*** SpamapS has left #openstack-swift | 00:00 | |
*** petertr7_away is now known as petertr7 | 00:03 | |
asettle | notmyname okay, I could keep typing emails for hours but just want to say that I think you're psychic cause I only just fired off that email to have you immediately, incidentally, respond with all the answers to my questions | 00:03 |
---|---|---|
notmyname | :-) | 00:03 |
notmyname | asettle: yeah, I just started a reply there | 00:04 |
asettle | Haha all good | 00:04 |
asettle | Also, I did not know I had to have a dev environment setup | 00:04 |
asettle | *facepalm* | 00:04 |
notmyname | yeah, I kinda stumbled into that myself by accident when I was helping someone else | 00:04 |
asettle | So, if you've got instructions on that - PLEASE send them over. I"m halfway through setting up my devstack. Which, btw, is impossible because that documentation kinda sucks. | 00:05 |
notmyname | I hadn't had an issue, because of course I have dev environment already | 00:05 |
asettle | Hahahaha | 00:05 |
asettle | Always hte way | 00:05 |
notmyname | yeah, just ignore devstack (when it comes to swift) | 00:05 |
notmyname | (we do) | 00:05 |
asettle | OH WHAT | 00:05 |
asettle | Damn | 00:05 |
notmyname | lol | 00:05 |
asettle | Back to the drawing board | 00:05 |
notmyname | don't worry, it's easy to get a swift dev environment. I'll send you something soon. | 00:05 |
asettle | Cool :) | 00:06 |
notmyname | I'll try to finish up my writeup this evening | 00:06 |
asettle | I'm running F21 here, OSX and I have an Ubuntu VM lying around somewhere - that okay? | 00:06 |
asettle | On of them okay* | 00:06 |
asettle | One* | 00:06 |
asettle | Christ Alex, just give up now. | 00:06 |
notmyname | totally | 00:07 |
asettle | Excellent, excellent. | 00:08 |
asettle | I'll just go remove my half hearted devstack in that case | 00:09 |
notmyname | ok, reply quickly done | 00:09 |
notmyname | now I've got to step out for a while. I'll try to get my instructions done tonight or tomorrow | 00:09 |
notmyname | probably your monday before you have them | 00:10 |
asettle | That works :) | 00:10 |
asettle | Thanks John | 00:10 |
notmyname | happy to help | 00:11 |
asettle | I think an audience analysis will be really important in understanding where you're heading here. I'll reply again in a second | 00:11 |
asettle | Emails for the win | 00:11 |
*** petertr7 is now known as petertr7_away | 00:12 | |
mattoliverau | asettle: I can walk you through a swift dev environment | 00:16 |
asettle | Ohhhhh you can? | 00:17 |
asettle | mattoliverau are you sure? | 00:17 |
mattoliverau | asettle: yeah, you got your rackspace dev cloud accout setup? | 00:18 |
asettle | THAT is debatable. Checking now that my dev account has been applied | 00:21 |
asettle | Yes! Huzzah. Penny added me to the thingy | 00:22 |
*** gyee has quit IRC | 00:27 | |
asettle | I'm good to go mattoliverau | 00:28 |
asettle | I *think* | 00:29 |
ho | interesting (kota_'s co-worker right?): http://events.linuxfoundation.org/sites/events/files/slides/linuxcon15_bando.pdf | 00:29 |
*** dmorita has joined #openstack-swift | 00:30 | |
*** km__ has joined #openstack-swift | 00:31 | |
*** km has quit IRC | 00:31 | |
*** haomaiwa_ has quit IRC | 00:32 | |
openstackgerrit | Merged openstack/swift: Allow default reseller prefix in domain_remap middleware https://review.openstack.org/142067 | 00:43 |
mattoliverau | asettle: lol, let me relocated to a cafenation station (cafe hack), then we'll take it to a private channel and I'll walk you through building an SAIO. | 00:45 |
asettle | mattoliverau YAY! :) thanks man | 00:45 |
*** openstack has joined #openstack-swift | 00:56 | |
*** haomaiwang has joined #openstack-swift | 01:11 | |
*** janonymous_ has quit IRC | 01:19 | |
*** ianbrown has quit IRC | 01:22 | |
*** ianbrown has joined #openstack-swift | 01:25 | |
*** haomaiwang has quit IRC | 01:28 | |
*** haomaiwang has joined #openstack-swift | 01:34 | |
*** wer has joined #openstack-swift | 01:39 | |
*** jrichli has joined #openstack-swift | 01:41 | |
*** jlhinson has joined #openstack-swift | 01:56 | |
*** aix has quit IRC | 02:02 | |
*** asettle is now known as asettle-afk | 02:07 | |
*** proteusguy has quit IRC | 02:11 | |
*** steve_mc has quit IRC | 02:17 | |
*** proteusguy has joined #openstack-swift | 02:24 | |
*** jlhinson has quit IRC | 02:25 | |
*** steve_mc has joined #openstack-swift | 02:30 | |
*** marzif has quit IRC | 02:36 | |
*** jamielennox is now known as jamielennox|away | 02:46 | |
*** bhakta has left #openstack-swift | 02:48 | |
*** jamielennox|away is now known as jamielennox | 02:53 | |
*** zaitcev has quit IRC | 02:59 | |
*** kei_yama has quit IRC | 03:22 | |
*** asettle-afk is now known as asettle | 03:29 | |
*** jrichli has quit IRC | 04:37 | |
*** ianbrown has quit IRC | 04:48 | |
*** ianbrown has joined #openstack-swift | 05:00 | |
*** SkyRocknRoll has joined #openstack-swift | 05:07 | |
*** haomaiwang has quit IRC | 05:33 | |
*** haomaiwang has joined #openstack-swift | 05:35 | |
*** SkyRocknRoll_ has joined #openstack-swift | 05:43 | |
*** nadeem has joined #openstack-swift | 05:44 | |
*** Kennan2 has joined #openstack-swift | 06:07 | |
*** Kennan has quit IRC | 06:07 | |
*** ianbrown has quit IRC | 06:11 | |
*** silor has joined #openstack-swift | 06:12 | |
*** remix_tj has quit IRC | 06:29 | |
*** chlong has quit IRC | 07:06 | |
*** nadeem has quit IRC | 07:22 | |
*** remix_tj has joined #openstack-swift | 07:33 | |
openstackgerrit | Michael Barton proposed openstack/swift: go: always asynchronous finalize PUT and DELETE https://review.openstack.org/190788 | 07:36 |
*** proteusguy_ has joined #openstack-swift | 07:38 | |
*** proteusguy has quit IRC | 07:40 | |
*** proteusguy_ has quit IRC | 07:45 | |
*** proteusguy has joined #openstack-swift | 07:45 | |
*** rledisez has joined #openstack-swift | 07:53 | |
*** joeljwright has joined #openstack-swift | 07:56 | |
*** ChanServ sets mode: +v joeljwright | 07:56 | |
*** jamielennox is now known as jamielennox|away | 08:00 | |
*** jordanP has joined #openstack-swift | 08:08 | |
*** mmcardle has joined #openstack-swift | 08:12 | |
*** mmcardle has quit IRC | 08:13 | |
*** mmcardle1 has joined #openstack-swift | 08:13 | |
*** aix has joined #openstack-swift | 08:32 | |
*** haomaiwang has quit IRC | 09:11 | |
*** jasondotstar has joined #openstack-swift | 09:41 | |
*** Kennan2 is now known as Kennan | 09:49 | |
*** ho has quit IRC | 10:13 | |
*** haomaiwang has joined #openstack-swift | 10:45 | |
*** dmorita has quit IRC | 11:16 | |
*** jasondotstar has quit IRC | 11:43 | |
*** joeljwright1 has joined #openstack-swift | 11:44 | |
*** joeljwright has quit IRC | 11:46 | |
openstackgerrit | Victor Stinner proposed openstack/swift: Replace xrange() with six.moves.xrange() https://review.openstack.org/185460 | 11:55 |
*** km__ has quit IRC | 12:18 | |
*** yuan has quit IRC | 12:35 | |
*** silor has quit IRC | 12:36 | |
*** jasondotstar has joined #openstack-swift | 12:43 | |
*** cdelatte has joined #openstack-swift | 12:47 | |
*** haomaiwang has quit IRC | 12:48 | |
*** jkugel has joined #openstack-swift | 12:56 | |
*** haomaiwang has joined #openstack-swift | 12:56 | |
*** thurloat_isgone is now known as thurloat | 13:04 | |
*** HenryG has quit IRC | 13:05 | |
*** HenryG has joined #openstack-swift | 13:05 | |
*** HenryG has quit IRC | 13:06 | |
*** petertr7_away is now known as petertr7 | 13:06 | |
*** HenryG has joined #openstack-swift | 13:07 | |
*** jrichli has joined #openstack-swift | 13:19 | |
*** mwheckmann has joined #openstack-swift | 13:22 | |
*** jlhinson has joined #openstack-swift | 13:31 | |
*** SkyRocknRoll_ has quit IRC | 13:42 | |
*** haomaiwang has quit IRC | 13:46 | |
*** acampbell has joined #openstack-swift | 13:51 | |
*** acampbel11 has joined #openstack-swift | 13:52 | |
*** acampbell has quit IRC | 13:52 | |
*** acampbel11 has quit IRC | 13:52 | |
*** acampbell has joined #openstack-swift | 13:52 | |
openstackgerrit | Michael Barton proposed openstack/swift: go: fix x-delete-at updates https://review.openstack.org/193557 | 13:55 |
openstackgerrit | Michael Barton proposed openstack/swift: go: fix x-delete-at updates https://review.openstack.org/193557 | 13:58 |
*** acampbell has quit IRC | 13:58 | |
*** jlhinson has quit IRC | 14:04 | |
*** morganfainberg is now known as morgan-devserver | 14:06 | |
*** blmartin has joined #openstack-swift | 14:07 | |
*** morgan-devserver is now known as CaptainMorgan | 14:08 | |
*** wbhuber has joined #openstack-swift | 14:17 | |
*** wbhuber_ has joined #openstack-swift | 14:20 | |
*** wbhuber has quit IRC | 14:23 | |
*** leopoldj has joined #openstack-swift | 14:27 | |
*** archers has joined #openstack-swift | 14:29 | |
*** blmartin_ has joined #openstack-swift | 14:31 | |
blmartin_ | Good Morning all | 14:31 |
pgbridge | morning | 14:32 |
*** jasondotstar has quit IRC | 14:35 | |
*** jasondotstar has joined #openstack-swift | 14:37 | |
*** mwheckmann has quit IRC | 14:41 | |
*** mwheckmann has joined #openstack-swift | 14:47 | |
*** archers has quit IRC | 14:50 | |
*** minwoob has joined #openstack-swift | 14:55 | |
*** esker has joined #openstack-swift | 15:11 | |
*** wbhuber has joined #openstack-swift | 15:21 | |
*** wbhuber has quit IRC | 15:22 | |
*** petertr7 is now known as petertr7_away | 15:24 | |
*** barker has joined #openstack-swift | 15:24 | |
*** fthiagogv has joined #openstack-swift | 15:24 | |
*** fthiagogv_ has joined #openstack-swift | 15:24 | |
*** jasondotstar has quit IRC | 15:26 | |
*** jasondotstar has joined #openstack-swift | 15:28 | |
*** jordanP has quit IRC | 15:36 | |
*** thurloat is now known as thurloat_isgone | 15:41 | |
notmyname | good morning, world | 15:43 |
*** barker has quit IRC | 15:43 | |
*** leopoldj has quit IRC | 15:49 | |
*** marzif has joined #openstack-swift | 15:50 | |
*** marzif has quit IRC | 15:51 | |
*** marzif has joined #openstack-swift | 15:52 | |
*** thurloat_isgone is now known as thurloat | 15:52 | |
*** nadeem has joined #openstack-swift | 15:56 | |
*** jrichli has quit IRC | 15:56 | |
*** barker has joined #openstack-swift | 15:57 | |
*** marzif has quit IRC | 15:58 | |
*** esker has quit IRC | 16:05 | |
*** mwheckmann has quit IRC | 16:07 | |
*** SkyRocknRoll_ has joined #openstack-swift | 16:09 | |
*** jrichli has joined #openstack-swift | 16:10 | |
*** jrichli_ has joined #openstack-swift | 16:10 | |
*** jrichli has left #openstack-swift | 16:12 | |
*** jrichli_ has left #openstack-swift | 16:13 | |
*** jrichli has joined #openstack-swift | 16:14 | |
*** gyee has joined #openstack-swift | 16:14 | |
*** SkyRocknRoll_ has quit IRC | 16:15 | |
*** rledisez has quit IRC | 16:23 | |
*** wer has quit IRC | 16:27 | |
*** wer has joined #openstack-swift | 16:27 | |
*** petertr7_away is now known as petertr7 | 16:28 | |
*** thurloat is now known as thurloat_isgone | 16:42 | |
*** barker has quit IRC | 16:46 | |
*** barker has joined #openstack-swift | 16:47 | |
*** thurloat_isgone is now known as thurloat | 16:55 | |
*** zhill has joined #openstack-swift | 17:02 | |
*** mfalatic has joined #openstack-swift | 17:03 | |
*** tmckay has joined #openstack-swift | 17:07 | |
*** joeljwright1 has quit IRC | 17:07 | |
tmckay | hi folks. So if I have a partition table with logical volumes, and tmpfs, etc etc, is there a way to just carve out part of the /home directory structure and have swift write to that? | 17:08 |
tmckay | I want to turn some machines into storage nodes, but they're already partitioned with fancy /dev/mapper/xxx, etc | 17:08 |
tmckay | machines are not directly installed by me, so I don't know if I can control the partition table | 17:09 |
notmyname | yes, it's possible. all swift is really asking for is a mount point. so yeah, you can take one physical drive and put a few different mount points on it | 17:09 |
ajiang | I think you can use bind mount to mind the directory in /home to the mount points swift is expecting. | 17:10 |
notmyname | however, it's also generally a bad idea to do that. I'd never recommend it for product data | 17:10 |
notmyname | *production | 17:10 |
notmyname | for prod, you should give swift mount points where 1 mount point = 1 physical drive | 17:10 |
tmckay | this is just a dev setup. I just need space, it doesn't have to be fast space :) | 17:10 |
ajiang | bad idea as notmyname pointed out to do such thing for production | 17:10 |
tmckay | notmyname, understood | 17:10 |
notmyname | it's not a matter of speed | 17:10 |
tmckay | what is it a matter of then? | 17:11 |
notmyname | it's a matter of hiding the physical failure domains from swift | 17:11 |
notmyname | swfit assumes that drives are distinct physical failure domains | 17:11 |
tmckay | ah, gotcha | 17:11 |
notmyname | that affects durability and availability if you have two mount points on one physical device | 17:12 |
tmckay | my machines have 1 drive. So, is it okay to have swift mounted on a portion of the drive? | 17:12 |
*** peluse has joined #openstack-swift | 17:12 | |
tmckay | (assuming I can get control of the partitioning) | 17:12 |
notmyname | "okay" will all the caveats above ;-) | 17:12 |
*** ChanServ sets mode: +v peluse | 17:13 | |
notmyname | the docs we have for setting up a dev environment do talk about simulating multiple drives with loopback devices. so yeah, it works. just something that is only useful in very specific circumstances (ie a local dev environment) | 17:13 |
tmckay | notmyname, so in the ideal case, swift always wants a machine to have > 1 drive (1 drive for OS, others for swift) | 17:13 |
notmyname | tmckay: I wouldn't say "ideal". I'd say swift always needs that | 17:14 |
peluse | clayg, patch 191521 (EC recon) is ready for another peek | 17:14 |
patchbot | peluse: https://review.openstack.org/#/c/191521/ | 17:14 |
tmckay | hmm, I suppose I could make really big loopbacks. But that seems lame | 17:14 |
tmckay | beggars can't be choosers for hardware, these single drive machines are what I have :) | 17:15 |
notmyname | yeah. I understand | 17:15 |
peluse | acoles_away, if you're stil lurking, I made the changes you suggested in patch 185459 but I wasn't positive where you meant wrt the man pages comment so please take a look when you get a chance | 17:15 |
patchbot | peluse: https://review.openstack.org/#/c/185459/ | 17:15 |
tmckay | alright, thanks, you've given me something to think about | 17:15 |
notmyname | tmckay: just note that the maximum object size you can have in swift will be min(configured max object size, size of drive) | 17:16 |
notmyname | eg don't try to upload a 3GB object with a 1Gb drive ;-) | 17:16 |
tmckay | good tip :) | 17:16 |
wbhuber_ | notmyname: peluse: i should be getting some performance benchmarking numbers out on tuesday (via github outlet) or some sort. right now i'm getting numbers crunched in a literal manner. stay tuned. | 17:18 |
notmyname | wbhuber_: great! | 17:18 |
*** barker has quit IRC | 17:18 | |
peluse | wbhuber_, yeah that's awesome. this week we had to replace our LB and one proxy but we're in data collection mode now so we expent initial numbers at about the same time | 17:19 |
wbhuber_ | notmyname: peluse: did you put performance numbers out recently (not the ones before the summit)? | 17:19 |
wbhuber_ | ok, i didn't miss anything yet. | 17:19 |
ajiang | I got a strange partition delete behavior on the storage nodes. every midnight for about 30-40 minutes, there is a spike of partition.delete.count (in object.py update_delete()). The jump is on average about 50 vs 1-2 when this doesn't happen. what could be the possible reason? | 17:19 |
notmyname | I haven't yet, but I'm hoping to get to it today | 17:20 |
notmyname | ajiang: expiring objects? | 17:20 |
ajiang | Not sure, the number I got is from the statsd partition.delete.count. How to confirm expiring objects is the cause? | 17:22 |
*** idevops has joined #openstack-swift | 17:22 | |
*** idevops has quit IRC | 17:23 | |
ajiang | notmyname: actually, this doesn't happen to all the storage nodes, just the newly added nodes that is still doing ring rebalance to increase the disk weight | 17:24 |
*** zul has quit IRC | 17:25 | |
*** zul has joined #openstack-swift | 17:25 | |
*** barker has joined #openstack-swift | 17:29 | |
*** breitz has quit IRC | 17:42 | |
*** breitz has joined #openstack-swift | 17:42 | |
*** mmcardle1 has quit IRC | 17:45 | |
*** haomaiwa_ has joined #openstack-swift | 17:48 | |
*** marzif has joined #openstack-swift | 17:53 | |
*** jasondotstar has quit IRC | 17:56 | |
*** acampbell has joined #openstack-swift | 17:57 | |
*** jasondot_ has joined #openstack-swift | 17:59 | |
*** elmiko has joined #openstack-swift | 18:03 | |
*** barker has quit IRC | 18:03 | |
elmiko | notmyname: hey, i'm trying to gain a better grasp on the usage patterns for this 1xx stuff. might you have a few minutes to talk about it? | 18:03 |
*** jith_ has quit IRC | 18:05 | |
*** thurloat is now known as thurloat_isgone | 18:07 | |
*** esker has joined #openstack-swift | 18:12 | |
*** esker has quit IRC | 18:12 | |
*** esker has joined #openstack-swift | 18:13 | |
*** thurloat_isgone is now known as thurloat | 18:16 | |
*** marzif has quit IRC | 18:18 | |
notmyname | elmiko: later. out of one meeting now and into another | 18:23 |
*** portante has quit IRC | 18:23 | |
*** ndk has quit IRC | 18:23 | |
*** barker has joined #openstack-swift | 18:25 | |
elmiko | notmyname: ack, i'll be around for awhile. thanks! | 18:30 |
*** ndk has joined #openstack-swift | 18:32 | |
*** portante has joined #openstack-swift | 18:33 | |
*** esker has quit IRC | 18:46 | |
*** zhill_ has joined #openstack-swift | 18:53 | |
*** zhill has quit IRC | 18:53 | |
*** petertr7 is now known as petertr7_away | 19:04 | |
*** zaitcev has joined #openstack-swift | 19:11 | |
*** ChanServ sets mode: +v zaitcev | 19:11 | |
*** petertr7_away is now known as petertr7 | 19:15 | |
*** zhill_ has quit IRC | 19:18 | |
*** zhill has joined #openstack-swift | 19:18 | |
ekarlso | hey guys, what's the reason for 2015-06-19 19:35:06.361 TRACE pidgey.mta.service ClientException: Container PUT failed: http://192.168.27.101:8080/v1/AUTH_ace8741694ad4d85b6d35af1bbe98325/pidgey 404 Not Found [first 60 chars of response] <html><h1>Not Found</h1><p>The resource could not be found.< | 19:35 |
ekarlso | when doing a rejoin_stach.sh ? | 19:35 |
minwoob | clayg: It seems like it would be better to abort the rebuild and wait until the handoff that has the existing fragment, to eventually push it to the correct location. | 19:37 |
minwoob | clayg: It saves computation and network resources, at least. | 19:38 |
clayg | minwoob: yeah that'd be pretty sweet - a little more code - we already need to drop the duplicates (not helpful) so dropping any response that finds the frag on the handoff is pretty cheap | 19:38 |
clayg | minwoob: yup - only reason to do the thing thats' cheaper to type is because not doing that can segfault | 19:38 |
clayg | minwoob: but if you see some cheap easy way to skip that object rebuilt without hozing things up that'd be great | 19:39 |
clayg | minwoob: I honestly didn't go looking for it | 19:39 |
minwoob | clayg: I see. I wasn't aware that even just aborting the rebuild process itself could cause a segfault. | 19:40 |
minwoob | clayg: I thought it was just when the rebuild process was initiated, with the fragment existing in the list. | 19:41 |
clayg | minwoob: i must have mis spoke - what you just said is right | 19:43 |
clayg | minwoob: if you can make a patch to do the abort that's better - the thing i had in laucnhpad was just to prevent the segfaults - it still does potentially useless work | 19:44 |
clayg | well not useless - just less efficient - the frag needs to get back on the primary and rebuilding it is one way to do that | 19:44 |
*** petertr7 is now known as petertr7_away | 19:45 | |
minwoob | clayg: Yeah, I agree that eventually it would be good to avoid that route :) | 19:46 |
*** marzif has joined #openstack-swift | 19:49 | |
*** barker has quit IRC | 19:51 | |
*** barker has joined #openstack-swift | 19:52 | |
*** blmartin__ has joined #openstack-swift | 19:56 | |
*** blmartin_ has quit IRC | 19:59 | |
*** blmartin has quit IRC | 20:00 | |
*** blmartin has joined #openstack-swift | 20:00 | |
*** petertr7_away is now known as petertr7 | 20:01 | |
*** cdelatte has quit IRC | 20:02 | |
*** thurloat is now known as thurloat_isgone | 20:11 | |
*** marzif has quit IRC | 20:14 | |
*** SkyRocknRoll has quit IRC | 20:14 | |
*** barker has quit IRC | 20:15 | |
*** barker has joined #openstack-swift | 20:19 | |
*** esker has joined #openstack-swift | 20:30 | |
*** marzif has joined #openstack-swift | 20:30 | |
*** marzif has quit IRC | 20:38 | |
*** CaptainMorgan is now known as morgan | 20:41 | |
*** blmartin__ has quit IRC | 20:43 | |
*** blmartin has quit IRC | 20:43 | |
*** acampbell has quit IRC | 20:49 | |
openstackgerrit | Victor Stinner proposed openstack/swift: Replace xrange() with six.moves.xrange() https://review.openstack.org/185460 | 20:50 |
*** marzif has joined #openstack-swift | 20:51 | |
*** esker has quit IRC | 20:58 | |
*** marzif has quit IRC | 21:00 | |
*** petertr7 is now known as petertr7_away | 21:07 | |
*** jkugel has quit IRC | 21:27 | |
*** fthiagogv has quit IRC | 21:35 | |
*** fthiagogv_ has quit IRC | 21:35 | |
*** barker has quit IRC | 21:39 | |
clayg | we need haypo hanging out in here ;) | 21:48 |
clayg | I can not fathom why the python3 json module seems so dead set on not letting you acctualy use it as a data exchange format without first converting the damn python3 str to bytes so you can ... you know ... use it to exchange the damn data! | 21:50 |
*** esker has joined #openstack-swift | 21:51 | |
clayg | torgomatic: i bet after you finish pulling simplejson out for python 2.7 we'll have to add it back in for python 3.4 | 21:51 |
swifterdarrell | clayg: lol | 21:51 |
clayg | who's gunna merge https://review.openstack.org/#/c/184189/ !? I ain't skerred | 21:52 |
clayg | you guys merged that ring crazyness I figure it's my turn to go cowboy | 21:53 |
clayg | I'm going to review p 191521 - if you guys don't merge or -1 1841189 by the time i'm done i'ma +A it | 21:54 |
*** esker has quit IRC | 21:54 | |
*** wbhuber_ has quit IRC | 21:55 | |
notmyname | patch 191521 | 21:55 |
patchbot | notmyname: https://review.openstack.org/#/c/191521/ | 21:55 |
*** esker has joined #openstack-swift | 21:57 | |
*** marzif has joined #openstack-swift | 21:58 | |
torgomatic | clayg: if you wanna be pedantic, then python3's json is correct; a JSON document isn't a string, but a sequence of Unicode codepoints. How it's serialized isn't part of JSON. | 22:00 |
torgomatic | If you *don't* want to be pedantic, it's a pain in your ass. | 22:00 |
clayg | torgomatic: idk man - https://tools.ietf.org/html/rfc7159#section-8.1 | 22:00 |
torgomatic | clayg: yup; if you're going to have the choice, then the serializer needs to hand out a unicode string, not a byte string, so you can make the choice | 22:02 |
torgomatic | if it were me, I'd have just said UTF-8 and screw everyone else, but that's probably why I don't write standards | 22:03 |
clayg | torgomatic: py2's json.dumps has an encoding kwarg - py3's does not :'( | 22:03 |
torgomatic | clayg: boo | 22:03 |
*** mwheckmann has joined #openstack-swift | 22:03 | |
torgomatic | encoding='utf8' by default is pretty much what I'd want there, and pass encoding=None if you want to do it yourself | 22:04 |
clayg | torgomatic: I mean the whole thing is for convience - but you hope the cost of that utility isn't doing a bunch of slow useless overhead? | 22:04 |
clayg | torgomatic: *see* you might not should write standards - but you be a dman good libary author | 22:05 |
*** mwheckmann has quit IRC | 22:08 | |
*** minwoob has quit IRC | 22:09 | |
*** marzif has quit IRC | 22:10 | |
*** jkugel has joined #openstack-swift | 22:23 | |
*** jasondot_ has quit IRC | 22:32 | |
*** peluse has quit IRC | 22:32 | |
*** marzif has joined #openstack-swift | 22:33 | |
*** gyee has quit IRC | 22:39 | |
*** peluse has joined #openstack-swift | 22:43 | |
*** jkugel has left #openstack-swift | 22:49 | |
*** mfalatic has quit IRC | 23:03 | |
*** thurloat_isgone has quit IRC | 23:05 | |
*** zigo has quit IRC | 23:05 | |
*** marzif has quit IRC | 23:06 | |
*** ondergetekende has quit IRC | 23:06 | |
openstackgerrit | Janie Richling proposed openstack/swift: Trivial Key Master for encyrption https://review.openstack.org/193749 | 23:06 |
*** zigo has joined #openstack-swift | 23:07 | |
zaitcev | Ouch | 23:07 |
*** ondergetekende has joined #openstack-swift | 23:07 | |
*** thurloat has joined #openstack-swift | 23:07 | |
jrichli | all: ^^ That review is a new one that consists of only a trivial key master. It is the first of what will be several patches to review for encryption. | 23:09 |
notmyname | thanks | 23:09 |
swifterdarrell | clayg: zaitcev: does https://review.openstack.org/#/c/184189 need something else for the +1? | 23:09 |
zaitcev | swifterdarrell: you mean +A? | 23:09 |
swifterdarrell | zaitcev: ya; 'cause I could add another class in there if that'll help | 23:09 |
clayg | torlorolrorlorrlololololo | 23:10 |
zaitcev | swifterdarrell: I dunno, it's just big, and things happen like that port/tier patch. | 23:10 |
clayg | [22:01] + clayg | I'm going to review p 191521 - if you guys don't merge or -1 1841189 by the time i'm done i'ma +A it | 23:11 |
zaitcev | I don't see anything objectionable, so +2, but usually we want more scrutiny for something potentially disruptive. | 23:11 |
swifterdarrell | zaitcev: "things happen like that port/tier patch." eh? | 23:11 |
zaitcev | okay, Clay is it then | 23:11 |
clayg | zaitcev: +2!!! if folks are going to look at it we want to give them a chance! | 23:11 |
clayg | if not - let's merge it | 23:11 |
swifterdarrell | clayg: zaitcev: I guess mattoliverau said he was installing it to test, but never heard back from him? | 23:12 |
clayg | notmyname: torgomatic: mattoliverau: I'm not sure who's still up - it's probably mostly saturday at this point!? | 23:12 |
notmyname | it's also saturday for him | 23:12 |
clayg | swifterdarrell: i'm sure it went fine | 23:12 |
zaitcev | Matthew is not Jewish though, or is he? | 23:12 |
clayg | notmyname: well tough shit - i'm going to +A it before I go home | 23:12 |
clayg | zaitcev: LOL | 23:12 |
clayg | notmyname: well - unless someon says -1 or asks me not to +A it :P | 23:13 |
zaitcev | clayg: Fine by me. | 23:13 |
clayg | zaitcev: it's *FASTER* | 23:13 |
*** jrichli has quit IRC | 23:13 | |
clayg | if you don't merge FASTER it's criminal | 23:13 |
clayg | it's like you HATE swift | 23:13 |
notmyname | swifterdarrell: clayg: no harm in waiting until monday is there? gives mattoliverau and cschwede and kota_ a chance to see it on their monday | 23:13 |
swifterdarrell | and hate FREEDOM | 23:13 |
clayg | nah probalby not | 23:13 |
mattoliverau | Sorry, yeah have it running on a cluster.. Looks good.. Ive got to spend the last 2 days doing swift related Rackspace work for customers :( | 23:14 |
notmyname | customers!?!?! | 23:14 |
zaitcev | Well... You can merge *swifter* a patch that makes Swift go *faster* | 23:14 |
clayg | so funny - ears must have been burning | 23:14 |
mattoliverau | Yeah I know right :( | 23:14 |
swifterdarrell | clayg: notmyname: makes sense; it's not like I can package new swift bits this weekend anyway ;) | 23:14 |
clayg | swifterdarrell: can't or *won't* - maybe YOU hate freedom | 23:14 |
swifterdarrell | clayg: maybe | 23:14 |
clayg | swifterdarrell: my point is that waiting till monday does help either unless people are working on it over the weekend - or there are some concerns | 23:15 |
zaitcev | I spent a week doing basically QA for our new super-duper Tripple-O based thing | 23:15 |
swifterdarrell | clayg: *nod* | 23:15 |
clayg | *I* think we beat all the stupid out and I don't have anything else I can think of worth testing that I haven't already done on has been done | 23:15 |
swifterdarrell | clayg: notmyname: zaitcev: well, whatever--as long as it merges by the time all the EC fixes get in :) | 23:15 |
clayg | if cschwede and mattoliverau are testing on it great - we'll wait till they done | 23:15 |
clayg | if we just wanted one more round of let's look at the code and waffle about this and that - nah - let's merge it | 23:15 |
notmyname | mattoliverau: are you doing anything that you'd expect to have a real review on your monday? | 23:16 |
notmyname | or will nothing happen in the next 2 days to do that? | 23:16 |
mattoliverau | Yeah, I'll have a review in on your Sunday :) maybe even today cause I'm a geek (and not Jewish) | 23:16 |
clayg | notmyname: i'm not nessecarily *looking* for anything new either - I think we covered our bases pretty well - not everyone has to turn it on the lab before it can merge | 23:17 |
clayg | if it's basically sound and moving in the right direction - we can all get in the lab eventually and make improvements | 23:17 |
clayg | it's not like it's backwards incompatible or breaks something - i mean - at least I couldn't find anything that got broke :) | 23:17 |
mattoliverau | How about if its +2ed I'll +A today or your Sunday if I like it ;) (which I do, just testing) | 23:18 |
notmyname | clayg: give mattoliverau a chance since he's been looking at it and expects to have that wrapped up in 24 hours | 23:18 |
notmyname | then mattoliverau does his +2 and a +A | 23:18 |
clayg | mattoliverau: perfect! yeah that's exactly the kind of feedback I was looking for earlier when I started threating to +A | 23:19 |
clayg | notmyname: thanks! | 23:19 |
mattoliverau | clayg: lol love it when a plan comes together :p | 23:19 |
clayg | mattoliverau: with any luck you'll find some sort of critical bug and look like a rock star! | 23:19 |
mattoliverau | Or miss it... And say oops :p | 23:20 |
clayg | heh | 23:20 |
clayg | well you can't say we didn't *try* | 23:20 |
*** esker has quit IRC | 23:28 | |
*** esker has joined #openstack-swift | 23:28 | |
clayg | notmyname: you still here? | 23:29 |
clayg | if I --amend and git review peluse's gerrit change - does it make it screw up authorship? | 23:30 |
openstackgerrit | Clay Gerrard proposed openstack/swift: EC Ssync: Update parms to include node and frag indices https://review.openstack.org/191521 | 23:30 |
clayg | nm, easiest thing was apparently just to try it | 23:30 |
zaitcev | no | 23:31 |
clayg | zaitcev: thanks :P | 23:31 |
zaitcev | clayg: yeah, too late I see | 23:32 |
clayg | peluse: i pushed over patch 191521 if you wanna double check me maybe ho, mattoliverau or notmyname can give it the final +A | 23:32 |
patchbot | clayg: https://review.openstack.org/#/c/191521/ | 23:32 |
zaitcev | But git review -d --> git commit --amend is basically the way to screw with people's patches in gerrit. | 23:32 |
clayg | mwuahahahaha! | 23:32 |
clayg | screw 'em | 23:32 |
zaitcev | It's so convenient that it sucks in. | 23:33 |
*** petertr7_away is now known as petertr7 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!