| *** shri has quit IRC | 00:02 | |
| *** acoles has quit IRC | 00:04 | |
| *** kevinc_ has joined #openstack-swift | 00:11 | |
| openstackgerrit | John Dickinson proposed a change to openstack/swift: added process pid to the end of storage node log lines https://review.openstack.org/105309 | 00:11 |
|---|---|---|
| *** dmorita has joined #openstack-swift | 00:15 | |
| *** kevinc_ has quit IRC | 00:22 | |
| mattoliverau | notmyname: lol, couldn't help your self huh :) ^^ | 00:29 |
| *** tsg has quit IRC | 00:33 | |
| *** mwstorer has quit IRC | 00:38 | |
| *** tsg has joined #openstack-swift | 00:41 | |
| *** fifieldt has joined #openstack-swift | 00:55 | |
| openstackgerrit | A change was merged to openstack/python-swiftclient: Removed now unnecesary workaround for PyPy https://review.openstack.org/97072 | 01:00 |
| *** cpen has joined #openstack-swift | 01:01 | |
| *** cpen has quit IRC | 01:01 | |
| *** Tyger has quit IRC | 01:09 | |
| *** nosnos has joined #openstack-swift | 01:42 | |
| *** haomai___ has quit IRC | 01:42 | |
| *** haomaiwa_ has joined #openstack-swift | 01:42 | |
| openstackgerrit | A change was merged to openstack/swift: Sync netifaces with global-requirements https://review.openstack.org/91879 | 01:54 |
| *** haomai___ has joined #openstack-swift | 01:58 | |
| *** haomaiwa_ has quit IRC | 01:59 | |
| *** zhiyan_ is now known as zhiyan | 02:08 | |
| *** robsparker has quit IRC | 02:29 | |
| tsg | torgomatic: zero-copy stuff is looking great - clever :) .. can't wait to see the benchmarks! | 02:44 |
| *** robsparker has joined #openstack-swift | 02:45 | |
| *** Edward-Zhang has joined #openstack-swift | 02:45 | |
| notmyname | tsg: he's doing the zero copy just for you. that is, let's get the cpu usage as low as possible before we go adding ec ;-) | 02:48 |
| tsg | notmyname: :) ya we need that cpu time for ec ;) | 02:53 |
| tsg | tee + AF_ALG implementation for etags is very well thought of | 02:55 |
| *** robsparker has quit IRC | 02:56 | |
| *** robsparker has joined #openstack-swift | 03:07 | |
| notmyname | mattoliverau: bgmccollum: I updated my logflow thing, including new logs and pictures based on the patch I submitted to swift earlier https://github.com/notmyname/logflow | 03:11 |
| notmyname | especially check out the .png files there | 03:11 |
| notmyname | https://raw.githubusercontent.com/notmyname/logflow/master/all.png is fun to look at I think | 03:12 |
| notmyname | bgmccollum: here's what you were looking for earlier: what updates the container servers when there is an object put? https://raw.githubusercontent.com/notmyname/logflow/master/object_put.png | 03:14 |
| mattoliverau | notmyname: thats awesome! definitely going to add it my SAIO and have a play. | 03:14 |
| notmyname | mattoliverau: I'd need to look into graphviz or dot a little more, but I think it would be interesting to weight the edges of the graph according to how much it's called. | 03:15 |
| notmyname | I'm not sure yet on how to effectively use it for a system monitoring thing, but I think it's an excellent tool for teaching | 03:15 |
| *** tsg has quit IRC | 03:15 | |
| *** tsg has joined #openstack-swift | 03:16 | |
| *** chandan_kumar has joined #openstack-swift | 03:17 | |
| notmyname | heh. now doing something crazy | 03:19 |
| mattoliverau | adding a weight or something to know how any calls would make it even more sueful. But teaching yes, maybe not continuing monitoring, unless you run it on daily split logs or something (or a periodic job on a syslog server). One could use it when wanting to diagnose a potention problem account/container/object server though :) | 03:20 |
| mattoliverau | *potential | 03:20 |
| notmyname | I'm running the `swift-bench --saio` now and I'll run that through logflow. I expect it to be ... messy | 03:21 |
| mattoliverau | man my typing is bad today.. I blame my phone cause I'm at lunch :P | 03:21 |
| mattoliverau | haha. I'm looking forward to it tho :) | 03:21 |
| *** nosnos has quit IRC | 03:22 | |
| notmyname | ok. don't do that with the xprofile middleware. bad things happen. I should look in to that at some point.... | 03:25 |
| notmyname | 24798 log lines. what could go wrong? | 03:29 |
| mattoliverau | haha, nothing could possibly.. I wonder how big the resulting png will be :) | 03:30 |
| notmyname | still processing and pegging one core | 03:32 |
| notmyname | I wonder ho long this will take | 03:34 |
| *** chandan_kumar has quit IRC | 03:36 | |
| *** chandan_kumar has joined #openstack-swift | 03:36 | |
| notmyname | eh, who knows. I'll leave it running overnight and check in the morning. | 03:42 |
| *** ppai has joined #openstack-swift | 03:43 | |
| *** ppai has quit IRC | 03:50 | |
| mattoliverau | Lol, well it hasn't died, so that's promising :) | 03:55 |
| *** Edward-Zhang has quit IRC | 04:00 | |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 04:04 |
| openstackgerrit | OpenStack Proposal Bot proposed a change to openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 04:05 |
| *** ppai has joined #openstack-swift | 04:08 | |
| *** aswadr has joined #openstack-swift | 04:20 | |
| *** ajc_ has joined #openstack-swift | 04:25 | |
| *** nosnos has joined #openstack-swift | 04:25 | |
| *** Edward-Zhang has joined #openstack-swift | 04:38 | |
| *** bkopilov has quit IRC | 04:38 | |
| *** anand_ts has joined #openstack-swift | 04:49 | |
| *** chandan_kumar has quit IRC | 04:58 | |
| *** chandan_kumar has joined #openstack-swift | 05:04 | |
| *** chandan_kumar has quit IRC | 05:04 | |
| *** chandankumar has joined #openstack-swift | 05:04 | |
| *** gyee has quit IRC | 05:14 | |
| *** psharma has joined #openstack-swift | 05:15 | |
| Edward-Zhang | notmyname, what is happened for xprofile middleware? :-( | 05:16 |
| Edward-Zhang | notmyname: let me know the bad thing so that I can fix it. | 05:17 |
| *** zaitcev has quit IRC | 05:19 | |
| *** tsg is now known as tsg_away | 05:33 | |
| *** ppai has quit IRC | 05:47 | |
| *** ppai has joined #openstack-swift | 06:00 | |
| *** tsg_away has quit IRC | 06:11 | |
| Edward-Zhang | John: what's wrong with XProfile middleware? let me know the detail so that I can fix it. :-) | 06:12 |
| *** bkopilov has joined #openstack-swift | 06:17 | |
| *** zigo has quit IRC | 06:27 | |
| *** zigo has joined #openstack-swift | 06:31 | |
| *** ppai has quit IRC | 06:43 | |
| *** nacim has quit IRC | 06:51 | |
| *** mmcardle has joined #openstack-swift | 07:04 | |
| mattoliverau | I'm calling it a day, time to go cook some dinner, night all! | 07:17 |
| *** haomai___ has quit IRC | 07:17 | |
| *** haomaiwang has joined #openstack-swift | 07:18 | |
| mattoliverau | Edward-Zhang: notmyname was paying with a log processing script, which I think doesn't work too well with the middleware. So don't think anything is actually wrong :) | 07:19 |
| *** haomai___ has joined #openstack-swift | 07:33 | |
| *** haomaiwang has quit IRC | 07:35 | |
| *** nacim has joined #openstack-swift | 08:00 | |
| *** andyandy has joined #openstack-swift | 08:10 | |
| Edward-Zhang | mattoliverau: thanks for your clarification. :-) | 08:46 |
| *** mmcardle has quit IRC | 08:59 | |
| *** acoles_ has joined #openstack-swift | 09:06 | |
| *** ChanServ sets mode: +v acoles_ | 09:06 | |
| *** bvandenh has joined #openstack-swift | 09:06 | |
| *** haomai___ has quit IRC | 09:13 | |
| *** haomaiwa_ has joined #openstack-swift | 09:14 | |
| *** ajc_ has quit IRC | 09:20 | |
| *** mmcardle has joined #openstack-swift | 09:28 | |
| *** Edward-Zhang has quit IRC | 09:29 | |
| *** haomaiw__ has joined #openstack-swift | 09:30 | |
| *** haomaiwa_ has quit IRC | 09:32 | |
| *** foexle has joined #openstack-swift | 09:41 | |
| *** ppai has joined #openstack-swift | 09:49 | |
| *** haomaiw__ has quit IRC | 09:59 | |
| *** haomaiwang has joined #openstack-swift | 09:59 | |
| *** kopparam has joined #openstack-swift | 10:04 | |
| *** haomaiw__ has joined #openstack-swift | 10:14 | |
| *** haomaiwang has quit IRC | 10:18 | |
| openstackgerrit | A change was merged to openstack/swift: Use except x as y instead of except x, y https://review.openstack.org/96610 | 10:22 |
| *** ho has quit IRC | 10:39 | |
| *** zhiyan is now known as zhiyan_ | 10:46 | |
| *** zhiyan_ is now known as zhiyan | 10:47 | |
| *** nacim has quit IRC | 10:53 | |
| openstackgerrit | A change was merged to openstack/swift: Parallel object auditor https://review.openstack.org/59778 | 10:58 |
| *** anand_ts has quit IRC | 11:06 | |
| *** mkerrin has quit IRC | 11:07 | |
| *** dmorita has quit IRC | 11:09 | |
| *** nosnos has quit IRC | 11:23 | |
| *** mmcardle has quit IRC | 11:23 | |
| *** bvandenh has quit IRC | 11:35 | |
| openstackgerrit | Alistair Coles proposed a change to openstack/swift: Restrict keystone cross-tenant ACLs to IDs https://review.openstack.org/86430 | 11:49 |
| *** bvandenh has joined #openstack-swift | 11:57 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Document the unnecessary method invoking in FakeRing https://review.openstack.org/103407 | 12:00 |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Remove fake _get_part and use the real thing https://review.openstack.org/103962 | 12:00 |
| *** tdasilva has joined #openstack-swift | 12:04 | |
| *** foexle has quit IRC | 12:22 | |
| *** kopparam has quit IRC | 12:23 | |
| *** mmcardle has joined #openstack-swift | 12:23 | |
| *** kopparam has joined #openstack-swift | 12:25 | |
| *** nacim has joined #openstack-swift | 12:30 | |
| *** kopparam has quit IRC | 12:37 | |
| *** acoles_ has quit IRC | 12:42 | |
| *** acoles_ has joined #openstack-swift | 12:42 | |
| *** ChanServ sets mode: +v acoles_ | 12:42 | |
| *** ppai has quit IRC | 12:42 | |
| *** foexle has joined #openstack-swift | 12:46 | |
| *** acoles_ has quit IRC | 12:50 | |
| *** acoles_ has joined #openstack-swift | 12:50 | |
| *** ChanServ sets mode: +v acoles_ | 12:50 | |
| *** ppai has joined #openstack-swift | 12:56 | |
| *** miqui has joined #openstack-swift | 13:01 | |
| *** acoles_ has quit IRC | 13:04 | |
| *** tsg has joined #openstack-swift | 13:13 | |
| *** madhuri has joined #openstack-swift | 13:13 | |
| *** psharma has quit IRC | 13:17 | |
| *** ppai has quit IRC | 13:22 | |
| madhuri | Hi, can anyone tell me how this method can be called if it is publically accessible? | 13:23 |
| madhuri | https://github.com/openstack/swift/blob/stable/icehouse/swift/proxy/controllers/base.py#L1225 | 13:23 |
| *** mrsnivvel has quit IRC | 13:23 | |
| madhuri | and moreover what is its use? | 13:25 |
| madhuri | As per my knowledge it just shows information about cors. | 13:26 |
| notmyname | eg `curl -i http://your.swift.cluster/ -XOPTIONS` | 13:26 |
| madhuri | Can anyone please correct my understanding? | 13:26 |
| notmyname | yes, it's mostly used for CORS, but it can also be used to see what verbs are allowed for a particular request | 13:26 |
| notmyname | eg `curl -i http://your.swift.cluster/account/container/object -XOPTIONS` | 13:27 |
| madhuri | Above request gives error "The resource could not be found." | 13:28 |
| goodes | notmyname: in one of RC releases I remember seeing something about improvements in ssync but its not in the official release notes. What happened with that? | 13:28 |
| notmyname | goodes: https://github.com/openstack/swift/blob/master/CHANGELOG#L223 | 13:29 |
| madhuri | notmyname: I was missing something | 13:29 |
| madhuri | It worked thanks | 13:29 |
| * fifieldt witnesses notmyname, having just released swift 2.0, living out the phrase "no rest for the wicked" | 13:29 | |
| *** robsparker has quit IRC | 13:29 | |
| bgmccollum | notmyname: awesome stuff! pretty much exactly what i wanted... | 13:30 |
| notmyname | fifieldt: :-) | 13:30 |
| notmyname | fifieldt: I sent a message to -announce. can you unstick it? | 13:31 |
| fifieldt | maybe! | 13:31 |
| * fifieldt checks | 13:31 | |
| notmyname | thanks | 13:31 |
| fifieldt | done | 13:31 |
| notmyname | yay! | 13:32 |
| fifieldt | :D | 13:32 |
| notmyname | it's fun seeing my twitter feed with "openstack swift" rather than that silly thing from apple ;-) | 13:33 |
| fifieldt | for sure | 13:34 |
| hugokuo | ：） | 13:34 |
| notmyname | fifieldt: hey you might like this. check out what bgmccollum and I were talking about yesterday: https://github.com/notmyname/logflow Look at the .png files | 13:34 |
| peluse_ | nottrobin: FYI I submitted a proposal for linux conf au - the thing I sent you in email regarding Swift 2.0. Will see where it goes if anywhere... | 13:34 |
| notmyname | peluse_: I'll submit one too, I think let's see if I can get it typed up before the end of the week :-) | 13:35 |
| fifieldt | whoa! | 13:35 |
| fifieldt | that's pretty awesome | 13:36 |
| notmyname | fifieldt: it takes a log file as input and the makes the png. I think it could be pretty cool for teaching/training about the data flow in swift | 13:36 |
| ahale | thats pretty cool notmyname | 13:37 |
| fifieldt | not to mention the debugging when you're by yourself setting up a cluster for the first time before you call swiftstack :) | 13:37 |
| notmyname | lol | 13:37 |
| fifieldt | I can actually see the tutorial excercise for this | 13:38 |
| peluse_ | notmyname: cool, you've got the abstract I submitted already, if you don't want to go with another topic we can always do that one together | 13:38 |
| fifieldt | first you get the students to draw their own manually | 13:38 |
| notmyname | peluse_: cool | 13:38 |
| fifieldt | then copy/paste the log into the tool | 13:38 |
| fifieldt | and see if they got it correct | 13:38 |
| peluse_ | notmyname: course, still have to get funding :) | 13:38 |
| fifieldt | if they did not, discuss | 13:38 |
| notmyname | peluse_: ya, I have no idea on that front either :-) | 13:39 |
| peluse_ | notmyname: nice blog post this morning BTW! | 13:39 |
| notmyname | fifieldt: ya, that's pretty cool | 13:39 |
| notmyname | peluse_: thanks for your help on both the feature and getting the intel quote :-) | 13:39 |
| notmyname | ahale: it also makes you realize just how expensive some operations are :-) | 13:40 |
| peluse_ | notmyname: sure, I haven't seen the full press release that was going out via business wire yet (with the quote), gues that's coming later today... | 13:40 |
| notmyname | peluse_: nope it's out already (6am pacific--thus why I'm awake and at the computer) | 13:40 |
| ahale | notmyname: indeed - some ops can be insane - especially when you start adding origin management via SOS | 13:40 |
| notmyname | peluse_: http://www.businesswire.com/news/home/20140708005491/en/OpenStack-Object-Storage-Swift-Adds-Storage-Policies#.U7v0p41dU_M | 13:41 |
| peluse_ | ahh, thanks | 13:41 |
| notmyname | ahale: ya, and I'm not tracking any daemon processes yet either. | 13:42 |
| *** robsparker has joined #openstack-swift | 13:45 | |
| madhuri | How does OPTIONS request gets resolved at proxy level for acount, container, objects? | 13:50 |
| madhuri | when the Controllers obtained from get_controller method is one of them and none of the storage server has the OPTIONS method defined? | 13:50 |
| *** foexle has quit IRC | 13:50 | |
| notmyname | madhuri: it's done just at the proxy level. eg look at the Allows header in the response to see the allowed verbs for that resource | 13:51 |
| bgmccollum | notmyname fifieldt: one other nice addition would be the ordering of the calls. would probably require more state to be passed around than desired, although the transaction ID is already being passed around...hmm | 13:54 |
| notmyname | bgmccollum: maybe the high-res timestamps could be used for that | 13:55 |
| bgmccollum | is that a log configuration? | 13:56 |
| madhuri | Yes, I have got that. But how is the request is processed itself by the proxy server, when the controller is one of the storage servers(account, container, object) in handle_request method? | 13:56 |
| madhuri | https://github.com/openstack/swift/blob/stable/icehouse/swift/proxy/server.py#L360 | 13:56 |
| madhuri | notmyname: Sorry, but I couldn't get it | 13:57 |
| *** zul has quit IRC | 14:08 | |
| *** zul has joined #openstack-swift | 14:08 | |
| *** tsg has quit IRC | 14:14 | |
| *** mwstorer has joined #openstack-swift | 14:15 | |
| *** acoles_ has joined #openstack-swift | 14:18 | |
| *** ChanServ sets mode: +v acoles_ | 14:18 | |
| *** tsg has joined #openstack-swift | 14:41 | |
| *** zhiyan is now known as zhiyan_ | 15:00 | |
| *** kevinc_ has joined #openstack-swift | 15:05 | |
| *** chandankumar has quit IRC | 15:07 | |
| *** bkopilov has quit IRC | 15:12 | |
| openstackgerrit | gholt proposed a change to openstack/swift: This setting should've been set to False back ... https://review.openstack.org/105483 | 15:18 |
| gholt | notmyname: https://review.openstack.org/105483 :/ | 15:23 |
| omame | ftw | 15:25 |
| notmyname | gholt: :-( | 15:33 |
| notmyname | fifieldt: still awake? | 15:33 |
| fifieldt | yup | 15:33 |
| fifieldt | in staff meeting :) | 15:33 |
| notmyname | fifieldt: pm | 15:33 |
| gholt | notmyname: I'm amazed it took 2 years to find out that spot I missed; but there it was, lurking. | 15:36 |
| *** acoles_ has quit IRC | 15:38 | |
| notmyname | gholt: looks like it was being set if the config var is explicitly set, though. just the default wasn't actually doing the right thing | 15:40 |
| notmyname | hmm...well, even the default sets it to false | 15:41 |
| gholt | notmyname: Yeah, it was fine for all core Swift stuff; just any new tools written outside the core that weren't aware they needed to set that false would get bit. | 15:42 |
| notmyname | ya, that :-) | 15:42 |
| notmyname | gholt: +2, would approve again | 15:42 |
| gholt | That was /not/ my intention, lol. "Missed a spot!" | 15:42 |
| gholt | I did consider a test, but the test would just be: Did Python correctly interpret = False? Which seemed silly but I'll add it if folks want. :) | 15:44 |
| *** kevinc_ has quit IRC | 15:45 | |
| *** robsparker has quit IRC | 15:51 | |
| *** kevinc_ has joined #openstack-swift | 15:53 | |
| notmyname | gholt: cool, looks like portante just approved it (welcome back portante!). and I see you've already rolled it into your prod repo | 15:55 |
| gholt | notmyname: Yeah, it attacked one of our clusters so we had to get going on it. | 15:56 |
| openstackgerrit | A change was merged to openstack/swift: Document the unnecessary method invoking in FakeRing https://review.openstack.org/103407 | 15:58 |
| *** madhuri has quit IRC | 15:59 | |
| notmyname | so, gentle reminder: don't host user-generated content on the same domain as sensitive info (eg login cookies). this includes hosting swift at the same domain as your web app | 16:00 |
| notmyname | today's proof of concept (not swift-specific): http://miki.it/blog/2014/7/8/abusing-jsonp-with-rosetta-flash/ | 16:00 |
| *** swat30 has joined #openstack-swift | 16:00 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Remove fake _get_part and use the real thing https://review.openstack.org/103962 | 16:00 |
| gholt | notmyname: Can't load that here (probably something blocking) but I assume it's the cookies are good for more than you'd expect thing. | 16:06 |
| *** cpen has joined #openstack-swift | 16:06 | |
| hugokuo | Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 Swift 2.0 !!!!!! | 16:07 |
| notmyname | gholt: it's an attack on jsonp endpoints with swf files | 16:07 |
| notmyname | hugokuo: seems you are excited ;-) | 16:07 |
| hugokuo | notmyname: I did .... am writing the ChangeLog of it in Chinese on my blog.... | 16:07 |
| peluse_ | here here | 16:08 |
| notmyname | gholt: really cool hack, from a technical perspective. he crafted a file so that the huffman encoding generates just alphnumeric values, which allows it to be *something*something* witht he jsonp endpoints to grab data at sensitive domains | 16:08 |
| gholt | notmyname: Hmm, maybe I'll load it later when I'm on my home network then. | 16:08 |
| peluse_ | hugokuo: nice updates to the latest version of the book BTW - just sent some feedback to the publisher this morning on the installation chapter | 16:09 |
| notmyname | ya, definitely worth reading, if just from a "wow he's really smart and that's really cool" perspective | 16:09 |
| gholt | The thing I was talking about was I was running two copies of a service, https, on two different ports. But the cookies are not separate by port, so each was sharing cookies with the other. Caused great hilarity. | 16:10 |
| notmyname | gholt: right. I remember when we got lucky with cloud files since it's at clouddrive.com and not the same domain as the control panel | 16:11 |
| hugokuo | peluse_: good to know that. Thanks for reviewing it ....... | 16:15 |
| peluse_ | hugokuo: you bet, shoot me an email or catch me here if you have questions on the comments | 16:16 |
| *** thurloat has joined #openstack-swift | 16:16 | |
| *** chandankumar has joined #openstack-swift | 16:17 | |
| *** gyee has joined #openstack-swift | 16:18 | |
| *** infotection has quit IRC | 16:23 | |
| * notmyname commutes | 16:30 | |
| * torgomatic is non-Abelian | 16:31 | |
| *** infotection has joined #openstack-swift | 16:32 | |
| *** bkopilov has joined #openstack-swift | 16:39 | |
| *** infotection has quit IRC | 16:39 | |
| *** aswadr has quit IRC | 16:42 | |
| *** andyandy has quit IRC | 16:44 | |
| *** mmcardle has quit IRC | 16:46 | |
| *** infotection has joined #openstack-swift | 16:52 | |
| tsg | notmyname: should we rebase the feature/ec branch against 2.0.0 now that master is tagged | 17:01 |
| openstackgerrit | Tushar Gohad proposed a change to openstack/swift: EC: Add support for policy types, 'erasure_coding' policy https://review.openstack.org/103644 | 17:02 |
| peluse_ | tsg: ^ nice! | 17:13 |
| *** chandankumar has quit IRC | 17:13 | |
| *** acoles has joined #openstack-swift | 17:14 | |
| *** ChanServ sets mode: +v acoles | 17:14 | |
| tsg | peluse_: a lot of rework on the other 2 patches .. coming up soon | 17:19 |
| notmyname | tsg: yes, like with the SP work, we should frequently merge master into feature/ec | 17:20 |
| peluse_ | tsg: great, looking forward to it! | 17:20 |
| peluse_ | notmyname: yup, once a week I think worked out nicely | 17:20 |
| tsg | notmyname: if I am able to submit the merge commit for review, I can do that regularly | 17:20 |
| notmyname | tsg: I think on core can submit a merge commit | 17:21 |
| peluse_ | tsg: has to be done by core for permissions reasons | 17:21 |
| peluse_ | jinx | 17:21 |
| notmyname | tsg: peluse_: but this is something we should discuss at the paris summit. consider supporting a feature-branch-style workflow | 17:22 |
| tsg | notmyname, peluse_: that'd be cool | 17:22 |
| openstackgerrit | John Dickinson proposed a change to openstack/swift: Merge branch 'master' into feature/ec https://review.openstack.org/105536 | 17:36 |
| notmyname | peluse_: tsg: done ^^ | 17:37 |
| notmyname | easy this time since there are zero conflicts :-) | 17:37 |
| tsg | notmyname: thanks! :) hopefully when I rebase my patches, there are zero conflicts as well ;) | 17:37 |
| tsg | less likely :) | 17:38 |
| *** kevinc_ has quit IRC | 17:39 | |
| openstackgerrit | A change was merged to openstack/swift: This setting should've been set to False back ... https://review.openstack.org/105483 | 17:42 |
| notmyname | gholt: ^^ done :-) | 17:45 |
| *** mmcardle has joined #openstack-swift | 17:51 | |
| *** kevinc_ has joined #openstack-swift | 17:58 | |
| *** Tyger has joined #openstack-swift | 18:05 | |
| *** mmcardle has quit IRC | 18:11 | |
| tsg | notmyname: didn't know that the Jenkins environment was Centos 6 | 18:39 |
| * tsg goes to validate pyeclib on Centos 6 | 18:40 | |
| * tsg just found out about that on #openstack-infra | 18:40 | |
| torgomatic | tsg: yeah, the python 2.6 test boxes are Cent6... I have a whole bunch of SkipTest stuff in the zero-copy patches because of that :) | 18:45 |
| *** shri has joined #openstack-swift | 18:48 | |
| tsg | torgomatic: ah, didn't notice that earlier in test_diskfile.py :) | 18:50 |
| tsg | torgomatic: in my case, it is probably pyeclib reporting false success on "pip install" | 18:51 |
| torgomatic | tsg: that sounds like it could cause problems; is it a case of stuff compiling but being unable to run, or just a failure to propagate a non-zero exit upwards? | 18:52 |
| tsg | torgomatic: only a guess thus far. setting up a centos6 VM to test as I type .. | 18:53 |
| tsg | mwstorer, elambert: looks like Jenkins for pyeclib/liberasurecode is in fact going to be useful - and we should get Centos6 in there! | 18:54 |
| *** gyee has quit IRC | 19:16 | |
| *** kevinc_ has quit IRC | 19:24 | |
| *** kevinc_ has joined #openstack-swift | 19:58 | |
| *** Tyger has quit IRC | 19:59 | |
| *** shri1 has joined #openstack-swift | 20:03 | |
| *** shri1 has quit IRC | 20:03 | |
| *** shri has quit IRC | 20:05 | |
| *** Tyger has joined #openstack-swift | 20:06 | |
| *** fifieldt has quit IRC | 20:15 | |
| *** gyee has joined #openstack-swift | 20:22 | |
| *** gyee has quit IRC | 20:27 | |
| *** fifieldt has joined #openstack-swift | 20:28 | |
| *** gyee has joined #openstack-swift | 20:32 | |
| openstackgerrit | Thiago da Silva proposed a change to openstack/swift: Refactoring functional tests https://review.openstack.org/92643 | 20:37 |
| *** jasond has left #openstack-swift | 20:38 | |
| openstackgerrit | paul luse proposed a change to openstack/swift: Initial Erasure Code Docs https://review.openstack.org/104713 | 20:58 |
| *** tsg has quit IRC | 21:04 | |
| *** Tyger has quit IRC | 21:06 | |
| *** Tyger has joined #openstack-swift | 21:07 | |
| *** acoles has quit IRC | 21:39 | |
| *** acoles has joined #openstack-swift | 21:44 | |
| *** ChanServ sets mode: +v acoles | 21:44 | |
| *** cpen has quit IRC | 22:01 | |
| mattoliverau | Morning all | 22:05 |
| *** tsg has joined #openstack-swift | 22:06 | |
| peluse_ | good morning! | 22:07 |
| *** tusharsg has joined #openstack-swift | 22:08 | |
| *** tgohad has joined #openstack-swift | 22:09 | |
| *** tsg has quit IRC | 22:11 | |
| *** tusharsg has quit IRC | 22:12 | |
| notmyname | mattoliverau: hi | 22:13 |
| notmyname | mattoliverau: I grabbed your patch to logflow (thanks!) and added a small edit | 22:14 |
| mattoliverau | notmyname: you're welcome, I got side tracked yesterday and played with it, obviously :) Did your swift bench complete? | 22:17 |
| notmyname | mattoliverau: ya, eventually. but either the png was corrupted or just too big to open (it was 1.8 million pixels wide) | 22:17 |
| torgomatic | so 5.4 MB RAM for each row once decoded... I think it might have been a little big ;) | 22:19 |
| mattoliverau | notmyname: with the new change there should be less edges :) | 22:20 |
| notmyname | ya | 22:20 |
| notmyname | torgomatic: ? | 22:21 |
| notmyname | mattoliverau: whoa!! | 22:22 |
| notmyname | mattoliverau: running it again with your changes, and it finished in 0.7 seconds | 22:22 |
| notmyname | 8195 x 431 | 22:22 |
| notmyname | http://d.not.mn/saio_bench_logflow.png | 22:23 |
| mattoliverau | Much better! G.add_edge must be really slow, with mine we only do it once per unique edge :) | 22:24 |
| notmyname | ya | 22:24 |
| notmyname | actually, no | 22:24 |
| notmyname | it's not add_edge | 22:24 |
| notmyname | it was the layout | 22:24 |
| openstackgerrit | A change was merged to openstack/swift: Merge branch 'master' into feature/ec https://review.openstack.org/105536 | 22:24 |
| notmyname | hmm... `swift-bench --saio` ends up making 1138 calls to auth | 22:25 |
| notmyname | seems....excessive | 22:25 |
| mattoliverau | Ahh, yes looks like there are just a few more calls! | 22:25 |
| mattoliverau | I like how you moved the count to the front, great idea | 22:25 |
| mattoliverau | notmyname: I like the Head (204) to the account server that does a loop (near the auth) | 22:28 |
| notmyname | ya. I'm impressed with dot | 22:28 |
| *** kevinc_ has quit IRC | 22:43 | |
| *** kevinc_ has joined #openstack-swift | 22:43 | |
| *** tgohad has quit IRC | 22:57 | |
| *** sungju has joined #openstack-swift | 23:00 | |
| *** zaitcev has joined #openstack-swift | 23:13 | |
| *** ChanServ sets mode: +v zaitcev | 23:13 | |
| *** dencaval has joined #openstack-swift | 23:15 | |
| *** kevinc_ has quit IRC | 23:18 | |
| *** kevinc_ has joined #openstack-swift | 23:19 | |
| creiht | notmyname: heh, it is only supposed to reauth on 401 | 23:19 |
| creiht | but it does use the python-swiftclient, so who knows? :) | 23:19 |
| notmyname | :-) | 23:19 |
| notmyname | mattoliverau: I just pushed a change to correctly set the edge weight. instead of setting "weight", you set "penwidth" (of course!) | 23:20 |
| zaitcev | Of course. :-) | 23:20 |
| notmyname | creiht: on the other hand, it's 1138, so you know it's got good geek cred (http://en.wikipedia.org/wiki/THX_1138) | 23:21 |
| creiht | heh | 23:21 |
| creiht | oh yeah it is an easter egg :) | 23:21 |
| mattoliverau | notmyname: lol, of course! :P | 23:22 |
| creiht | zaitcev: so are you seriously going to start working on manilla? | 23:23 |
| notmyname | creiht: speaking of the client, torgomatic was telling me that once he groked https://review.openstack.org/#/c/85453/ it's actually a pretty good thing. makes it so the CLI and SDK of swiftclient can share functionality (like uploading manifests) | 23:25 |
| openstackgerrit | John Dickinson proposed a change to openstack/swift: Merge tag '2.0.0' https://review.openstack.org/105222 | 23:32 |
| notmyname | seems that wasn't auto-merged in. ^^ | 23:32 |
| notmyname | landing it now | 23:32 |
| *** wer has quit IRC | 23:33 | |
| *** shri has joined #openstack-swift | 23:34 | |
| zaitcev | creiht: I have... BTW, Jason (one of your RAX guys) told me that Manila was #1 request from customers. | 23:47 |
| zaitcev | creiht: well I hope his name was Jason. he's manager other than Scott. | 23:47 |
| zaitcev | brb | 23:47 |
| zaitcev | or actuall ttyl | 23:48 |
| *** mwstorer has quit IRC | 23:56 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
