| *** kei_yama has quit IRC | 00:05 | |
| *** kei_yama has joined #openstack-swift | 00:08 | |
| *** shakamunyi has joined #openstack-swift | 00:09 | |
| *** NM has quit IRC | 00:11 | |
| *** chlong has quit IRC | 00:12 | |
| *** cryptsting has quit IRC | 00:12 | |
| *** asettle is now known as asettle-bfast | 00:17 | |
| *** lyrrad_ has joined #openstack-swift | 00:31 | |
| *** lyrrad has quit IRC | 00:32 | |
| *** lyrrad_ is now known as lyrrad | 00:32 | |
| *** rcernin has quit IRC | 00:37 | |
| openstackgerrit | Merged openstack/swift: Added unit tests to recon.py. https://review.openstack.org/239534 | 00:57 |
|---|---|---|
| *** kei_yama has quit IRC | 01:00 | |
| *** kei_yama has joined #openstack-swift | 01:01 | |
| *** shakamunyi has quit IRC | 01:02 | |
| *** shakamunyi has joined #openstack-swift | 01:03 | |
| *** lyrrad has quit IRC | 01:04 | |
| *** blmartin has quit IRC | 01:05 | |
| *** NM has joined #openstack-swift | 01:07 | |
| *** lyrrad has joined #openstack-swift | 01:09 | |
| *** chlong has joined #openstack-swift | 01:48 | |
| *** sgundur has joined #openstack-swift | 01:51 | |
| *** chlong has quit IRC | 01:51 | |
| *** sgundur has left #openstack-swift | 01:53 | |
| *** kei_yama_ has joined #openstack-swift | 01:57 | |
| *** kei_yama has quit IRC | 01:57 | |
| *** chlong has joined #openstack-swift | 02:00 | |
| *** tsg has quit IRC | 02:02 | |
| *** lyrrad has quit IRC | 02:03 | |
| *** chlong has quit IRC | 02:06 | |
| *** klrmn1 has quit IRC | 02:08 | |
| *** chlong has joined #openstack-swift | 02:19 | |
| mattoliverau | phew... I think I found the cause of the testr functests in_process failure! | 02:28 |
| mattoliverau | Also if we use os-testr we don't need to the testr wrapper, which keeps it a little cleaner. | 02:28 |
| *** chlong has quit IRC | 02:35 | |
| *** NM has quit IRC | 02:39 | |
| *** asettle-bfast is now known as asettle | 02:44 | |
| *** chlong has joined #openstack-swift | 02:48 | |
| *** sanchitmalhotra has joined #openstack-swift | 02:50 | |
| *** mahatic has joined #openstack-swift | 03:05 | |
| *** dmorita has quit IRC | 03:06 | |
| *** nadeem has joined #openstack-swift | 03:06 | |
| *** venkat_p has joined #openstack-swift | 03:08 | |
| *** vinsh has quit IRC | 03:17 | |
| *** jerrygb has quit IRC | 03:19 | |
| *** janonymous has quit IRC | 03:19 | |
| *** janonymous has joined #openstack-swift | 03:20 | |
| *** badari has quit IRC | 03:21 | |
| *** chlong has quit IRC | 03:21 | |
| *** badari has joined #openstack-swift | 03:23 | |
| *** jome has quit IRC | 03:48 | |
| *** jome has joined #openstack-swift | 03:52 | |
| *** ppai has joined #openstack-swift | 04:07 | |
| *** silor has joined #openstack-swift | 04:13 | |
| *** jerrygb has joined #openstack-swift | 04:20 | |
| *** jerrygb has quit IRC | 04:26 | |
| *** janonymous has quit IRC | 04:43 | |
| *** janonymous has joined #openstack-swift | 04:44 | |
| *** trifon has joined #openstack-swift | 04:51 | |
| venkat_p | Hi all | 04:55 |
| venkat_p | I have one specific problem in swift | 04:55 |
| venkat_p | how to extract account db contents from the container-server? | 04:55 |
| venkat_p | or object-server | 04:56 |
| venkat_p | assume that I have a new table in account-db | 04:56 |
| *** jome has quit IRC | 04:58 | |
| zaitcev | I don't know of any way to do it, but I'm not entirely sure nobody tried it before. In theory, the informaton is out there in the container server, I think. | 05:01 |
| zaitcev | The first problem I see is how to find all the container for the emptied account. You have to scan the whole cluster. | 05:01 |
| *** jome has joined #openstack-swift | 05:03 | |
| *** dmorita has joined #openstack-swift | 05:05 | |
| *** links has joined #openstack-swift | 05:06 | |
| *** changbl has quit IRC | 05:13 | |
| *** ianbrown has joined #openstack-swift | 05:14 | |
| openstackgerrit | Merged openstack/python-swiftclient: Change ignore-errors to ignore_errors https://review.openstack.org/251591 | 05:15 |
| *** haomaiwang has joined #openstack-swift | 05:16 | |
| gmmaha | mattoliverau: woohoo.. | 05:19 |
| gmmaha | i was trying to reproduce the issue locally, but no luck | 05:19 |
| mattoliverau | gmmaha: you need to move the test.conf file, and not run services so it can use in_process testing, or set some env varaibles. I've stopped it from freezing, but just debugging a bit more. Will push up a patchset soon to see how it goes. | 05:21 |
| *** ianbrown has quit IRC | 05:21 | |
| gmmaha | mattoliverau: ohhh.. | 05:22 |
| gmmaha | i think there was a similar discussion on the same that notmyname debugged to reproduce a different issue | 05:23 |
| gmmaha | thanks.. will check it out and jenkins will definitely beat me to it | 05:23 |
| *** badari has quit IRC | 05:29 | |
| *** changbl has joined #openstack-swift | 05:30 | |
| *** badari has joined #openstack-swift | 05:36 | |
| *** venkat_p has quit IRC | 05:52 | |
| *** venkat_p has joined #openstack-swift | 05:53 | |
| *** gyee has quit IRC | 06:00 | |
| *** janonymous has quit IRC | 06:01 | |
| *** janonymous has joined #openstack-swift | 06:19 | |
| *** jerrygb has joined #openstack-swift | 06:22 | |
| *** jerrygb has quit IRC | 06:26 | |
| *** nadeem has quit IRC | 06:52 | |
| *** mahatic has quit IRC | 07:02 | |
| *** eranrom has joined #openstack-swift | 07:10 | |
| *** rcernin has joined #openstack-swift | 07:19 | |
| openstackgerrit | pei ran wu proposed openstack/swift: Fix POST-as-COPY mess up the DLO manifest https://review.openstack.org/252791 | 07:22 |
| *** rcernin_ has joined #openstack-swift | 07:24 | |
| *** silor has quit IRC | 07:25 | |
| *** zaitcev has quit IRC | 07:30 | |
| *** flaper87 has quit IRC | 07:44 | |
| *** flaper87 has joined #openstack-swift | 07:49 | |
| *** chlong has joined #openstack-swift | 07:55 | |
| *** arnox has joined #openstack-swift | 08:05 | |
| *** dmorita has quit IRC | 08:15 | |
| *** chlong has quit IRC | 08:29 | |
| *** geaaru has joined #openstack-swift | 08:38 | |
| *** Trixboxer has joined #openstack-swift | 08:40 | |
| *** SkyRocknRoll has joined #openstack-swift | 08:42 | |
| *** jordanP has joined #openstack-swift | 08:46 | |
| *** mahatic has joined #openstack-swift | 08:47 | |
| *** rcernin has quit IRC | 08:50 | |
| *** rcernin_ has quit IRC | 08:51 | |
| *** ho has quit IRC | 08:51 | |
| *** jome has quit IRC | 08:52 | |
| *** jome has joined #openstack-swift | 08:54 | |
| *** jome_ has joined #openstack-swift | 09:16 | |
| *** jome has quit IRC | 09:16 | |
| *** jome has joined #openstack-swift | 09:28 | |
| *** jome_ has quit IRC | 09:28 | |
| *** aix has joined #openstack-swift | 09:29 | |
| *** rcernin has joined #openstack-swift | 09:59 | |
| *** jistr has joined #openstack-swift | 10:09 | |
| *** venkat_p has quit IRC | 10:10 | |
| *** mahatic has quit IRC | 10:20 | |
| *** venkat_p has joined #openstack-swift | 10:28 | |
| *** jmccarthy has joined #openstack-swift | 10:31 | |
| *** haomaiwang has quit IRC | 10:38 | |
| *** hseipp has joined #openstack-swift | 10:41 | |
| *** silor has joined #openstack-swift | 10:45 | |
| *** silor1 has joined #openstack-swift | 10:49 | |
| *** silor has quit IRC | 10:49 | |
| *** silor1 is now known as silor | 10:49 | |
| *** mahatic has joined #openstack-swift | 10:50 | |
| *** chlong has joined #openstack-swift | 11:01 | |
| *** aix has quit IRC | 11:18 | |
| *** mahatic has quit IRC | 11:37 | |
| *** NM has joined #openstack-swift | 11:43 | |
| *** mahatic has joined #openstack-swift | 11:45 | |
| *** SkyRocknRoll has quit IRC | 11:55 | |
| mahatic | acoles: hi, are you here? | 11:56 |
| *** NM has quit IRC | 12:05 | |
| *** flaper87 has quit IRC | 12:17 | |
| *** flaper87 has joined #openstack-swift | 12:17 | |
| *** venkat_p has quit IRC | 12:22 | |
| *** openstackgerrit has quit IRC | 12:32 | |
| *** openstackgerrit has joined #openstack-swift | 12:32 | |
| *** venkat_p has joined #openstack-swift | 12:35 | |
| *** aix has joined #openstack-swift | 12:39 | |
| *** haomaiwa_ has joined #openstack-swift | 12:43 | |
| *** proteusguy has quit IRC | 12:56 | |
| openstackgerrit | Eran Rom proposed openstack/swift: Container-Sync to iterate only over synced containers https://review.openstack.org/205803 | 13:05 |
| *** proteusguy has joined #openstack-swift | 13:09 | |
| *** kei_yama_ has quit IRC | 13:14 | |
| *** openstackgerrit has quit IRC | 13:17 | |
| *** openstackgerrit has joined #openstack-swift | 13:17 | |
| acoles | hi mahatic | 13:21 |
| *** jerrygb has joined #openstack-swift | 13:21 | |
| mahatic | acoles: hello, did you get a chance to take a look at my comments on 248377? No problem if you havent, just wanted to check and discuss on it | 13:27 |
| acoles | mahatic: i looked briefly | 13:28 |
| acoles | i think the goal should be to iterate over the body, decrypting pieces as you go (each mime part and its content-type header) | 13:29 |
| acoles | the harder piece if calculating the correct length of the plaintext response *before* iterating through the body | 13:29 |
| mahatic | right | 13:29 |
| mahatic | acoles: in you suggestions as well, even if I do have content-type (either way through sysmeta or otherwise), I would still need to know how many of those headers are there to calculate body length | 13:31 |
| acoles | so my thoughts were either to do that by reading just the start of the body, until the first boundary plus encrypted content-type, decrypt just that header and from there predict the new content length (shorter by whatever the content-type reduces by) | 13:31 |
| acoles | mahatic: you're right :/ | 13:31 |
| mahatic | hmmm :/ | 13:31 |
| acoles | doesn't the content-range header tell us how many parts to expect? | 13:32 |
| * acoles goes to read the spec | 13:32 | |
| *** jome has quit IRC | 13:33 | |
| mahatic | acoles: I dont think so | 13:34 |
| acoles | no it doesn't | 13:34 |
| acoles | hmmm | 13:36 |
| * mahatic is trying to think | 13:36 | |
| * acoles has given up thinking :) | 13:36 | |
| mahatic | lol. Noo that cant be possible! | 13:37 |
| acoles | I wonder if its possible that swift obj server could tell us how many parts it is sending in a new response header? | 13:38 |
| *** janonymous_ has joined #openstack-swift | 13:38 | |
| mahatic | yeah, thats where i was heading too :) | 13:38 |
| *** janonymous_ has quit IRC | 13:38 | |
| mahatic | acoles: i think it should be able to, lemme grab the code line - but we should persist that value in some way to the middleare | 13:40 |
| *** haomaiwa_ has quit IRC | 13:40 | |
| acoles | mahatic: the swob response knows how many ranges because it calculates the content-length in _prepare_for_ranges. | 13:41 |
| mahatic | acoles: https://github.com/openstack/swift/blob/master/swift/proxy/controllers/base.py#L1105 | 13:41 |
| acoles | ^^ thats an iter, we don't know how may items it has up-front. | 13:43 |
| acoles | i was thinking the obj server could tell us in a header | 13:43 |
| mahatic | yeah, I was trying to point out that, maybe poking around inside that (that iter reads through the body and we can caculate the ńumber parts along) | 13:43 |
| mahatic | and in swob, https://github.com/openstack/swift/blob/master/swift/common/swob.py#L1183 - that also iterated through the given ranges and calculates length | 13:44 |
| mahatic | *iterates | 13:44 |
| mahatic | I kinda gave it a shot before - capture ranges in request (GET) and proceed to calculate length | 13:45 |
| * mahatic cant remember why that idea was dropped | 13:45 | |
| acoles | mahatic: does EC need to do something similar with multi-range responses i.e. does the content-length change after recontruction? | 13:46 |
| *** Guest36914 has quit IRC | 13:46 | |
| acoles | mahatic: i don't think you can assume that the number of ranges requested is the number of parts returned | 13:46 |
| acoles | the obj server can not return some ranges, and can merge ranges. | 13:47 |
| mahatic | acoles: ah, and so it was throwing an error intermittently and I couldnt pin point why :) | 13:47 |
| acoles | so we need to know what got returned, not what was requested. | 13:47 |
| mahatic | ok, thanks for explaining that | 13:47 |
| acoles | mahatic: it might be instructive to understand EC obj controller handling of multi-range responses, in case that throws up any ideas. | 13:48 |
| mahatic | I havent touched upon the EC stuff, I can poke around it and see | 13:48 |
| mahatic | (or maybe I did and cant recall, will take a quick look) | 13:48 |
| mahatic | sure | 13:48 |
| * acoles emphasizes 'might' | 13:49 | |
| mahatic | yeah, doesnt hurt :) | 13:49 |
| acoles | good luck! | 13:50 |
| *** links has quit IRC | 13:51 | |
| *** tsg has joined #openstack-swift | 13:54 | |
| mahatic | acoles: thanks, i need that | 13:54 |
| *** ppai has quit IRC | 14:00 | |
| openstackgerrit | Tushar Gohad proposed openstack/swift: Bump PyECLib requirement to >=1.0.7 https://review.openstack.org/244287 | 14:00 |
| *** ir2ivps8_ has quit IRC | 14:03 | |
| *** breitz has quit IRC | 14:05 | |
| *** breitz has joined #openstack-swift | 14:05 | |
| *** venkat_p has quit IRC | 14:10 | |
| *** tsg has quit IRC | 14:12 | |
| *** ir2ivps8 has joined #openstack-swift | 14:27 | |
| *** CaioBrentano has joined #openstack-swift | 14:31 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:31 | |
| *** mahatic has quit IRC | 14:45 | |
| *** mahatic has joined #openstack-swift | 14:46 | |
| *** dustins has joined #openstack-swift | 14:50 | |
| *** blmartin has joined #openstack-swift | 14:59 | |
| *** vinsh has joined #openstack-swift | 15:00 | |
| *** rohit_ has joined #openstack-swift | 15:09 | |
| *** petertr7_away is now known as petertr7 | 15:22 | |
| *** NM has joined #openstack-swift | 15:31 | |
| *** lcurtis has joined #openstack-swift | 15:33 | |
| *** jordanP has quit IRC | 15:39 | |
| *** lcurtis has quit IRC | 15:41 | |
| *** lcurtis has joined #openstack-swift | 15:43 | |
| *** arnox has quit IRC | 15:48 | |
| *** jome has joined #openstack-swift | 15:49 | |
| *** haomaiwa_ has joined #openstack-swift | 15:52 | |
| *** jordanP has joined #openstack-swift | 15:58 | |
| *** jome_ has joined #openstack-swift | 16:00 | |
| *** jome has quit IRC | 16:00 | |
| *** haomaiwa_ has quit IRC | 16:00 | |
| *** 18VAAEQWZ has joined #openstack-swift | 16:01 | |
| hurricanerix | notmyname when you say 'file a bug and set it to "wishlist"', you mean, add a tag with the value 'wishlist' right? did i do this correctly? https://bugs.launchpad.net/swift/+bug/1522450 | 16:01 |
| openstack | Launchpad bug 1522450 in OpenStack Object Storage (swift) "only install requirements you use" [Undecided,New] | 16:01 |
| *** aix has quit IRC | 16:01 | |
| *** 18VAAEQWZ has quit IRC | 16:02 | |
| *** haomaiwa_ has joined #openstack-swift | 16:02 | |
| *** haomaiwa_ has quit IRC | 16:03 | |
| *** haomaiwa_ has joined #openstack-swift | 16:03 | |
| *** haomaiwa_ has quit IRC | 16:04 | |
| *** haomaiwang has joined #openstack-swift | 16:04 | |
| *** nadeem has joined #openstack-swift | 16:04 | |
| openstackgerrit | Peter Lisák proposed openstack/swift: Add info about state of ring file to default command. https://review.openstack.org/253038 | 16:04 |
| *** haomaiwang has quit IRC | 16:05 | |
| *** haomaiwa_ has joined #openstack-swift | 16:05 | |
| *** haomaiwa_ has quit IRC | 16:06 | |
| *** haomaiwang has joined #openstack-swift | 16:06 | |
| *** haomaiwang has quit IRC | 16:07 | |
| *** haomaiwang has joined #openstack-swift | 16:07 | |
| *** haomaiwang has quit IRC | 16:08 | |
| *** haomaiwa_ has joined #openstack-swift | 16:08 | |
| *** haomaiwa_ has quit IRC | 16:09 | |
| *** haomaiwa_ has joined #openstack-swift | 16:09 | |
| *** haomaiwa_ has quit IRC | 16:10 | |
| *** 5EXAADMHT has joined #openstack-swift | 16:10 | |
| *** 5EXAADMHT has quit IRC | 16:11 | |
| *** haomaiwang has joined #openstack-swift | 16:11 | |
| *** diogogmt has joined #openstack-swift | 16:11 | |
| *** haomaiwang has quit IRC | 16:12 | |
| *** 5EXAADMKT has joined #openstack-swift | 16:12 | |
| *** rcernin has quit IRC | 16:12 | |
| *** 5EXAADMKT has quit IRC | 16:13 | |
| *** 5EXAADML5 has joined #openstack-swift | 16:13 | |
| *** 5EXAADML5 has quit IRC | 16:13 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 16:16 |
| *** jordanP has quit IRC | 16:17 | |
| *** siva_krishnan has joined #openstack-swift | 16:18 | |
| *** jordanP has joined #openstack-swift | 16:22 | |
| *** diazjf has joined #openstack-swift | 16:29 | |
| *** dmorita has joined #openstack-swift | 16:39 | |
| *** mahatic_ has joined #openstack-swift | 16:48 | |
| *** zaitcev has joined #openstack-swift | 16:49 | |
| *** ChanServ sets mode: +v zaitcev | 16:49 | |
| *** petertr7 is now known as petertr7_away | 16:50 | |
| *** jome_ has quit IRC | 16:52 | |
| *** petertr7_away is now known as petertr7 | 16:53 | |
| *** diazjf has quit IRC | 16:53 | |
| *** rcernin has joined #openstack-swift | 16:53 | |
| *** diazjf has joined #openstack-swift | 16:53 | |
| *** thumpba has joined #openstack-swift | 16:54 | |
| *** diazjf has quit IRC | 16:58 | |
| *** mahatic__ has joined #openstack-swift | 17:03 | |
| *** mahatic has quit IRC | 17:05 | |
| *** gyee has joined #openstack-swift | 17:07 | |
| *** mahatic__ has quit IRC | 17:08 | |
| *** dmorita has quit IRC | 17:17 | |
| *** dmorita has joined #openstack-swift | 17:17 | |
| *** mahatic_ has quit IRC | 17:17 | |
| *** mahatic has joined #openstack-swift | 17:18 | |
| *** jistr has quit IRC | 17:21 | |
| mahatic | there we go, znc set up done! thank you gmmaha ! | 17:24 |
| gmmaha | my pleasure mahatic :) | 17:24 |
| acoles | yay mahatic has a bouncer | 17:29 |
| mahatic | yup :) | 17:30 |
| gmmaha | mahatic: the least i can do for fending all my stupid Qs. ;) | 17:34 |
| *** petertr7 is now known as petertr7_away | 17:35 | |
| notmyname | good morning | 17:36 |
| notmyname | mahatic: yay IRC bouncers! | 17:36 |
| notmyname | (and gmmaha!) | 17:36 |
| gmmaha | good morning notmyname | 17:37 |
| notmyname | hurricanerix: no, not a tag for "wishlist". set the status to wishlist. eg like this one https://bugs.launchpad.net/swift/+bug/1521350 | 17:37 |
| openstack | Launchpad bug 1521350 in OpenStack Object Storage (swift) "policy descriptions" [Wishlist,New] | 17:37 |
| notmyname | oh, not status. the "importance" field | 17:37 |
| *** dmorita has quit IRC | 17:38 | |
| *** petertr7_away is now known as petertr7 | 17:39 | |
| *** thumpba_ has joined #openstack-swift | 17:39 | |
| *** thumpba has quit IRC | 17:40 | |
| *** dmorita has joined #openstack-swift | 17:42 | |
| *** tsg has joined #openstack-swift | 17:43 | |
| *** lyrrad has joined #openstack-swift | 17:44 | |
| *** dmorita has quit IRC | 17:44 | |
| *** tsg_ has joined #openstack-swift | 17:45 | |
| *** tsg has quit IRC | 17:48 | |
| *** aix has joined #openstack-swift | 17:49 | |
| *** dmorita has joined #openstack-swift | 17:50 | |
| *** geaaru has quit IRC | 17:56 | |
| *** klrmn1 has joined #openstack-swift | 17:58 | |
| hurricanerix | notmyname i don't think i have permission to edit that one | 17:58 |
| hurricanerix | it has a icon of a pencil with a red line through it | 17:58 |
| notmyname | ah ok | 17:58 |
| notmyname | so file it as new and it will get put as wishlist :-) | 17:59 |
| hurricanerix | newer than this morning? | 17:59 |
| openstackgerrit | Sivasathurappan Radhakrishnan proposed openstack/swift: Added unit test cases for common/bufferedhttp.py https://review.openstack.org/244678 | 18:01 |
| *** jordanP has quit IRC | 18:04 | |
| hurricanerix | notmyname or do you mean, someone will see it, and change the status for me to wishlist? (just trying to verify if it should auto-set to wishlist, or if this is a manual process) | 18:04 |
| hurricanerix | because if it is supposed to be auto, i don't think it is working. | 18:05 |
| notmyname | it's a manual process | 18:05 |
| *** klrmn1 has quit IRC | 18:06 | |
| *** klrmn1 has joined #openstack-swift | 18:08 | |
| *** aix has quit IRC | 18:09 | |
| *** aix has joined #openstack-swift | 18:13 | |
| *** cmar_eurotux has joined #openstack-swift | 18:19 | |
| *** cmar_eurotux has left #openstack-swift | 18:19 | |
| *** cmar_eurotux has joined #openstack-swift | 18:19 | |
| *** cmar_eurotux has left #openstack-swift | 18:21 | |
| *** hseipp has quit IRC | 18:23 | |
| *** rohit_ has quit IRC | 18:23 | |
| *** rohit_ has joined #openstack-swift | 18:23 | |
| *** petertr7 is now known as petertr7_away | 18:25 | |
| *** dmorita has quit IRC | 18:33 | |
| *** dmorita has joined #openstack-swift | 18:34 | |
| *** dmorita has quit IRC | 18:50 | |
| *** rohit_ has quit IRC | 18:52 | |
| *** acoles is now known as acoles_ | 18:56 | |
| *** blmartin has quit IRC | 18:56 | |
| *** blmartin has joined #openstack-swift | 19:00 | |
| *** dmorita has joined #openstack-swift | 19:00 | |
| *** dmsimard is now known as dmsimard|afk | 19:01 | |
| *** dmorita has quit IRC | 19:04 | |
| *** dmorita has joined #openstack-swift | 19:06 | |
| *** petertr7_away is now known as petertr7 | 19:08 | |
| *** dmorita has quit IRC | 19:13 | |
| hurricanerix | ahh ok, just wanted to make sure | 19:15 |
| *** dmorita has joined #openstack-swift | 19:15 | |
| openstackgerrit | Tushar Gohad proposed openstack/swift: Bump PyECLib requirement to >=1.0.7 https://review.openstack.org/244287 | 19:18 |
| *** dmorita has quit IRC | 19:19 | |
| *** dmorita has joined #openstack-swift | 19:19 | |
| notmyname | hurricanerix: I set it to wishlist. thanks for submitting it! | 19:28 |
| hurricanerix | notmyname thanks! | 19:28 |
| *** diazjf has joined #openstack-swift | 19:29 | |
| *** dmorita_ has joined #openstack-swift | 19:32 | |
| *** dmorita has quit IRC | 19:32 | |
| *** dmorita_ has quit IRC | 19:40 | |
| *** dmorita has joined #openstack-swift | 19:45 | |
| *** gyee has quit IRC | 19:46 | |
| *** dmorita has quit IRC | 19:47 | |
| *** dmorita has joined #openstack-swift | 19:50 | |
| *** eranrom has quit IRC | 19:50 | |
| *** eranrom has joined #openstack-swift | 19:50 | |
| *** [1]eranrom has joined #openstack-swift | 19:51 | |
| *** eranrom has quit IRC | 19:51 | |
| *** [1]eranrom is now known as eranrom | 19:51 | |
| *** dmsimard|afk is now known as dmsimard | 19:52 | |
| *** dmorita has quit IRC | 19:54 | |
| *** dmorita has joined #openstack-swift | 19:55 | |
| *** chsc has joined #openstack-swift | 19:56 | |
| *** chlong has quit IRC | 19:57 | |
| *** petertr7 is now known as petertr7_away | 19:59 | |
| *** petertr7_away is now known as petertr7 | 20:03 | |
| *** chlong has joined #openstack-swift | 20:16 | |
| *** diogogmt has quit IRC | 20:24 | |
| *** bill_az has joined #openstack-swift | 20:25 | |
| *** diogogmt has joined #openstack-swift | 20:25 | |
| *** changbl has quit IRC | 20:26 | |
| clayg | notmyname: I hadn't really considered that adding new bugs as wishlist would require some launchpad permission - that's an annoying overhead :'( | 20:31 |
| notmyname | yeah | 20:31 |
| notmyname | could use tags | 20:31 |
| notmyname | or we could say that core (who does have permission) should keep on top of it ;-) | 20:32 |
| *** NM has quit IRC | 20:33 | |
| clayg | notmyname: if we can't fix launchpad rules to allow that priority to be set (maybe ask thierry) maybe ask people to file with tag and look at maybe a script that sets "status==new and has_tag(wishlist)" => "status=wishlist" | 20:34 |
| clayg | but yeah we should have a link to launchpad bugs with status==new that we always try to keep clean | 20:35 |
| notmyname | yeah, I had the same thought (script that changes tag wishlist to priority wishlist) | 20:35 |
| *** dmorita has quit IRC | 20:35 | |
| *** dmorita has joined #openstack-swift | 20:35 | |
| *** acoles_ is now known as acoles | 20:36 | |
| *** silor has quit IRC | 20:37 | |
| *** diazjf has quit IRC | 20:38 | |
| clayg | acoles: i'd like to get ahold of the replica2part2dev tables that are hitting the "zero weight device won't fully drain" problem | 20:42 |
| clayg | acoles: or ask locrin if he could to a min_part_hours/rebalance cycle a few times and find out if they *eventually* drain - or if they just will not shed those parts (and how many parts!?) until you remove the device? | 20:43 |
| *** changbl has joined #openstack-swift | 20:44 | |
| acoles | clayg: sure. you saw he posted a link to all his results, right? | 20:44 |
| clayg | acoles: yes that's what i'm looking at | 20:45 |
| acoles | clayg: the small cluster ring data we can share i think | 20:45 |
| acoles | and that has one of those results | 20:45 |
| acoles | clayg: i'll ask Lorcan to do both - re-run test and share the data | 20:45 |
| acoles | cycle i mean, as you say | 20:46 |
| clayg | well if I can get a small ring in the state of test4 I think i'm golden - it's quite possible it'd be trivial to convert his notes to a ring_analyizer file | 20:46 |
| clayg | i might start there - just tell him thanks | 20:47 |
| acoles | clayg: he's not on irc, and i am ooo tomorrow so i'll introduce you via email | 20:47 |
| *** openstackgerrit has quit IRC | 20:47 | |
| *** openstackgerrit has joined #openstack-swift | 20:47 | |
| *** zul has quit IRC | 20:48 | |
| *** gyee has joined #openstack-swift | 20:57 | |
| *** gyee has quit IRC | 20:57 | |
| acoles | clayg: you still there? | 20:59 |
| clayg | acoles: sure | 20:59 |
| acoles | clayg: i've been ack on fast-post today, to sort out the inconsistent hashing of suffix dirs... | 21:00 |
| clayg | acoles: baddass! | 21:00 |
| acoles | clayg: if I change the hashng algorithm to use timestamps, rather than filenames (wot you said i believe) , i am trying to convince myself that would not cause a replication storm during upgrade? | 21:01 |
| clayg | acoles: i asked torgomatic one time and he also thought our best bet was to just make sure that the hashing is consistent (more than the file system) and for bonus points would come up with the same answer if the fs only had single .data file in the suffix dir | 21:02 |
| clayg | acoles: it may not cause tons of replication (it would *while* the nodes are upgrading) - but it would *definately* cause a bunch of hash recalculation. | 21:02 |
| *** lpabon has joined #openstack-swift | 21:02 | |
| acoles | clayg: yeah, thats my plan, and that what i thought/hoped - recalcs everywhere but not replication traffic storm | 21:03 |
| clayg | acoles: is there anyway that if a directory doesn't have a ctype time that doesn't match it's data/meta ts (like is always the case toay) we could just update the new shizzy to come up with the same answer the current code would? | 21:03 |
| clayg | i'm sure it would be a cludge - but it would be worth it not to melt all swift clusters when the upgrade? | 21:03 |
| clayg | well... but recalcs are balls :\ | 21:04 |
| *** dmorita has quit IRC | 21:04 | |
| *** dmorita has joined #openstack-swift | 21:04 | |
| clayg | i mean it'd only happen on the parts that get updated I guess - so maybe it just rolls its way along due to normal activity as it always would | 21:04 |
| *** gyee has joined #openstack-swift | 21:04 | |
| clayg | be nice to not have to have the cludge i 'spose - and it depends on how long it takes to upgrade | 21:05 |
| clayg | doing a bunch of sync work that doesn't ship any data just because the two nodes can't come up with the same hash for the same data is annoying | 21:05 |
| acoles | clayg: yup, that would be a goal - ie ctype time is somehow handled seamlessly, I think thats do-able. | 21:05 |
| *** rcernin has quit IRC | 21:05 | |
| clayg | acoles: so you think you can have a hash function that would give the same answer for a .data .meta as the old code *unless* it has a ctype_ts it needs to slip in!? | 21:06 |
| acoles | clayg: oh wait i just caught up with you | 21:06 |
| *** delatte has quit IRC | 21:07 | |
| acoles | hmmm, interesting. so if no distinct ctype timestamp, use filenames, else use new shiny timestamp based hashing. | 21:07 |
| clayg | well - anyway - it's a bonus goal - we could test what happens if we just up and change the hash recalculation full tilt - but it seems scary | 21:07 |
| clayg | acoles: maybe, or like *use* timestamps but convert them to a "filename looking timestamp" idk | 21:08 |
| acoles | of course, EC already uses timestamps. | 21:08 |
| clayg | acoles: wtg EC! | 21:08 |
| acoles | just not the ctype one | 21:08 |
| clayg | acoles: stupid EC! | 21:10 |
| acoles | OK, good suggestions, thanks. now its just a matter of writing the code :) I banged something together today, mainly to inform a test strategy for it, but this^^ will be better | 21:10 |
| clayg | torgomatic: acoles says you're smart | 21:10 |
| * acoles is always stating the obvious ;) | 21:11 | |
| clayg | acoles: FWIW my attempt to trivially reproduce Lorcan's results with ring-builder-analyzer did not fly | 21:11 |
| acoles | clayg: i sent email. he is in Ireland so same timezone as me. | 21:11 |
| acoles | I'll be there too next week | 21:12 |
| acoles | clayg: ttfn | 21:14 |
| *** acoles is now known as acoles_ | 21:15 | |
| *** petertr7 is now known as petertr7_away | 21:16 | |
| *** breitz has quit IRC | 21:18 | |
| *** breitz has joined #openstack-swift | 21:18 | |
| *** klrmn1 has quit IRC | 21:21 | |
| *** petertr7_away is now known as petertr7 | 21:22 | |
| *** klrmn1 has joined #openstack-swift | 21:23 | |
| mattoliverau | Morning | 21:24 |
| *** dustins has quit IRC | 21:26 | |
| *** changbl has quit IRC | 21:34 | |
| *** cdelatte has joined #openstack-swift | 21:38 | |
| *** petertr7 is now known as petertr7_away | 21:39 | |
| *** siva_krishnan has quit IRC | 21:44 | |
| *** petertr7_away is now known as petertr7 | 21:46 | |
| *** trifon has quit IRC | 21:47 | |
| *** thumpba_ has quit IRC | 21:47 | |
| *** changbl has joined #openstack-swift | 21:47 | |
| *** sivasath has joined #openstack-swift | 21:49 | |
| *** ChubYann has joined #openstack-swift | 21:58 | |
| *** sivasath has quit IRC | 22:02 | |
| *** siva_krishnan has joined #openstack-swift | 22:03 | |
| tsg_ | notmyname, clayg: https://review.openstack.org/#/c/244287/ looking good! waiting for your OK | 22:08 |
| *** tsg_ is now known as tsg | 22:08 | |
| notmyname | tsg: done | 22:08 |
| *** siva_krishnan has left #openstack-swift | 22:08 | |
| notmyname | tsg: looks like something's up with the qa cluster | 22:09 |
| tsg | notmyname: yeah "Fatal error: sudo() received nonzero return code 1 while executing! | 22:09 |
| tsg | " | 22:09 |
| *** siva_krishnan has joined #openstack-swift | 22:09 | |
| tsg | deploy_swift step is erroring out | 22:09 |
| tsg | notmyname: does the qa cluster have requirements preinstalled? | 22:11 |
| tsg | notmyname: ie, are they not installed/refreshed on every test run? | 22:11 |
| notmyname | yeah, I want to ask charz about that | 22:12 |
| *** siva_krishnan has left #openstack-swift | 22:13 | |
| jrichli | I was reviewing patch 252791 that fixes bug https://bugs.launchpad.net/swift/+bug/1514317 | 22:13 |
| openstack | Launchpad bug 1514317 in OpenStack Object Storage (swift) "POST-as-COPY to DLO will compresses manifest" [High,In progress] - Assigned to pei ran wu (wupeiran) | 22:13 |
| patchbot | jrichli: https://review.openstack.org/#/c/252791/ - Fix POST-as-COPY mess up the DLO manifest | 22:13 |
| jrichli | I am trying to reproduce the issue with the manifest file, but i dont understand the issue: what does it mean that the manifest was "compressed"? its an empty file, right? | 22:14 |
| *** siva_krishnan has joined #openstack-swift | 22:14 | |
| *** dmorita has quit IRC | 22:15 | |
| *** dmorita has joined #openstack-swift | 22:18 | |
| *** petertr7 is now known as petertr7_away | 22:24 | |
| * jrichli will check back later | 22:26 | |
| torgomatic | jrichli: probably means that you POST to it and it's no longer a manifest, but the concatenation of the segments | 22:33 |
| * torgomatic is just guessing though | 22:33 | |
| openstackgerrit | Travis Nguyen proposed openstack/swift: skip object versioning tests when allow_versioned_writes = False https://review.openstack.org/249354 | 22:33 |
| zaitcev | Sure looks lke it. | 22:33 |
| hrou | Any different between that and bug 1487791 (both referring to non fast-post on dlo) ? Kinda makes sense given its a GET -> PUT to manifest ? | 22:36 |
| openstack | bug 1487791 in OpenStack Object Storage (swift) "POST to DLO squashes data without fast-POST" [Undecided,New] https://launchpad.net/bugs/1487791 | 22:36 |
| *** lpabon has quit IRC | 22:42 | |
| *** klrmn1 has quit IRC | 22:47 | |
| *** [1]eranrom has joined #openstack-swift | 22:55 | |
| *** eranrom has quit IRC | 22:56 | |
| *** [1]eranrom is now known as eranrom | 22:56 | |
| *** dmorita has quit IRC | 23:00 | |
| *** dmorita has joined #openstack-swift | 23:03 | |
| *** chlong has quit IRC | 23:08 | |
| *** jerrygb has quit IRC | 23:12 | |
| *** sanchitmalhotra has quit IRC | 23:17 | |
| *** sanchitmalhotra has joined #openstack-swift | 23:17 | |
| *** blair has quit IRC | 23:19 | |
| *** badari has quit IRC | 23:30 | |
| *** badari has joined #openstack-swift | 23:34 | |
| *** blmartin has quit IRC | 23:39 | |
| *** chsc has quit IRC | 23:42 | |
| *** km has joined #openstack-swift | 23:42 | |
| *** km is now known as Guest62837 | 23:43 | |
| *** kei_yama has joined #openstack-swift | 23:43 | |
| openstackgerrit | John Dickinson proposed openstack/python-swiftclient: authors and changelog update for 2.6.1 release https://review.openstack.org/253266 | 23:46 |
| notmyname | timburke: ^ | 23:46 |
| *** dmorita has quit IRC | 23:47 | |
| notmyname | once that patch lands, i'll submit the patch to do the 2.6.1 release | 23:48 |
| *** tsg has quit IRC | 23:59 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!