| *** hurricanerix has quit IRC | 00:09 | |
| notmyname | from the openstack mailing list: "Also, I have added a section noting crucial bugs/patches that are blocking Minesweeper." | 00:10 |
|---|---|---|
| notmyname | can I respond with, "just put flags around them and move on"? | 00:10 |
| torgomatic | YES | 00:11 |
| torgomatic | in fact, I think you're legally obligated to do so | 00:11 |
| *** _bluev has quit IRC | 00:19 | |
| *** _bluev has joined #openstack-swift | 00:20 | |
| *** matsuhashi has joined #openstack-swift | 00:20 | |
| *** jergerber has quit IRC | 00:22 | |
| *** joeljwright has joined #openstack-swift | 00:38 | |
| *** joeljwright has quit IRC | 00:42 | |
| *** nosnos has joined #openstack-swift | 00:56 | |
| *** tdasilva has left #openstack-swift | 01:12 | |
| *** igor has joined #openstack-swift | 01:15 | |
| *** igor____ has quit IRC | 01:16 | |
| *** zackf has joined #openstack-swift | 01:27 | |
| *** zackf has quit IRC | 01:35 | |
| *** shri has quit IRC | 01:38 | |
| zaitcev | Guys | 01:38 |
| zaitcev | Imagine I want to review what's up with storage policies. | 01:38 |
| zaitcev | Since it has its own branch... I cannot review what's already merged into that branch. | 01:39 |
| *** joeljwright has joined #openstack-swift | 01:40 | |
| portante | what about a query against gerrit for all merged commits against feature/ec? | 01:40 |
| torgomatic | alternately, pull the branch and look at the commits; they're all in git | 01:42 |
| zaitcev | I know how to use git in this case. But I do not see other option except post an old-fashioned review to openstack-dev and cc: peluse | 01:43 |
| torgomatic | true; you could propose patches to fix things up if you see problems | 01:43 |
| torgomatic | does Gerrit not let you post comments on merged reviews/ | 01:43 |
| torgomatic | ? | 01:43 |
| portante | it does | 01:43 |
| torgomatic | Gerrit query: project:openstack/swift branch:feature/ec | 01:44 |
| zaitcev | ah, okay. I can drop "status:open" from gerrit URL | 01:44 |
| *** joeljwright has quit IRC | 01:44 | |
| *** bada_ has quit IRC | 01:45 | |
| torgomatic | status:merged will even let you skip the abandoned ones | 01:45 |
| zaitcev | wooow the first commit is "update .gitreview with defaultbranch=feature/ec" | 01:46 |
| zaitcev | no fans of Blitzkrieg I see | 01:47 |
| *** joeljwright has joined #openstack-swift | 02:40 | |
| *** joeljwright has quit IRC | 02:45 | |
| *** jasondotstar has joined #openstack-swift | 02:50 | |
| *** briancline has quit IRC | 03:10 | |
| *** briancline has joined #openstack-swift | 03:14 | |
| briancline | just catching up in the scrollback from earlier | 03:16 |
| briancline | portante: torgomatic: trust me, as a mostly-sane adult, I'm quite aware that some utopian ideal of fairness doesn't exist | 03:18 |
| briancline | just voicing a simple observation is all -- it was to some degree a slightly related topic wrt one of the final meeting topics today | 03:19 |
| *** _bluev has quit IRC | 03:23 | |
| portante | briancline: okay | 03:23 |
| zaitcev | Man, I'm like an hour into the policies and I'm already mad. | 03:31 |
| zaitcev | I'm going to ask peluse tomorrow why he ever thought that exposing policy index might be a good idea. | 03:32 |
| portante | zaitcev: careful | 03:32 |
| zaitcev | portante: Sure. | 03:32 |
| portante | there have been many design discussions on this topic for 6 months | 03:32 |
| zaitcev | daaarn | 03:32 |
| portante | it is not all on peluse | 03:32 |
| zaitcev | But what are they for, however? | 03:33 |
| portante | so let's walk through it a bit | 03:33 |
| portante | where are you seeing the policy index exposed that you consider it a problem? | 03:33 |
| *** odyssey4me has joined #openstack-swift | 03:35 | |
| zaitcev | I'm building the graph of all the info leakage through all the get_by_index, get_diskfile_through_hash, and so on. | 03:35 |
| portante | info leakage? | 03:35 |
| zaitcev | call it knowledge leakage | 03:36 |
| portante | isn't the policy index passed as an argument to the object server? | 03:36 |
| zaitcev | I haven't established that user on the other side of proxy knows anything about indexes | 03:36 |
| *** odyssey4me2 has joined #openstack-swift | 03:39 | |
| *** odyssey4me has quit IRC | 03:40 | |
| zaitcev | Maybe I'm being worked up over nothing. Say, the indexing is in swift.conf. But it also has the main hash suffix, which you cannot afford to lose anyway. So it's not like one could recover any data if anything happened to that information. | 03:40 |
| *** joeljwright has joined #openstack-swift | 03:42 | |
| *** joeljwright has quit IRC | 03:46 | |
| zaitcev | https://review.openstack.org/70497 - okay that's a good one | 03:48 |
| notmyname | zaitcev: the storage policy index is the canonical identifier for the policy. and yes, I agree that it leaked too much. that's what prompted me to talk to peluse about it, and all of the patches related to that haven't landed yet | 03:50 |
| zaitcev | notmyname: thanks. | 03:51 |
| notmyname | but the index is used to allow the provider to migrate (ie everyone today is using storage policy 0), and it will allow for a policy to have one or more names | 03:51 |
| zaitcev | BTW, does anyone know why it happens like this... If I look at the git log, it says change Change-Id: I71eab0788.... added policies to DiskFile. But if I look at Gerrit, it's a different commit at all and Change-Id is I70189013. | 03:52 |
| zaitcev | notmyname: I don't understand how anything can be migrated if indexes are in container meta. Once they get stuck there, you can never change policy index of a given policy, or can you? | 03:54 |
| notmyname | correct. and that's why the storage policy can only be set at container creation and can't be changed | 03:54 |
| notmyname | the policy reconciler that torgomatic is working on handles the split-brain case when the same container is created with 2 different policies | 03:55 |
| zaitcev | hmm. Once thing is clear, I tuned out your discussions due to PBE and other things and you probably talked about all this. | 03:55 |
| zaitcev | portante: in case you have a second, how do you know that get_hashes() is not going to recurse endlessly? | 04:01 |
| portante | get_hashes in diskfile.py? | 04:02 |
| zaitcev | yes | 04:02 |
| * portante looks | 04:02 | |
| * portante assumes ec branch ... | 04:02 | |
| notmyname | zaitcev: I'm going to wander away from my computer now. if you've got more questions, portante will answer them ;-) | 04:02 |
| zaitcev | I think it does in case there was some kind of conflict | 04:02 |
| notmyname | (or jsut ask me tomorrow) | 04:02 |
| portante | ;) | 04:02 |
| *** zackf has joined #openstack-swift | 04:06 | |
| portante | zaitcev: so regarding the module level get_hashes() method | 04:07 |
| portante | It can recurse for ever | 04:08 |
| portante | and if there is a python implementation that does not do tail-call optimizations, this will be a problem | 04:08 |
| zaitcev | but it's very unlikely, right? | 04:08 |
| portante | CPython, for better or worse (in some cases) will optimize that call away so that there is no effective stack usage for the call | 04:09 |
| zaitcev | Looks like it's completely in service of replicator, which ought to retry if someone overflows its stack and crashes. | 04:09 |
| zaitcev | Nice. | 04:09 |
| portante | but it won't overflow the stack | 04:09 |
| zaitcev | Thanks! | 04:09 |
| portante | now, run that under pypy, or jython or something else, that fails to do tail call opts, and boom | 04:09 |
| *** joeljwright has joined #openstack-swift | 04:43 | |
| openstackgerrit | Amala Basha proposed a change to openstack/python-swiftclient: swift_client needs to close http connection https://review.openstack.org/55132 | 04:44 |
| *** basha has joined #openstack-swift | 04:44 | |
| basha | zaitcev: chmouel : have rebased change https://review.openstack.org/#/c/55132/3. Please can you review it before it runs off to abandoned again. :) | 04:45 |
| zaitcev | basha: I'll look. | 04:46 |
| basha | thanks zaitcev :) | 04:46 |
| *** hurricanerix has joined #openstack-swift | 04:46 | |
| *** joeljwright has quit IRC | 04:47 | |
| *** hurricanerix has quit IRC | 04:47 | |
| *** zackf has quit IRC | 04:51 | |
| *** hurricanerix has joined #openstack-swift | 04:52 | |
| *** hurricanerix has quit IRC | 04:52 | |
| *** jasondotstar has quit IRC | 04:57 | |
| *** nosnos_ has joined #openstack-swift | 05:02 | |
| *** nosnos has quit IRC | 05:06 | |
| *** nshaikh has joined #openstack-swift | 05:19 | |
| *** madhuri has joined #openstack-swift | 05:23 | |
| *** zaitcev has quit IRC | 05:33 | |
| *** joeljwright has joined #openstack-swift | 05:44 | |
| *** joeljwright has quit IRC | 05:49 | |
| *** matsuhashi has quit IRC | 06:07 | |
| *** matsuhashi has joined #openstack-swift | 06:07 | |
| *** matsuhas_ has joined #openstack-swift | 06:09 | |
| *** matsuhashi has quit IRC | 06:09 | |
| *** igor has quit IRC | 06:18 | |
| *** igor has joined #openstack-swift | 06:18 | |
| *** ppai has joined #openstack-swift | 06:19 | |
| *** nosnos has joined #openstack-swift | 06:40 | |
| *** nosnos_ has quit IRC | 06:42 | |
| *** joeljwright has joined #openstack-swift | 06:45 | |
| *** joeljwright has quit IRC | 06:49 | |
| *** mfactor has joined #openstack-swift | 06:54 | |
| *** pheadron has joined #openstack-swift | 06:55 | |
| *** kragniz has quit IRC | 07:03 | |
| *** mfactor has quit IRC | 07:04 | |
| *** kragniz has joined #openstack-swift | 07:05 | |
| *** pheadron has quit IRC | 07:09 | |
| *** joeljwright has joined #openstack-swift | 07:10 | |
| *** bvandenh has joined #openstack-swift | 07:12 | |
| *** joeljwright has left #openstack-swift | 07:28 | |
| *** joeljwright has joined #openstack-swift | 07:28 | |
| *** mfactor has joined #openstack-swift | 07:29 | |
| *** foexle has joined #openstack-swift | 07:49 | |
| *** xga has joined #openstack-swift | 08:07 | |
| *** joeljwright has quit IRC | 08:14 | |
| *** joeljwright has joined #openstack-swift | 08:15 | |
| *** mlipchuk has joined #openstack-swift | 08:22 | |
| *** ben_roeder has joined #openstack-swift | 08:31 | |
| *** ben_roeder has quit IRC | 08:39 | |
| *** mmcardle has joined #openstack-swift | 08:42 | |
| *** nacim has joined #openstack-swift | 08:51 | |
| *** fbo is now known as fbo_away | 09:01 | |
| *** odyssey4me2 is now known as odyssey4me | 09:01 | |
| *** fbo_away is now known as fbo | 09:06 | |
| *** mlipchuk has quit IRC | 09:08 | |
| *** mlipchuk has joined #openstack-swift | 09:22 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Limit logged headers in proxy_logging middleware https://review.openstack.org/64363 | 09:27 |
| *** igor_ has joined #openstack-swift | 09:30 | |
| *** igor has quit IRC | 09:32 | |
| *** ben_roeder has joined #openstack-swift | 09:44 | |
| openstackgerrit | Christian Schwede proposed a change to openstack/swift: Remove swiftclient dependency https://review.openstack.org/65660 | 09:45 |
| *** xga has quit IRC | 09:57 | |
| *** ben_roeder has quit IRC | 09:58 | |
| *** jasondotstar has joined #openstack-swift | 10:00 | |
| *** xga has joined #openstack-swift | 10:05 | |
| *** xga_ has joined #openstack-swift | 10:07 | |
| *** xga has quit IRC | 10:08 | |
| *** ben_roeder has joined #openstack-swift | 10:10 | |
| openstackgerrit | Eamonn O'Toole proposed a change to openstack/swift: Parallel object auditor https://review.openstack.org/59778 | 10:12 |
| *** mlipchuk has quit IRC | 10:17 | |
| *** ppai has quit IRC | 10:32 | |
| *** mlipchuk has joined #openstack-swift | 10:32 | |
| *** odyssey4me2 has joined #openstack-swift | 10:33 | |
| *** odyssey4me has quit IRC | 10:34 | |
| *** odyssey4me2 has quit IRC | 10:38 | |
| *** psharma has joined #openstack-swift | 10:44 | |
| *** ppai has joined #openstack-swift | 10:49 | |
| *** jasondotstar has quit IRC | 11:00 | |
| *** gvernik_ has joined #openstack-swift | 11:10 | |
| openstackgerrit | Prashanth Pai proposed a change to openstack/swift: Make .expiring_objects account name configurable https://review.openstack.org/70978 | 11:11 |
| *** mef has joined #openstack-swift | 11:25 | |
| *** mfactor has quit IRC | 11:25 | |
| *** fbo is now known as fbo_away | 11:35 | |
| *** jasondotstar has joined #openstack-swift | 11:50 | |
| *** pheadron has joined #openstack-swift | 12:03 | |
| *** basha has quit IRC | 12:05 | |
| *** ben_roeder has quit IRC | 12:12 | |
| *** Anju has joined #openstack-swift | 12:19 | |
| *** igor has joined #openstack-swift | 12:45 | |
| *** igor_ has quit IRC | 12:46 | |
| *** ben_roeder has joined #openstack-swift | 12:47 | |
| *** Trixboxer has joined #openstack-swift | 12:50 | |
| *** fbo_away is now known as fbo | 12:55 | |
| *** mlipchuk has quit IRC | 12:56 | |
| *** jasondotstar has quit IRC | 13:00 | |
| *** psharma has quit IRC | 13:10 | |
| *** mlipchuk has joined #openstack-swift | 13:10 | |
| *** jasondotstar has joined #openstack-swift | 13:19 | |
| *** basha has joined #openstack-swift | 13:47 | |
| *** mfactor has joined #openstack-swift | 13:58 | |
| *** basha has quit IRC | 14:00 | |
| *** ppai has quit IRC | 14:01 | |
| *** mef has quit IRC | 14:01 | |
| openstackgerrit | A change was merged to openstack/swift: Fix a couple ifs in account quota middleware. https://review.openstack.org/71405 | 14:08 |
| *** gvernik_ has quit IRC | 14:08 | |
| *** mmcardle has quit IRC | 14:15 | |
| *** ppai has joined #openstack-swift | 14:18 | |
| *** ben_roeder has quit IRC | 14:18 | |
| *** xga has joined #openstack-swift | 14:24 | |
| *** ppai has quit IRC | 14:25 | |
| *** mvenesio has joined #openstack-swift | 14:26 | |
| *** mmcardle has joined #openstack-swift | 14:33 | |
| *** mkollaro has joined #openstack-swift | 14:34 | |
| *** mfactor has quit IRC | 14:36 | |
| *** mef has joined #openstack-swift | 14:36 | |
| mvenesio | portante, notmyname, creiht: Hi guys we are choosing hardware to expand our swift pictures cluster, do you have any recommendations regarding HDDs ? , we are thinking in the seagate 6Gb SAS constellation es3, but also could be some SATA3 drive. Do you have some experience to suggest us ? | 14:37 |
| *** ben_roeder has joined #openstack-swift | 14:42 | |
| *** mvenesio_ has joined #openstack-swift | 14:45 | |
| *** tongli has joined #openstack-swift | 14:46 | |
| *** mvenesio has quit IRC | 14:48 | |
| *** zackf has joined #openstack-swift | 14:55 | |
| *** zackf has quit IRC | 15:00 | |
| *** zackf has joined #openstack-swift | 15:08 | |
| portante | mvenesio: do you need disk speed, or capacity density? | 15:15 |
| *** matsuhas_ has quit IRC | 15:17 | |
| *** nosnos has quit IRC | 15:21 | |
| *** ben_roeder has quit IRC | 15:24 | |
| mvenesio_ | portante: disk speed and capacity density :P | 15:24 |
| mvenesio_ | portante: but at this time, disk speed | 15:24 |
| portante | what do you have for disk speeds on your drives now? | 15:25 |
| portante | 54, 72, 100, ... 150? | 15:25 |
| portante | how much money can I help you spend?! | 15:25 |
| *** nshaikh has left #openstack-swift | 15:26 | |
| mvenesio_ | portante: nothing, i'm not using flashcache or something like that, we just have SATA3 disks and SSDs for account & container ops | 15:27 |
| portante | I meant, 5,400 RPM, 7,200, etc. | 15:27 |
| portante | are your objects large? and what do you have for network between clients and proxy, proxy and obj store? | 15:28 |
| mvenesio_ | portante: 7,2k rpm toshiba SATA3 2TB drives with 64MB cache | 15:28 |
| mvenesio_ | portante: we use 10GB interfaces, and our user case is billons of small objects | 15:29 |
| *** xga has quit IRC | 15:29 | |
| mvenesio_ | portante: i'm from mercadolibre | 15:30 |
| *** xga has joined #openstack-swift | 15:30 | |
| *** xga__ has joined #openstack-swift | 15:30 | |
| mvenesio_ | portante: we store all our site pictures in swift | 15:30 |
| *** xga_ has quit IRC | 15:30 | |
| mvenesio_ | portante: just for know if you have some disk model that recommend us, or you tested SATA3 vs SAS in your environment. | 15:33 |
| mvenesio_ | portante: the disks that we are considering now are seagate SAS 2TB with 128MB cache, constellation es3 series | 15:34 |
| mvenesio_ | portante: money is not an issue with this mission critical clusters, but we must to find a good balance | 15:37 |
| *** ben_roeder has joined #openstack-swift | 15:41 | |
| *** lpabon has joined #openstack-swift | 15:46 | |
| *** ben_roeder has quit IRC | 15:46 | |
| portante | mvenesio_: let me look ... | 15:46 |
| mvenesio_ | portante: thankss | 15:47 |
| *** xga_ has joined #openstack-swift | 15:50 | |
| *** xga__ has quit IRC | 15:50 | |
| *** xga has quit IRC | 15:50 | |
| *** xga has joined #openstack-swift | 15:50 | |
| *** xga_ has quit IRC | 15:52 | |
| *** xga has quit IRC | 15:52 | |
| *** xga_ has joined #openstack-swift | 15:52 | |
| *** xga has joined #openstack-swift | 15:52 | |
| *** mvenesio_ has quit IRC | 16:02 | |
| *** zackf has quit IRC | 16:02 | |
| *** zackf has joined #openstack-swift | 16:03 | |
| portante | mvenesio_: I have seagate's datasheet, at http://www.seagate.com/files/www-content/product-content/_cross-product/en-us/docs/storage-solution-guide-oct-13-ssg1351-14-1310us.pdf | 16:04 |
| portante | is that what you using for the model names? | 16:04 |
| portante | I want to be sure I am seeing the same drive information specs that you are | 16:04 |
| notmyname | good morning | 16:07 |
| *** mvenesio has joined #openstack-swift | 16:08 | |
| portante | notmyname: mornin' | 16:08 |
| notmyname | swift meeting time updated to be every week (same bat time, same bat channel) | 16:08 |
| portante | mvenesio: I have seagate's datasheet, at | 16:08 |
| portante | http://www.seagate.com/files/www-content/product-content/_cross-product/en-us/docs/storage-solution-guide-oct-13-ssg1351-14-1310us.pdf | 16:08 |
| portante | notmyname: great, thanks | 16:08 |
| notmyname | mvenesio: FWIW, we've had some customers use seagate constellation drives (I'm not sure of the exact model number) | 16:08 |
| notmyname | mvenesio: in general, the "SSDs for account + container and SATA for objects" is a good strategy. are you seeing the disk IO as a bottleneck now? what's prompting you to look at new/different drives? | 16:09 |
| mvenesio | notmyname: humm there is not a special reason, our vendors are quoting us storage nodes with SAS 7.2rpm drives and we have no experience with this kind of drives in swift | 16:12 |
| notmyname | ah. I don't think I know of anyone using SAS drives with Swift. not that it won't work, but it's just not too cost effective | 16:13 |
| mvenesio | notmyname: okkk, seems that there is not a significant difference between prices | 16:17 |
| notmyname | mvenesio: are you just looking at new stuff for your capacity expansion? | 16:17 |
| mvenesio | notmyname: yes | 16:18 |
| mvenesio | notmyname: we'll expand the pictures cluster to the maximun size | 16:18 |
| notmyname | mvenesio: cool | 16:18 |
| notmyname | mvenesio: what version of swift are you currently running? | 16:19 |
| notmyname | I know you've had it in prod for a long time | 16:19 |
| mvenesio | notmyname: yes we hope | 16:19 |
| notmyname | mvenesio: did you see the backblaze post on drives a few weeks ago? http://blog.backblaze.com/2014/01/21/what-hard-drive-should-i-buy/ | 16:20 |
| notmyname | mvenesio: something else you may consider as you run later versions of swift is to test out the per-drive threadpool instead of your current "one object server process per drive" pattern | 16:23 |
| *** byeager has joined #openstack-swift | 16:25 | |
| portante | mvenesio: sorry I dropped off, client issues to my znc server | 16:30 |
| *** xga__ has joined #openstack-swift | 16:31 | |
| *** xga has quit IRC | 16:32 | |
| *** xga_ has quit IRC | 16:32 | |
| *** mlipchuk has quit IRC | 16:36 | |
| *** xga has joined #openstack-swift | 16:36 | |
| mvenesio | notmyname: sorry john, i was in a meeting | 16:37 |
| mvenesio | notmyname: i'll check that post | 16:37 |
| notmyname | talking in IRC during a meeting!?! I'm sure none of the rest of us have ever done that ;-) | 16:37 |
| mvenesio | notmyname: toxic meetings :P | 16:38 |
| notmyname | running funk tests http://www.youtube.com/watch?v=camNwgGXb4s | 16:41 |
| *** ben_roeder has joined #openstack-swift | 16:42 | |
| *** bvandenh has quit IRC | 16:44 | |
| *** ben_roeder has quit IRC | 16:44 | |
| tristanC | hehe, that is a nice funky tune notmyname :) | 16:44 |
| notmyname | I wonder if I can trigger that to play every time I run functests in swift? | 16:45 |
| mvenesio | notmyname: we are running 1.8.0 grizzly release, and we are planning to migrate it to the 1.9 this year | 16:47 |
| notmyname | don't stop at 1.9. we've already released 1.12 :-) | 16:47 |
| mvenesio | notmyname: ok we can go there, and believe me when i say that you will have new about that migration :P | 16:48 |
| mvenesio | notmyname: practically we will going to live to your home | 16:49 |
| *** gyee has joined #openstack-swift | 16:50 | |
| notmyname | come on! I've got some room downstairs ;-) | 16:50 |
| notmyname | mvenesio: just as long as you're ok with a 3 year-old waking up at 5:30am every day | 16:51 |
| mvenesio | notmyname: no problem we'll be coming back for a breakfast at that hour | 16:53 |
| *** CrackerJackMack has quit IRC | 16:54 | |
| *** sudorandom has quit IRC | 16:55 | |
| *** zackf has quit IRC | 16:55 | |
| mvenesio | notmyname, portante: Thanks for your help guys | 16:55 |
| *** CrackerJackMack has joined #openstack-swift | 16:55 | |
| *** sudorandom has joined #openstack-swift | 16:56 | |
| *** zackf has joined #openstack-swift | 17:03 | |
| notmyname | hmm, so it turns out that jenkins reports stopped jobs as failures to graphite | 17:03 |
| *** portante has left #openstack-swift | 17:04 | |
| notmyname | "sometimes it is reported as ABORTED, sometimes as FAILURE" | 17:04 |
| notmyname | so eg the recent dip in swift-gate-docs pass rate seen on http://not.mn/swift_gate_status.html is actually just the result of jobs being cancelled because of a gate reset | 17:05 |
| *** xga has quit IRC | 17:05 | |
| *** xga__ has quit IRC | 17:05 | |
| *** xga__ has joined #openstack-swift | 17:05 | |
| *** xga has joined #openstack-swift | 17:05 | |
| *** portante has joined #openstack-swift | 17:05 | |
| *** ChanServ sets mode: +v portante | 17:05 | |
| *** portante has left #openstack-swift | 17:05 | |
| notmyname | I'm seeing many links to http://java.dzone.com/articles/openstack-object-store-base on twitter this morning. all about using swift as the base for a video CDN | 17:09 |
| *** tuxninja has joined #openstack-swift | 17:11 | |
| openstackgerrit | Greg Lange proposed a change to openstack/swift: Fix a comment in bulk middleware. https://review.openstack.org/71600 | 17:12 |
| *** tuxninja has quit IRC | 17:17 | |
| briancline | is 1.13 planned to be the release coinciding with Icehouse just prior to summit? | 17:26 |
| notmyname | briancline: either 1.13 or 1.14. depending on how the storage policy stuff goes these next 2 weeks, we may choose to do a release with everything just before storage policies lands (ie 1.13) and then another for icehouse (ie 1.14). or if storage policies is going really smoothly and lands more quickly than I expect, we may just have 1.13 of icehouse | 17:28 |
| *** hurricanerix has joined #openstack-swift | 17:28 | |
| *** basha has joined #openstack-swift | 17:35 | |
| *** mvenesio has quit IRC | 17:42 | |
| *** basha has quit IRC | 17:42 | |
| *** mmcardle has quit IRC | 17:58 | |
| *** davidhadas_ has joined #openstack-swift | 18:01 | |
| *** xga has quit IRC | 18:02 | |
| *** xga__ has quit IRC | 18:02 | |
| *** davidhadas has quit IRC | 18:02 | |
| *** fbo is now known as fbo_away | 18:03 | |
| *** xga has joined #openstack-swift | 18:03 | |
| *** xga__ has joined #openstack-swift | 18:03 | |
| *** pberis has quit IRC | 18:13 | |
| *** pberis has joined #openstack-swift | 18:17 | |
| *** shri has joined #openstack-swift | 18:18 | |
| openstackgerrit | A change was merged to openstack/swift: Make .expiring_objects account name configurable https://review.openstack.org/70978 | 18:23 |
| *** tuxninja has joined #openstack-swift | 18:27 | |
| *** bvandenh has joined #openstack-swift | 18:34 | |
| *** igor has quit IRC | 18:40 | |
| *** bvandenh has quit IRC | 18:40 | |
| *** shri1 has joined #openstack-swift | 18:42 | |
| *** igor has joined #openstack-swift | 18:44 | |
| *** shri has quit IRC | 18:44 | |
| *** shri1 has quit IRC | 18:45 | |
| briancline | ok. thanks | 18:45 |
| *** nacim has quit IRC | 18:53 | |
| *** byeager has quit IRC | 18:58 | |
| *** krtaylor has quit IRC | 19:01 | |
| *** mvenesio has joined #openstack-swift | 19:10 | |
| *** pberis has quit IRC | 19:13 | |
| *** pberis has joined #openstack-swift | 19:14 | |
| *** mef has quit IRC | 19:34 | |
| *** xga__ has quit IRC | 19:36 | |
| *** xga has quit IRC | 19:36 | |
| *** byeager has joined #openstack-swift | 19:39 | |
| *** byeager_ has joined #openstack-swift | 19:41 | |
| *** byeager has quit IRC | 19:44 | |
| *** foexle has quit IRC | 19:44 | |
| *** bsdkurt has quit IRC | 19:49 | |
| *** zackf has quit IRC | 19:54 | |
| *** lpabon has quit IRC | 19:59 | |
| *** mkollaro has quit IRC | 20:04 | |
| openstackgerrit | A change was merged to openstack/swift: Fix a comment in bulk middleware. https://review.openstack.org/71600 | 20:11 |
| *** mkollaro has joined #openstack-swift | 20:16 | |
| *** mvenesio has quit IRC | 20:29 | |
| *** krast has quit IRC | 20:30 | |
| notmyname | redbo: have you taken a look at the profiling middleware yet? https://review.openstack.org/#/c/53270/ kinda cool info | 20:36 |
| *** krtaylor has joined #openstack-swift | 20:39 | |
| redbo | nope! my guess is it's going to say swift is slow. | 20:41 |
| *** krtaylor has quit IRC | 21:24 | |
| peluse | notmyname: one of our guys here started messing with it, pretty cool | 21:26 |
| notmyname | indeed :-) | 21:28 |
| peluse | we looked at a sample from some large obj runs yesterday - still some things to figure out wrt getting the output in a totally useful format but what was readable made sense. He's going to keep messing around with it and I think using it/interpretting a run will be a great activity at the next hackathon | 21:30 |
| *** byeager_ has quit IRC | 21:33 | |
| *** byeager has joined #openstack-swift | 21:44 | |
| notmyname | at what point should we get in on the "replace eventlet" mailing list thread? | 21:45 |
| torgomatic | I think my least favorite commit in Swift is 4bc1b1f, but that's only because I'm really amused by the idea of a container borker. | 21:45 |
| *** bsdkurt has joined #openstack-swift | 21:45 | |
| notmyname | aren't container borkers another name for "users" | 21:46 |
| torgomatic | ah, users... imagine the systems we could build if it weren't for those pesky users ;) | 21:49 |
| *** portante has joined #openstack-swift | 21:53 | |
| *** ChanServ sets mode: +v portante | 21:53 | |
| *** shri has joined #openstack-swift | 22:02 | |
| *** bsdkurt has quit IRC | 22:03 | |
| *** bsdkurt has joined #openstack-swift | 22:06 | |
| notmyname | cschwede: around? | 22:10 |
| peluse | notmyname: I've got a net 2 line change just waiting for your approval :) https://review.openstack.org/#/c/70265/ | 22:14 |
| *** usvyatsky has joined #openstack-swift | 22:20 | |
| *** Trixboxer has quit IRC | 22:25 | |
| usvyatsky | hi there | 22:29 |
| openstackgerrit | John Dickinson proposed a change to openstack/swift: Limit logged headers in proxy_logging middleware https://review.openstack.org/64363 | 22:29 |
| notmyname | usvyatsky: hi | 22:29 |
| usvyatsky | I am looking for wire spec for kinetic protocol - is there such a thing? | 22:30 |
| usvyatsky | the reason being, I'd like to write a wireshark plugin for kinetic | 22:30 |
| usvyatsky | and all seagate supplies at the moment is just a binary library | 22:31 |
| notmyname | cool | 22:32 |
| notmyname | usvyatsky: you've looked at their simulator? | 22:32 |
| usvyatsky | yes, it's pretty neat | 22:32 |
| usvyatsky | unfortunately, it comes with either python bytecode API or Java bytecode jar | 22:32 |
| usvyatsky | or some sort of erlang API, of which I know nothing | 22:33 |
| usvyatsky | so the wire protocol remains hidden | 22:33 |
| usvyatsky | ideally, I'd like to see a protocol spec or a source-code API library | 22:34 |
| notmyname | my understanding is that it's "just protocol buffers", but I know that doesn't help too much | 22:35 |
| notmyname | you still need the actual packet structure | 22:35 |
| usvyatsky | yes, as well as the protocol details w.r.t. authentication, handshakes and such | 22:35 |
| notmyname | usvyatsky: do you know anyone at seagate working on kinetic? | 22:40 |
| usvyatsky | unfortunately I don't | 22:40 |
| notmyname | usvyatsky: send me an email at me@not.mn and I'll introduce you to the guy writing the kinetic api | 22:40 |
| usvyatsky | great, thank you very much! | 22:41 |
| *** krtaylor has joined #openstack-swift | 22:43 | |
| usvyatsky | done | 22:43 |
| openstackgerrit | John Dickinson proposed a change to openstack/swift: Replace Policy Index with Policy Name in Response Headers https://review.openstack.org/70265 | 22:48 |
| notmyname | peluse: done ^ (with the spelling error fixed) | 22:49 |
| notmyname | usvyatsky: fun with graylists. I havne't gotten it yet :-) | 22:50 |
| notmyname | peluse: the constants at the top of storage_policy.py. I don't really like them because they are the names of headers, not a particular policy or policy index (which is what I assume every time I see them in the code somewhere). | 22:52 |
| notmyname | peluse: appending "_HEADER_NAME" would make it explicit, but that of course will make things longer. obviously not a functional difference. just something I see every time I switch over to the ec branch | 22:53 |
| peluse | notmyname: OK, I can change those names in a separeate patch. thanks | 23:00 |
| notmyname | peluse: thanks. I guess I could have submitted a patch too, but now that you said you would...I don't have to! ;-) | 23:02 |
| peluse | hehe... BTW I do know how to spell, I just can't type for sh^*&t :) | 23:02 |
| notmyname | usvyatsky: done | 23:03 |
| openstackgerrit | Greg Lange proposed a change to openstack/swift: Fix a comment in SLO middleware https://review.openstack.org/71682 | 23:14 |
| notmyname | glange: whew. good thing we got that one take care of | 23:15 |
| dfg | dammit notmyname | 23:15 |
| notmyname | lol | 23:15 |
| glange | there are no small changes :) | 23:16 |
| glange | they all have to run the gates | 23:16 |
| notmyname | looks like the gate's only about 2 hours deep right now | 23:18 |
| peluse | notmyname: big box of stuff coming your way finally... check email | 23:25 |
| *** mmcardle has joined #openstack-swift | 23:25 | |
| notmyname | peluse: just saw it! awesome. | 23:28 |
| notmyname | they will go nicely with those 72 6TB helium drives I just got in the mail :-) | 23:28 |
| *** mmcardle has joined #openstack-swift | 23:29 | |
| peluse | can you crack those open and suck out the gas I wonder? | 23:29 |
| notmyname | negative shipping costs because they float? | 23:29 |
| *** mmcardle has quit IRC | 23:29 | |
| peluse | hehe... guess I cc'd Martin by accident - menat to add Mario.... | 23:30 |
| notmyname | no worries. martin (mlanner) will be helping be rack and wire them up, I'm sure :-) | 23:31 |
| peluse | awesome.... no idea if there are any docs in the boxes but just let me know if there are hw questions and I'll get you to the right guys | 23:31 |
| notmyname | thanks | 23:31 |
| usvyatsky | notmyname: thank you very much! | 23:32 |
| notmyname | for those of you wondering what's going on, new swift cluster for public community testing coming soon :-) | 23:32 |
| *** byeager has quit IRC | 23:32 | |
| notmyname | imagine! public CI tests on a real cluster! | 23:32 |
| notmyname | more details when we get it all set up | 23:33 |
| notmyname | usvyatsky: np. hope you're able to get something | 23:33 |
| peluse | now we just need to invent virtual tests to run on the real hw :) | 23:33 |
| *** byeager has joined #openstack-swift | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!