |
| jrichli | tdasilva acoles: so basically, we may as well keep it on ec | 19:34 |
| tdasilva | jrichli: it makes sense to me, but I could be over thinking this :P | 19:36 |
| *** morganfainberg is now known as needslesscoffee | 19:39 | |
| acoles | tdasilva: got it. but we could merge jrichli patch before ec copy is fixed. the tests will just fail *if* SWIFT_TEST_POLICY=<ec-policy>. correct? | 19:40 |
| *** Nadeem has joined #openstack-swift | 19:41 | |
| tdasilva | acoles: yes, true | 19:41 |
| acoles | tdasilva: i have to go, thanks for your help! | 19:44 |
| acoles | jrichli: tdasilva : have a good afternoon/evening! | 19:44 |
| jrichli | acoles: you too | 19:44 |
| tdasilva | acoles: thank you! have a good one | 19:44 |
| clayg | peluse: looks like we're still waiting on jenkins for the ec-probe branch - but otherwise looking good - thanks! | 19:45 |
| *** acoles is now known as acoles_away | 19:46 | |
| clayg | let's merge something! jrichli's func tests? ec fix copy? something related to torgomatic's range fixes? | 19:49 |
| clayg | oh oh or maybe patch 167097?! | 19:49 |
| patchbot | clayg: https://review.openstack.org/#/c/167097/ | 19:49 |
| clayg | just throwin' that out there | 19:49 |
| tdasilva | lol | 19:49 |
| jrichli | :-) | 19:49 |
| *** needslesscoffee is now known as morganfainberg | 19:50 | |
| jrichli | clayg: I still have one more bit to do on the func tests. I probably wont get to work on that tonight, tho. | 19:51 |
| clayg | jrichli: sounds like a good reason to merge it now! | 19:51 |
| jrichli | lol! | 19:51 |
| petertr7 | notmyname: I'm interested in discussing an idea at tomorrow's weekly meeting. Is it okay if I just add it as an item on the (wiki) agenda? | 19:55 |
| *** lpabon has quit IRC | 19:56 | |
| notmyname | petertr7: what's up? | 19:56 |
| petertr7 | Hi! I just had an idea I wanted to discuss, I mentioned it here yesterday and figured maybe the meeting was a good place to open it up | 19:56 |
| petertr7 | It's regarding python-swiftclient | 19:57 |
| petertr7 | And how large files are handled | 19:57 |
| petertr7 | I wanted to allow users, say uploading a 10gb file, to stop their upload and continue their upload at a later time without reuploading successfully uploaded segements | 19:57 |
| notmyname | petertr7: ya, that sounds pretty useful. I'd guess the client would need to be smart enough to detect that previously uploaded chunks are available and not re-upload them | 19:58 |
| petertr7 | I just wanted to get a feel for whether the community would find it useful or not | 19:58 |
| petertr7 | I implemented something locally | 19:58 |
| notmyname | cool | 19:58 |
| petertr7 | outside of python-swiftclient | 19:59 |
| petertr7 | which does an md5 checksum and compares it to the etag of potentially uploaded segments | 19:59 |
| petertr7 | So for example, a 10gb file segmented into 1gb chunks are uploaded and interrupted | 19:59 |
| petertr7 | I'd attempt to look at etags for the first 1gb, then the 2gb and stop when the etags don't exist or are not matching | 20:00 |
| petertr7 | and then continue uploading at that chunk | 20:00 |
| notmyname | right, makes sense | 20:01 |
| petertr7 | Would this be something I could work on and contribute to python-swiftclient? | 20:01 |
| notmyname | petertr7: so, yes, I think it would be great. is it something you're planning on working on? | 20:01 |
| notmyname | yes! :-) | 20:01 |
| petertr7 | Awesome, thanks! | 20:02 |
| petertr7 | One thing I haven't ironed out in my head is when to do these checks | 20:02 |
| clayg | notmyname: how do i close a bug -> https://bugs.launchpad.net/swift/+bug/667956 | 20:03 |
| openstack | Launchpad bug 667956 in OpenStack Object Storage (swift) "chunked transfer client disconnect causes unhandled value error in object server" [Low,Confirmed] | 20:03 |
| petertr7 | In my local script I do the check every time I attempt to upload any large file | 20:03 |
| joeljwright | petertr7: it's possible that the continue upload logic would be closely related to a patch that's currently being reviewed for skip-identical and large objects | 20:03 |
| joeljwright | petertr7: https://review.openstack.org/#/c/161043/ | 20:04 |
| notmyname | petertr7: definitely joeljwright is the one to talk to :-) | 20:04 |
| petertr7 | joeljwright: Neat! Thank you! | 20:04 |
| petertr7 | notmyname: Thanks for your time! I appreciate it! | 20:05 |
| notmyname | joeljwright: also, I've been getting bugged about doing a new release of swiftclient | 20:05 |
| joeljwright | notmyname: that would be good - but it seem like everyone is too focused on EC patches at the moment :) | 20:05 |
| clayg | peluse: sweet looks like jenkins is cool with the probe tests too - are they passing? | 20:06 |
| notmyname | joeljwright: last time I asked about that, the answer was "wait until https://review.openstack.org/#/c/130339/ lands" | 20:06 |
| *** agentle has joined #openstack-swift | 20:06 | |
| joeljwright | I'd definitely like to see that land | 20:06 |
| notmyname | joeljwright: yes, there's a huge focus on ec at the moment, but the actual release is pretty simple (trivial, actually) | 20:06 |
| notmyname | joeljwright: all i have to do is make a singed tag and push it up. that's it. takes about 10 seconds | 20:07 |
| joeljwright | notmyname: and the skip-identical patch I just mentioned would be nice to have | 20:07 |
| notmyname | ya | 20:07 |
| clayg | notmyname: peluse: so like if we can't get the current master (including the PUT method extraction) merged to feature/ec like by EOB - we *literally* have a problem. | 20:07 |
| notmyname | joeljwright: so the question is if we wait until those land or just tag what's there now | 20:07 |
| clayg | notmyname: peluse: please let me know if there is anything more useful I could do to help with that than say just whining like a little baby | 20:07 |
| * notmyname goes to get the grease for the squeaky wheel ;-) | 20:08 | |
| joeljwright | notmyname: I think the problem is not the effort in creating the release, but the effort required to review the nice pending patches | 20:08 |
| clayg | joeljwright: stupid patches, always needing reviews | 20:08 |
| joeljwright | notmyname: might be better to release what's there now | 20:08 |
| notmyname | joeljwright: yup. exactly | 20:08 |
| clayg | just *one more thing* | 20:08 |
| clayg | FOREVAR | 20:08 |
| joeljwright | clayg: :D | 20:08 |
| peluse | clayg, I am more than willing to do it | 20:09 |
| joeljwright | notmyname: there is one patch already merged that I'd love to see released (fixes a bug that leaves orphaned segments) | 20:09 |
| peluse | notmyname, do you know whats up with the requirements.txt problem, should I just try a fresh one w/latest master? | 20:09 |
| notmyname | peluse: it's getting sorted. they had a bug in their fix | 20:10 |
| peluse | notmyname, ahh OK. | 20:10 |
| peluse | clayg, once we can do that (a simple merge again with a requirements change) I'll take the big one - well, I'll attack it and whether its a clean tackle or not is yet to be seen | 20:10 |
| notmyname | which is what clayg was just whining about. but it looks like the current patch needs to be replaces with a new one | 20:10 |
| notmyname | "whining" said with love | 20:11 |
| peluse | notmyname, yeah, we can land it then a new one jsut to make sure the requiments chnnge takes | 20:11 |
| peluse | since the current one is super simple | 20:11 |
| peluse | but I guess in the mantime I will shelve my unit test work and start a merge so its ready when things "work" on the jenkins side... starting now | 20:11 |
| notmyname | peluse: thanks. that's what I was about to say :-) | 20:12 |
| *** cdelatte has quit IRC | 20:16 | |
| notmyname | peluse: ok, I'm told it takes about an hour to rebuild and deploy the worker images so the fix is out | 20:17 |
| notmyname | peluse: and fungi is working on it | 20:17 |
| peluse | cool | 20:17 |
| notmyname | so that gives time for getting a new master->ec merge | 20:18 |
| peluse | fixing merge conflicts now... only 32 of them so not horrible I guess :) | 20:18 |
| clayg | peluse: so... we're all in a holding pattern waiting on infra before we attempt the merge? oh.. or we should work on a new merge now? | 20:18 |
| clayg | peluse: wow :) | 20:18 |
| peluse | clayg, I'm resolving conflicts now and will push it when told its ready | 20:18 |
| * clayg hugs peluse | 20:19 | |
| peluse | "it" meaning the jenkins fix of course | 20:19 |
| notmyname | peluse: actually push it whenever it's ready. and if -infra isn't ready by then, we'll simply recheck it | 20:20 |
| peluse | OK | 20:21 |
| * peluse wonders if clayg will be in a huging mood when he has to rebase the patch chain on feature/ec :) | 20:21 | |
| *** trunk has joined #openstack-swift | 20:21 | |
| clayg | oh right :\ | 20:21 |
| clayg | well maybe it's no so bad - that change should mostly be out of the proxy - well hell I guess I don't really know what's coming in with master | 20:22 |
| clayg | peluse: anyway you're right - we have too much un-merged code | 20:22 |
| clayg | peluse: but I can't find anything I didn't write worth merging ;P (j/k) | 20:23 |
| peluse | heh | 20:29 |
| peluse | I do indeed agree though | 20:29 |
| clayg | see tdasilva gets it | 20:33 |
| tdasilva | lol | 20:33 |
| tdasilva | clayg: go merge obj. versioning :P | 20:33 |
| *** reed has quit IRC | 20:42 | |
| notmyname | joeljwright: I'm about to go to a meeting, but I'm not opposed to tagging/releasing python-swiftclient 2.3.2 today | 20:46 |
| clayg | tdasilva: oh good one | 20:46 |
| tdasilva | ;-) | 20:47 |
| *** nellysmitt has joined #openstack-swift | 20:47 | |
| clayg | tdasilva: i sorta think i was either putting that off because functionally testing the upgrade was a pita - or there was something about container vs. object request routing that I keep trying to refactor everytime I look at it | 20:47 |
| notmyname | joeljwright: I know it's getting late for you. please let me know if you have any concerns if a 2.3.2 release were to happen | 20:48 |
| clayg | tdasilva: oh wait - what happend - it's like way simpler now? | 20:48 |
| tdasilva | clayg: what do you mean? | 20:48 |
| clayg | tdasilva: like maybe everything goes through the context thing now? | 20:49 |
| clayg | tdasilva: idk, just seems simpler since the last time I looked at it? | 20:50 |
| tdasilva | clayg: I think that was based on a suggestion you made | 20:50 |
| tdasilva | clayg: so now there's container_request and object_request | 20:51 |
| tdasilva | remember __call__ was very long?? | 20:51 |
| clayg | tdasilva: I think you're just plying to my egomania trying to get reviews | 20:56 |
| tdasilva | lol | 20:56 |
| clayg | tdasilva: ... could work | 20:56 |
| tdasilva | clayg: no, i actually just found the comment too | 20:56 |
| tdasilva | https://review.openstack.org/#/c/134347/21/swift/common/middleware/versioned_writes.py | 20:56 |
| tdasilva | classicsnail: L347 | 20:56 |
| tdasilva | clayg: L347 | 20:57 |
| tdasilva | classicsnail: sorry | 20:57 |
| joeljwright | notmyname: sorry I was picking my wife up from the station | 20:57 |
| *** mahatic has quit IRC | 20:59 | |
| joeljwright | I don't think I'd have any concerns about a 2.3.2 release | 21:00 |
| joeljwright | there are certainly quite a few fixes that would be nice to have in the released version… I just might be bugging you for a 2.3.3 release sooner than you'd like | 21:01 |
| *** joeljwright1 has joined #openstack-swift | 21:05 | |
| *** joeljwright has quit IRC | 21:06 | |
| joeljwright1 | notmyname: and my laptop chose that moment to lock up | 21:06 |
| *** mmcardle has joined #openstack-swift | 21:06 | |
| *** openstackgerrit has quit IRC | 21:07 | |
| *** openstackgerrit has joined #openstack-swift | 21:07 | |
| *** kevinc_ has joined #openstack-swift | 21:11 | |
| tdasilva | jrichli: still around? | 21:13 |
| jrichli | ye | 21:13 |
| jrichli | s | 21:13 |
| tdasilva | jrichli: hi, i started playing with that functional test change that acoles_away was talking about, will try to send you something later, ok? | 21:14 |
| jrichli | sounds good, thanks! | 21:15 |
| peluse | looks like I picked the wrong week to stop sniffin' glue... | 21:17 |
| jrichli | lol | 21:17 |
| tdasilva | haha | 21:17 |
| peluse | tdasilva, I'm merging into feature/ec if you hadn't gathered (you PUT refactor)... fun stuff! | 21:19 |
| *** reed has joined #openstack-swift | 21:20 | |
| tdasilva | peluse: mmm...i can imagine...I almost wish we had landed that other patch on feature/ec first | 21:20 |
| tdasilva | then it would be easy to merge | 21:20 |
| peluse | bah, it'll sort out :) | 21:21 |
| tdasilva | ok, let me know if I can help with anything | 21:21 |
| clayg | man, our test suite is going to get so much better when we don't care about python 2.6 anymore | 21:22 |
| clayg | who cares about python 2.6 again? | 21:22 |
| * torgomatic sure doesn't | 21:25 | |
| mattoliverau | Morning, wow, that was a lot of scroll back to read! | 21:25 |
| openstackgerrit | Samuel Merritt proposed openstack/swift: EC: support multiple ranges for GET requests https://review.openstack.org/166576 | 21:28 |
| openstackgerrit | Samuel Merritt proposed openstack/swift: EC: better failure handling for GETs https://review.openstack.org/167406 | 21:28 |
| *** jrichli has quit IRC | 21:31 | |
| peluse | hey cores: quick request, please don't land anything on feature/ec til I'm done the the current merge from master. Gracias! | 21:38 |
| mattoliverau | hai | 21:39 |
| *** jkugel has quit IRC | 21:41 | |
| *** jkugel has joined #openstack-swift | 21:41 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Update contianer sync to use internal client https://review.openstack.org/143791 | 21:43 |
| *** mmcardle has quit IRC | 21:43 | |
| clayg | mattoliverau: yeah idk why everyone is so chatty today | 21:43 |
| *** jkugel has quit IRC | 21:46 | |
| *** agentle has quit IRC | 21:48 | |
| *** Cipher45 has quit IRC | 21:48 | |
| *** wolsen has quit IRC | 21:48 | |
| *** tristanC has quit IRC | 21:48 | |
| *** agentle has joined #openstack-swift | 21:49 | |
| *** Cipher45 has joined #openstack-swift | 21:49 | |
| *** wolsen has joined #openstack-swift | 21:49 | |
| *** tristanC has joined #openstack-swift | 21:49 | |
| *** sc has quit IRC | 21:50 | |
| *** agentle has quit IRC | 21:50 | |
| *** sc has joined #openstack-swift | 21:52 | |
| *** zhill has quit IRC | 21:53 | |
| *** dfg has quit IRC | 22:06 | |
| *** dfg has joined #openstack-swift | 22:07 | |
| *** hurricanerix has quit IRC | 22:07 | |
| *** hurricanerix_ has joined #openstack-swift | 22:07 | |
| *** joeljwright1 has quit IRC | 22:12 | |
| notmyname | peluse: what's the status of a new merge from master? | 22:14 |
| *** zhill has joined #openstack-swift | 22:14 | |
| peluse | notmyname, its going pretty good if you can believe it | 22:20 |
| notmyname | cool. -infra says things should be ready from their side | 22:20 |
| peluse | GET/PUT working fine. half dozen or so tests failing with one signature and another set with just one other signature. Digging into those now | 22:21 |
| clayg | torgomatic's good idea of the hour - X-If-Not-Match-Fragment-Index-List | 22:27 |
| clayg | also, i feel like the commit message on patch 167406 could be a popular theme this week | 22:28 |
| patchbot | clayg: https://review.openstack.org/#/c/167406/ | 22:28 |
| clayg | tdasilva: I really wanted to review versioned writes - but I just can't - i need to keep looking at some ec stuff - sorry - maybe next week between rebase's of the merge to master chain! | 22:30 |
| tdasilva | clayg: no worries | 22:31 |
| notmyname | ok, I'm tagging python-swiftclient 2.4.0 | 22:32 |
| notmyname | (there hasn't been a release since september!) | 22:33 |
| *** shri has joined #openstack-swift | 22:35 | |
| notmyname | done | 22:35 |
| shri | Hi, I have a question about the upstart scripts for various swift processes that are installed with the debian packages. | 22:37 |
| shri | I see that these scripts don't have a respawn clause in case the process dies. Any particular reason for that? | 22:38 |
| peluse | OK, bout to do last tox and then if it works "fire in the hole" | 22:41 |
| peluse | oh crap... have to go pick up son from school... BRB :) | 22:41 |
| *** zhill has quit IRC | 22:44 | |
| mattoliverau | peluse: Lol, it's like going to an ad break just before the cliff hanger :P | 22:49 |
| *** Nadeem has quit IRC | 22:55 | |
| peluse | heh, back and tox'ing... | 22:58 |
| openstackgerrit | paul luse proposed openstack/swift: Merge master to feature/ec https://review.openstack.org/167429 | 23:05 |
| notmyname | peluse: yay | 23:05 |
| notmyname | I'll abandon the other one (if you haven't already) | 23:06 |
| peluse | just did! | 23:06 |
| notmyname | That Was Easy | 23:06 |
| notmyname | I'm tracking the jobs on http://status.openstack.org/zuul/ | 23:07 |
| peluse | heh, good luck trying to read the obj.py diff on that bad boy in gerrit! | 23:07 |
| peluse | me too | 23:08 |
| *** vinsh has joined #openstack-swift | 23:09 | |
| *** welldannit has joined #openstack-swift | 23:11 | |
| peluse | BAH! requirements failed | 23:12 |
| notmyname | peluse: on it | 23:12 |
| peluse | cool | 23:13 |
| notmyname | peluse: but the good news is that we have the merge commit done now | 23:13 |
| notmyname | so it should "just" be a recheck | 23:14 |
| clayg | yeah that proxy.controller.obj diff looks wonky :\ | 23:15 |
| clayg | torgomatic: can you look at patch 167429 when you get a chance and guess at if the ObjController changes look sane? | 23:16 |
| patchbot | clayg: https://review.openstack.org/#/c/167429/ | 23:16 |
| torgomatic | clayg: sure, give me 15 minutes | 23:16 |
| peluse | indeed | 23:16 |
| peluse | clayg, well, its wonky because of all the movement | 23:17 |
| peluse | I didn't bring over the repl specific _store_object() so what will land will be the putter/mime method (refactored) for both | 23:18 |
| clayg | peluse: oh i know - but the line diffs seem strange too - i guess because maybe - did we already merge something close to that refactoring onto feature/ec? | 23:18 |
| peluse | if you really want to separate the paths we *could* have different _store_object() methods | 23:18 |
| clayg | peluse: oic - I had taken a different approach in patch 164950 - yeah i split ec stuff into ECObjectController next to GET | 23:19 |
| patchbot | clayg: https://review.openstack.org/#/c/164950/ | 23:19 |
| clayg | peluse: I still think that's the way to go - but it can come after we merge them together then pull them back apart | 23:19 |
| peluse | ahh geeze, didn't even know that was there :) | 23:20 |
| *** chlong has joined #openstack-swift | 23:20 | |
| clayg | peluse: oh it's fine - i don't think it would have helped anything | 23:21 |
| peluse | well its just software, we can always change it again later :) | 23:21 |
| notmyname | peluse: I've got to head home, so I'll be offline for about an hour. what do you need from me for the merge? I want to see that land today | 23:28 |
| peluse | I guess just lean on infra dudes to fix whatever is messed up. the code I think is fine | 23:29 |
| notmyname | ok | 23:29 |
| notmyname | peluse: "zuul estimates 5 minutes until it merges, then roughly an hour for me to get the worker images replaced again" | 23:31 |
| notmyname | from fungi in -infra | 23:31 |
| peluse | OK, do we just recheck no bug in a few hrs? | 23:32 |
| notmyname | peluse: yup | 23:32 |
| peluse | BTW, just to confirm. All unit, functional and probetests pass on the merge patch | 23:32 |
| *** chlong has quit IRC | 23:35 | |
| *** chlong has joined #openstack-swift | 23:35 | |
| InAnimaTe | anyone know how long python-swiftclient stores the auth url? | 23:42 |
| InAnimaTe | sigh, i mean the x-storage-url | 23:42 |
| *** trunk has quit IRC | 23:47 | |
| *** dencaval has joined #openstack-swift | 23:48 | |
| *** ho has joined #openstack-swift | 23:49 | |
| *** gyee has joined #openstack-swift | 23:54 | |
| *** zhill has joined #openstack-swift | 23:55 | |
| *** zaitcev has joined #openstack-swift | 23:56 | |
| *** ChanServ sets mode: +v zaitcev | 23:56 | |
| *** vinsh has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!