| *** chlong has quit IRC | 00:01 | |
| *** dfflanders has joined #openstack-swift | 00:02 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Move documented reclaim_age option to correct location https://review.openstack.org/374419 | 00:09 |
|---|---|---|
| openstackgerrit | Clay Gerrard proposed openstack/swift: Test current reclaim_age handling https://review.openstack.org/419696 | 00:09 |
| clayg | kota_: it's *never* quick is it :'( | 00:09 |
| openstackgerrit | Merged openstack/swift: Support last modified on listing containers https://review.openstack.org/198634 | 00:12 |
| clayg | mahatic: acoles_: well shoot - was https://review.openstack.org/#/c/410311 already attempting to document current behavior? looks like it went a totally different route than https://review.openstack.org/419696 - and was posted as a follow-up the to the [DEFAULT] reclaim_age fix instead of as a dependency :\ | 00:15 |
| patchbot | patch 410311 - swift - Add test for configured reclaim_age being used | 00:15 |
| patchbot | patch 419696 - swift - Test current reclaim_age handling | 00:15 |
| *** furlongm has quit IRC | 00:34 | |
| *** furlongm has joined #openstack-swift | 00:36 | |
| *** furlongm has quit IRC | 00:38 | |
| *** klrmn has quit IRC | 00:42 | |
| *** klrmn has joined #openstack-swift | 00:43 | |
| *** tqtran has joined #openstack-swift | 00:49 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Move documented reclaim_age option to correct location https://review.openstack.org/374419 | 00:56 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Test current reclaim_age handling https://review.openstack.org/410311 | 00:56 |
| *** furlongm has joined #openstack-swift | 00:56 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Fix download resumption for new SLOs. https://review.openstack.org/419664 | 01:13 |
| kota_ | morning | 01:19 |
| kota_ | clayg: don't mind! Ah, thanks for addressing that reclaim_age patch. | 01:21 |
| kota_ | clayg: i didn't notice that has not landed yet. | 01:21 |
| kota_ | acoles_, timburke, zaitcev: thanks for reviewing, merging https://review.openstack.org/#/c/198634/ that is one of the oldest patches in my list. | 01:23 |
| patchbot | patch 198634 - swift - Support last modified on listing containers (MERGED) | 01:23 |
| clayg | kota_: pay it forward ? pach 419579 | 01:26 |
| kota_ | patch 419579 | 01:28 |
| patchbot | https://review.openstack.org/#/c/419579/ - swift - Trivial follow up to addition of last modified in ... | 01:28 |
| kota_ | thank you patchbot | 01:28 |
| kota_ | okay, it's follo up to me from acoles_. I'll take a look, at least in today's work. | 01:29 |
| kota_ | hmm | 01:29 |
| kota_ | looks like very small. do that now. | 01:29 |
| zaitcev | too late | 01:43 |
| kota_ | zaitcev: exactly | 01:43 |
| kota_ | and i found one more thing I'd address my mistake in follow-up :P | 01:43 |
| kota_ | will push soon | 01:44 |
| zaitcev | oh dear | 01:45 |
| zaitcev | What did I miss this time? | 01:45 |
| zaitcev | brb dojo | 01:46 |
| kota_ | 1 sec | 01:46 |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: Make timestamp setting to be consistent in test loop https://review.openstack.org/419709 | 01:46 |
| kota_ | that one | 01:47 |
| kota_ | zaitcev: ^^ | 01:47 |
| *** sams-gleb has joined #openstack-swift | 01:53 | |
| *** sams-gleb has quit IRC | 01:57 | |
| zaitcev | my session is canceled | 02:01 |
| kota_ | zaitcev: thanks for your quick review ;-) | 02:06 |
| *** vint_bra has joined #openstack-swift | 02:14 | |
| *** vint_bra has quit IRC | 02:14 | |
| *** manij has joined #openstack-swift | 02:18 | |
| *** manij has quit IRC | 02:22 | |
| *** winggundamth has joined #openstack-swift | 02:22 | |
| *** dmorita has quit IRC | 02:25 | |
| *** Jeffrey4l_ has joined #openstack-swift | 02:30 | |
| *** Jeffrey4l has quit IRC | 02:34 | |
| *** JimCheung has quit IRC | 02:42 | |
| clayg | what is pavel's handle? | 02:51 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Test current reclaim_age handling https://review.openstack.org/410311 | 03:08 |
| *** tqtran has quit IRC | 03:09 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Move documented reclaim_age option to correct location https://review.openstack.org/374419 | 03:10 |
| *** Jeffrey4l_ has quit IRC | 03:19 | |
| *** Jeffrey4l_ has joined #openstack-swift | 03:21 | |
| *** sams-gleb has joined #openstack-swift | 03:25 | |
| *** sams-gleb has quit IRC | 03:26 | |
| jrichli | clayg: I think it is PavelK | 03:39 |
| *** dfflanders has quit IRC | 03:40 | |
| *** links has joined #openstack-swift | 03:54 | |
| mahatic | good morning | 03:57 |
| mahatic | looks like lots to catch up on scroll back | 03:58 |
| mahatic | clayg: hi! if you're still around - do we intend to abandon either of these patches - patch 410311 or 419696? | 03:59 |
| patchbot | https://review.openstack.org/#/c/410311/ - swift - Test current reclaim_age handling | 03:59 |
| mahatic | ugh. patch 410311 or patch 419696 | 03:59 |
| patchbot | https://review.openstack.org/#/c/410311/ - swift - Test current reclaim_age handling | 03:59 |
| patchbot | https://review.openstack.org/#/c/419696/ - swift - Test current reclaim_age handling | 03:59 |
| clayg | yeah one of those has to go | 03:59 |
| mahatic | clayg: ok thanks, saw your update. thanks for your work! | 04:01 |
| * mahatic is yet to look over the actual changes (plus the scroll back discussion) | 04:02 | |
| *** tqtran has joined #openstack-swift | 04:09 | |
| *** tqtran has quit IRC | 04:11 | |
| *** _JZ_ has quit IRC | 04:13 | |
| *** psachin has joined #openstack-swift | 04:14 | |
| *** sams-gleb has joined #openstack-swift | 04:26 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:29 | |
| *** sams-gleb has quit IRC | 04:32 | |
| *** ppai has joined #openstack-swift | 04:36 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Extract test pattern to helper https://review.openstack.org/418689 | 04:43 |
| *** sanchitmalhotra has quit IRC | 04:55 | |
| *** sanchitmalhotra has joined #openstack-swift | 04:56 | |
| *** ppai has quit IRC | 04:58 | |
| *** ppai has joined #openstack-swift | 05:14 | |
| *** klrmn has quit IRC | 05:51 | |
| *** klrmn has joined #openstack-swift | 06:00 | |
| *** dmorita has joined #openstack-swift | 06:04 | |
| *** dmorita has quit IRC | 06:09 | |
| *** sams-gleb has joined #openstack-swift | 06:29 | |
| *** sams-gleb has quit IRC | 06:34 | |
| *** ppai has quit IRC | 06:37 | |
| *** ppai has joined #openstack-swift | 06:54 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Extract test pattern to helper https://review.openstack.org/418689 | 07:06 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Better optimistic lock in get_hashes https://review.openstack.org/419787 | 07:06 |
| *** tqtran has joined #openstack-swift | 07:11 | |
| *** tesseract has joined #openstack-swift | 07:11 | |
| *** tqtran has quit IRC | 07:13 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Optimize noop case for suffix rehash https://review.openstack.org/418692 | 07:26 |
| *** CrackerJackMack has quit IRC | 07:28 | |
| *** supamatt has quit IRC | 07:29 | |
| *** dfg_ has quit IRC | 07:30 | |
| *** rsFF has quit IRC | 07:30 | |
| *** philipw_ has quit IRC | 07:30 | |
| *** philipw has joined #openstack-swift | 07:31 | |
| *** CrackerJackMack has joined #openstack-swift | 07:35 | |
| *** dfg_ has joined #openstack-swift | 07:35 | |
| *** rsFF has joined #openstack-swift | 07:35 | |
| *** supamatt has joined #openstack-swift | 07:36 | |
| *** ChubYann has quit IRC | 07:39 | |
| openstackgerrit | Merged openstack/swift: Trivial follow up to addition of last modified in container listings https://review.openstack.org/419579 | 07:40 |
| *** CrackerJackMack has quit IRC | 07:57 | |
| *** oshritf has joined #openstack-swift | 07:59 | |
| *** fgemein has joined #openstack-swift | 07:59 | |
| fgemein | Hi out there | 08:01 |
| fgemein | I’m experiencing some problems after upgrading swift to liberty. | 08:03 |
| fgemein | precisely, ceilometer is not regognizing the storage.objects.outgoing.bytes for multipart objects anymore since the upgrade. | 08:03 |
| *** CrackerJackMack has joined #openstack-swift | 08:04 | |
| fgemein | Any hints would be highly apreciated | 08:04 |
| *** rledisez has joined #openstack-swift | 08:09 | |
| *** hseipp has joined #openstack-swift | 08:13 | |
| *** oshritf has quit IRC | 08:13 | |
| *** silor has joined #openstack-swift | 08:14 | |
| clayg | fgemein: I think most of the folks that hang around in here are using some form of log processing to calculate network transfer utilization reporting | 08:15 |
| clayg | you might check with the ceilometer team - i'm not sure why they capture network transfer exactly? | 08:15 |
| fgemein | well, they did untill kilo. | 08:16 |
| clayg | sorry not *why* - I mean *how* | 08:16 |
| clayg | typo | 08:16 |
| fgemein | I’m asking here, because it seems, that the wsgi writing for this packets has changed. | 08:16 |
| rledisez | fgemein: clayg: just arriving, so i may have missed the context, but i'm using ceilometer to bill network usage in swift | 08:17 |
| clayg | the... "packets" like tcp packets? I mean... the *LO manifests still download - the client still gets the bytes | 08:17 |
| fgemein | I’m still able to get all storage.objects.incoming.bytes, but for the outgoing I only see an initial PUT request once and no subsequent entries in mongo | 08:18 |
| clayg | rledisez: hoooray! do you have any idea how it works? I heard once they drop a message on a bus from middleware ... | 08:18 |
| clayg | is it snarfing values out of proxy logging? | 08:18 |
| rledisez | no, it's a middleware that count the bytes that pass in the wsgi iterator (both way), and then send a message on rabbit. then ceilometer-collector get the message from rabbit and store in mongo or gnocchi | 08:19 |
| fgemein | well ceilometer is inside the pipeline short before proxy-server | 08:19 |
| clayg | rledisez: seems like the counting bytes in the iterator part would be pretty stright forward - how much is there to go wrong really? | 08:21 |
| fgemein | There has been a bug in 2005 which led to the missing of resouce_id or project_id entries with storage.objects.outgoing.bytes. But the patch seems to be part of the liberts wsgi | 08:21 |
| *** silor1 has joined #openstack-swift | 08:21 | |
| clayg | how does it associate the stream with the user? :) | 08:21 |
| fgemein | by project_id or resource_id, which is the same with swift | 08:22 |
| rledisez | it associates the project id, not user id. it extract it from account, by removing the reseller prefix | 08:22 |
| clayg | yeah, but I mean is that like a special key in the environ? | 08:22 |
| rledisez | so, from the URL | 08:22 |
| clayg | could be subrequests don't copy in the magic keys | 08:22 |
| clayg | ... to the environ | 08:22 |
| fgemein | I did some searching inside the database (ceilo) , there aren’t any entries for the segment downloads, even if I search for path or user etc. | 08:23 |
| *** silor has quit IRC | 08:23 | |
| *** silor1 is now known as silor | 08:23 | |
| rledisez | that's why positioning is very important. you'll have to put it before slo/dlo/... | 08:23 |
| clayg | is "before" to the left or to the right ;) | 08:23 |
| clayg | like before in the request I guess? to the left? | 08:23 |
| rledisez | yes :) | 08:23 |
| clayg | after in the response | 08:23 |
| clayg | fgemein: you catch that? it's a good tip. if it works you owe rledisez a quarter. Thanks rledisez. | 08:24 |
| fgemein | i’m not shure, I have to change the ceilometer inside the oipeline? | 08:25 |
| clayg | yeah just move it left of the *lo wares | 08:25 |
| fgemein | I’ll try. THX a million | 08:25 |
| clayg | proxy emits some log message about "final pipeline order" or something - just make sure ceilomter is to the left | 08:25 |
| clayg | fgemein: looks like they got it figured out by mitaka - maybe have just been a doc bug - http://docs.openstack.org/mitaka/install-guide-rdo/ceilometer-swift.html | 08:28 |
| rledisez | fgemein: FYI, here is how our pipeline start : pipeline = catch_errors gatekeeper healthcheck proxy-logging cache container_sync ceilometer | 08:28 |
| rledisez | everything else is right of ceilometer | 08:28 |
| clayg | i like rledisez' better | 08:28 |
| clayg | aren't we on Newton already? What was O? | 08:28 |
| rledisez | ocata ? | 08:29 |
| clayg | oh heck yeah - octocat is going to be *AWEsome* | 08:29 |
| *** cbartz has joined #openstack-swift | 08:29 | |
| fgemein | I would love being able to upgrade to mitaka by now, but it’s mainly me to care about the whole cloud production | 08:29 |
| fgemein | we’re running swift since grizzly | 08:30 |
| openstackgerrit | Merged openstack/swift: Make timestamp setting to be consistent in test loop https://review.openstack.org/419709 | 08:30 |
| fgemein | And Openstack since juno | 08:30 |
| openstackgerrit | Merged openstack/swift: Fix download resumption for new SLOs. https://review.openstack.org/419664 | 08:30 |
| clayg | fgemein: oldskewl | 08:31 |
| clayg | fgemein: has it gotten any better along the way? | 08:31 |
| fgemein | a lot! | 08:31 |
| *** sams-gleb has joined #openstack-swift | 08:31 | |
| fgemein | In the beginning keystone was a big problem with swift | 08:33 |
| *** sams-gleb has quit IRC | 08:36 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Fix race in new partitions detecting new/invalid suffixes. https://review.openstack.org/418690 | 08:38 |
| *** sams-gleb has joined #openstack-swift | 08:39 | |
| onovy | hi guys. can anyone look into stable/newton patches? #416090 #419816 thanks :] | 08:39 |
| patchbot | Error: Spurious "]". You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands. | 08:39 |
| clayg | onovy: we were just discussing that the octocat release is where it's at now ;) | 08:40 |
| clayg | onovy: tell PavelK I can see his comments - he needs to get in IRC - unless he's still on leave - them tell him to kick back and stop doing reviews | 08:41 |
| *** geaaru has joined #openstack-swift | 08:43 | |
| *** glb has quit IRC | 08:44 | |
| *** glb has joined #openstack-swift | 08:47 | |
| onovy | clayg: he is still at home | 08:50 |
| *** glb has quit IRC | 08:53 | |
| *** glb has joined #openstack-swift | 08:54 | |
| clayg | tsk tsk tsk | 08:57 |
| clayg | he's acctually been really helpful | 08:58 |
| clayg | I should be humbled right? That he can still spot my bugs sleep deprived from home? | 08:58 |
| onovy | :))) | 09:04 |
| onovy | btw: i don't understand what you mean by 'octocat release is where it's at now'? so you are not going to support stable/newton, or it's not priority now, ...? | 09:04 |
| clayg | i recently came across this method is os.path -> https://docs.python.org/3.1/library/os.path.html#os.path.sameopenfile | 09:05 |
| clayg | i didn't know it was down there | 09:05 |
| clayg | i'm curious the underlying implementation | 09:05 |
| onovy | https://github.com/python/cpython/blob/master/Lib/genericpath.py#L103 | 09:06 |
| *** cbartz has joined #openstack-swift | 09:06 | |
| clayg | onovy: we were picking on poor fgemein earlier becauase he was running Liberty | 09:06 |
| onovy | still curious? :) | 09:06 |
| clayg | i was just trying to be funny | 09:07 |
| clayg | onovy: ummm dunno - let me look | 09:07 |
| clayg | yup, that's what I thought :D | 09:07 |
| clayg | nice | 09:07 |
| onovy | same inode + same dev | 09:08 |
| clayg | oic - there's a samefile *and* a sameopenfile - k that makes more sense | 09:08 |
| onovy | yep | 09:09 |
| onovy | but be carefull: Availability: Unix. // so no Swift on Windows? :(( | 09:09 |
| clayg | now you're trying to be funny | 09:09 |
| onovy | yep, me too | 09:10 |
| clayg | btw - do you have a like browser hotlink greeseonkey script or something to the cpython repo on github? I feel like it would have taken me at least a minute or two just to remmeber where I have cpython checked out - much less grep? | 09:12 |
| onovy | i'm using google for it | 09:13 |
| onovy | https://www.google.cz/webhp?sourceid=chrome-instant&ion=1&espv=2&ie=UTF-8#q=python%20github%20sameopenfile // second link :] | 09:15 |
| patchbot | Error: Spurious "]". You may want to quote your arguments with double quotes in order to prevent extra brackets from being evaluated as nested commands. | 09:15 |
| clayg | yeah - that works pretty well - python github - nice | 09:17 |
| clayg | it doesn't work as well for like "eventlet github server" - maybe that method really is obscure and the only index is the damn source? | 09:18 |
| onovy | yep, obscure name of method is problem imho | 09:21 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Optimize noop case for suffix rehash https://review.openstack.org/418692 | 09:23 |
| *** fgemein has quit IRC | 09:23 | |
| *** klrmn has quit IRC | 09:24 | |
| *** mvk has quit IRC | 09:25 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Better optimistic lock in get_hashes https://review.openstack.org/419787 | 09:32 |
| *** klrmn has joined #openstack-swift | 09:33 | |
| *** ppai has quit IRC | 09:35 | |
| *** cbartz has quit IRC | 09:36 | |
| *** klrmn has quit IRC | 09:38 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 09:42 |
| *** klrmn has joined #openstack-swift | 09:47 | |
| clayg | nope, I can't do the conslidated patch tonight - so five seperate it is p 418689 418690 418691 419787 418692 | 09:49 |
| patchbot | https://review.openstack.org/#/c/418689/ - swift - Extract test pattern to helper | 09:49 |
| *** jordanP has joined #openstack-swift | 09:50 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Extract test pattern to helper https://review.openstack.org/418689 | 09:52 |
| *** ppai has joined #openstack-swift | 09:52 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Fix race in new partitions detecting new/invalid suffixes. https://review.openstack.org/418690 | 09:52 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Fix performance regression with hash invalidations https://review.openstack.org/418691 | 09:53 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Better optimistic lock in get_hashes https://review.openstack.org/419787 | 09:53 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Optimize noop case for suffix rehash https://review.openstack.org/418692 | 09:53 |
| *** mvk has joined #openstack-swift | 09:56 | |
| *** sheel has joined #openstack-swift | 09:56 | |
| *** cbartz has joined #openstack-swift | 09:57 | |
| openstackgerrit | Jeremy Liu proposed openstack/python-swiftclient: Add missing requirements https://review.openstack.org/419863 | 10:00 |
| openstackgerrit | Christian Schwede proposed openstack/swift: Add support to increase object ring partition power https://review.openstack.org/337297 | 10:06 |
| *** sams-gleb has quit IRC | 10:17 | |
| *** openstackgerrit has quit IRC | 10:18 | |
| *** ppai has quit IRC | 10:34 | |
| *** sams-gleb has joined #openstack-swift | 10:34 | |
| *** ppai has joined #openstack-swift | 10:50 | |
| *** kei_yama has quit IRC | 10:53 | |
| *** kei_yama has joined #openstack-swift | 10:57 | |
| *** acoles_ is now known as acoles | 11:21 | |
| *** openstackgerrit has joined #openstack-swift | 11:35 | |
| openstackgerrit | Gábor Antal proposed openstack/swift: Reduced complexity of _response_iter() method https://review.openstack.org/247560 | 11:35 |
| acoles | clayg: I had forgotten about patch 410311 - sorry. my only excuse is that it was last year :/ | 11:35 |
| patchbot | https://review.openstack.org/#/c/410311/ - swift - Test current reclaim_age handling | 11:35 |
| *** openstackgerrit has quit IRC | 11:47 | |
| *** SkyRocknRoll has quit IRC | 11:54 | |
| *** kei_yama has quit IRC | 12:13 | |
| *** catintheroof has joined #openstack-swift | 12:30 | |
| *** catintheroof has quit IRC | 12:31 | |
| *** catintheroof has joined #openstack-swift | 12:32 | |
| *** psachin has quit IRC | 13:02 | |
| *** asettle__ has joined #openstack-swift | 13:19 | |
| *** asettle has quit IRC | 13:19 | |
| *** asettle__ is now known as asettle | 13:33 | |
| tdasilva | good morning | 13:46 |
| *** hseipp has quit IRC | 13:59 | |
| *** winggundamth has quit IRC | 14:05 | |
| *** openstackgerrit has joined #openstack-swift | 14:09 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: EC Fragment Duplication - Foundational Global EC Cluster Support https://review.openstack.org/219165 | 14:09 |
| kota_ | tdasilva: morning | 14:10 |
| kota_ | and leaving because finish up to update the global ec patch to address timburke's great comment. | 14:12 |
| kota_ | comments | 14:12 |
| * kota_ is going to offline | 14:12 | |
| tdasilva | kota_: have a good weekend! | 14:18 |
| *** JimCheung has joined #openstack-swift | 14:20 | |
| *** JimCheung has quit IRC | 14:24 | |
| *** sams-gleb has quit IRC | 14:25 | |
| *** dmorita has joined #openstack-swift | 14:26 | |
| *** sams-gleb has joined #openstack-swift | 14:26 | |
| *** sams-gleb has quit IRC | 14:30 | |
| *** dmorita has quit IRC | 14:30 | |
| *** tqtran has joined #openstack-swift | 14:39 | |
| *** tqtran has quit IRC | 14:41 | |
| *** joeljwright has joined #openstack-swift | 14:41 | |
| *** ChanServ sets mode: +v joeljwright | 14:41 | |
| *** chlong has joined #openstack-swift | 14:45 | |
| *** sheel has quit IRC | 14:57 | |
| *** ppai has quit IRC | 14:59 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Fix object server tests to include content-type headers https://review.openstack.org/420011 | 15:02 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Fix object server tests to include content-type headers https://review.openstack.org/420011 | 15:11 |
| *** sams-gleb has joined #openstack-swift | 15:20 | |
| timburke | clayg: ...that seems to be more patches than before? if the goal was a "consolidated patch" i think you might be moving in the wrong direction :-P | 15:33 |
| *** bkeller` has joined #openstack-swift | 15:38 | |
| *** sams-gleb has quit IRC | 15:40 | |
| *** sams-gleb has joined #openstack-swift | 15:41 | |
| *** chsc has joined #openstack-swift | 15:50 | |
| *** chsc has joined #openstack-swift | 15:50 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift: add swift_source to PUT request part of COPY https://review.openstack.org/420045 | 15:51 |
| *** mvk has quit IRC | 15:51 | |
| acoles | timburke: hi. re. symlink target/etag etc, yep, if target is an extra header then container server presumably would need to tack it onto etag (or put it somewhere) to store it. I don't yet have a strong opinion on it, I can see pros/cons of both approaches. | 15:52 |
| acoles | timburke: if we made container server smart enough then it would parse etag and populate extra items into the listing dict as it iters over the container listing, so middleware wouldn't need to intercept container GETs and parse the etags. But, it does mean enhancing the container server. | 15:55 |
| *** bkeller` has quit IRC | 16:05 | |
| tdasilva | acoles: the way you put it, it seems like enhancing the container server seems like a more generic approach?? although for middleware like encryption it would still need to intercept so that it can decrypt the data | 16:13 |
| *** cbartz has quit IRC | 16:15 | |
| *** cdelatte has joined #openstack-swift | 16:15 | |
| tdasilva | is github down? | 16:16 |
| joeljwright | tdasilva: looks like it :( | 16:18 |
| *** bkeller` has joined #openstack-swift | 16:25 | |
| *** pcaruana has joined #openstack-swift | 16:36 | |
| *** philipw has quit IRC | 16:37 | |
| *** philipw has joined #openstack-swift | 16:37 | |
| *** chlong has quit IRC | 16:38 | |
| *** chlong has joined #openstack-swift | 16:43 | |
| *** chlong has quit IRC | 16:48 | |
| *** chlong has joined #openstack-swift | 16:52 | |
| *** chlong has quit IRC | 16:56 | |
| *** chlong has joined #openstack-swift | 16:59 | |
| *** jordanP has quit IRC | 17:01 | |
| *** joeljwright has quit IRC | 17:03 | |
| *** chsc has quit IRC | 17:08 | |
| *** rledisez has quit IRC | 17:08 | |
| *** chlong has quit IRC | 17:09 | |
| *** chsc has joined #openstack-swift | 17:11 | |
| *** chsc has joined #openstack-swift | 17:11 | |
| *** asettle has quit IRC | 17:23 | |
| *** _JZ_ has joined #openstack-swift | 17:27 | |
| *** klrmn has quit IRC | 17:29 | |
| *** tqtran has joined #openstack-swift | 17:33 | |
| *** mvk has joined #openstack-swift | 17:38 | |
| *** JimCheung has joined #openstack-swift | 17:50 | |
| *** ChubYann has joined #openstack-swift | 17:58 | |
| *** oshritf has joined #openstack-swift | 17:59 | |
| *** dmorita has joined #openstack-swift | 18:01 | |
| timburke | good morning | 18:06 |
| *** links has quit IRC | 18:12 | |
| *** oshritf has quit IRC | 18:12 | |
| *** oshritf has joined #openstack-swift | 18:15 | |
| *** acoles is now known as acoles_ | 18:16 | |
| clayg | are we freaking out!? | 18:19 |
| *** oshritf has quit IRC | 18:19 | |
| *** oshritf has joined #openstack-swift | 18:20 | |
| *** oshritf has quit IRC | 18:23 | |
| *** oshritf has joined #openstack-swift | 18:23 | |
| *** oshritf has quit IRC | 18:25 | |
| *** pcaruana has quit IRC | 18:26 | |
| *** klrmn has joined #openstack-swift | 18:27 | |
| notmyname | OpenStack Board of Directors election is over: http://lists.openstack.org/pipermail/foundation/2017-January/002449.html | 18:39 |
| *** pcaruana has joined #openstack-swift | 18:39 | |
| *** manij has joined #openstack-swift | 18:41 | |
| *** oshritf has joined #openstack-swift | 18:46 | |
| *** oshritf has quit IRC | 18:46 | |
| *** oshritf has joined #openstack-swift | 18:49 | |
| *** tesseract has quit IRC | 18:53 | |
| *** geaaru has quit IRC | 18:53 | |
| *** oshritf has quit IRC | 18:56 | |
| timburke | bah. knew there was something i meant to do this week :-/ | 18:57 |
| *** openstack has joined #openstack-swift | 19:00 | |
| timburke | clayg: Eat Arby's :arbys: | 19:01 |
| *** pcaruana has quit IRC | 19:02 | |
| *** caiobrentano_ has quit IRC | 19:07 | |
| *** bkeller` has quit IRC | 19:09 | |
| *** jamielennox is now known as jamielennox|away | 19:12 | |
| *** chlong has joined #openstack-swift | 19:14 | |
| *** asettle has joined #openstack-swift | 19:19 | |
| clayg | can anyone think of a middleware that sets swift.source on the users request - or is swift.source only for additional subrequests? | 19:22 |
| clayg | TempURL - tempurl would totally set swift.source on the users request after installing it's authorize callback? | 19:22 |
| clayg | oh hrmm... maybe only for fetching the account/container info? | 19:23 |
| tdasilva | yep | 19:23 |
| timburke | it seems reasonable for copy to set both halves, though -- like, there isn't a COPY verb, we made that up. user sent a request to the proxy, and to service it we needed to make two subrequests to the backend | 19:26 |
| clayg | tdasilva: yeah idk man, it seems like maybe we only set swift.source on new requests that middlewares make up - i'm not sure the PUT side of a COPY request should say SSC or not | 19:27 |
| timburke | i don't know that either of those is really the "user's request" still, even if we largely take the PUT response and send it back to the user | 19:27 |
| clayg | I sorta think yes - because the body of the request did *not* come from the client? | 19:27 |
| clayg | timburke: yeah I was thinking PUT with X-Copy-From - but I think we all agree - the middleware takes the user request - turns it into to subrequests - then it makes sends a response | 19:28 |
| clayg | ok, so patch 420045 is pretty straight forward - ya'll wanna +A it or stew on it over the weekend? | 19:28 |
| patchbot | https://review.openstack.org/#/c/420045/ - swift - add swift_source to PUT request part of COPY | 19:28 |
| tdasilva | clayg: yeah, i've always viewed it as two sub_requests, had not considered the PUT half being the same as the user request | 19:29 |
| tdasilva | I thought the lack of swift_source was just something that had been overlooked and not actually done on purpose | 19:31 |
| notmyname | it matters in terms of billing | 19:36 |
| notmyname | in that the general principle is "filter out stuff with swift.source set" so you don't double-bill or otherwise mess up a customer's bill | 19:37 |
| notmyname | right? | 19:37 |
| clayg | do people do that? | 19:39 |
| clayg | probably not a good idea to charge for transfer bandwidth on a server-side-copy anyway? | 19:39 |
| clayg | it's too late now - i clicked +A - you're going to have to do another patch to take it out - it's much harder to take code out | 19:40 |
| tdasilva | lol | 19:40 |
| clayg | although... https://review.openstack.org/#/c/418690/5/swift/obj/diskfile.py | 19:40 |
| patchbot | patch 418690 - swift - Fix race in new partitions detecting new/invalid s... | 19:40 |
| *** bkeller` has joined #openstack-swift | 19:45 | |
| *** openstackstatus has quit IRC | 19:57 | |
| *** openstack has joined #openstack-swift | 20:00 | |
| *** david-lyle is now known as baling-wire | 20:00 | |
| clayg | jrichli: hey this one is pretty lightweight I think - patch 405134 - between you & mahatic and I - we can probably merge it - it's just test cleanup really | 20:07 |
| patchbot | https://review.openstack.org/#/c/405134/ - swift - Cleanup tests from empty suffix quarantined db fix | 20:07 |
| jrichli | i was already on this on this morning. great minds and all ;-) | 20:09 |
| clayg | wha?? no whey | 20:09 |
| clayg | TMW git branch | wc -l == 76 | 20:10 |
| clayg | I bet someone in channel here is worse and I bet it's timburke | 20:10 |
| *** psachin has joined #openstack-swift | 20:10 | |
| timburke | clayg: which checkout? :P | 20:11 |
| timburke | either one, i think you're still way ahead of me | 20:11 |
| timburke | go clayg, go! | 20:11 |
| jrichli | i keep my vms for a long time ... but i still only had 37 | 20:11 |
| clayg | 37 vms? And do most of them have swift installed on them (!?) | 20:13 |
| tdasilva | heh | 20:13 |
| jrichli | no - 37 branches! | 20:13 |
| clayg | oh phew | 20:14 |
| jrichli | they build up when you re-use the same vm for a long time is all ... | 20:14 |
| clayg | my code is on my host | 20:14 |
| jrichli | ah, ic | 20:14 |
| jrichli | newly elected directors from HPE? if they remove important contributors ... seems like this is not the appropriate role | 20:16 |
| jrichli | sorry ... i felt the need | 20:17 |
| *** openstackstatus has quit IRC | 20:27 | |
| *** openstack has joined #openstack-swift | 20:30 | |
| timburke | why stay away? afraid i'll insist on more word-smithing? | 20:33 |
| clayg | oh, no - the whole thing feels very much like we're got the sharpest crayon's in the bin all discussing the color of the nuclear facilities new employee handbook cover | 20:34 |
| clayg | i guess you may have already looked at it? Maybe it's just better to get it over with. | 20:35 |
| *** cdelatte has quit IRC | 20:43 | |
| clayg | did anyone besides cschwede look at patch 311226 | 20:45 |
| patchbot | https://review.openstack.org/#/c/311226/ - swift - Pretend *some* parts min_part_hours_passed | 20:45 |
| clayg | I wrote it for dfg_ and ahale - cschwede i guess said he liked it - but I was sorta thinking about abandoning it? | 20:45 |
| *** JimCheung has quit IRC | 20:46 | |
| *** silor has quit IRC | 20:50 | |
| *** JimCheung has joined #openstack-swift | 21:02 | |
| *** JimCheung has quit IRC | 21:06 | |
| *** JimCheung has joined #openstack-swift | 21:32 | |
| *** Jeffrey4l__ has joined #openstack-swift | 21:34 | |
| *** Jeffrey4l_ has quit IRC | 21:35 | |
| *** chosafine has joined #openstack-swift | 21:37 | |
| *** sams-gleb has quit IRC | 21:57 | |
| *** jamielennox|away is now known as jamielennox | 21:59 | |
| tdasilva | clayg: hi, still around? | 22:00 |
| zaitcev | what again dude, have a mercy, he's landed the part++ thing :-) | 22:04 |
| openstackgerrit | Merged openstack/swift: add swift_source to PUT request part of COPY https://review.openstack.org/420045 | 22:15 |
| clayg | tdasilva: let's go with yes; but I may poof if i don't like the where things go | 22:22 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Rebalance with min_part_seconds_left > 0 https://review.openstack.org/326967 | 22:24 |
| *** baling-wire is now known as david-lyle | 22:26 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 22:30 |
| *** chosafine has quit IRC | 22:34 | |
| *** chosafine has joined #openstack-swift | 22:36 | |
| *** catintheroof has quit IRC | 22:38 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Cleanup tests from empty suffix quarantined db fix https://review.openstack.org/405134 | 22:42 |
| clayg | tdasilva: well hmph - where'd YOU go then? | 22:43 |
| *** vatlas has joined #openstack-swift | 22:44 | |
| clayg | zaitcev: what landed now? part power increase is done!? | 22:44 |
| zaitcev | yes | 22:44 |
| clayg | zaitcev: patch 337297? | 22:45 |
| patchbot | https://review.openstack.org/#/c/337297/ - swift - Add support to increase object ring partition power | 22:45 |
| zaitcev | oh... I looked at 155107 by mistake | 22:46 |
| zaitcev | Neever mind then, let's have Thiago poke you more | 22:46 |
| clayg | zaitcev: oh that was a long time ago | 22:47 |
| clayg | yeah poke at me!? | 22:47 |
| clayg | tdasilva: did you just need a review? on what? is it quick/easy!? Is it DONE!!!!??? LETS MERGE IT! | 22:47 |
| clayg | sigh, i guess reading the commit message of patch 311226 it maybe *was* a good idea - I supose I'll try to rebase it - :sigh: | 22:50 |
| patchbot | https://review.openstack.org/#/c/311226/ - swift - Pretend *some* parts min_part_hours_passed | 22:50 |
| clayg | oh you know what - i'll rebase it on patch 326967 - that's be super hawt | 22:51 |
| patchbot | https://review.openstack.org/#/c/326967/ - swift - Rebalance with min_part_seconds_left > 0 | 22:51 |
| *** wasmum has joined #openstack-swift | 22:52 | |
| *** sams-gleb has joined #openstack-swift | 22:58 | |
| MooingLemur | can a host or set of devices have its zone number changed easily? | 22:59 |
| MooingLemur | seems like it should be trivial to do, ring-wise, and then rebalance should evict duplicated parts within the zone if it can. | 23:01 |
| *** sams-gleb has quit IRC | 23:02 | |
| clayg | MooingLemur: there's no interface to do it in the CLI | 23:04 |
| clayg | MooingLemur: you should see if there's an open bug | 23:04 |
| MooingLemur | aha, but there's nothing preventing me from programmatically setting it though, right? | 23:04 |
| clayg | wehhhhhellll | 23:04 |
| * MooingLemur chuckles | 23:04 | |
| clayg | If *I* had a good reason to do it I'd cirtainly *try* | 23:05 |
| clayg | FLW "it should be fine" | 23:05 |
| MooingLemur | yeah, I have a single-zone region which formerly only had max 1 replica, but now it's becoming the larger region | 23:05 |
| MooingLemur | and it seems prudent to rezone it based on the switch the host is connected to | 23:05 |
| clayg | So i suggest to people that when they add capacity they should *not* put some hard drives from this batch into zone 1, some into zone 2 etc | 23:06 |
| clayg | they should add a zone and put all the new capacity in there | 23:06 |
| clayg | group devices that will fail together in the same failure domain | 23:06 |
| clayg | that way when the drives start dropping like flies you KNOW the ring would have *tried* to only have at max on replica of some parts in that zone | 23:06 |
| clayg | when you spread new capacity around the ring doesn't know they they are acctually in share a domain in which failures will be correlated | 23:07 |
| clayg | but ... that's just a little personal thing for me becase ... NO REASON ;) | 23:07 |
| MooingLemur | aha, yeah, good point wrt HDD batches | 23:07 |
| MooingLemur | :) | 23:07 |
| clayg | maybe not related to your toplogy situation | 23:08 |
| clayg | having more zones is a good thing | 23:08 |
| clayg | but they should represent failure domains | 23:08 |
| clayg | if you have regions you don't really have to worry about zones across regions being the wrong size | 23:08 |
| MooingLemur | the failure domain I'm most concerned about is the ToR switch in the larger region | 23:09 |
| clayg | if you ever pigeon hole replicanths assigned in a tier to < #failure domains in subtier you have to watch your overload/dispersion - but generally "split a big zone into two" is - idk, there's probably good topology regions to do it | 23:10 |
| clayg | i'm sure you'll do the right thing | 23:10 |
| clayg | so if you have a two regions and one has more weight it's going to get more part-replicas - if in that region you have two zones and one of those zones is larger it's going to get more of the part-replicas assigned to the region | 23:11 |
| clayg | I mean obviously a big of planning can go a long way - but who's got a crystal ball! | 23:11 |
| clayg | My opinion is that if you tell the ring what you've got it should put the part-replicas where they go and if it doesn't it's a bug ;) | 23:11 |
| clayg | you shouldn't have to make up zones to trick it into doing the right thing | 23:11 |
| clayg | but if you *have* two faiure domains (and accidently told the ring they were the same failure domain) - by all means - edit your builder and see what happens with a rebalance (keep a backup) | 23:12 |
| *** lespaul has quit IRC | 23:13 | |
| clayg | also... i'm not saying there's no bugs in there - i'm sure there's a totally reasonable topology that leads to wrong placement (for some definition of wrong) - but I think if it's objectively bad (as opposed to just; but that's not what I *want*) - we could file a bug, look for a workaround, look for fix, etc | 23:14 |
| clayg | we had lots of "this ring produces not good placement" over the years | 23:15 |
| clayg | my new favorite wish-list item lp bug #1656440 | 23:15 |
| openstack | Launchpad bug 1656440 in OpenStack Object Storage (swift) "test.unit.cli.test_ringbuilder should consisently use run_srb" [Wishlist,New] https://launchpad.net/bugs/1656440 | 23:15 |
| MooingLemur | yeah, I might end up being on 10 different switches, so I'll probably want 10 zones in my bigger region rather than 1 | 23:16 |
| MooingLemur | so both of my replicas don't end up in the same rack | 23:17 |
| MooingLemur | thanks :) | 23:17 |
| clayg | If you *do* end up doing the zone editing by hand and think you have a real use-case that some operator is going to run into ... and especially if you can think of a command syntax that would make sense - i'd love to have upstream bug to add the feature | 23:19 |
| clayg | *think you have and could describe a use-case ... | 23:19 |
| *** asettle has quit IRC | 23:19 | |
| MooingLemur | for a syntax example, perhaps: set_info 10.1.2.3 r2z10 | 23:20 |
| MooingLemur | or --change-zone | 23:20 |
| *** asettle has joined #openstack-swift | 23:20 | |
| *** chsc has quit IRC | 23:21 | |
| *** asettle is now known as Guest11517 | 23:21 | |
| clayg | oh yeah... can you just plumb that in? it's like we update all the other metadata but didn't add that one? | 23:21 |
| MooingLemur | yeah, possibly | 23:21 |
| *** bkeller` has quit IRC | 23:21 | |
| clayg | since it's a feature if you decide to work on it you don't *have* to file an upstream bug - but enumerating the use-case in the commit message would still be helpful | 23:22 |
| openstackgerrit | Joel Wright proposed openstack/swift: Add Preamble and Postamble to SLO and SegmentedIterable https://review.openstack.org/365371 | 23:34 |
| *** chlong has quit IRC | 23:38 | |
| *** joeljwright has joined #openstack-swift | 23:39 | |
| *** ChanServ sets mode: +v joeljwright | 23:39 | |
| joeljwright | timburke: you still around? | 23:40 |
| timburke | joeljwright: yup | 23:40 |
| joeljwright | great! | 23:40 |
| joeljwright | I've just pushed up another revision of the SLO preamble/postamble patch | 23:40 |
| joeljwright | it covers all the comments on the changed code (thanks for those comments by the way, I think the new version is an improvement) | 23:41 |
| timburke | i saw. doing a quick look at the diff w/ ps13 now | 23:41 |
| joeljwright | I still need to improve the request helpers coverage and add docs | 23:43 |
| joeljwright | would appreciate thoughts on the former | 23:43 |
| *** _JZ_ has quit IRC | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!