| kota_ | indeed | 00:00 |
|---|---|---|
| kota_ | notmyname: from the experience, i'd like to vote humming bird into swift upstream repo :/ | 00:01 |
| notmyname | kota_: yes!! | 00:01 |
| notmyname | this is exactly why I've been advocating for that. we'll see what happens | 00:02 |
| kota_ | separate repo is ofenly my frusration in pyeclib/liberasurecode/swift3 etc... | 00:02 |
| kota_ | oftenly | 00:02 |
| notmyname | *commonly | 00:02 |
| *** david_c has quit IRC | 00:02 | |
| timburke | or just "often" | 00:02 |
| kota_ | thanks teaher | 00:02 |
| kota_ | teacher | 00:02 |
| *** clu_ has quit IRC | 00:05 | |
| *** clu_ has joined #openstack-swift | 00:06 | |
| * kota_ is going to offline for going to his office, will be back in 1 or 2 hours | 00:07 | |
| *** klamath has quit IRC | 00:13 | |
| *** StraubTW has quit IRC | 00:13 | |
| *** david_c has joined #openstack-swift | 00:18 | |
| *** chlong has joined #openstack-swift | 00:23 | |
| *** diogogmt has quit IRC | 00:23 | |
| *** asettle has joined #openstack-swift | 00:31 | |
| *** asettle has quit IRC | 00:35 | |
| *** vint_bra has quit IRC | 00:45 | |
| *** jerrygb has quit IRC | 00:48 | |
| *** jerrygb has joined #openstack-swift | 00:49 | |
| *** jerrygb has quit IRC | 00:49 | |
| *** jerrygb has joined #openstack-swift | 00:49 | |
| mattoliverau | sorry I missed the meeting this morning, the wife and I have managed to get food poisoning, and lets just say the last day or so hasn't been pleasent. Still feel like crap so not working today either, I hope to be back at it tomorrow. | 00:58 |
| *** vint_bra has joined #openstack-swift | 01:12 | |
| *** StraubTW has joined #openstack-swift | 01:14 | |
| clayg | mattoliverau: get well soon - don't forget ptg tickets! | 01:15 |
| *** tsg has quit IRC | 01:16 | |
| *** diogogmt has joined #openstack-swift | 01:17 | |
| *** StraubTW has quit IRC | 01:19 | |
| clayg | notmyname: thanks for doing that release thing hommie! | 01:21 |
| notmyname | clayg: I hope you're unblocked. I want to continue to figure out the rest of it tomorrow | 01:25 |
| clayg | notmyname: yah i'm good - i'm testing bro ;) | 01:28 |
| *** vint_bra has quit IRC | 01:38 | |
| *** asettle has joined #openstack-swift | 02:00 | |
| *** asettle has quit IRC | 02:05 | |
| charz_ | morning | 02:05 |
| jrichli | mattoliverau: I hope that you both feel better soon | 02:09 |
| *** klrmn has quit IRC | 02:13 | |
| *** david_c has quit IRC | 02:17 | |
| *** abalfour has quit IRC | 02:20 | |
| *** abalfour has joined #openstack-swift | 02:20 | |
| *** clu_ has quit IRC | 02:40 | |
| *** tsg has joined #openstack-swift | 02:49 | |
| openstackgerrit | Cheng Li proposed openstack/swift: prevent object-server from start if no devices https://review.openstack.org/395911 | 02:53 |
| *** StraubTW has joined #openstack-swift | 02:54 | |
| *** klrmn has joined #openstack-swift | 03:13 | |
| *** klrmn has quit IRC | 03:25 | |
| *** asettle has joined #openstack-swift | 03:31 | |
| *** winggundamth has quit IRC | 03:31 | |
| *** jerrygb has quit IRC | 03:33 | |
| *** asettle has quit IRC | 03:35 | |
| *** klrmn has joined #openstack-swift | 03:40 | |
| *** takashi has joined #openstack-swift | 03:48 | |
| *** klrmn has quit IRC | 03:51 | |
| *** klrmn has joined #openstack-swift | 03:54 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/liberasurecode: ISA-L Cauchy support https://review.openstack.org/393263 | 04:06 |
| *** StraubTW has quit IRC | 04:06 | |
| *** dikonoor has joined #openstack-swift | 04:07 | |
| *** chlong has quit IRC | 04:13 | |
| *** dmorita has quit IRC | 04:14 | |
| *** dmorita has joined #openstack-swift | 04:16 | |
| *** dmorita has quit IRC | 04:19 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/pyeclib: ISA-L Cauchy support https://review.openstack.org/393276 | 04:19 |
| *** links has joined #openstack-swift | 04:23 | |
| *** klrmn has quit IRC | 04:27 | |
| *** jerrygb has joined #openstack-swift | 04:34 | |
| *** jerrygb has quit IRC | 04:39 | |
| kota_ | tdasilva, notmyname: As I thought, the bindep deosn't work for the dependency management if we built the binary from source, https://gist.github.com/bloodeagle40234/fd81c65c0ee9bb2513ed00e78a23899b | 04:44 |
| kota_ | tdasilva, notmyname: probably the bindep refers the package repo (e.g. apt, yum) version | 04:45 |
| * kota_ is going to try his idea, hard-coded dependency limitation which may be painful needle for devstack gate | 04:47 | |
| *** dmorita has joined #openstack-swift | 04:56 | |
| *** dmorita has quit IRC | 04:58 | |
| *** dmorita has joined #openstack-swift | 04:59 | |
| mahatic | mattoliverau: get well soon | 05:00 |
| *** dikonoor has quit IRC | 05:08 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:31 | |
| kota_ | mattolivearu: please take care of yourself | 05:46 |
| *** dmorita has quit IRC | 05:49 | |
| *** dmorita has joined #openstack-swift | 05:50 | |
| *** dmorita has quit IRC | 05:54 | |
| *** dmorita has joined #openstack-swift | 05:58 | |
| *** dfflanders has quit IRC | 05:59 | |
| charz_ | mattoliverau: sorry to hear that, hope you fell better soon. | 06:05 |
| *** dmorita has quit IRC | 06:06 | |
| *** dmorita has joined #openstack-swift | 06:06 | |
| *** StraubTW has joined #openstack-swift | 06:07 | |
| *** dmorita has quit IRC | 06:08 | |
| *** dmorita has joined #openstack-swift | 06:08 | |
| *** ChubYann has quit IRC | 06:09 | |
| charz_ | s/fell/fell | 06:10 |
| charz_ | s/fell/feel | 06:10 |
| *** StraubTW has quit IRC | 06:12 | |
| *** asettle has joined #openstack-swift | 06:16 | |
| *** rcernin has joined #openstack-swift | 06:17 | |
| *** asettle has quit IRC | 06:21 | |
| *** dikonoor has joined #openstack-swift | 06:23 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/liberasurecode: Need to bump the version to 1.3.0 also erasurecode_version.h https://review.openstack.org/395966 | 06:30 |
| kota_ | notmyname: i noticed we should bump the header version in liberasurecode too. sorry i forgot it in this morning. | 06:33 |
| notmyname | kota_: ah? so another point release for liberasurecode? | 06:33 |
| *** jerrygb has joined #openstack-swift | 06:35 | |
| kota_ | notmyname: not sure if we need to make another release point but the change in the erasurecode_version.h is need for pyeclib to know which liberasurecode version is used | 06:35 |
| zaitcev | already tagged? | 06:35 |
| zaitcev | can you do git push -f --tags? | 06:36 |
| notmyname | kota_: ah, I see. sorry about that. now I know for next time | 06:37 |
| notmyname | zaitcev: not sure, but I don't think that's the right answer anyway | 06:37 |
| notmyname | we can do a simple 1.3.1 | 06:37 |
| zaitcev | ooooooh, that's actually superb. It now matches pyeclib | 06:38 |
| zaitcev | you can pretend that it was planned | 06:38 |
| notmyname | heh. except we'll be releasing pyeclib, too, right? to use the 1.3.1 of liberasurecode | 06:38 |
| notmyname | hmm...I wonder if this would be a convenient opportunity to combine them | 06:39 |
| zaitcev | Oh, right. In RDO it's all in RPM spec, so I don't really need to bump python-pyeclib, as long as it's compatible with the new liberasurecode. | 06:39 |
| notmyname | true, but you *really* want the new liberasurecode to avoid bad things. the dependency version bump is IMO our way of saying "do it now or else" | 06:40 |
| notmyname | zaitcev: but if you've got a better way, especially from the perspective of the distro packager, on what to do, I'm all ears | 06:41 |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/liberasurecode: Prepare version bump to 1.3.1 https://review.openstack.org/395966 | 06:41 |
| kota_ | notmyname: addressed to 1.3.1 for the header change. | 06:42 |
| *** jerrygb has quit IRC | 06:42 | |
| kota_ | do we need changlog? i don't think for the qucik release... | 06:42 |
| openstackgerrit | John Dickinson proposed openstack/liberasurecode: Version bump to 1.3.1 https://review.openstack.org/395966 | 06:44 |
| notmyname | kota_: yeah. I'd like to push over it, though | 06:44 |
| notmyname | yeah, that | 06:44 |
| notmyname | kota_: I pushed over with the changelog. look ok? | 06:45 |
| notmyname | kota_: if so, land it and I'll tag it | 06:45 |
| kota_ | notmyname: looks good | 06:45 |
| notmyname | kota_: ok. do the honors (+2/+A) | 06:46 |
| kota_ | notmyname: thanks for working on your late night. | 06:50 |
| notmyname | kota_: click the button, and i'll be able to go to bed ;-) | 06:50 |
| notmyname | I'm glad you found this and we'll be able to get it finished quickly :-) | 06:51 |
| *** pcaruana has joined #openstack-swift | 06:51 | |
| kota_ | done | 06:51 |
| notmyname | thanks. should merge in ~5 minutes | 06:52 |
| openstackgerrit | Merged openstack/liberasurecode: Version bump to 1.3.1 https://review.openstack.org/395966 | 06:55 |
| notmyname | done. 1.3.1 tag is upstream now | 06:56 |
| *** rcernin has quit IRC | 06:56 | |
| notmyname | kota_: thank you very much for your work on getting the EC issue fixed | 06:56 |
| *** rcernin has joined #openstack-swift | 06:56 | |
| kota_ | notmyname: you're welcom :-) and not finished yet ;-) | 06:57 |
| notmyname | you're doing good work | 06:58 |
| notmyname | but i'm finished for today. good night :-) | 06:58 |
| *** chlong has joined #openstack-swift | 06:58 | |
| kota_ | notmyname: have a good sleep | 06:58 |
| *** dmorita has quit IRC | 06:59 | |
| *** dmorita has joined #openstack-swift | 07:00 | |
| *** sams-gleb has joined #openstack-swift | 07:02 | |
| *** sams-gleb has quit IRC | 07:02 | |
| *** sams-gleb has joined #openstack-swift | 07:02 | |
| *** dmorita has quit IRC | 07:04 | |
| *** dmorita has joined #openstack-swift | 07:04 | |
| *** chlong has quit IRC | 07:09 | |
| *** dmorita has quit IRC | 07:18 | |
| *** tesseract has joined #openstack-swift | 07:20 | |
| *** tesseract is now known as Guest40216 | 07:20 | |
| *** dmorita has joined #openstack-swift | 07:21 | |
| *** oshritf has joined #openstack-swift | 07:23 | |
| *** dmorita has quit IRC | 07:26 | |
| *** oshritf has quit IRC | 07:26 | |
| *** oshritf has joined #openstack-swift | 07:27 | |
| *** oshritf has quit IRC | 07:30 | |
| *** sams-gleb has quit IRC | 07:35 | |
| *** sams-gleb has joined #openstack-swift | 07:35 | |
| *** oshritf has joined #openstack-swift | 07:36 | |
| *** sams-gleb has quit IRC | 07:40 | |
| *** asettle has joined #openstack-swift | 07:47 | |
| *** openstackgerrit has quit IRC | 07:48 | |
| *** openstackgerrit has joined #openstack-swift | 07:48 | |
| *** oshritf has quit IRC | 07:49 | |
| *** oshritf has joined #openstack-swift | 07:51 | |
| *** asettle has quit IRC | 07:52 | |
| *** sams-gleb has joined #openstack-swift | 07:52 | |
| *** SkyRocknRoll has quit IRC | 07:52 | |
| *** sams-gleb has quit IRC | 07:54 | |
| *** cbartz has joined #openstack-swift | 07:58 | |
| *** oshritf has quit IRC | 07:59 | |
| *** SkyRocknRoll has joined #openstack-swift | 08:04 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/pyeclib: Make hard-coded dependency for liberasurecode>=1.3.1 https://review.openstack.org/395998 | 08:06 |
| kota_ | hmm... I'm realizing patch 395998 may be bad way. | 08:06 |
| patchbot | https://review.openstack.org/#/c/395998/ - pyeclib - Make hard-coded dependency for liberasurecode>=1.3.1 | 08:06 |
| kota_ | because it seems too painful to manage. | 08:07 |
| kota_ | however, right now we need to build liberasurecode from source to use the newest. | 08:07 |
| kota_ | to track the master or github tags, we need a few of hard-coded things but it cannot be detected in the packager maintainance which is used in gerrit gate vm's | 08:09 |
| kota_ | from source and patches (fast?) vs waiting packagers (slow? not slow? not sure) | 08:10 |
| *** oshritf has joined #openstack-swift | 08:11 | |
| kota_ | and the last item is whether we need to consider user overwritten binary built from source over the packager's :/ | 08:11 |
| kota_ | too complex dependency management :'( | 08:12 |
| *** tsg has quit IRC | 08:12 | |
| *** tsg has joined #openstack-swift | 08:12 | |
| *** sams-gleb has joined #openstack-swift | 08:13 | |
| *** rledisez has joined #openstack-swift | 08:13 | |
| *** oshritf has quit IRC | 08:14 | |
| *** oshritf has joined #openstack-swift | 08:17 | |
| *** oshritf has quit IRC | 08:21 | |
| *** amoralej|off is now known as amoralej | 08:26 | |
| *** acoles_ is now known as acoles | 08:35 | |
| *** oshritf has joined #openstack-swift | 08:38 | |
| *** jerrygb has joined #openstack-swift | 08:38 | |
| *** geaaru has joined #openstack-swift | 08:39 | |
| acoles | hi kota | 08:43 |
| kota_ | hi acoles | 08:43 |
| *** jerrygb has quit IRC | 08:44 | |
| acoles | kota_: looks like you're doing great work | 08:44 |
| *** jordanP has joined #openstack-swift | 08:44 | |
| kota_ | acoles: thank you | 08:44 |
| kota_ | acoles: but sorry i couldn't have enough time to think of admin6's auditor renamer problem. | 08:45 |
| acoles | notmyname: clayg cschwede (I just glanced over scrollback...) we don't have any gating test job using EC policy AFAIK. We have the community CI job using EC. If we could add an EC policy to in process tests then it would be easy to drop in another in process gate jon configured to use EC, jrichli put some framework in to config in process tests for the encryption job. | 08:46 |
| acoles | kota_: NP I could not find any explanation in code. maybe was two auditor processes?? | 08:48 |
| kota_ | acoles: maybe?? | 08:50 |
| kota_ | acoles: and sounds great to add EC policy testing to the gate anyway. | 08:50 |
| acoles | ntata: were you looking into oslo.config for swift? I have been discussing some ideas with lxsli who knows a bit about oslo.config, see https://etherpad.openstack.org/p/swift_oslo-config | 08:51 |
| *** geaaru has quit IRC | 09:06 | |
| *** oshritf has quit IRC | 09:15 | |
| *** geaaru has joined #openstack-swift | 09:20 | |
| *** oshritf has joined #openstack-swift | 09:20 | |
| *** oshritf has quit IRC | 09:26 | |
| *** asettle has joined #openstack-swift | 09:37 | |
| *** geaaru has quit IRC | 09:41 | |
| *** pcaruana has quit IRC | 09:49 | |
| *** geaaru has joined #openstack-swift | 09:55 | |
| *** pcaruana has joined #openstack-swift | 10:01 | |
| *** oshritf has joined #openstack-swift | 10:06 | |
| *** oshritf has quit IRC | 10:06 | |
| *** StraubTW has joined #openstack-swift | 10:09 | |
| *** StraubTW has quit IRC | 10:15 | |
| *** dikonoor has quit IRC | 10:16 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: WIP read conf options from alternate ini file https://review.openstack.org/396089 | 10:40 |
| *** mvk has quit IRC | 10:40 | |
| *** jerrygb has joined #openstack-swift | 10:40 | |
| *** jerrygb has quit IRC | 10:45 | |
| mahatic | acoles: ^ nice start! | 10:45 |
| *** kei_yama has quit IRC | 10:46 | |
| acoles | mahatic: hi. fyi https://etherpad.openstack.org/p/swift_oslo-config | 10:46 |
| * acoles afk for a while | 10:46 | |
| mahatic | acoles: thanks | 10:47 |
| *** dikonoor has joined #openstack-swift | 11:19 | |
| *** smatzek has joined #openstack-swift | 11:23 | |
| *** mvk has joined #openstack-swift | 11:24 | |
| *** jerrygb has joined #openstack-swift | 11:25 | |
| *** SkyRocknRoll has quit IRC | 11:38 | |
| *** qwertyco has joined #openstack-swift | 11:42 | |
| *** qwertyco has quit IRC | 11:42 | |
| *** qwertyco has joined #openstack-swift | 11:42 | |
| *** chlong has joined #openstack-swift | 11:43 | |
| *** chlong has quit IRC | 11:44 | |
| *** dikonoor has quit IRC | 11:49 | |
| *** _JZ_ has joined #openstack-swift | 12:01 | |
| *** dikonoor has joined #openstack-swift | 12:05 | |
| *** vint_bra has joined #openstack-swift | 12:14 | |
| *** qwertyco has quit IRC | 12:22 | |
| *** bkopilov has quit IRC | 12:31 | |
| *** Guess456787654 has joined #openstack-swift | 12:35 | |
| *** vint_bra has quit IRC | 12:51 | |
| *** takashi has quit IRC | 13:01 | |
| *** catintheroof has joined #openstack-swift | 13:08 | |
| *** amoralej is now known as amoralej|lunch | 13:30 | |
| *** jerrygb has quit IRC | 13:35 | |
| *** jerrygb has joined #openstack-swift | 13:36 | |
| *** jerrygb has quit IRC | 13:40 | |
| *** vint_bra has joined #openstack-swift | 13:42 | |
| *** jerrygb has joined #openstack-swift | 13:47 | |
| *** silor has joined #openstack-swift | 13:54 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/pyeclib: Make hard-coded dependency for liberasurecode>=1.3.1 https://review.openstack.org/395998 | 13:56 |
| openstackgerrit | Alexis Lee proposed openstack/swift: WIP PoC to add oslo-config-generator to Swift https://review.openstack.org/396208 | 13:56 |
| kota_ | hmmm... it looks like we are back to those days how to install liberasurecode from source in the gate. But it may be worse way? | 14:00 |
| *** silor1 has joined #openstack-swift | 14:01 | |
| *** silor has quit IRC | 14:03 | |
| *** silor1 is now known as silor | 14:03 | |
| *** amoralej|lunch is now known as amoralej | 14:05 | |
| *** StraubTW has joined #openstack-swift | 14:09 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/pyeclib: Make hard-coded dependency for liberasurecode>=1.3.1 https://review.openstack.org/395998 | 14:11 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Refactor container sync stats reporting https://review.openstack.org/395094 | 14:12 |
| acoles | jrichli: ^^ thx for review, also I found some more places where self.cs could be used to remove some lines | 14:13 |
| *** jerrygb has quit IRC | 14:19 | |
| *** dmorita has joined #openstack-swift | 14:19 | |
| *** sams-gleb has quit IRC | 14:23 | |
| *** sams-gleb has joined #openstack-swift | 14:23 | |
| *** smatzek has quit IRC | 14:23 | |
| *** dmorita has quit IRC | 14:24 | |
| *** klamath has joined #openstack-swift | 14:27 | |
| *** sams-gleb has quit IRC | 14:28 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:29 | |
| *** links has quit IRC | 14:31 | |
| *** sams-gleb has joined #openstack-swift | 14:43 | |
| *** dikonoor has quit IRC | 14:49 | |
| *** tsg has quit IRC | 14:57 | |
| *** gabor_antal has joined #openstack-swift | 14:58 | |
| *** acoles is now known as acoles_ | 15:01 | |
| *** mvk has quit IRC | 15:01 | |
| *** diogogmt has quit IRC | 15:07 | |
| *** smatzek has joined #openstack-swift | 15:11 | |
| *** mvk has joined #openstack-swift | 15:14 | |
| *** SkyRocknRoll has quit IRC | 15:27 | |
| *** tsg has joined #openstack-swift | 15:29 | |
| *** jerrygb has joined #openstack-swift | 15:31 | |
| *** jerrygb has quit IRC | 15:31 | |
| notmyname | acoles_: lxsli: very interesting on the config stuff. I'll look at it more later today | 15:31 |
| *** jerrygb has joined #openstack-swift | 15:32 | |
| *** jerrygb has quit IRC | 15:39 | |
| openstackgerrit | Alexis Lee proposed openstack/swift: WIP PoC to add oslo-config-generator to Swift https://review.openstack.org/396208 | 15:42 |
| openstackgerrit | Alexis Lee proposed openstack/swift: WIP PoC to add oslo-config-generator to Swift https://review.openstack.org/396208 | 15:43 |
| *** vint_bra1 has joined #openstack-swift | 15:46 | |
| *** vint_bra has quit IRC | 15:46 | |
| *** SkyRocknRoll has joined #openstack-swift | 15:56 | |
| *** Guest40216 has quit IRC | 15:56 | |
| *** diogogmt has joined #openstack-swift | 15:58 | |
| *** klrmn has joined #openstack-swift | 16:00 | |
| *** stradling has joined #openstack-swift | 16:08 | |
| *** links has joined #openstack-swift | 16:09 | |
| *** sgundur_ has joined #openstack-swift | 16:09 | |
| *** rcernin has quit IRC | 16:11 | |
| *** sgundur_ has quit IRC | 16:16 | |
| *** acoles_ is now known as acoles | 16:17 | |
| *** stradling has quit IRC | 16:21 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Add process level concurrency to container sync https://review.openstack.org/210099 | 16:22 |
| acoles | jrichli: rebased onto the stats refactor ^^ but no other changes | 16:24 |
| jrichli | acoles: great! I'll check it out. thx! | 16:25 |
| *** silor has quit IRC | 16:27 | |
| *** acoles is now known as acoles_ | 16:31 | |
| *** rcernin has joined #openstack-swift | 16:32 | |
| *** mvk has quit IRC | 16:32 | |
| *** eranrom has quit IRC | 16:33 | |
| *** links has quit IRC | 16:37 | |
| *** StraubTW_ has joined #openstack-swift | 16:44 | |
| *** catinthe_ has joined #openstack-swift | 16:45 | |
| *** dikonoor has joined #openstack-swift | 16:45 | |
| *** catintheroof has quit IRC | 16:46 | |
| *** StraubTW has quit IRC | 16:47 | |
| *** tongli has joined #openstack-swift | 16:51 | |
| notmyname | good morning | 16:53 |
| notmyname | the ML thread "Embracing new languages in OpenStack" is getting interesting. I really like a lot of what cdent has been advocating for. | 16:53 |
| *** tsg_ has joined #openstack-swift | 16:59 | |
| *** tsg has quit IRC | 17:00 | |
| notmyname | timur: I think it would be great if you could add some links to your current work on https://bugs.launchpad.net/swift/+bug/1319097 | 17:05 |
| openstack | Launchpad bug 1319097 in OpenStack Object Storage (swift) "Filter results when querying a container" [Wishlist,New] - Assigned to Christian Hugo (christianhugo) | 17:05 |
| *** cbartz has left #openstack-swift | 17:12 | |
| *** asettle has quit IRC | 17:16 | |
| *** rledisez has quit IRC | 17:19 | |
| *** jordanP has quit IRC | 17:21 | |
| *** david_c_ has joined #openstack-swift | 17:25 | |
| *** tsg__ has joined #openstack-swift | 17:29 | |
| *** tsg_ has quit IRC | 17:30 | |
| *** _JZ_ has quit IRC | 17:32 | |
| *** tsg__ has quit IRC | 17:35 | |
| clayg | liberasure 1.3.*1* - what happened!? | 17:40 |
| *** dikonoor has quit IRC | 17:41 | |
| notmyname | clayg: we forgot to bump the #defines in the code to say 1.3.0 | 17:41 |
| *** acoles_ is now known as acoles | 17:42 | |
| *** _JZ_ has joined #openstack-swift | 17:43 | |
| clayg | yup - scrollback explains all | 17:43 |
| clayg | acoles: no gate ec jobs!? | 17:44 |
| acoles | clayg: not afaict | 17:44 |
| acoles | just the community cluster job | 17:44 |
| *** tsg__ has joined #openstack-swift | 17:45 | |
| *** _JZ_ has quit IRC | 17:46 | |
| acoles | notmyname: lxsli and I discussed a bit the single conf file vs multiple swift conf files issue, only really as far as identifying it as an issue | 17:47 |
| notmyname | acoles: ok | 17:47 |
| *** dmorita has joined #openstack-swift | 17:48 | |
| acoles | having more than one instance of the same service on the same node is perhaps quite different from the norm e.g. multiple object server instances with differing configs | 17:48 |
| acoles | as well as different conf files for different service types | 17:49 |
| *** tsg__ has left #openstack-swift | 17:50 | |
| acoles | re. auto generated conf files... my hope is that one "win" would be reducing the maintenance of conf options (some of them are documented in four places, with a good chance that >0 are therefore not up to date!) | 17:50 |
| acoles | but like you I like having a sample conf in the repo tree | 17:50 |
| acoles | (just like a changelog...) | 17:50 |
| acoles | notmyname: ^^ | 17:51 |
| notmyname | yeah, I get the concern about having more than one and getting out of date | 17:51 |
| notmyname | I think it's totally reasonable to be able to generate docs for defined config options. but I'd kinda like to have static sample configs too | 17:52 |
| acoles | yup | 17:52 |
| *** sgundur_ has joined #openstack-swift | 17:53 | |
| acoles | it's also another step to do in a good review - run the doc generator, compile html maybe and check all looks good. | 17:53 |
| acoles | question is if we had something like this, could it replace the tables in the deployment guide?? http://docs.openstack.org/developer/nova/sample_config.html | 17:54 |
| notmyname | hmm...inline of a config file doesn't seem an obvious better choice than http://docs.openstack.org/developer/swift/deployment_guide.html#object-server-configuration | 17:56 |
| *** sgundur_ has quit IRC | 17:56 | |
| *** sgundur_ has joined #openstack-swift | 17:57 | |
| acoles | agree, I'm hoping that oslo config might generate something prettier than a sample conf file, i.e. tabulated, but IDK | 17:57 |
| notmyname | acoles: did you learn anything about the bootstrapping problem? | 18:00 |
| zaitcev | notmyname: Interesting thread indeed. | 18:00 |
| *** SkyRocknRoll has quit IRC | 18:01 | |
| *** LZ has joined #openstack-swift | 18:01 | |
| *** LZ has quit IRC | 18:02 | |
| acoles | notmyname: yes, spent yesterday afternoon discussing it with lxsli, latest thoughts are on the etherpad. So, the key realisation is that wsgi loader is the bootstrapper - we already have it! the wsgi loader already passes each middleware a dict of opts which the middleware can use to create an oslo configopts instance with its option schema checking magic. And we could (if wanted) teach the wsgi loader to read opts from a | 18:07 |
| acoles | regular ini file (like all other oslo configured projects) and merge those with the dict it passes to each middleware. See patch 396089. | 18:07 |
| patchbot | https://review.openstack.org/#/c/396089/ - swift - WIP read conf options from alternate ini file | 18:07 |
| timburke | 88 page-downs. that's a big config file! | 18:07 |
| acoles | timburke: rotate your monitor 90 degrees, you may get to down to ~50 ;) | 18:08 |
| timburke | i'm not rotating my 4k. i think that may require duct tape | 18:09 |
| acoles | notmyname: so IF you want an ini conf that is separate from your paste ini, then we know how do do that and be backwards compatible. | 18:09 |
| acoles | timburke: now I'm jealous | 18:09 |
| *** Renich has joined #openstack-swift | 18:28 | |
| *** jamielennox|away is now known as jamielennox | 18:37 | |
| *** ChubYann has joined #openstack-swift | 18:53 | |
| zaitcev | so we have liberasurecode 1.3.1 and not picked upt https://review.openstack.org/#/c/386890/ ;_; | 18:55 |
| patchbot | patch 386890 - liberasurecode - Beautify fill_buffer() | 18:55 |
| *** sgundur_ has quit IRC | 18:56 | |
| *** amoralej is now known as amoralej|off | 18:58 | |
| *** geaaru has quit IRC | 19:01 | |
| timburke | oh, hey! i have commit bits there now! | 19:02 |
| notmyname | acoles: intesting | 19:03 |
| *** asettle has joined #openstack-swift | 19:07 | |
| openstackgerrit | Merged openstack/liberasurecode: Beautify fill_buffer() https://review.openstack.org/386890 | 19:07 |
| *** asettle has quit IRC | 19:09 | |
| *** asettle has joined #openstack-swift | 19:09 | |
| *** asettle has quit IRC | 19:14 | |
| notmyname | tdasilva: http://lists.openstack.org/pipermail/openstack-dev/2016-November/107133.html | 19:37 |
| notmyname | if my assumption is correct, I'm concerned that it's only the stuff in the left column that is counted towards "cross-project" | 19:38 |
| notmyname | well, that's the case anyway (cross-project vs inter-project) | 19:38 |
| clayg | timburke: zaitcev: what's the biggest size_t we ever use? are we counting on integer overflow instead of explicit modulo? | 19:39 |
| notmyname | but that "cross-project" proposals would have to be gated by a small group | 19:39 |
| timburke | clayg: yeah, stuffing a too-large number into a char does the '% 256' for us | 19:40 |
| zaitcev | clayg: Since it's unsigned, overflows are probably harmless, as long as you expect them to happen... E.g. know they can wrap your address space. | 19:40 |
| zaitcev | oh, the fill_buff | 19:40 |
| clayg | timburke: oh kay - I guess I sort of thought it was bad form to "ignore/use" overflow like (and it's *much* bigger than 256 right?) that w/o sort of owning it (a comment or something) - i guess I just need to get used to modern c | 19:42 |
| zaitcev | That one relies on truncation, yes. But I have to admit, I didn't want to get too ashamed and I created a test program that ran old and new code side by side and compared that fill remains the same. Even though it does not really matter how you fill, the fill values aren't used for anything. | 19:42 |
| zaitcev | Note that it's a kind of fibonacchi fill, not sequential. No idea why, maybe a bug in the original :-) | 19:43 |
| *** jerrygb has joined #openstack-swift | 19:43 | |
| zaitcev | I have this book https://www.amazon.com/Hackers-Delight-Henry-Warren-Jr/dp/0201914654/ | 19:44 |
| *** openstackgerrit has quit IRC | 19:48 | |
| *** openstackgerrit has joined #openstack-swift | 19:49 | |
| tdasilva | notmyname: thanks for the email, hopefully there will be some clarification | 19:57 |
| *** cppforlife_ has quit IRC | 20:00 | |
| *** cppforlife_ has joined #openstack-swift | 20:03 | |
| *** mgagne has joined #openstack-swift | 20:07 | |
| clayg | zaitcev: the original never had values > 256 - the new will go up to unsigned int - unless we never created like... idk 128K buffers? | 20:08 |
| zaitcev | yeah but it's properly clipped | 20:08 |
| zaitcev | by assignment | 20:08 |
| *** sgundur_ has joined #openstack-swift | 20:09 | |
| clayg | oh wow, unsigned ints or sorta big on modern hadware :P more like > 4GiB ;P | 20:10 |
| zaitcev | This will work on anything, even sparc | 20:11 |
| clayg | i'm sure it will "work" - you know - i guess I just don't know what the buffer is used for :P | 20:11 |
| clayg | the % 256 seemed to hint maybe thought something like ascii/char/byte buffer or something | 20:12 |
| zaitcev | It's just garbage, to make sure spurious zero-fill or spurious use are caught | 20:12 |
| zaitcev | The best is to fill it with some pseudo-random noise, maybe plus-and-mult. | 20:13 |
| mgagne | is there a way to make swift-init work for humans (not upstart or init scripts) if swift binaries aren't in the default PATH (without having to update PATH in .bashrc) ? | 20:13 |
| clayg | oh, but there's the whole fibonacci thing - i really don't know what size_t will overflow! fascinating | 20:13 |
| clayg | mgagne: not really - it uses subprocess - which uses os.execvp | 20:16 |
| mgagne | ok, so there is no known way to override PATH or whatever in swift.conf | 20:16 |
| clayg | mgagne: not currently - but it seems like there is possibly a reasonable use-case there | 20:17 |
| mgagne | right, will figure out a way to do it for now, maybe updating PATH in /etc/environment | 20:18 |
| clayg | mgagne: some of the work I'm doing to support the transistion to hummingbird will have swift-init finding /etc/swift/<service>.conf[.d] and parsing that before forking the implementation script/binary | 20:18 |
| mgagne | =) | 20:18 |
| clayg | it's possible that + ntata's work to make the /etc/swift root configurable via some kind of env could lead to swift_dir = and swift_bin = sorta of config options | 20:18 |
| clayg | or idk, something else | 20:19 |
| clayg | i don't acctually need to do a swift_bin kind of thing - i'm just choosing which binary to load based on the config option - my point was there may at some point be infrastructure that would make plubming for that trivial assuming there was use-case for a swift_dir | 20:19 |
| clayg | what's so bad about updating PATH ;) | 20:19 |
| *** gyee has joined #openstack-swift | 20:27 | |
| openstackgerrit | Pete Zaitcev proposed openstack/swift: Add dist-collect and dist: target https://review.openstack.org/396394 | 20:32 |
| zaitcev | I update PYTHONPATH | 20:33 |
| zaitcev | sometimes | 20:33 |
| mgagne | clayg: someone has to manage the config =) | 20:34 |
| *** Guess456787654 has quit IRC | 20:34 | |
| *** rcernin has quit IRC | 20:36 | |
| *** acoles is now known as acoles_ | 20:36 | |
| clayg | zaitcev: how does that help find the binaries? | 20:36 |
| mgagne | will create symlinks in usr/bin for convenience for now | 20:36 |
| clayg | mgagne: yeah sure or that - but if you have root anyways - why not just install there? | 20:37 |
| mgagne | building custom packages with dh-virtualenv | 20:37 |
| zaitcev | clayg: it collects the source, which I need to build the binaries, because Fedora/RDO/RHOS buildroots/containers do not have networking and thus cannot use git. And we want signed tarballs from a trusted source anyway. | 20:37 |
| mgagne | initial goal was to isolate the install in /opt/openstack/swift so I can stage an update and flip a symlink | 20:38 |
| zaitcev | clayg: The etherpad says "current thinking is that nobody expects tarballs of go source code". Precisely because of that thinking, I had to set about making my own tarballs. | 20:38 |
| zaitcev | mgagne: That is very interesting! I never did that, just upgraded node by node. Stop services, "dnf update", swift-init main start, swift-init rest start, go to next node. Nobody ever notices there was an upgrade going on. | 20:41 |
| clayg | mgagne: oic - interesting | 20:41 |
| mgagne | it's not an original idea, rackspace is/was using this method years ago | 20:41 |
| *** sams-gleb has quit IRC | 20:46 | |
| *** sams-gleb has joined #openstack-swift | 20:47 | |
| *** sams-gleb has quit IRC | 20:48 | |
| *** sgundur_ has quit IRC | 20:53 | |
| *** gyee has quit IRC | 20:54 | |
| *** clu_ has joined #openstack-swift | 20:54 | |
| openstackgerrit | Merged openstack/python-swiftclient: 3.2.0 release notes https://review.openstack.org/395768 | 21:20 |
| *** smatzek has quit IRC | 21:21 | |
| *** gyee has joined #openstack-swift | 21:32 | |
| *** Jeffrey4l has quit IRC | 21:33 | |
| openstackgerrit | Tim Burke proposed openstack/swift: adding reno sphinx tree https://review.openstack.org/381382 | 21:34 |
| *** sgundur_ has joined #openstack-swift | 21:34 | |
| *** vint_bra1 has left #openstack-swift | 21:37 | |
| *** sgundur__ has joined #openstack-swift | 21:46 | |
| *** sgundur_ has quit IRC | 21:48 | |
| clayg | zaitcev: so is the problem you're trying to solve with dist-collect is that python setup.py sdist isn't puttig the golang code in the source tarball? | 21:48 |
| zaitcev | clayg: It does, and in fact I use that currently. But I need the dependencies too. | 21:49 |
| clayg | zaitcev: I don't understand who is saying that nobody (developers? packagers? deployers?) expects tarballs of go source code | 21:49 |
| zaitcev | Because I can't run go get when building | 21:49 |
| clayg | i thought the expectation was that you would buildrequres the golang-XXX-devel; which would install into GOPATH and then the `go get` would not be required? | 21:50 |
| clayg | https://fedoraproject.org/wiki/PackagingDrafts/Go#Packaging_a_binary <- sorta looking at this | 21:50 |
| clayg | or http://pkg-go.alioth.debian.org/packaging.html | 21:51 |
| clayg | depending on your posion | 21:51 |
| zaitcev | I need to experiment with that BuildRequires: golang(github.com/owner/project) one day. | 21:52 |
| zaitcev | Hopefuly then we can abandon that dist-collect thing. | 21:58 |
| *** catinthe_ has quit IRC | 21:59 | |
| notmyname | FYI swiftclient release request https://review.openstack.org/#/c/396418/ | 21:59 |
| patchbot | patch 396418 - releases - 3.2.0 swiftclient release | 21:59 |
| *** smatzek has joined #openstack-swift | 22:01 | |
| *** smatzek has quit IRC | 22:02 | |
| clayg | zaitcev: yes, let's hope! | 22:02 |
| mattoliverau | morning | 22:03 |
| timburke | mattoliverau: feeling better? | 22:14 |
| *** tongli has quit IRC | 22:17 | |
| *** jerrygb_ has joined #openstack-swift | 22:21 | |
| *** jerrygb has quit IRC | 22:24 | |
| *** jerrygb has joined #openstack-swift | 22:25 | |
| *** jerrygb_ has quit IRC | 22:26 | |
| timur | notmyname: good idea. Once I wrap up the current work items at work, I'll submit a review to pull in what SwiftStack is doing for metadata search into Swift, to at least get some discussion going around it | 22:26 |
| notmyname | timur: thanks | 22:26 |
| *** jerrygb_ has joined #openstack-swift | 22:27 | |
| openstackgerrit | Drew Balfour proposed openstack/swift: add byteorder information and logic to ring files https://review.openstack.org/395237 | 22:28 |
| *** jerrygb has quit IRC | 22:30 | |
| *** jerrygb_ has quit IRC | 22:31 | |
| notmyname | abalfour: heh. for some reason your comment on that patch made me laugh. "got way more complicated that originally thought" :-) | 22:32 |
| *** Renich has quit IRC | 22:33 | |
| abalfour | yeah... once I started doing things like this: | 22:33 |
| abalfour | fmt = '>H' if sys.byteorder == 'little' else '<H' | 22:33 |
| abalfour | dev_len = len(ring_dict['devs']) | 22:33 |
| abalfour | byteswapped_dev_len = struct.unpack(fmt, struct.pack('=H', dev_len-1))[0] | 22:33 |
| abalfour | I figured I should step back from the edge. | 22:34 |
| clayg | whoa | 22:34 |
| *** mvk has joined #openstack-swift | 22:34 | |
| abalfour | byteswap the length of the array and check to see if that value appears in the serialized version, and make sure nothing bigger does, and ... yeah. | 22:35 |
| *** mmotiani2 has joined #openstack-swift | 22:38 | |
| *** mmotiani2 has quit IRC | 22:39 | |
| *** abalfour has quit IRC | 22:41 | |
| openstackgerrit | Merged openstack/swift: adding reno sphinx tree https://review.openstack.org/381382 | 23:00 |
| *** catintheroof has joined #openstack-swift | 23:05 | |
| *** sgundur__ has quit IRC | 23:06 | |
| *** catintheroof has quit IRC | 23:09 | |
| *** catintheroof has joined #openstack-swift | 23:17 | |
| *** klamath has quit IRC | 23:27 | |
| mattoliverau | timburke: feeling much better. Having my first coffee in a few days.. and its amazing :) | 23:29 |
| *** StraubTW_ has quit IRC | 23:31 | |
| *** Jeffrey4l has joined #openstack-swift | 23:31 | |
| *** Guess456787654 has joined #openstack-swift | 23:33 | |
| *** diogogmt has quit IRC | 23:33 | |
| *** kei_yama has joined #openstack-swift | 23:34 | |
| *** diogogmt has joined #openstack-swift | 23:43 | |
| *** Guess456787654 has quit IRC | 23:48 | |
| timburke | apply last [shuffle WTF!? "\"Sorry, what was that?\"" "\"Re-write me in Go.\""] | 23:55 |
| patchbot | Sorry, what was that? | 23:55 |
| notmyname | Do not taunt happy fun patchbot | 23:58 |
| *** Guess456787654 has joined #openstack-swift | 23:58 | |
| timburke | but it's so fun to do! | 23:59 |
| *** Guess456787654 has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!