| *** chlong has joined #openstack-swift | 00:00 | |
| *** hrou has joined #openstack-swift | 00:12 | |
| *** kota_ has joined #openstack-swift | 01:00 | |
| *** ChanServ sets mode: +v kota_ | 01:00 | |
| kota_ | good morning | 01:01 |
|---|---|---|
| ho | kota_: good morning! | 01:03 |
| kota_ | ho: yeah, how's going? | 01:04 |
| ho | kota_: i'm exhausted. too too too hot :-) | 01:05 |
| kota_ | ho: ya, it's gettting hotter in Japan. | 01:08 |
| *** wbhuber has joined #openstack-swift | 01:10 | |
| *** Kennan has quit IRC | 01:13 | |
| *** Kennan has joined #openstack-swift | 01:16 | |
| *** jasondotstar has joined #openstack-swift | 01:26 | |
| *** wbhuber_ has joined #openstack-swift | 01:31 | |
| *** wbhuber has quit IRC | 01:34 | |
| *** wbhuber_ has quit IRC | 01:48 | |
| jrichli | mahatic: there has been a merge from master. start now by picking up the latest patch on https://review.openstack.org/#/c/203454/ | 02:08 |
| jrichli | you won't have to rebase what is there now | 02:09 |
| *** wbhuber has joined #openstack-swift | 02:28 | |
| *** jrichli has quit IRC | 02:48 | |
| notmyname | hello, world | 02:50 |
| *** wbhuber has quit IRC | 02:50 | |
| ho | notmyname: good evening! | 02:50 |
| notmyname | how are things? | 02:55 |
| ho | great! except the weather (too hot for me) :-) | 02:57 |
| *** tsubic has quit IRC | 02:59 | |
| notmyname | I went to seattle today (where I am now). similar temperature as home, but it's been raining lightly. very nice :-) | 03:02 |
| notmyname | ho: what's the weather going to be in tokyo at the summit? | 03:06 |
| *** kcaj has quit IRC | 03:07 | |
| *** kcaj has joined #openstack-swift | 03:14 | |
| kota_ | it could be good weather in October in Tokyo. | 03:17 |
| kota_ | perhaps a bit colder than bay area. | 03:17 |
| kota_ | but almost of days in Octover will be sunny. | 03:18 |
| *** changbl has quit IRC | 03:19 | |
| *** jasondotstar has quit IRC | 03:20 | |
| notmyname | nice! | 03:22 |
| *** sanchitmalhotra has joined #openstack-swift | 03:24 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 03:31 | |
| *** changbl has joined #openstack-swift | 03:32 | |
| *** sanchitmalhotra has quit IRC | 03:33 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:38 | |
| *** sanchitmalhotra1 has quit IRC | 03:40 | |
| ho | i was out for lunch. yeah, +1 for weather in october in tokyo | 03:50 |
| *** zul has quit IRC | 03:52 | |
| *** zul has joined #openstack-swift | 03:53 | |
| *** nexusz99 has joined #openstack-swift | 04:00 | |
| *** haomaiwang has quit IRC | 04:02 | |
| *** haomaiwang has joined #openstack-swift | 04:04 | |
| *** proteusguy has quit IRC | 04:06 | |
| ho | i didn't check all yet but meta data search has many entries. https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/SearchForm | 04:07 |
| *** hrou has quit IRC | 04:08 | |
| *** ppai has joined #openstack-swift | 04:13 | |
| *** proteusguy has joined #openstack-swift | 04:18 | |
| *** Kennan2 has joined #openstack-swift | 04:20 | |
| *** Kennan has quit IRC | 04:20 | |
| *** DericHorn-HP has joined #openstack-swift | 04:26 | |
| notmyname | I'm going back through old local branches I have in my swift tree | 04:29 |
| notmyname | so far it seems like most are just stuff from reviews | 04:29 |
| notmyname | but I did just fine one from Oct 2012 that I had never pushed up | 04:30 |
| *** ho_ has joined #openstack-swift | 04:38 | |
| *** marzif has joined #openstack-swift | 04:41 | |
| *** ho_ has left #openstack-swift | 04:51 | |
| *** jome has joined #openstack-swift | 04:59 | |
| *** marzif has quit IRC | 05:11 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 05:25 | |
| *** sanchitmalhotra has quit IRC | 05:27 | |
| *** nakagawamsa has joined #openstack-swift | 05:29 | |
| *** DericHorn-HP has quit IRC | 05:35 | |
| *** sanchitmalhotra has joined #openstack-swift | 05:55 | |
| *** sanchitmalhotra1 has quit IRC | 05:57 | |
| *** fifieldt has joined #openstack-swift | 06:00 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 06:22 | |
| *** sanchitmalhotra has quit IRC | 06:24 | |
| *** ppai has quit IRC | 06:59 | |
| *** ppai has joined #openstack-swift | 07:09 | |
| ho | acoles: around? are you back? | 07:22 |
| *** chlong has quit IRC | 07:28 | |
| *** rledisez has joined #openstack-swift | 07:41 | |
| *** fifieldt has quit IRC | 07:42 | |
| *** geaaru has joined #openstack-swift | 07:43 | |
| *** jistr has joined #openstack-swift | 07:44 | |
| *** jome has quit IRC | 07:52 | |
| *** jome has joined #openstack-swift | 07:52 | |
| *** jome_ has joined #openstack-swift | 08:01 | |
| *** jome has quit IRC | 08:02 | |
| *** jome_ has quit IRC | 08:08 | |
| *** haypo has joined #openstack-swift | 08:16 | |
| *** jome has joined #openstack-swift | 08:19 | |
| *** jome_ has joined #openstack-swift | 08:22 | |
| *** silor has joined #openstack-swift | 08:24 | |
| *** jome has quit IRC | 08:24 | |
| ho | today i have a party so i will leave now. have a nice day! | 08:27 |
| *** ho has quit IRC | 08:30 | |
| *** aix has joined #openstack-swift | 08:31 | |
| *** silor1 has joined #openstack-swift | 08:44 | |
| *** silor has quit IRC | 08:46 | |
| *** silor1 is now known as silor | 08:46 | |
| *** wbhuber has joined #openstack-swift | 08:46 | |
| *** wbhuber has quit IRC | 08:50 | |
| *** jome_ has quit IRC | 08:59 | |
| *** eandersson has joined #openstack-swift | 09:00 | |
| *** sanchitmalhotra has joined #openstack-swift | 09:21 | |
| openstackgerrit | Victor Stinner proposed openstack/swift: Fix warnings of hacking 0.10.2 https://review.openstack.org/205964 | 09:22 |
| *** sanchitmalhotra1 has quit IRC | 09:23 | |
| openstackgerrit | Victor Stinner proposed openstack/swift: Don't format using self.__dict__ https://review.openstack.org/205966 | 09:26 |
| *** jordanP has joined #openstack-swift | 09:32 | |
| *** jasondotstar has joined #openstack-swift | 09:32 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 09:37 | |
| *** sanchitmalhotra has quit IRC | 09:39 | |
| *** jome has joined #openstack-swift | 09:40 | |
| openstackgerrit | Victor Stinner proposed openstack/swift: Fix warnings of hacking 0.10.2 https://review.openstack.org/205964 | 09:42 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Update hacking to 0.10.0 or newer https://review.openstack.org/205977 | 09:42 |
| *** jome has quit IRC | 09:46 | |
| *** jome has joined #openstack-swift | 09:49 | |
| *** jome has quit IRC | 09:54 | |
| *** jome has joined #openstack-swift | 09:56 | |
| *** jome has quit IRC | 10:00 | |
| *** jome has joined #openstack-swift | 10:02 | |
| *** jome has quit IRC | 10:02 | |
| *** kota_ has quit IRC | 10:04 | |
| *** jome has joined #openstack-swift | 10:09 | |
| *** ppai has quit IRC | 10:12 | |
| *** jome_ has joined #openstack-swift | 10:18 | |
| *** jome_ has quit IRC | 10:20 | |
| *** cppforlife__ has joined #openstack-swift | 10:21 | |
| *** jome has quit IRC | 10:21 | |
| *** cppforlife_ has quit IRC | 10:21 | |
| *** acoles has quit IRC | 10:21 | |
| *** blair has quit IRC | 10:21 | |
| *** tdasilva has quit IRC | 10:21 | |
| *** cppforlife__ is now known as cppforlife_ | 10:22 | |
| *** jome has joined #openstack-swift | 10:23 | |
| *** ppai has joined #openstack-swift | 10:24 | |
| jith_ | hi all, how to upload data with more than 5GB | 10:33 |
| jith_ | in swift | 10:33 |
| nexusz99 | @jith_ you can use multipart upload (SLO, DLO) | 10:34 |
| jith_ | nexusz99: Thanks... can u suggest some documentation for that | 10:35 |
| nexusz99 | jith_, see http://docs.openstack.org/developer/swift/overview_large_objects.html | 10:35 |
| jith_ | nexusz99: thanks.. i was seeing that... i am not uploading via swift client but through API... means i have configured swift as backend for glance... so when i upload images in glance, swift throws an error like "Request entity too large". So i should change in some conf files?? | 10:38 |
| nexusz99 | jith_, you mean api is the swift REST API? | 10:40 |
| jith_ | yes i think so.. i am a beginner in this... if i upload images less than 5 GB its working fine.. but throwing error if it exceds 5gb... | 10:42 |
| nexusz99 | jith_, then you have to choice which middleware will you use to multipart upload. swift support 2 types of middleware (SLO , DLO) | 10:44 |
| *** kei_yama has quit IRC | 10:44 | |
| nexusz99 | jith_, if you want to use dlo middleware, see this documents (http://docs.openstack.org/developer/swift/overview_large_objects.html#direct-api) | 10:45 |
| nexusz99 | jith_, fix your client code when upload object exceed 5GB splite the object into segment( min size 1MB) and upload segments. | 10:46 |
| nexusz99 | jith_, when upload segments finished, create manifest file with 'X-Object-Manifest'. | 10:50 |
| jith_ | nexusz99: Thanks... but i dont have client code... i am using glance client with swift options | 10:52 |
| jith_ | in glance api.conf i stored swift credentials and regarding the size limits i could see only two parameters in that.. they are What size, in MB, should Glance start chunking image files | 10:53 |
| jith_ | # and do a large object manifest in Swift? By default, this is | 10:53 |
| jith_ | # the maximum object size in Swift, which is 5GB | 10:53 |
| jith_ | swift_store_large_object_size = 5120 | 10:53 |
| jith_ | # When doing a large object manifest, what size, in MB, should | 10:53 |
| jith_ | # Glance write chunks to Swift? This amount of data is written | 10:53 |
| jith_ | # to a temporary disk buffer during the process of chunking | 10:53 |
| jith_ | # the image file, and the default is 200MB | 10:53 |
| jith_ | swift_store_large_object_chunk_size = 200 | 10:53 |
| nexusz99 | jith_, could you show your swift proxy-server's pipeline in config file? | 10:54 |
| *** sanchitmalhotra has joined #openstack-swift | 10:55 | |
| *** sanchitmalhotra1 has quit IRC | 10:57 | |
| jith_ | sure | 11:05 |
| jith_ | nexusz99: http://paste.openstack.org/show/405908/ | 11:07 |
| *** ppai has quit IRC | 11:08 | |
| jith_ | nexusz99: Sorry for late reply... thanks for guiding | 11:09 |
| nexusz99 | jith_, hmm.. i have to research about glance. i think glance does not upload object by multipart .. | 11:10 |
| jith_ | nexusz99: Thanks no error is from swift side... if possible check the log.. http://paste.openstack.org/show/405909/ | 11:13 |
| *** acoles_ has joined #openstack-swift | 11:13 | |
| *** ChanServ sets mode: +v acoles_ | 11:13 | |
| *** acoles has joined #openstack-swift | 11:17 | |
| *** ChanServ sets mode: +v acoles | 11:17 | |
| *** tdasilva has joined #openstack-swift | 11:18 | |
| *** acoles- has joined #openstack-swift | 11:19 | |
| *** acoles has quit IRC | 11:19 | |
| *** acoles- is now known as acoles | 11:19 | |
| *** ChanServ sets mode: +v acoles | 11:19 | |
| *** ppai has joined #openstack-swift | 11:20 | |
| *** acoles_ has quit IRC | 11:20 | |
| *** jome has quit IRC | 11:21 | |
| *** marzif has joined #openstack-swift | 11:31 | |
| *** blair has joined #openstack-swift | 11:33 | |
| *** sanchitmalhotra has quit IRC | 11:41 | |
| *** nexusz99 has quit IRC | 11:42 | |
| *** nexusz99 has joined #openstack-swift | 11:43 | |
| *** marcusvrn has joined #openstack-swift | 12:02 | |
| *** ppai has quit IRC | 12:05 | |
| openstackgerrit | Victor Stinner proposed openstack/swift: Replace deprecated assert_() https://review.openstack.org/206031 | 12:09 |
| *** km has quit IRC | 12:10 | |
| *** jasondotstar has quit IRC | 12:10 | |
| *** thurloat_isgone is now known as thurloat | 12:18 | |
| *** jome has joined #openstack-swift | 12:18 | |
| *** ppai has joined #openstack-swift | 12:20 | |
| *** chlong has joined #openstack-swift | 12:30 | |
| *** jasondotstar has joined #openstack-swift | 12:32 | |
| *** vinsh_ has quit IRC | 12:38 | |
| *** hrou has joined #openstack-swift | 12:45 | |
| *** ppai has quit IRC | 12:47 | |
| openstackgerrit | Victor Stinner proposed openstack/swift: Update hacking to 0.10.0 or newer https://review.openstack.org/205977 | 12:54 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Fix warnings of hacking 0.10.2 https://review.openstack.org/205964 | 12:54 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Don't format using self.__dict__ https://review.openstack.org/205966 | 12:54 |
| *** annegentle has joined #openstack-swift | 12:54 | |
| *** silor has quit IRC | 12:55 | |
| openstackgerrit | janonymous proposed openstack/swift: Python3: do not use im_self/im_func/func_closure https://review.openstack.org/203066 | 12:59 |
| *** ppai has joined #openstack-swift | 12:59 | |
| *** _hrou_ has joined #openstack-swift | 13:00 | |
| *** hrou has quit IRC | 13:01 | |
| *** tongli has joined #openstack-swift | 13:08 | |
| *** annegentle has quit IRC | 13:10 | |
| *** annegentle has joined #openstack-swift | 13:11 | |
| *** janonymous_ has joined #openstack-swift | 13:19 | |
| *** charlesw has joined #openstack-swift | 13:21 | |
| janonymous_ | Hi Guys, it has not been a long time since i have been a part of this community but i have always tried to do as much as i can, with this effort only i have submitted a small talk which might get approved with your votes: https://www.openstack.org/summit/tokyo-2015/vote-for-speakers/presentation/4570 ; Kindly vote if you feel if it's worth giving a chance ;) | 13:24 |
| *** ppai has quit IRC | 13:30 | |
| *** annegent_ has joined #openstack-swift | 13:38 | |
| *** annegentle has quit IRC | 13:41 | |
| *** chlong has quit IRC | 13:42 | |
| *** ppai has joined #openstack-swift | 13:45 | |
| *** jasondotstar has quit IRC | 13:48 | |
| *** jistr has quit IRC | 13:52 | |
| *** blmartin has joined #openstack-swift | 13:53 | |
| *** jistr has joined #openstack-swift | 13:54 | |
| *** jrichli has joined #openstack-swift | 13:54 | |
| *** janonymous_ has quit IRC | 13:57 | |
| *** nexusz99 has quit IRC | 14:07 | |
| *** jome_ has joined #openstack-swift | 14:10 | |
| *** jome has quit IRC | 14:10 | |
| *** jome_ has quit IRC | 14:10 | |
| *** jome has joined #openstack-swift | 14:13 | |
| *** wbhuber has joined #openstack-swift | 14:17 | |
| *** jlhinson has joined #openstack-swift | 14:18 | |
| jrichli | mahatic: I will be offline this morning and the earlier part of tomorrow morning. | 14:19 |
| *** jasondotstar has joined #openstack-swift | 14:28 | |
| *** changbl has quit IRC | 14:32 | |
| *** vince_ has joined #openstack-swift | 14:51 | |
| vince_ | hi guys, is there a way to post dynamic large objects from the python APIs of python-swiftclient? | 14:53 |
| *** annegent_ has quit IRC | 14:54 | |
| vince_ | it's possible from the command line with the -S option, to split the object into segments and get the manifest automatically created, but I'd like a Python equivalent | 14:54 |
| *** minwoob has joined #openstack-swift | 14:58 | |
| *** pgbridge has joined #openstack-swift | 14:58 | |
| openstackgerrit | Nicolas Trangez proposed openstack/swift: Use entrypoints for storage policy implementation lookups https://review.openstack.org/206105 | 15:02 |
| eikke | notmyname: some time ago we talked about this ^^ I went ahead and created a patch | 15:02 |
| *** charlesw has quit IRC | 15:06 | |
| *** nadeem has joined #openstack-swift | 15:17 | |
| *** aix has quit IRC | 15:22 | |
| *** silor has joined #openstack-swift | 15:23 | |
| *** acoles is now known as acoles_away | 15:24 | |
| *** ppai has quit IRC | 15:24 | |
| *** acoles_away is now known as acoles | 15:25 | |
| *** early` has quit IRC | 15:29 | |
| acoles | ho: i'm back. catch you tomorrow maybe. | 15:29 |
| peluse | acoles, hey how are ya? | 15:30 |
| jrichli | acoles! How was your vaca? | 15:32 |
| *** early has joined #openstack-swift | 15:32 | |
| *** wer has quit IRC | 15:32 | |
| jrichli | I didn't have to be offline as long as I thought I would | 15:32 |
| acoles | peluse: jrichli hi guys! vacation was great thanks | 15:32 |
| jrichli | good to hear it. And good to have you back! I guess you have seen that there is an etherpad with encryption issues? | 15:33 |
| acoles | jrichli: not seen that yet, got a link? | 15:33 |
| acoles | jrichli: still catching up... | 15:33 |
| jrichli | its in the topic, actually | 15:34 |
| jrichli | https://etherpad.openstack.org/p/swift_encryption_issues | 15:35 |
| acoles | oic thx | 15:35 |
| jrichli | acoles: mahatic has joined the effort. She is working on the probetests that you had made trello cards for. | 15:37 |
| acoles | jrichli: nice, thanks mahatic | 15:37 |
| peluse | jrichli, ahh cool. I'll read it too. Need to get spun up on the crypto branch myself | 15:40 |
| jrichli | peluse: sounds great. would love your input | 15:41 |
| peluse | jrichli, so what's the current status of the feature branch wrt what should work/should fail in terms of unit/func/probe? | 15:41 |
| *** thurloat is now known as thurloat_isgone | 15:41 | |
| * peluse will BRB | 15:42 | |
| jrichli | peluse: 2 unittests fail (POST as copy and containerSync), functests errors=9, failures=17, probetests errors=2, failures=1 | 15:45 |
| notmyname | good morning | 15:45 |
| notmyname | acoles is back! | 15:45 |
| acoles | notmyname: remind me what we do here? :P | 15:46 |
| notmyname | acoles: I think we mostly roll our face against the keyboard until clayg stops griping at us. ;-) | 15:47 |
| jrichli | lol | 15:49 |
| notmyname | eikke: interesting. that will be worth a look. thanks | 15:49 |
| acoles | notmyname: lol! | 15:49 |
| eikke | notmyname: I'll experiment with our diskfile impl to find out whether it's as useful as I expect it to be :-p | 15:49 |
| notmyname | eikke: I had originally been thinking about diskfile implementations instead of "Storage policies" as the pluggable piece. | 15:50 |
| notmyname | eikke: ah, that'd be great | 15:50 |
| eikke | notmyname: well, I guess it makes sense to have a 'storage policy' define the underlying diskfile being used, otherwise there's some other strange indirection | 15:50 |
| *** thurloat_isgone is now known as thurloat | 15:51 | |
| notmyname | it depends :-) | 15:51 |
| eikke | notmyname: in the end, diskfiles define which policy types they support, or something along those lines. not sure how the lines should be drawn here | 15:51 |
| notmyname | hmm | 15:51 |
| eikke | but as noted in the commit message, having the split-in-3 (storage policy, proxy obj controller and diskfilemanager) makes not much sense | 15:52 |
| eikke | it's unclear to me how tightly (or configurable) coupling should be | 15:52 |
| eikke | which combos make sense and which dont | 15:53 |
| eikke | in the current swift codebase it's fairly coupled: [repl SP, repl objcontroller, repl diskfile] or [ec SP, ec objcontroller, ec diskfile] | 15:54 |
| eikke | although I could imagine a PACO/in-process objcontroller to be something untied to SP or diskfile | 15:54 |
| notmyname | yeah. torgomatic and clayg and I have talked about it some. | 15:58 |
| notmyname | on the one hand, EC diskfile vs replication diskfile is important because they do different things. on the other, when it comes down to just talking to storage media, it shouldn't matter what cluster durability scheme is used | 15:58 |
| peluse | jrichli, thanks. I see the same errors so guess we're ready to ship it eh? :) | 15:59 |
| jrichli | peluse: guess so! | 15:59 |
| *** rledisez has quit IRC | 16:00 | |
| * notmyname is hearing that encryption is done. good job, team ;-) | 16:00 | |
| *** mgarza_ has joined #openstack-swift | 16:00 | |
| peluse | who is buying t-shirts? | 16:01 |
| jrichli | LOL. only if you don't want to have a handful of features - you know, like COPY. and you want some middleware with FAKE in the name | 16:01 |
| *** wer has joined #openstack-swift | 16:01 | |
| jrichli | peluse: IBM. You will get them at Hackathon :-) | 16:02 |
| peluse | damn, so we have an entire two weeeks to get things done. that's awesome :) Feels like policies all over again! | 16:03 |
| *** vinsh has joined #openstack-swift | 16:06 | |
| *** thurloat is now known as thurloat_isgone | 16:07 | |
| *** acoles is now known as acoles_away | 16:07 | |
| *** vince_ has quit IRC | 16:08 | |
| jrichli | since acoles already left, I guess I will go for an early lunch ;-) | 16:08 |
| * peluse thinks maybe an early beer sounds good now for some reason. must be Monday... | 16:09 | |
| *** thurloat_isgone is now known as thurloat | 16:16 | |
| *** mragupat has joined #openstack-swift | 16:18 | |
| clayg | acoles_away: ! | 16:30 |
| *** acoles_away is now known as acoles | 16:30 | |
| clayg | notmyname: last year I clicked through every openstack summit to vote - it was terrible " | 16:31 |
| acoles | clayg: ! | 16:31 |
| peluse | clayg, yeah, I was wondering if it was me or is it really a PITA to go through all of them (even jsut storage)?? | 16:31 |
| notmyname | wow, there's a lot https://swiftstack.com/blog/2015/07/27/cast-your-vote-openstack-tokyo-summit-2015-presentations-on-swift/ | 16:32 |
| clayg | notmyname: yeah I think that's only the ones mario and his his team picked | 16:33 |
| clayg | there's probably tons more | 16:33 |
| clayg | *we're* doing a lot | 16:33 |
| notmyname | there's more at the bottom | 16:33 |
| clayg | well - i guess we'll *see* how many we'll do | 16:33 |
| clayg | notmyname: yeah but that's not all "storage" - they've dont some filtering | 16:33 |
| notmyname | oh, yeah. I keep forgetting. storage != just swift ;-) | 16:33 |
| peluse | what?? | 16:34 |
| clayg | lol | 16:34 |
| *** jasondotstar has quit IRC | 16:34 | |
| haypo | notmyname: hi. for my change "Update hacking to 0.10.0 or newer", i now understand the consequence of my change on tox.ini: i disabled *all* Hxxx warnings, whereas before 6 warnings were enabled | 16:36 |
| haypo | notmyname: i can only blacklist Hxxx warnings which were disabled if you prefer | 16:36 |
| haypo | notmyname: the patches related to coding style are quite huge, so they are likely to conflict | 16:37 |
| haypo | notmyname: i'm not sure that i will want to spend too much time on them if you are not ready to review them | 16:37 |
| notmyname | haypo: we settled on the current way as an explicit whitelist of stuff we wanted to test on. that way as stuff is added, it won't suddenly cause errors | 16:37 |
| haypo | notmyname: so what's the opinion on them? | 16:37 |
| *** janonymous_ has joined #openstack-swift | 16:39 | |
| notmyname | I want to keep testing the ones that are explicitly called out. if there are others that we should consider adding, let's discuss it | 16:39 |
| notmyname | does the hacking version update still allow that? having just a whitelist of stuff that is checked? | 16:39 |
| *** thurloat is now known as thurloat_isgone | 16:40 | |
| *** jasondotstar has joined #openstack-swift | 16:40 | |
| haypo | notmyname: i don't think that the behaviour of hacking (flake8) changed | 16:40 |
| notmyname | ok (good :-) | 16:40 |
| notmyname | so update hacking, if needed, but leave the checks as-is | 16:40 |
| haypo | notmyname: the "ignore =" was already ignored before (but i'm not 100% sure) | 16:40 |
| notmyname | ah | 16:41 |
| haypo | notmyname: my bet is more that pep8 & flake8 have more tests | 16:41 |
| notmyname | oh, I know they do. actually, it's hacking that has a lot more checks | 16:41 |
| haypo | anyway, swift cannot hack pep8 or flake8, we have to adopt the new behaviour anyway ;) | 16:41 |
| notmyname | pep8 should still all be checked | 16:41 |
| notmyname | it's the hacking checks we've got a whitelist on | 16:42 |
| *** geaaru has quit IRC | 16:42 | |
| haypo | notmyname: it's amazing how much stuff should be fixed first to enable python3 tests on Swift... | 16:44 |
| haypo | notmyname: dnspython3 (WIP), PyEClib (done!), hacking, what's next? :-) | 16:44 |
| notmyname | haypo: which is why I'm impressed you and janonymous are sticking with it! | 16:45 |
| haypo | haha | 16:45 |
| haypo | notmyname: it's almost my job to port OpenStack to Python 3 | 16:45 |
| haypo | notmyname: would you be interested by patches fixing some flake8 warnings, like the one on assert_? | 16:46 |
| notmyname | only if they can also be elevated to an error so that future me won't re-introduce it. ie it would have to break something obviously | 16:47 |
| haypo | notmyname: i see | 16:48 |
| *** janonymous_ has quit IRC | 16:49 | |
| *** darrenc_ has joined #openstack-swift | 16:51 | |
| *** janonymous_ has joined #openstack-swift | 16:52 | |
| *** darrenc has quit IRC | 16:52 | |
| openstackgerrit | Victor Stinner proposed openstack/swift: Update hacking to 0.10.0 or newer https://review.openstack.org/205977 | 16:52 |
| haypo | notmyname: this new change enables _more_ tests than before, and keep existing Hxxx warnings: https://review.openstack.org/#/c/205977/3/tox.ini | 16:53 |
| haypo | notmyname: it's better than ignoring all hackings tests (Hxxx) | 16:53 |
| *** nexusz99 has joined #openstack-swift | 16:53 | |
| *** jistr has quit IRC | 16:55 | |
| notmyname | haypo: why not stick with the selected list of things to check? | 16:56 |
| haypo | notmyname: it's not possible to use select _and_ ignore | 16:56 |
| notmyname | ok, so why not just use select? won't that implicitly ignore everything else? | 16:56 |
| haypo | notmyname: it's a dummy limitation of pep8 (so also of flake8): https://github.com/PyCQA/pep8/blob/d48eef071003952ed0f98445dc3dc6bc29618f9c/pep8.py#L1808 | 16:56 |
| haypo | notmyname: i want to get the ability to ignore specific tests | 16:57 |
| notmyname | why? | 16:57 |
| haypo | notmyname: IHMO "F812: list comprehension redefines ..." warnings are really stupid | 16:57 |
| notmyname | sure. so don't put them in the select line :-) | 16:57 |
| notmyname | assuming that flake8 would actually allow both select and ignore, then the current version of our tox.ini says "ignore everything and then select these few". right? | 16:58 |
| haypo | notmyname: one example: https://github.com/openstack/swift/blob/master/swift/common/middleware/x_profile/html_viewer.py#L415 | 16:58 |
| haypo | notmyname: "F812 list comprehension redefines '__' from line 410" | 16:58 |
| haypo | notmyname: the __ variable is replaced... well.. ok, so what? should I use the same "___" or write "del __" ? | 16:58 |
| notmyname | ugh. yeah, that's annoying | 16:58 |
| notmyname | ah, ok, I think I see | 16:59 |
| haypo | notmyname: the scope of the variable is just fine, it's ok to reuse variables, we don't do SSA in Python (Static single assignment form) | 16:59 |
| notmyname | we had the select line F, E, W, and the Hxxx warnings | 17:00 |
| haypo | notmyname: with hacking 0.8, we had much less warnings, so it wasn't so much importable to ignore a single check | 17:00 |
| notmyname | which means that we can't have select = F and also ignore F812 | 17:00 |
| haypo | notmyname: since you use the "select =" line, you cannot use "ignore =" anymore | 17:00 |
| *** nexusz99 has quit IRC | 17:00 | |
| haypo | notmyname: it's annoying | 17:00 |
| notmyname | what's the impact of the warning on F812? does it cause flake8 to fail? | 17:01 |
| haypo | notmyname: without ignore = F812, "tox -e pep8" fails | 17:02 |
| notmyname | wow | 17:02 |
| *** jordanP has quit IRC | 17:02 | |
| notmyname | and you're working on updating the hacking version because that's required if you update pbr which is required for py3? | 17:02 |
| notmyname | 'cause so far I'm not seeing any advantage ot update hacking ;-) | 17:03 |
| haypo | notmyname: ahaha, that's insane | 17:03 |
| haypo | notmyname: but yes, you rebuilt my puzzle | 17:03 |
| haypo | notmyname: well, hacking helps you to have a better coding style | 17:03 |
| haypo | notmyname: at the end, the code should be more readable | 17:03 |
| notmyname | what's the upstream opinion on not being able to both select and ignore with flake8? | 17:05 |
| notmyname | bug or working as designed? | 17:05 |
| haypo | notmyname: i didn't ask | 17:05 |
| haypo | notmyname: for me, it's more a bug, but from flake8 point of view, it can be seen as a new feature :-D | 17:06 |
| haypo | notmyname: i'm not sure that pep8 is designed to use select, but more to use ignore | 17:06 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Fix hacking H232 warnings (octal) https://review.openstack.org/206149 | 17:07 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Fix hacking H403 warnings (docstrings) https://review.openstack.org/206150 | 17:07 |
| notmyname | the problem I have with the ignore methodology is that the next time something is added that we don't pass, our gate will fail | 17:07 |
| notmyname | hmm..seems like it's time for our yearly "why are we actually using pbr because it just seems to get in the way" arguments | 17:09 |
| notmyname | ;-) | 17:09 |
| haypo | notmyname: i'm checking nova. it also uses "hacking<0.11,>=0.10.0" in test-requirements.txt and "ignore = ..." in tox.ini | 17:09 |
| notmyname | haypo: IIRC most project do only have a few they select and then ignore the rest | 17:09 |
| redbo | yay | 17:09 |
| notmyname | redbo: you must have an alert on "pbr" ;-) | 17:09 |
| haypo | notmyname: nova dev hates other developers when they introduce a regression, so you can trust them that it's the well usage of flake8 :) | 17:09 |
| haypo | notmyname: "most project": ah? which ones? | 17:10 |
| janonymous_ | I think https://review.openstack.org/#/c/204093/ will fix H235 | 17:10 |
| *** nexusz99 has joined #openstack-swift | 17:10 | |
| notmyname | anecdotal based on my own hallway conversations a while (years) back | 17:10 |
| haypo | cinder, glance: same than nova (ignore) | 17:10 |
| haypo | nova: same than nova (ignore) | 17:11 |
| haypo | notmyname: i trust hacking developers to *not* break cinder, glance, neutron, nova by adding new random tests just to be pedantic | 17:11 |
| notmyname | yet here we are ;-) | 17:11 |
| haypo | notmyname: that's exactly why my patch uses "hacking>=0.10.0,<0.11" | 17:12 |
| haypo | notmyname: hacking must not add new tests in minor versions (0.10.x), only in major versions (0.11). well, that's what I understood | 17:12 |
| haypo | notmyname: FYI i wrote two patches removing checks from the ignore list, to test this approach (fixing a warning and ensure that it doesn't come back): https://review.openstack.org/206149 & https://review.openstack.org/206150 | 17:13 |
| notmyname | I'm sure the hacking devs will do their best to do the right thing. I'm leery of planting a time bomb for ourselves. comes down to explicit vs implicit | 17:15 |
| notmyname | I need to step away for a fe minutes | 17:16 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Update hacking to 0.10.0 https://review.openstack.org/205977 | 17:23 |
| openstackgerrit | Victor Stinner proposed openstack/swift: Fix hacking H232 warnings (octal) https://review.openstack.org/206149 | 17:26 |
| haypo | notmyname: ok. i simplified my patch serie into one unique patch: https://review.openstack.org/#/c/205977/ (+ example: https://review.openstack.org/#/c/206149/) | 17:27 |
| haypo | the idea is then to fix issues reported by flake8 and each time enable one more check in tox.ini (flake8) | 17:28 |
| haypo | notmyname: i have to go, bye! | 17:28 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Refactor diskfile https://review.openstack.org/198429 | 17:28 |
| notmyname | haypo: thanks | 17:28 |
| * notmyname just got back | 17:28 | |
| *** acoles is now known as acoles_away | 17:30 | |
| notmyname | haypo: yes! that pattern of tacking it I like. get base functionality then add them each one at a time | 17:30 |
| *** zaitcev has joined #openstack-swift | 17:31 | |
| *** ChanServ sets mode: +v zaitcev | 17:31 | |
| *** ChanServ sets mode: +v tdasilva | 17:31 | |
| *** mgarza_ has quit IRC | 17:44 | |
| *** nakagawamsa has quit IRC | 17:47 | |
| *** NM has joined #openstack-swift | 17:47 | |
| *** NM has quit IRC | 17:47 | |
| *** NM has joined #openstack-swift | 17:48 | |
| clayg | well that was an interesting scrollback | 17:49 |
| notmyname | clayg: on hacking? | 17:51 |
| *** jome_ has joined #openstack-swift | 17:57 | |
| *** annegentle has joined #openstack-swift | 17:58 | |
| *** jome has quit IRC | 18:01 | |
| *** blmartin has quit IRC | 18:09 | |
| *** changbl has joined #openstack-swift | 18:12 | |
| *** annegentle has quit IRC | 18:15 | |
| *** jome has joined #openstack-swift | 18:17 | |
| *** jome_ has quit IRC | 18:19 | |
| *** mahatic has joined #openstack-swift | 18:24 | |
| *** jasondot_ has joined #openstack-swift | 18:25 | |
| portante | Are folks aware of this PEP coming in Python 3.5? https://www.python.org/dev/peps/pep-0475/ | 18:25 |
| *** janonymous_ has quit IRC | 18:26 | |
| portante | retrying system calls in the python stdlib at the lowest levels | 18:26 |
| notmyname | portante: nope | 18:26 |
| *** annegentle has joined #openstack-swift | 18:27 | |
| *** jasondotstar has quit IRC | 18:27 | |
| hurricanerix | notmyname: I remember seeing a blueprint for proposed /v2 stuff, but can't seem to find it now. Do you know where it is? | 18:31 |
| notmyname | hurricanerix: auth or api? | 18:32 |
| hurricanerix | api | 18:32 |
| notmyname | hurricanerix: like a v2 swift api? | 18:32 |
| notmyname | yeah | 18:32 |
| hurricanerix | yeah | 18:32 |
| *** eandersson has quit IRC | 18:32 | |
| notmyname | yeah, where was that? | 18:32 |
| hurricanerix | lol, I know right, I have been looking around for it, but haven't been able to find it. | 18:32 |
| zaitcev | Heh. | 18:32 |
| zaitcev | I suspect there may be dozens. | 18:33 |
| hurricanerix | zaitcev: yeah, I know I saw one, once. But now when I look, all I find is auth stuff or people asking questions about when swift's version got bumped to 2.* and asking why the URL was still /v1 | 18:34 |
| zaitcev | I talked about it w/ Yehuda, the RGW leader. He has this idea to create tenantized buckets, which are containers in our case. But he insists that we use '/', e.g. "tenant/container". That offsets everything by a level in Swift. | 18:34 |
| *** ohno13 has joined #openstack-swift | 18:34 | |
| zaitcev | So I said, "okay, worst come to worst, I'll propose a /v2, then we'll know there's an extra slash" | 18:35 |
| notmyname | there's been lots of conversations about it. not sure what's been written down, if anything | 18:35 |
| notmyname | most recent I idea I've heard is to have a middleware that translates stuff back to v1 for older clients | 18:36 |
| hurricanerix | notmyname: I know I saw at least one thing written down once. Maybe torgomatic or clayg gave it to me before, but I don't remember for sure who it was. | 18:36 |
| notmyname | hurricanerix: theres https://wiki.openstack.org/wiki/SwiftNextAPI and https://wiki.openstack.org/wiki/Swift/API which look like really old scratch pads | 18:37 |
| *** proteusguy has quit IRC | 18:38 | |
| hurricanerix | notmyname: the next api one was the one I was looking for. Thanks | 18:40 |
| notmyname | hurricanerix: should I read anything into your question? ;-) | 18:40 |
| notmyname | hurricanerix: FWIW, I found that linked at the bottom of https://wiki.openstack.org/wiki/Swift | 18:41 |
| hurricanerix | notmyname: not really, I just had an idea that might fit a /v2 change, and wanted to re-read what existing things were proposed. | 18:41 |
| hurricanerix | Ahh, I kept searching for "v2" which is probably why I didn't find it. =) | 18:41 |
| *** annegentle has quit IRC | 18:43 | |
| hurricanerix | notmyname: I was thinking about how there is some middleware that gets auto enabled in the pipeline because it was migrated from swift to middleware. And it seemed like maybe the /v2 API might be a good opportunity to remove that restriction so that if someone wanted to setup a vanilla swift cluster with no extra features, they are not forced into it. | 18:44 |
| notmyname | hurricanerix: depends on what you consider "extra" ;-) | 18:44 |
| clayg | notmyname: hurricanerix was wondering if we could make DLO's optional ;) | 18:45 |
| notmyname | yeah. I'm going the other way ;-) | 18:45 |
| torgomatic | so someone wants an extra URL path component of /tenant/ for reasons? boo | 18:45 |
| torgomatic | that's nonsense when talking about non-Keystone auth systems | 18:45 |
| notmyname | middleware is a way to organize code and control flow. it doesn't mean something is optional | 18:45 |
| hurricanerix | True, then why require it be configured in the pipeline to begin with? | 18:46 |
| hurricanerix | well i guess it isn't required | 18:46 |
| clayg | notmyname: i think when defining a v2 api we should have the *option* of not including some v1 features :P | 18:46 |
| notmyname | sure | 18:47 |
| hurricanerix | I was looking at it from the point that the only reason DLOs are required, is because they were written before we put everything in middleware, no? | 18:47 |
| hurricanerix | (or anything that is a feature in swift, and has been migrated to middleware) | 18:47 |
| notmyname | I just have knee-jerk reaction to middleware==pluggable after trying to correct sales people a dozen times ;-) | 18:47 |
| clayg | hurricanerix: i think they're required because of clients mostly :\ | 18:47 |
| hurricanerix | yeah, so a client wouldn't have an expectation that they would be supported in /v2 | 18:48 |
| hurricanerix | and would have to check /info or somthing | 18:48 |
| clayg | hurricanerix: +1 I think it's totally debatable what's included in the v2 api feature wise | 18:48 |
| notmyname | or rather that middleware==optional and every config option is middleware. (of course middleware is pluggable) | 18:48 |
| clayg | hurricanerix: it's all about covering the use cases | 18:48 |
| notmyname | hurricanerix: yeah, I definitely thing any v2 api should be tied to what's in /info | 18:48 |
| hurricanerix | yeah, I was thinking about brining it up at the hackathon, but then couldn't find that link to read over. | 18:48 |
| openstackgerrit | Ben Martin proposed openstack/swift: Document method to avoid rsync filling root drive https://review.openstack.org/199203 | 18:49 |
| clayg | hurricanerix: awesome! good idea! | 18:49 |
| hurricanerix | clayg: thanks! | 18:49 |
| notmyname | the https://etherpad.openstack.org/p/swift-midcycle-aug-2015 link in the topic is where we should put stuff like this :-) | 18:50 |
| clayg | notmyname: oh nice! thanks! | 18:50 |
| *** mragupat_ has joined #openstack-swift | 18:50 | |
| notmyname | actually, hurricanerix created it and it was his idea! | 18:50 |
| hurricanerix | notmyname: thanks! I had lost the link... =) | 18:50 |
| hurricanerix | (always back up your bookmarks when reinstalling your OS...) | 18:51 |
| redbo | or quit jumping between linux and osx every 5 minutes | 18:51 |
| notmyname | I gotta run. I'll be out for a few hours, then back on tonight | 18:51 |
| hurricanerix | redbo: that has only happened two or three times... | 18:52 |
| *** mragupat has quit IRC | 18:54 | |
| *** proteusguy has joined #openstack-swift | 18:55 | |
| redbo | how about we take out middleware configuration entirely and put in feature flags that then set up the pipeline in the right order. because it's getting pretty goofballs. | 18:56 |
| *** mfalatic has joined #openstack-swift | 18:58 | |
| redbo | with like 25 middlewares that have to be in the right order to not break things in weird ways or cause security problems | 18:59 |
| hurricanerix | redbo: Do you have an example of how that would work, because I am not sure I understand. Are you proposing that it is just a "on/off" type thing? | 18:59 |
| torgomatic | redbo: that's probably not a bad idea | 19:00 |
| redbo | Yeah. slo = true, then it puts that middleware in the right spot. which might not be simple, but we're making people do it right now. | 19:00 |
| hurricanerix | redbo: so do they loose the ability to tweak them? or they would just do it a different way? | 19:00 |
| *** annegentle has joined #openstack-swift | 19:02 | |
| redbo | I don't know. Maybe toss in a feature config section if it's really necessary. Are we all really changing the defaults on lots of middlewares? | 19:02 |
| hurricanerix | redbo: min_segment_size helps a lot in testing envs | 19:02 |
| hurricanerix | Also, how would third party middlewares work with that setup? | 19:04 |
| redbo | oh, true | 19:05 |
| redbo | Still. goofballs. | 19:06 |
| hurricanerix | maybe slo = true could just be a shorthand to add things to the pipeline. and modifying the pipeline yourself would become "use at your own risk". =) | 19:06 |
| *** _hrou_ is now known as hrou | 19:06 | |
| *** blmartin has joined #openstack-swift | 19:08 | |
| *** mahatic has quit IRC | 19:09 | |
| *** mahatic has joined #openstack-swift | 19:09 | |
| mahatic | anyone know what this is supposed to mean? (although I do have that in entry_points.txt) | 19:10 |
| mahatic | "LookupError: Entry point 'decrypter' not found in egg 'swift' (dir: /home/mahati/swift; protocols: paste.filter_factory, paste.filter_app_factory; entry_points:" | 19:10 |
| *** robefran has joined #openstack-swift | 19:11 | |
| *** annegentle has quit IRC | 19:11 | |
| mahatic | I am getting that on "swift-init start main" | 19:12 |
| *** annegentle has joined #openstack-swift | 19:12 | |
| clayg | slo = true is great - but how do I say *my* middleware needs to be after slo? | 19:13 |
| clayg | mahatic: you need to checkout the crypto change (I think the tail review is on the priority review wiki) and then setup.py develop again (pip install -e .) | 19:13 |
| clayg | hurricanerix: I think use at your own risk is sort of what we have now - only there's no other option | 19:15 |
| clayg | hurricanerix: making it somewhat easier to do the stock thing (like copying what's there is "too hard") doesn't really solve a general way for middleware authors to not just *describe* but *maintain* a set of rules that will allow the computer to build the pipeline *correctly* | 19:16 |
| *** robefran_ has joined #openstack-swift | 19:18 | |
| *** robefran has quit IRC | 19:21 | |
| *** mahatic has quit IRC | 19:21 | |
| *** mahatic has joined #openstack-swift | 19:21 | |
| openstackgerrit | Ben Martin proposed openstack/swift: +Document method to avoid rsync filling root drive https://review.openstack.org/199203 | 19:25 |
| *** ccavanna has joined #openstack-swift | 19:28 | |
| mahatic | clayg: thanks. I did that. I was trying to run that on "feature/crypto" branch | 19:31 |
| *** silor has quit IRC | 19:32 | |
| jrichli | mahatic: are you still using the config file from the paste the other day? | 19:36 |
| jrichli | mahatic: although, I agree with clayg. sounds like the code you have installed doesn't include the new stuff | 19:37 |
| mahatic | jrichli: hi, yes | 19:37 |
| mahatic | um, I took that code again | 19:38 |
| mahatic | jrichli: I did cross check the code on the tail review and my local | 19:40 |
| jrichli | its the develop and install step you might want to do again: and makes sure that is pointing to the right repo | 19:41 |
| jrichli | your "egg" doesn't have the new middleware in it | 19:41 |
| *** lpabon has joined #openstack-swift | 19:42 | |
| mahatic | jrichli: I am not on "feature/crypto" branch (I took the code and created a new branch) | 19:43 |
| jrichli | can you see the decrypter.py file in your repo? | 19:45 |
| mahatic | yeah | 19:45 |
| mahatic | I get that error when I run the startmain script on "feature/crypto" or master | 19:45 |
| mahatic | (I did do the install and develop) | 19:46 |
| jrichli | you shouldn't run against just feature/crypto. it doesn't have that file | 19:46 |
| jrichli | you have to start from the code on the latest review. The new middleware has never been accepted onto the feature branch | 19:47 |
| mahatic | jrichli: thanks! that's what I wanted to confirm | 19:48 |
| jrichli | make your new branch track the review branch | 19:48 |
| mahatic | yeah, that's there | 19:48 |
| mahatic | I'm still sweating out on pip issues! | 19:48 |
| *** breitz has quit IRC | 20:19 | |
| *** breitz has joined #openstack-swift | 20:19 | |
| *** ohno13 has quit IRC | 20:23 | |
| *** jome has quit IRC | 20:25 | |
| peluse | or maybe you mean "swearing" out on pip issues :) | 20:28 |
| *** robefran_ has quit IRC | 20:30 | |
| *** dmsimard has quit IRC | 20:30 | |
| *** dmsimard has joined #openstack-swift | 20:34 | |
| *** lpabon has quit IRC | 20:35 | |
| *** mragupat_ has quit IRC | 20:36 | |
| *** mragupat has joined #openstack-swift | 20:37 | |
| *** mragupat has quit IRC | 20:37 | |
| *** annegentle has quit IRC | 20:37 | |
| *** mragupat has joined #openstack-swift | 20:37 | |
| *** mragupat has quit IRC | 20:38 | |
| *** mragupat has joined #openstack-swift | 20:39 | |
| clayg | !@#$%ing pip | 20:49 |
| openstack | clayg: Error: "@#$%ing" is not a valid command. | 20:49 |
| clayg | openstack: !@#$% you | 20:49 |
| *** blmartin_ has joined #openstack-swift | 20:56 | |
| *** blmartin has quit IRC | 20:56 | |
| *** blmartin_ has quit IRC | 20:57 | |
| *** blmartin__ has joined #openstack-swift | 20:57 | |
| *** tongli has quit IRC | 20:57 | |
| haypo | notmyname: yeah, it took me some iterations to find the simple solution ;-) it's always like that | 20:57 |
| jrichli | clayg : LOL | 21:00 |
| mahatic | lol | 21:00 |
| mahatic | :D | 21:00 |
| *** annegentle has joined #openstack-swift | 21:06 | |
| *** elmiko has joined #openstack-swift | 21:07 | |
| elmiko | hi folks, has there been any talk(or blueprints/reviews) about adding keystone Session support to the swiftclient? | 21:07 |
| *** hrou has quit IRC | 21:16 | |
| *** CaioBrentano has joined #openstack-swift | 21:17 | |
| openstackgerrit | Merged openstack/swift: Use right most occurrence object dir in diskfile.extract_policy https://review.openstack.org/205641 | 21:26 |
| *** ccavanna has quit IRC | 21:34 | |
| *** jrichli has quit IRC | 21:44 | |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Rename WsgiStringIO -> WsgiBytesIO. https://review.openstack.org/203775 | 21:45 |
| *** blmartin__ has quit IRC | 21:46 | |
| *** mragupat has quit IRC | 21:57 | |
| *** mahatic has quit IRC | 21:57 | |
| *** mahatic has joined #openstack-swift | 22:05 | |
| *** chlong has joined #openstack-swift | 22:09 | |
| *** jlhinson has quit IRC | 22:09 | |
| *** jasondot_ has quit IRC | 22:15 | |
| *** CaioBrentano has quit IRC | 22:18 | |
| *** tongli has joined #openstack-swift | 22:18 | |
| *** CaioBrentano has joined #openstack-swift | 22:18 | |
| *** CaioBrentano has quit IRC | 22:22 | |
| *** marcusvrn has quit IRC | 22:25 | |
| clayg | elmiko: timburke is working on something like that | 22:25 |
| clayg | elmiko: well acctually wait - what is a keystone session | 22:26 |
| *** NM has quit IRC | 22:27 | |
| *** haypo has left #openstack-swift | 22:42 | |
| *** ccavanna has joined #openstack-swift | 22:43 | |
| *** hrou has joined #openstack-swift | 22:44 | |
| *** annegentle has quit IRC | 22:50 | |
| *** DericHorn-HP has joined #openstack-swift | 22:51 | |
| *** nadeem has quit IRC | 22:53 | |
| peluse | clayg, torgomatic - did you guys see my note on patch 201283? | 22:54 |
| patchbot | peluse: https://review.openstack.org/#/c/201283/ | 22:54 |
| clayg | peluse: yeah you like sent an email right - it was like "pls review, kthxbai" | 22:56 |
| peluse | pretty much :) | 22:57 |
| peluse | and now I'm doing the same thing here | 22:57 |
| clayg | :) | 22:57 |
| peluse | so basically the perf work we're trying to do with you guys on the 15 node cluster here in Phx is stalled on the EC side. We're updating code today to close to master (Swiftstack 2.50) in an attempt to avoid having to debug a lot of misc shit | 22:58 |
| peluse | like FI's missing where they shouldn't be, etc | 22:58 |
| *** DericHorn-HP has quit IRC | 22:58 | |
| clayg | peluse: right on | 22:59 |
| peluse | so not only do I want to get our stuff ugpraded, want to get the other EC bug fixes landed too, as does everyone else, or we won't get good data out of this thing. others don't seem to hit the bugs we're hitting but not sure anyone else has a full 15 node 12 disk per node all xeon cluster dedicated to beating on the Ec code | 23:00 |
| clayg | peluse: yes I agree these bugs are delaying a lot of testing | 23:01 |
| peluse | yeah, I'm guily of spending to much time on just that one, going to spend the rest of this week looking at the others as well (minus an all day Wed shin dig I'm stuck in) | 23:02 |
| *** DericHorn-HP has joined #openstack-swift | 23:05 | |
| *** km has joined #openstack-swift | 23:06 | |
| *** DericHorn-HP has quit IRC | 23:06 | |
| *** jasondotstar has joined #openstack-swift | 23:08 | |
| *** jasondotstar has quit IRC | 23:13 | |
| *** kei_yama has joined #openstack-swift | 23:13 | |
| *** darrenc_ is now known as darrenc | 23:18 | |
| *** jasondot_ has joined #openstack-swift | 23:28 | |
| *** onovy has quit IRC | 23:29 | |
| *** onovy has joined #openstack-swift | 23:31 | |
| peluse | hmmm, just went through the current list of Ec bugs. Not too bad I don't think... | 23:36 |
| *** jasondotstar has joined #openstack-swift | 23:44 | |
| *** DericHorn-HP has joined #openstack-swift | 23:44 | |
| *** jasondot_ has quit IRC | 23:47 | |
| *** tongli has quit IRC | 23:54 | |
| *** ho has joined #openstack-swift | 23:54 | |
| elmiko | clayg: i was referring to http://docs.openstack.org/developer/python-keystoneclient/using-sessions.html# | 23:56 |
| elmiko | keystoneclient.session.Session and the related authentication plugins to authenticate/authorize clients, instead of parameterized auth | 23:56 |
| elmiko | well, i shouldn't say authorize just authenticate | 23:57 |
| elmiko | but the Session objects allow for things like connection pooling to keystone and a centralized abstraction for security updates and the like | 23:57 |
| ctennis | peluse: 2.50 release is hinging on clayg, so whisper sweet thoughts into his ear | 23:57 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!