| *** minwoob has quit IRC | 00:05 | |
| *** kota_ has joined #openstack-swift | 00:25 | |
| *** ChanServ sets mode: +v kota_ | 00:25 | |
| * notmyname admits defeat for now on patch 215276 | 00:27 | |
| patchbot | notmyname: https://review.openstack.org/#/c/215276/ - Enable object server to return non-durable data | 00:27 |
|---|---|---|
| *** zhill has quit IRC | 00:29 | |
| kota_ | good morning | 00:30 |
| *** DericHorn-HP has joined #openstack-swift | 00:34 | |
| timburke | good morning kota_! | 00:34 |
| *** dmorita has joined #openstack-swift | 00:34 | |
| kota_ | timburke: hi tim :D | 00:35 |
| *** lyrrad has quit IRC | 00:36 | |
| *** aix has quit IRC | 00:39 | |
| *** aix has joined #openstack-swift | 00:40 | |
| *** m_kazuhiro has joined #openstack-swift | 00:41 | |
| *** amit213 has quit IRC | 00:45 | |
| *** klrmn has quit IRC | 00:45 | |
| *** thurloat_isgone has quit IRC | 00:47 | |
| *** _hrou_ has joined #openstack-swift | 00:47 | |
| *** zhill has joined #openstack-swift | 00:47 | |
| *** DericHorn-HP has quit IRC | 00:49 | |
| *** hrou has quit IRC | 00:51 | |
| *** zhill has quit IRC | 00:52 | |
| *** aix has quit IRC | 00:52 | |
| *** DericHorn-HP has joined #openstack-swift | 00:52 | |
| *** klrmn has joined #openstack-swift | 00:53 | |
| *** thurloat has joined #openstack-swift | 00:53 | |
| *** shakamunyi has joined #openstack-swift | 00:55 | |
| *** aix has joined #openstack-swift | 00:55 | |
| *** marzif has joined #openstack-swift | 00:56 | |
| *** klrmn1 has quit IRC | 01:08 | |
| mattoliverau | kota_: morning | 01:17 |
| mattoliverau | timburke: o/ | 01:17 |
| *** DericHorn-HP has quit IRC | 01:20 | |
| timburke | hi mattoliverau! | 01:28 |
| *** jkugel has joined #openstack-swift | 01:35 | |
| EmilienM | I'm doing puppet integration CI and have issues with tempest tests for Swift | 01:42 |
| EmilienM | "Access was denied to this resource" | 01:42 |
| EmilienM | all logs can be found here: http://logs.openstack.org/50/228450/16/check/gate-puppet-openstack-integration-dsvm-centos7/a8bb16d/ | 01:43 |
| EmilienM | the failure is here: http://logs.openstack.org/50/228450/16/check/gate-puppet-openstack-integration-dsvm-centos7/a8bb16d/console.html#_2015-09-29_01_16_19_397 | 01:46 |
| *** dustins has joined #openstack-swift | 01:47 | |
| *** DericHorn-HP has joined #openstack-swift | 01:48 | |
| *** zhill has joined #openstack-swift | 01:52 | |
| *** DericHorn-HP has quit IRC | 01:56 | |
| *** zhill has quit IRC | 01:57 | |
| *** duobei has joined #openstack-swift | 01:59 | |
| *** DericHorn-HP has joined #openstack-swift | 02:02 | |
| *** DericHorn-HP has quit IRC | 02:11 | |
| *** klrmn1 has joined #openstack-swift | 02:12 | |
| *** zhill has joined #openstack-swift | 02:28 | |
| *** zhill has quit IRC | 02:32 | |
| *** jkugel has quit IRC | 02:33 | |
| *** dmorita has quit IRC | 02:34 | |
| *** garthb has joined #openstack-swift | 02:44 | |
| *** haomaiwang has joined #openstack-swift | 02:51 | |
| *** zhill has joined #openstack-swift | 02:52 | |
| *** silor has joined #openstack-swift | 02:56 | |
| *** zhill has quit IRC | 02:56 | |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-swift | 03:01 | |
| *** zhill has joined #openstack-swift | 03:05 | |
| janonymous | morning | 03:06 |
| *** DericHorn-HP has joined #openstack-swift | 03:14 | |
| *** _hrou_ has quit IRC | 03:19 | |
| *** DericHorn-HP has quit IRC | 03:32 | |
| *** dustins has quit IRC | 03:33 | |
| *** links has joined #openstack-swift | 03:36 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:40 | |
| *** bill_az has quit IRC | 03:44 | |
| *** Kennan2 is now known as Kennnan | 03:45 | |
| *** DericHorn-HP has joined #openstack-swift | 03:46 | |
| *** DericHorn-HP has quit IRC | 03:49 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 03:54 | |
| *** sanchitmalhotra has quit IRC | 03:56 | |
| *** sanchitmalhotra has joined #openstack-swift | 03:59 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** sanchitmalhotra1 has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-swift | 04:01 | |
| *** mahatic has joined #openstack-swift | 04:03 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 04:08 | |
| *** sanchitmalhotra has quit IRC | 04:10 | |
| *** proteusguy has quit IRC | 04:14 | |
| *** proteusguy has joined #openstack-swift | 04:26 | |
| mattoliverau | janonymous: morning (sorry for slow reply been at lunch) | 04:29 |
| *** DericHorn-HP has joined #openstack-swift | 04:31 | |
| *** DericHorn-HP has quit IRC | 04:34 | |
| mahatic | mattoliverau: hello, is it noon for you? where do you stay again? :D I know it's Australia | 04:40 |
| *** chsc has joined #openstack-swift | 04:41 | |
| *** DericHorn-HP has joined #openstack-swift | 04:48 | |
| *** chsc has quit IRC | 04:54 | |
| *** DericHorn-HP has quit IRC | 04:56 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** haomaiwang has joined #openstack-swift | 05:01 | |
| *** klrmn1 has quit IRC | 05:06 | |
| mattoliverau | mahatic: well now its 3pm, when you asked it was 2:40pm, I just had a late lunch :) I'm in Melbourne Australia :) | 05:20 |
| mattoliverau | mahatic: oh and o/ | 05:20 |
| mahatic | mattoliverau: ah :) | 05:22 |
| mahatic | mattoliverau: yes o/ :D | 05:22 |
| *** garthb has quit IRC | 05:32 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift: Don't send commits for quorum *BAD* requests on EC https://review.openstack.org/220059 | 05:33 |
| *** sileht has quit IRC | 05:33 | |
| kota_ | mattoliverau: no worries | 05:37 |
| *** ppai has joined #openstack-swift | 05:39 | |
| mattoliverau | kota_: first time using the gerrit edit commit message feature, you'd think it'll be smart enough to not nuke votes now that it's editable. | 05:42 |
| *** mahatic has quit IRC | 05:44 | |
| kota_ | mattoliverau: yup, it's easy to understand what changed but hopefully i wish it would keep the votes :/ | 05:46 |
| *** haomaiwang has quit IRC | 06:01 | |
| *** haomaiwang has joined #openstack-swift | 06:01 | |
| *** jamielennox is now known as jamielennox|away | 06:12 | |
| *** openstackgerrit has quit IRC | 06:16 | |
| *** openstackgerrit has joined #openstack-swift | 06:16 | |
| *** jst_ has quit IRC | 06:21 | |
| janonymous | mattoliverau: :) it's morning here :) | 06:21 |
| *** SkyRocknRoll has joined #openstack-swift | 06:24 | |
| *** km has quit IRC | 06:30 | |
| *** km_ has joined #openstack-swift | 06:30 | |
| *** trifon has quit IRC | 06:36 | |
| *** trifon has joined #openstack-swift | 06:49 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** haomaiwang has joined #openstack-swift | 07:01 | |
| *** zhill has quit IRC | 07:17 | |
| *** rledisez has joined #openstack-swift | 07:24 | |
| *** mac_ified has quit IRC | 07:27 | |
| *** SkyRocknRoll has quit IRC | 07:45 | |
| *** geaaru has joined #openstack-swift | 07:48 | |
| *** jordanP has joined #openstack-swift | 07:49 | |
| *** eranrom has joined #openstack-swift | 07:57 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-swift | 08:01 | |
| *** jistr has joined #openstack-swift | 08:16 | |
| *** marzif has quit IRC | 08:18 | |
| *** acoles_ is now known as acoles | 08:42 | |
| acoles | notmyname: its *possible* ;) however, i'd need to tinker with object server dirs to delete .durable files, like clayg says, and i think its just as effective to do that in the unit tests in proxy/test_server.py | 08:48 |
| acoles | notmyname: my main focus this week is to work on the tests on patch 215276, and of course address review comments. | 08:49 |
| patchbot | acoles: https://review.openstack.org/#/c/215276/ - Enable object server to return non-durable data | 08:49 |
| *** Kennnan2 has joined #openstack-swift | 08:54 | |
| *** Kennnan has quit IRC | 08:54 | |
| acoles | kota_: you still around? | 08:55 |
| kota_ | yup | 09:00 |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-swift | 09:01 | |
| kota_ | acoles: ack, I'm working in progress for reviewing 215276. that's awesome but it still needs more time to look at deeply, I'll add comments in today. | 09:05 |
| acoles | kota_: fantastic, thank you, its a big review :/ did you see my question on your patch 220059? | 09:06 |
| patchbot | acoles: https://review.openstack.org/#/c/220059/ - Don't send commits for quorum *BAD* requests on EC | 09:06 |
| kota_ | sorry, I am going to be offline in 30mins or 1 hour for meeting. | 09:06 |
| kota_ | not yet, sorry. | 09:06 |
| acoles | kota_: ok np | 09:06 |
| kota_ | acoles: I hope I will do that asap. | 09:06 |
| *** kota_ has quit IRC | 09:06 | |
| acoles | kota_: i think there is no problem but I just wanted to be sure what you thought before voting. | 09:07 |
| *** silor has quit IRC | 09:07 | |
| *** geaaru has quit IRC | 09:12 | |
| *** nakagawamsa has joined #openstack-swift | 09:18 | |
| *** geaaru has joined #openstack-swift | 09:26 | |
| *** mahatic has joined #openstack-swift | 09:29 | |
| *** kota_ has joined #openstack-swift | 09:48 | |
| *** ChanServ sets mode: +v kota_ | 09:48 | |
| *** nakagawamsa has quit IRC | 09:49 | |
| kota_ | I'm back. | 09:49 |
| *** haomaiwang has quit IRC | 10:01 | |
| acoles | hi kota_ : so my query with patch 220059 was about the change to have_adequate_successes | 10:01 |
| patchbot | acoles: https://review.openstack.org/#/c/220059/ - Don't send commits for quorum *BAD* requests on EC | 10:01 |
| *** haomaiwa_ has joined #openstack-swift | 10:01 | |
| acoles | which you change to include 1xx responses as 'successes' | 10:01 |
| acoles | on master the proxy is waiting for all nodes to respond with 1xx (or timeout) - which may not be what we wanted to happen - and with your change the proxy waits for just 'adequate' 1xx before proceeding with the confirmation phase. | 10:03 |
| acoles | kota_: but it sends the confirmation to *all* nodes. | 10:04 |
| acoles | kota_: i think that's ok?? a node that was not counted in 'adequate successes' may timeout or may return 422 but in either case will ignore the confirmation phase, correct? | 10:05 |
| kota_ | yup | 10:05 |
| acoles | kota_: good! | 10:06 |
| *** haomaiwa_ has quit IRC | 10:07 | |
| *** m_kazuhiro has quit IRC | 10:08 | |
| acoles | kota_: another thing, when i ran test_PUT_ec_fragment_quorum_bad_request i noticed that the proxy tries to send chunks to a node that has responded 400 :/ that's not an issue with your patch but seems odd. | 10:08 |
| kota_ | sorry, got phone call | 10:09 |
| acoles | kota_: heh. busy evening for you! | 10:09 |
| kota_ | so sorry | 10:10 |
| kota_ | back, again. | 10:15 |
| kota_ | I had a crasy busy begining of this week :/ | 10:16 |
| kota_ | so... | 10:16 |
| *** T0m_ has joined #openstack-swift | 10:18 | |
| *** T0m_ has left #openstack-swift | 10:18 | |
| kota_ | oh, yeah, you're right. if we change the 1xx in _have_adequate_successes, get_put_response waits just quorum 1xx. | 10:24 |
| kota_ | but that's out side of my intention, i have to reconsider the effect, now. | 10:25 |
| kota_ | In my intention, just reusing existing method for my purpose. | 10:25 |
| kota_ | acoles: very nice catch for that. | 10:29 |
| acoles | kota_: i'm not sure it is a problem: once we have ec_ndata+1 1xx responses then go ahead with the confirmation phase. | 10:35 |
| *** km_ has quit IRC | 10:36 | |
| acoles | kota_: maybe the method name is a little unclear - "have_adequate_successes" == have adequate success or informational | 10:36 |
| acoles | kota_: but you have though about this more than I have so I'll wait for your conclusion | 10:37 |
| kota_ | acoles: thanks for notify me that, exactly, have_adequate_success should do about "success" responses not "informational" imo. | 10:39 |
| kota_ | currently I'm thinking another idea to get apply functions like partial, but not sure, now making sure for the feasibility. | 10:40 |
| kota_ | we might be able to send commit with quourum 1xx but i'm not sure too. | 10:45 |
| kota_ | thinking bad cases. | 10:45 |
| acoles | kota_: k. its getting late for you, if you make another version i can review tomorrow. | 10:45 |
| *** duobei has quit IRC | 10:47 | |
| kota_ | acoles: thanks, maybe I have to lean more the HTTP sequence because if we don't pull the 1xx response from the rest of putters, can we handle the final responses? | 10:47 |
| kota_ | i'm not sure. | 10:47 |
| kota_ | after sending commits. | 10:47 |
| *** jamielennox|away is now known as jamielennox | 10:49 | |
| *** silor has joined #openstack-swift | 10:50 | |
| *** SkyRocknRoll has joined #openstack-swift | 10:51 | |
| kota_ | acoles: anyways, thanks a lot! :D | 10:53 |
| acoles | kota_: hmmm, good question. yeah, if the proxy does not wait to read the 1xx response from a node then does it get the 2xx response after the commit? idk, sorry! it *might* be ok because if you look at conn.await_response it does a different thing for final_phase. | 10:53 |
| acoles | kota_: np. good luck! | 10:53 |
| *** silor has quit IRC | 10:54 | |
| *** silor has joined #openstack-swift | 10:54 | |
| *** 18VAARFFB has joined #openstack-swift | 10:56 | |
| *** silor has quit IRC | 10:59 | |
| *** silor1 has joined #openstack-swift | 10:59 | |
| *** 18VAARFFB has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-swift | 11:01 | |
| *** silor1 is now known as silor | 11:02 | |
| *** ppai has quit IRC | 11:04 | |
| *** haomaiwang has quit IRC | 11:11 | |
| *** ppai has joined #openstack-swift | 11:14 | |
| *** duobei has joined #openstack-swift | 11:20 | |
| *** delattec has quit IRC | 11:21 | |
| *** marzif has joined #openstack-swift | 11:28 | |
| *** ccavanna has quit IRC | 11:30 | |
| *** vinsh has joined #openstack-swift | 11:31 | |
| *** satyadevSVN has joined #openstack-swift | 11:35 | |
| *** satyadevSVN has quit IRC | 11:41 | |
| *** delattec has joined #openstack-swift | 11:43 | |
| *** delattec has quit IRC | 11:50 | |
| *** cdelatte has joined #openstack-swift | 11:53 | |
| *** delattec has joined #openstack-swift | 11:53 | |
| *** kei_yama has quit IRC | 12:09 | |
| *** mandarine has joined #openstack-swift | 12:19 | |
| mandarine | hello hello hello, it's been a while :) | 12:20 |
| mandarine | It appear I run once again into a delicate problem of swift generating "useless" load on deleted containers db files. | 12:20 |
| *** janonymous_ has joined #openstack-swift | 12:21 | |
| mandarine | In order to explain more: when my containers are deleted, the .db file on the disk stays and doesn't seem to be deleted any time soon by the container-replicator or by the container-reconciler. That would be all but some container related processes are touching those silly files and make my poor server use CPU on potentially useless operations. | 12:23 |
| *** delatte has joined #openstack-swift | 12:24 | |
| mandarine | Do any of you have an idea on what might happen or any pointer on what I ought to do ? | 12:24 |
| *** cdelatte has quit IRC | 12:27 | |
| *** delattec has quit IRC | 12:27 | |
| eranrom | mandarine,How mush time passed since you deleted the container via REST? | 12:28 |
| mandarine | eranrom: months | 12:28 |
| eranrom | mandarine, ok, I did not expect that much time... | 12:29 |
| mandarine | :| | 12:30 |
| eranrom | mandarine: So I assme you are aware of the reclaim_age | 12:30 |
| mandarine | Yup | 12:30 |
| mandarine | It does not seem to work | 12:30 |
| eranrom | mandarine: I am afraid then that best wuold be to open a bug, esp. if you can reproduce. Did you go over any ring changes? | 12:32 |
| mandarine | no ring change, nope :) | 12:32 |
| *** kairo has joined #openstack-swift | 12:32 | |
| mandarine | I higly doubt it's a bug. It looks more like something in my configuration or something I may have missed | 12:32 |
| eranrom | mandarine: Is that a production env? | 12:33 |
| mandarine | Yes it is | 12:33 |
| *** NM has joined #openstack-swift | 12:34 | |
| eranrom | Swift version? | 12:35 |
| mandarine | 2.2.0-1 | 12:36 |
| *** ppai has quit IRC | 12:36 | |
| *** amoturi has joined #openstack-swift | 12:36 | |
| *** aix has quit IRC | 12:37 | |
| *** mahatic has quit IRC | 12:38 | |
| *** mahatic has joined #openstack-swift | 12:40 | |
| eranrom | pardon the question, but I guess that heading the container returns 404 - right? | 12:44 |
| mandarine | probably. let me check | 12:47 |
| mandarine | Yes it does | 12:47 |
| mandarine | The file is still there and I can do a "swift-container-info" on it to see an amazing "Delete Timestamp" field in the past. | 12:50 |
| eranrom | right. Can you paste the other container info you see? | 12:51 |
| *** alejandrito has joined #openstack-swift | 12:52 | |
| eranrom | esp. do you have a "'put_timestamp" there? | 12:53 |
| mandarine | On my dev platform, I put the reclaim_age to 30. | 12:53 |
| mandarine | I do | 12:53 |
| *** SkyRocknRoll has quit IRC | 12:53 | |
| eranrom | is it also in the past? | 12:53 |
| mandarine | Yes | 12:53 |
| eranrom | which one is greater? | 12:54 |
| mandarine | The Delete Timestamp | 12:54 |
| *** breitz has quit IRC | 12:54 | |
| *** breitz has joined #openstack-swift | 12:54 | |
| mandarine | Put at 1443528214.63302 and Delete at 1443528345.32847 | 12:54 |
| mandarine | (again, on my dev tests) | 12:55 |
| eranrom | where the container does not get deleted (that is on your dev tests machine) | 12:56 |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: Don't send commits for quorum *BAD* requests on EC https://review.openstack.org/220059 | 12:57 |
| mandarine | in the filesystem : /srv/node/2/containers/58/388/01d0c60670da13e694796ad6fc108388/01d0c60670da13e694796ad6fc108388.db | 12:57 |
| mandarine | That file should be deleted | 12:57 |
| mandarine | And is not | 12:57 |
| kota_ | acoles: I updated patch set but I'm tired so I'll continue to review yours ;) | 12:57 |
| *** kota_ has quit IRC | 12:57 | |
| *** jrichli has joined #openstack-swift | 13:02 | |
| eranrom | ok, trying another direction: Do you see a .lock file in the directory? (/srv/node/2/containers/58/388/01d0c60670da13e694796ad6fc108388/) | 13:02 |
| *** jkugel has joined #openstack-swift | 13:03 | |
| acoles | kota: have a good night | 13:05 |
| *** hrou has joined #openstack-swift | 13:07 | |
| *** daemontool_ has joined #openstack-swift | 13:10 | |
| *** amoturi has quit IRC | 13:10 | |
| *** DericHorn-HP has joined #openstack-swift | 13:12 | |
| *** delatte has quit IRC | 13:13 | |
| *** CR7 has joined #openstack-swift | 13:13 | |
| *** eranrom has quit IRC | 13:15 | |
| mandarine | Damn | 13:17 |
| mandarine | Yes, there is a .lock file | 13:17 |
| mandarine | (I'll wait for the ping to untimeout :p) | 13:18 |
| *** satyadevSVN has joined #openstack-swift | 13:20 | |
| *** satyadevSVN has quit IRC | 13:20 | |
| *** dustins has joined #openstack-swift | 13:21 | |
| *** amoturi has joined #openstack-swift | 13:21 | |
| *** links has quit IRC | 13:22 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 13:25 |
| *** bill_az has joined #openstack-swift | 13:25 | |
| *** alejandrito has quit IRC | 13:28 | |
| *** ctrath has joined #openstack-swift | 13:31 | |
| *** trifon has quit IRC | 13:35 | |
| *** haomaiwang has joined #openstack-swift | 13:36 | |
| *** alejandrito has joined #openstack-swift | 13:40 | |
| *** haomaiwang has quit IRC | 13:40 | |
| *** aix has joined #openstack-swift | 13:40 | |
| *** haomaiwa_ has joined #openstack-swift | 13:40 | |
| *** CR7 has quit IRC | 13:41 | |
| *** cdelatte has joined #openstack-swift | 13:46 | |
| *** delattec has joined #openstack-swift | 13:46 | |
| *** marzif has quit IRC | 13:49 | |
| *** wbhuber has joined #openstack-swift | 13:50 | |
| *** ccavanna has joined #openstack-swift | 13:50 | |
| *** DericHorn-HP has quit IRC | 13:50 | |
| *** DericHorn-HP has joined #openstack-swift | 13:52 | |
| *** pberis has joined #openstack-swift | 13:53 | |
| *** haomaiwa_ has quit IRC | 14:01 | |
| *** haomaiwa_ has joined #openstack-swift | 14:01 | |
| *** lcurtis has joined #openstack-swift | 14:02 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:02 | |
| *** chsc has joined #openstack-swift | 14:03 | |
| *** slo_ has joined #openstack-swift | 14:09 | |
| *** DericHorn-HP has quit IRC | 14:10 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 14:20 |
| *** jlhinson has joined #openstack-swift | 14:24 | |
| *** minwoob has joined #openstack-swift | 14:28 | |
| *** juzuluag has joined #openstack-swift | 14:37 | |
| *** MVenesio has joined #openstack-swift | 14:38 | |
| *** duobei has quit IRC | 14:51 | |
| *** zaitcev has joined #openstack-swift | 14:56 | |
| *** ChanServ sets mode: +v zaitcev | 14:56 | |
| *** seandrummond has joined #openstack-swift | 15:00 | |
| *** oddsam91 has joined #openstack-swift | 15:00 | |
| *** haomaiwa_ has quit IRC | 15:01 | |
| *** seandrummond has left #openstack-swift | 15:01 | |
| *** haomaiwang has joined #openstack-swift | 15:01 | |
| *** oddsam91 has left #openstack-swift | 15:07 | |
| *** kairo has left #openstack-swift | 15:14 | |
| *** haomai___ has joined #openstack-swift | 15:16 | |
| *** haomaiwang has quit IRC | 15:18 | |
| *** DericHorn-HP has joined #openstack-swift | 15:31 | |
| *** haomai___ has quit IRC | 15:35 | |
| *** Fin1te has joined #openstack-swift | 15:39 | |
| *** marzif has joined #openstack-swift | 15:54 | |
| *** lyrrad has joined #openstack-swift | 16:01 | |
| *** sanchitmalhotra1 has quit IRC | 16:03 | |
| *** gyee has joined #openstack-swift | 16:04 | |
| *** marzif has quit IRC | 16:05 | |
| *** trifon has joined #openstack-swift | 16:14 | |
| *** fthiagogv has joined #openstack-swift | 16:14 | |
| *** fthiagogv_ has joined #openstack-swift | 16:14 | |
| *** Fin1te has quit IRC | 16:15 | |
| *** fthiagogv has quit IRC | 16:16 | |
| *** gyee has quit IRC | 16:19 | |
| *** jistr has quit IRC | 16:26 | |
| *** gyee has joined #openstack-swift | 16:29 | |
| *** pberis1 has joined #openstack-swift | 16:35 | |
| *** pberis has quit IRC | 16:36 | |
| *** pberis1 is now known as pberis | 16:36 | |
| *** rledisez has quit IRC | 16:36 | |
| *** pberis1 has joined #openstack-swift | 16:40 | |
| *** pberis has quit IRC | 16:40 | |
| *** pberis1 is now known as pberis | 16:40 | |
| *** klrmn1 has joined #openstack-swift | 16:44 | |
| *** blmartin has joined #openstack-swift | 16:51 | |
| *** zhill has joined #openstack-swift | 16:52 | |
| *** jordanP has quit IRC | 16:55 | |
| *** ctrath has quit IRC | 16:59 | |
| *** garthb has joined #openstack-swift | 17:00 | |
| *** cazino has joined #openstack-swift | 17:03 | |
| *** pgbridge has quit IRC | 17:09 | |
| *** janonymous_ has quit IRC | 17:11 | |
| openstackgerrit | David Goetz proposed openstack/swift: Add concurrent reads option to proxy https://review.openstack.org/117710 | 17:14 |
| *** ctrath has joined #openstack-swift | 17:17 | |
| *** marzif has joined #openstack-swift | 17:19 | |
| notmyname | good morning | 17:21 |
| *** Fin1te has joined #openstack-swift | 17:23 | |
| *** garthb has quit IRC | 17:27 | |
| *** garthb has joined #openstack-swift | 17:28 | |
| openstackgerrit | Merged openstack/swift: Fix inlines for test/unit/obj/test_server.py https://review.openstack.org/228442 | 17:29 |
| *** trifon has quit IRC | 17:38 | |
| *** sileht_ has joined #openstack-swift | 17:44 | |
| *** sileht_ is now known as sileht | 17:45 | |
| *** sileht has quit IRC | 17:45 | |
| *** sileht has joined #openstack-swift | 17:45 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 17:57 |
| *** ctrath has quit IRC | 18:01 | |
| acoles | timburke: thanks for reviewing that crypto patch | 18:02 |
| acoles | notmyname: good morning | 18:02 |
| hrou | timburke, +1 You made a lot of interesting points (particularly about the encoding), thanks! | 18:03 |
| *** ctrath has joined #openstack-swift | 18:04 | |
| timburke | acoles: happy to help! i'd starred it a while ago but hadn't gotten around to reviewing it. then the StopIteration stuff made me think of patch 227569, and by the time i was convinced that we *weren't* going to run into similar issues, i felt like i ought to just actually leave a review :) | 18:05 |
| patchbot | timburke: https://review.openstack.org/#/c/227569/ - Only yield the pending segment on error if it's a ... | 18:05 |
| *** ctrath has left #openstack-swift | 18:06 | |
| acoles | timburke: yeah i want to go _understand_ that patch ^^ looks like a really interesting catch | 18:07 |
| * acoles thinks unit tests run slower when he's hungry :( | 18:08 | |
| *** marzif has quit IRC | 18:09 | |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 18:09 |
| *** delatte has joined #openstack-swift | 18:09 | |
| *** delattec has quit IRC | 18:09 | |
| *** cdelatte has quit IRC | 18:09 | |
| *** acoles is now known as acoles_ | 18:11 | |
| *** mahatic has quit IRC | 18:12 | |
| timburke | acoles: it was weird. i'm not event entirely convinced that there was an issue that could affect a client. but my takeaway was "never yield from inside a finally" | 18:12 |
| clayg | weeee | 18:14 |
| hrou | timburke, at some point I didn't even think you can use a yield in a try/catch/finally (never mind yielding inside a finally)? Maybe that was changed in 2.6 ? | 18:14 |
| *** SkyRocknRoll has quit IRC | 18:16 | |
| notmyname | I wonder if we can get patch 220059 landed today | 18:19 |
| patchbot | notmyname: https://review.openstack.org/#/c/220059/ - Don't send commits for quorum *BAD* requests on EC | 18:19 |
| notmyname | hrou: https://gist.github.com/notmyname/bed03c9796bc9dcada1b | 18:22 |
| hrou | notmyname, yep I noticed when you pasted that and tried it my self : ) And it indeed throws the generator error on the break. | 18:22 |
| *** geaaru has quit IRC | 18:23 | |
| hrou | notmyname, there's some backstory https://www.python.org/dev/peps/pep-0342/ and https://bugs.python.org/issue22988 | 18:25 |
| *** delatte has quit IRC | 18:26 | |
| *** delatte has joined #openstack-swift | 18:29 | |
| *** delatte has quit IRC | 18:30 | |
| *** SkyRocknRoll has joined #openstack-swift | 18:31 | |
| *** SkyRocknRoll has quit IRC | 18:34 | |
| *** SkyRocknRoll has joined #openstack-swift | 18:34 | |
| *** trifon has joined #openstack-swift | 18:35 | |
| *** 32NAAGPES has joined #openstack-swift | 18:35 | |
| *** cdelatte has joined #openstack-swift | 18:35 | |
| *** haypo has joined #openstack-swift | 18:37 | |
| *** aix has quit IRC | 18:38 | |
| *** flwang has quit IRC | 18:38 | |
| notmyname | and it seems there's some follow-up needed for https://review.openstack.org/#/c/227133/ by cschwede | 18:38 |
| notmyname | mattoliverau: ^ in the gerrit comment, you said you were working on testing it with keystone. any results there yet? | 18:38 |
| *** flwang has joined #openstack-swift | 18:38 | |
| haypo | hi. what is the status of EC in swift? does it work now? :) | 18:40 |
| notmyname | haypo: we're working on some known edge cases now. I'd certainly recommend it for testing right now. I wouldn't yet recommend it for prod data | 18:41 |
| notmyname | we'll see what gets done by friday | 18:41 |
| notmyname | and of course it "works" now. it's "worked" for a long time. it's the last edge cases to properly handle certain failure scenarios that we're finishing up | 18:42 |
| haypo | notmyname: hum, so it will require one more version to be fully stable? | 18:44 |
| notmyname | haypo: I'm not sure what you're asking | 18:48 |
| haypo | notmyname: is EC announced as experimental/not ready for the production, in the doc? | 18:49 |
| notmyname | http://docs.openstack.org/developer/swift/overview_erasure_code.html | 18:49 |
| haypo | notmyname: so the "Beta: Not production ready" warning is kept, ok | 18:49 |
| notmyname | haypo: note that we haven't yet done our liberty release | 18:49 |
| *** eranrom has joined #openstack-swift | 18:52 | |
| *** SkyRocknRoll has quit IRC | 18:55 | |
| *** slo_ has quit IRC | 18:57 | |
| *** slo_ has joined #openstack-swift | 18:58 | |
| minwoob | acoles_: I noticed that you were able to get around the xattr copy issue by avoiding the copy in the first place ... I haven't found a copy solution that works, but I found some sources stating that the copy methods in shutil incorporate xattrs in going forward in py3. | 19:01 |
| *** DericHorn-HP has quit IRC | 19:01 | |
| *** DericHorn-HP has joined #openstack-swift | 19:02 | |
| *** ccavanna has quit IRC | 19:07 | |
| notmyname | torgomatic: did probe tests for https://review.openstack.org/#/c/220059/6 work for you? | 19:11 |
| torgomatic | notmyname: I didn't run them for 6; let me do that now | 19:12 |
| notmyname | I'm running them against master now to compare | 19:12 |
| zaitcev | I spent whole day today building a package after package for CVE-2015-5223 openstack-swift: Information leak via Swift tempurls | 19:13 |
| zaitcev | Feel like Charlie Chaplin in the factory. So many identical packages. F22, OSP5 RHEL6, OSP5 RHEL7, OSP6 | 19:14 |
| notmyname | zaitcev: thanks | 19:14 |
| notmyname | I'm surprised mitre still hasn't been updated http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2015-5223 | 19:14 |
| *** juzuluag has quit IRC | 19:15 | |
| zaitcev | BTW, it backports to Icehouse trivially | 19:15 |
| zaitcev | But the upstream somehow got rid of that branch, it appears. | 19:16 |
| notmyname | zaitcev: yeah, after a branch isn't supported any more (ie after 2 releases) -infra deletes those old stable releases | 19:16 |
| zaitcev | meeeh | 19:17 |
| zaitcev | Oh well. I guess this is what I get for cheating death again today. | 19:17 |
| zaitcev | here's a screencap from FlightAware https://twitter.com/ani_nouto/status/648923678630133760 | 19:18 |
| notmyname | zaitcev: you do packaging, so you get to cheat death fairly often. or, alternatively, you've already sold your soul to do packaging, so Death isn't too interested ;-) | 19:18 |
| notmyname | zaitcev: wow | 19:19 |
| notmyname | also, you're up in NY now? | 19:19 |
| zaitcev | In hotel in Rochester | 19:19 |
| zaitcev | Landed aroudn 9 a.m. and have a meeting with a guy around 5 p.m., so thought I could do somethiing for Swift for a change. It was all Ceph all the time for weeks now. | 19:20 |
| zaitcev | It's a nightmare. So many classes and redundant variables each set in mysterious ways and interacting with each other. I've seen more mantainable code in assembly. | 19:21 |
| notmyname | yikes | 19:21 |
| notmyname | zaitcev: in the swift world, if you haven't caught up yet, we're looking at these patches this week: https://wiki.openstack.org/wiki/Swift/PriorityReviews | 19:22 |
| *** silor has quit IRC | 19:26 | |
| vinsh | So... when configuring erasure coding.. are the account/container rings still a standard replication type? Do I need to assign a storage policy to them? or does the storage policy only apply to an object ring? | 19:26 |
| notmyname | vinsh: storage policies are only of object rings | 19:26 |
| vinsh | Ok cool. So swift will treat the account/container rings as they are built to be. | 19:26 |
| *** silor has joined #openstack-swift | 19:27 | |
| notmyname | right | 19:27 |
| zaitcev | Rings, yes | 19:27 |
| vinsh | notmyname: Got the erasure code trial hardware in finally. Starting the build. | 19:27 |
| notmyname | great | 19:27 |
| vinsh | keeping account/container on the proxy nodes. | 19:28 |
| vinsh | on NVMe drives :D | 19:28 |
| notmyname | cool | 19:28 |
| vinsh | Are zones treated specially with EC? | 19:30 |
| notmyname | no | 19:30 |
| vinsh | Huhm. I wonder how to decide on zones then. | 19:32 |
| zaitcev | It's orthogonal in a sense that zones are taken into account by the builder when it scatters partitions. It walks whone address space and for each #, asks, where is the best place? Is this zone good? This nore? etc. | 19:32 |
| vinsh | ah ok. good.. was thinking something like that. | 19:32 |
| notmyname | right. often times a zone == a rack | 19:32 |
| zaitcev | But EC is all in proxy and reconstructors. They load the map the builder has done and don't even care how it came up with the layout. | 19:32 |
| *** silor1 has joined #openstack-swift | 19:33 | |
| *** silor has quit IRC | 19:33 | |
| *** silor1 is now known as silor | 19:33 | |
| vinsh | maybe some choice of ec of 10:3.. given x nodes, split to y zones. | 19:34 |
| vinsh | something to further enforce a spread of chunks. | 19:34 |
| * vinsh will trust the builder. | 19:35 | |
| zaitcev | Builder can't know what racks are where in what datacenters, what busses they share. That's why you specify zones by hand. | 19:36 |
| vinsh | Which makes sense to me with replicas | 19:36 |
| vinsh | Just grocking what it means with ec | 19:37 |
| zaitcev | I think... it's like having 13 replicas | 19:37 |
| zaitcev | placement-wise | 19:37 |
| notmyname | yup | 19:37 |
| vinsh | where some number is needed to reconstruct a whole | 19:37 |
| zaitcev | no, it's the total m. you only need 10 to reconstruct | 19:38 |
| zaitcev | but you (the builder) have to place all 13 | 19:38 |
| *** gyee has quit IRC | 19:38 | |
| *** sileht has quit IRC | 19:39 | |
| *** pgbridge has joined #openstack-swift | 19:40 | |
| *** pberis has quit IRC | 19:40 | |
| zaitcev | I have 40 minutes... Feel like can't give Kota justice in 220059 but I can post some senseless comments | 19:42 |
| *** sileht has joined #openstack-swift | 19:45 | |
| *** sileht has quit IRC | 19:47 | |
| *** sileht has joined #openstack-swift | 19:48 | |
| *** pberis has joined #openstack-swift | 19:50 | |
| *** earlephilhower has quit IRC | 20:01 | |
| *** csmart has quit IRC | 20:32 | |
| *** csmart has joined #openstack-swift | 20:34 | |
| *** DericHorn-HP has quit IRC | 20:35 | |
| openstackgerrit | Michael Barton proposed openstack/swift: go: CopyN that re-uses buffers https://review.openstack.org/229141 | 20:36 |
| *** cdelatte has quit IRC | 20:40 | |
| *** 32NAAGPES has quit IRC | 20:40 | |
| *** zaitcev has quit IRC | 20:44 | |
| *** pberis has quit IRC | 20:52 | |
| *** pberis has joined #openstack-swift | 20:53 | |
| *** blmartin has quit IRC | 21:01 | |
| *** amoturi has left #openstack-swift | 21:03 | |
| *** silor has quit IRC | 21:04 | |
| briancline | are there any known issues that could cause questionable functest failures for test_object.TestObject when it creates its containers in setUp? | 21:11 |
| briancline | i ask because i'm seeing test failures when it gets a 202 rather than a 201.. seems to be issuing two successive puts, but it's only expecting a 201 | 21:12 |
| layzee__ | how does one run ssbench targetted against a specific swift node ? | 21:14 |
| *** layzee__ is now known as layzee | 21:14 | |
| layzee | i was looking for an option like --- ssbench-master run-scenario --targetserver=x.y.z.u | 21:15 |
| vinsh | target a specific proxy you mean? | 21:15 |
| layzee | yes | 21:15 |
| layzee | or do you run this locally only on the proxy node ? | 21:15 |
| vinsh | ah, there is a flag... lemme find my notes | 21:16 |
| vinsh | I run it from a powerful benchmark host | 21:16 |
| layzee | nice | 21:16 |
| vinsh | I target a proxy, or many proxies or a load balancer | 21:16 |
| layzee | ok | 21:16 |
| *** MVenesio has quit IRC | 21:17 | |
| vinsh | layzee: the -S flag | 21:18 |
| vinsh | for example -S http://69.134.198.60:8080/v1/KEY_83796bbb39734fe18e8e975aedfe639a | 21:19 |
| layzee | ha ! thanks. | 21:19 |
| vinsh | I wish you well. Benchmarking with that tool is fun but far from easy(when it comes to interpreting exactly what you are doing) | 21:20 |
| layzee | vinsh: thx | 21:22 |
| *** haypo has left #openstack-swift | 21:23 | |
| eranrom | briancline: I just updated master and got all func tests working. | 21:28 |
| eranrom | briancline: can you point pls to the 2 successive PUTs you mention? | 21:30 |
| eranrom | briancline: if these are the 2 create containers in setUP then it seems they are done for different accounts | 21:35 |
| *** Zyric2 has joined #openstack-swift | 21:35 | |
| briancline | eranrom: yeah, I think I did something terrifically dumb in my test config :X | 21:36 |
| *** jrichli has quit IRC | 21:38 | |
| briancline | eranrom: well, config was right, but the steps i did to set up the accounts had a typo. ack | 21:38 |
| briancline | yep, that was it. i need some sleep, it's been a long week already :P | 21:41 |
| eranrom | briancline: me too. :-) | 21:42 |
| *** garthb has quit IRC | 21:45 | |
| *** garthb has joined #openstack-swift | 21:46 | |
| *** eranrom has quit IRC | 21:47 | |
| *** slo_ has quit IRC | 21:50 | |
| *** devlaps has joined #openstack-swift | 21:52 | |
| *** fthiagogv_ has quit IRC | 21:54 | |
| *** ccavanna has joined #openstack-swift | 21:57 | |
| *** gyee has joined #openstack-swift | 21:59 | |
| *** hrou has quit IRC | 22:00 | |
| *** jkugel has quit IRC | 22:01 | |
| openstackgerrit | John Dickinson proposed openstack/swift: fix docstring: s/2xx/1xx/ https://review.openstack.org/229163 | 22:04 |
| notmyname | easiest review ever? ^ | 22:04 |
| torgomatic | yes | 22:05 |
| notmyname | if for whatever reason kota's patch doesn't merge cleanly, I'll squash that into his | 22:06 |
| *** NM has quit IRC | 22:09 | |
| *** marcusvrn_ has quit IRC | 22:15 | |
| *** wbhuber has quit IRC | 22:17 | |
| *** pberis has quit IRC | 22:17 | |
| *** CaioBrentano has quit IRC | 22:24 | |
| *** CaioBrentano has joined #openstack-swift | 22:24 | |
| *** CaioBrentano has quit IRC | 22:29 | |
| mattoliverau | morning | 22:40 |
| *** lcurtis has quit IRC | 22:41 | |
| notmyname | hello | 22:42 |
| *** proteusguy_ has joined #openstack-swift | 22:43 | |
| *** jlhinson has quit IRC | 22:43 | |
| *** proteusguy has quit IRC | 22:45 | |
| openstackgerrit | Merged openstack/swift: Don't send commits for quorum *BAD* requests on EC https://review.openstack.org/220059 | 22:46 |
| mattoliverau | notmyname: still mucking around with keystone. It wasn't working (keystone server) so got a little sidetracked with other reviews. Got it working last night while starting to look at bulk middleware placement.. then wife got home. Will look at the service token thing this morn :) | 22:46 |
| notmyname | mattoliverau: thanks | 22:46 |
| *** NM has joined #openstack-swift | 22:53 | |
| *** alejandrito has quit IRC | 22:56 | |
| *** trifon has quit IRC | 22:56 | |
| *** km has joined #openstack-swift | 23:02 | |
| *** cdelatte has joined #openstack-swift | 23:15 | |
| *** trifon has joined #openstack-swift | 23:23 | |
| *** kei_yama has joined #openstack-swift | 23:25 | |
| *** ccavanna_ has joined #openstack-swift | 23:27 | |
| *** hrou has joined #openstack-swift | 23:29 | |
| *** ccavanna has quit IRC | 23:30 | |
| *** xnox has quit IRC | 23:32 | |
| *** charz has quit IRC | 23:32 | |
| *** onovy has quit IRC | 23:32 | |
| *** amandap has quit IRC | 23:32 | |
| *** tristanC has quit IRC | 23:32 | |
| *** tristanC has joined #openstack-swift | 23:33 | |
| *** albertom has quit IRC | 23:33 | |
| *** charz has joined #openstack-swift | 23:34 | |
| *** albertom has joined #openstack-swift | 23:34 | |
| *** onovy has joined #openstack-swift | 23:34 | |
| *** xnox has joined #openstack-swift | 23:34 | |
| *** trifon has quit IRC | 23:35 | |
| *** amandap has joined #openstack-swift | 23:35 | |
| *** mac_ified has joined #openstack-swift | 23:44 | |
| *** garthb has quit IRC | 23:44 | |
| *** jrichli has joined #openstack-swift | 23:45 | |
| openstackgerrit | Merged openstack/swift: fix docstring: s/2xx/1xx/ https://review.openstack.org/229163 | 23:54 |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Improve SLO PUT error checking https://review.openstack.org/228099 | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!