| *** haomaiwang has quit IRC | 00:01 | |
| *** haomaiwang has joined #openstack-swift | 00:01 | |
| *** Kennan2 is now known as Kennan | 00:02 | |
| openstackgerrit | Matt Riedemann proposed openstack/python-swiftclient: Update from global requirements https://review.openstack.org/215786 | 00:28 |
|---|---|---|
| *** dmorita has joined #openstack-swift | 00:28 | |
| openstackgerrit | Matt Riedemann proposed openstack/python-swiftclient: Update from global requirements https://review.openstack.org/215786 | 00:44 |
| *** lyrrad has joined #openstack-swift | 00:54 | |
| *** lyrrad has quit IRC | 00:58 | |
| *** lyrrad has joined #openstack-swift | 00:59 | |
| *** haomaiwang has quit IRC | 01:01 | |
| *** haomaiwang has joined #openstack-swift | 01:02 | |
| *** ccavanna_ has joined #openstack-swift | 01:04 | |
| *** ccavanna has quit IRC | 01:06 | |
| *** haomaiwang has quit IRC | 01:16 | |
| *** abhirc has joined #openstack-swift | 01:43 | |
| *** haomaiwang has joined #openstack-swift | 01:54 | |
| *** haomaiwang has quit IRC | 01:54 | |
| *** haomaiwang has joined #openstack-swift | 01:54 | |
| *** haomaiwang has quit IRC | 02:01 | |
| *** haomaiwang has joined #openstack-swift | 02:02 | |
| *** Kennan has quit IRC | 02:02 | |
| *** Kennan has joined #openstack-swift | 02:02 | |
| openstackgerrit | Kazuhiro MIYAHARA proposed openstack/swift: Fix dispersion-reports error message https://review.openstack.org/213697 | 02:10 |
| *** jamielennox is now known as jamielennox|away | 02:20 | |
| *** abhirc has quit IRC | 02:23 | |
| *** jamielennox|away is now known as jamielennox | 02:32 | |
| *** nakagawamsa has joined #openstack-swift | 02:53 | |
| *** haomaiwang has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-swift | 03:02 | |
| *** dododo has quit IRC | 03:11 | |
| *** openstackgerrit has quit IRC | 03:31 | |
| *** openstackgerrit has joined #openstack-swift | 03:31 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** jrichli has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-swift | 04:02 | |
| *** ppai has joined #openstack-swift | 04:09 | |
| *** silor has joined #openstack-swift | 04:36 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:39 | |
| *** silor has joined #openstack-swift | 04:40 | |
| *** SkyRocknRoll has quit IRC | 04:41 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:42 | |
| *** silor1 has joined #openstack-swift | 04:43 | |
| *** silor has quit IRC | 04:45 | |
| *** silor1 is now known as silor | 04:45 | |
| *** SkyRocknRoll has quit IRC | 04:46 | |
| *** SkyRocknRoll has joined #openstack-swift | 04:46 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** haomaiwang has joined #openstack-swift | 05:02 | |
| *** hrou has quit IRC | 05:02 | |
| *** SkyRocknRoll has quit IRC | 05:08 | |
| jith_ | janonymous: Thanks | 05:18 |
| *** mahatic has joined #openstack-swift | 05:18 | |
| *** pberis has joined #openstack-swift | 05:30 | |
| jith_ | hi all i am getting an error "ImportError: No module named pbr.version " when i do swift-init all restart.. all services except proxy server got restarted.. can anyone guide me pls? | 05:31 |
| mahatic | jith_: did you run this? sudo pip install -r swift/test-requirements.txt | 05:44 |
| mahatic | and also the requirements? | 05:44 |
| *** jamielennox is now known as jamielennox|away | 05:47 | |
| *** haomaiwang has quit IRC | 06:01 | |
| jith_ | mahatic: yes i did | 06:01 |
| *** haomaiwang has joined #openstack-swift | 06:02 | |
| jith_ | when i try to install pbr again.. It populated a msg " sudo pip install pbr | 06:02 |
| jith_ | Requirement already satisfied (use --upgrade to upgrade): pbr in /home/stack/python-swiftclient/pbr-1.5.0-py2.7.egg | 06:02 |
| jith_ | Cleaning up..." | 06:02 |
| mahatic | which os are you using? | 06:03 |
| *** pberis has quit IRC | 06:15 | |
| jith_ | not ubuntu.. it is debian based flavor.. but its almost equivalent to ubuntu 14.04 | 06:22 |
| *** iepupp has joined #openstack-swift | 06:25 | |
| jith_ | mahatic: | 06:29 |
| jith_ | uname -a | 06:29 |
| jith_ | Linux ajitha 3.16.0-4-amd64 #1 SMP Debian 3.16.7-2 (2014-11-06) x86_64 GNU/Linux | 06:29 |
| *** xnox has joined #openstack-swift | 06:33 | |
| *** remix_tj has joined #openstack-swift | 06:35 | |
| openstackgerrit | Nakagawa Masaaki proposed openstack/swift: SAIO on Ubuntu 14.04 can not pass unittest. https://review.openstack.org/207790 | 06:44 |
| *** SkyRocknRoll has joined #openstack-swift | 06:47 | |
| *** akle has joined #openstack-swift | 06:51 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** 7JTAAJDYR has joined #openstack-swift | 07:02 | |
| *** rledisez has joined #openstack-swift | 07:04 | |
| *** haigang has joined #openstack-swift | 07:05 | |
| *** haigang has quit IRC | 07:06 | |
| *** I has joined #openstack-swift | 07:06 | |
| *** I is now known as Guest63057 | 07:07 | |
| *** Guest63057 has quit IRC | 07:07 | |
| *** haigang has joined #openstack-swift | 07:07 | |
| *** haigang has quit IRC | 07:09 | |
| *** haigang has joined #openstack-swift | 07:10 | |
| *** SkyRocknRoll has quit IRC | 07:16 | |
| onovy | hi, almost "one liner" change review, someone? :) https://review.openstack.org/#/c/210736/ thanks | 07:26 |
| openstackgerrit | Merged openstack/swift: Enable Object Replicator's failure count in recon https://review.openstack.org/138342 | 07:32 |
| *** jordanP has joined #openstack-swift | 07:51 | |
| *** mahatic has quit IRC | 07:52 | |
| *** 7JTAAJDYR has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-swift | 08:02 | |
| *** haigang has quit IRC | 08:06 | |
| *** mahatic has joined #openstack-swift | 08:15 | |
| *** jistr has joined #openstack-swift | 08:17 | |
| *** geaaru has joined #openstack-swift | 08:18 | |
| *** joeljwright has joined #openstack-swift | 08:25 | |
| *** ChanServ sets mode: +v joeljwright | 08:25 | |
| *** dmorita has quit IRC | 08:32 | |
| *** silor1 has joined #openstack-swift | 08:37 | |
| *** silor has quit IRC | 08:39 | |
| *** silor1 is now known as silor | 08:39 | |
| *** jistr has quit IRC | 08:45 | |
| *** jistr has joined #openstack-swift | 08:47 | |
| *** sanchitmalhotra has joined #openstack-swift | 08:48 | |
| *** sanchitmalhotra has quit IRC | 08:51 | |
| *** sanchitmalhotra has joined #openstack-swift | 08:57 | |
| mahatic | acoles: hello! around? | 09:00 |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-swift | 09:01 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 09:06 | |
| acoles | mahatic: hi | 09:07 |
| *** jistr has quit IRC | 09:08 | |
| *** sanchitmalhotra has quit IRC | 09:08 | |
| mahatic | acoles: where would I set a default value for a header? (for encrypt flag in keymaster). Right now, if I add a header in one of the failing tests and check in keymaster for the header, it works. But I need to set it to a defualt value - so that I don't have to modify existing tests (like you've suggested) | 09:10 |
| mahatic | does it sound out of the blue? :) | 09:12 |
| acoles | mahatic: i guess you would need to set the default value where you need to read the value, so keymaster perhaps? | 09:13 |
| acoles | mahatic: also, just a guess, but you might find the config_true_value method useful in utils | 09:14 |
| acoles | mahatic: swift/common/utils.py:272 | 09:16 |
| janonymous | jith_ : Sure, no problem. | 09:17 |
| acoles | mahatic: its a useful helper to accept any true-like string from a value, in this case possibly your header value | 09:17 |
| mahatic | acoles: right, thanks. Looked at that. I don't think I can set the header value in keymaster itself, I check the incoming headers from req and verify it's value | 09:19 |
| acoles | mahatic: so, if the header is not found in the req can you not then assume a default value | 09:20 |
| *** aix has joined #openstack-swift | 09:20 | |
| acoles | mahatic: something like req.headers.get('MY-Header-Name', my_default_value) ?? | 09:20 |
| mahatic | acoles: oh maybe my question should be, how do I add a default header to a req? or is not necessary? | 09:20 |
| mahatic | acoles: okay, I could try the default value you mentioned. I was just wondering, if I need to introduce a header - is there a specific common place for it? or is it just according to my need (like what I'm doing now) | 09:24 |
| acoles | mahatic: there is no canonical list of all swift headers. you don't need to add a default to the request class, just handle the default where you use the header. | 09:29 |
| mahatic | acoles: okay, great. thanks! yes, the default thing seems to be working (I'm running tests) | 09:31 |
| acoles | cool | 09:31 |
| mahatic | acoles: how's your monday coming by? Do you work out of an office or home? | 09:34 |
| acoles | mahatic: ok thanks, today i'm in office, most days the same | 09:36 |
| mahatic | ah okay | 09:37 |
| acoles | mattoliverau: you make it home? | 09:37 |
| mattoliverau | acoles: yup home safe and sound :) | 09:38 |
| mahatic | mattoliverau: good evening! | 09:38 |
| mattoliverau | mahatic: o/ | 09:39 |
| acoles | hmmm, swiftclient tox -e pep8 is no longer enforcing max line length ?!? | 09:41 |
| acoles | since hacking version got bumped | 09:42 |
| mahatic | oh, interesting (didn't know it was because of hacking) | 09:48 |
| * mahatic is off for a while - be back in an hour or so | 09:48 | |
| *** mahatic has quit IRC | 09:49 | |
| acoles | correlation != causation, but... | 09:49 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 09:53 |
| openstackgerrit | Alistair Coles proposed openstack/python-swiftclient: Add minimal working service token support. https://review.openstack.org/182640 | 09:54 |
| *** haomaiwang has quit IRC | 10:01 | |
| *** haomaiwang has joined #openstack-swift | 10:02 | |
| *** SkyRocknRoll has joined #openstack-swift | 10:05 | |
| *** Kennan has quit IRC | 10:09 | |
| *** Kennan has joined #openstack-swift | 10:10 | |
| *** kei_yama has quit IRC | 10:11 | |
| *** aix has quit IRC | 10:15 | |
| *** marzif has joined #openstack-swift | 10:31 | |
| *** ekarlso has quit IRC | 10:32 | |
| *** ekarlso has joined #openstack-swift | 10:32 | |
| *** km has quit IRC | 10:35 | |
| *** aix has joined #openstack-swift | 10:48 | |
| *** donagh has quit IRC | 10:50 | |
| *** jamielennox|away has quit IRC | 10:50 | |
| *** jistr has joined #openstack-swift | 10:52 | |
| *** sudorandom has quit IRC | 10:56 | |
| *** SkyRocknRoll has quit IRC | 10:56 | |
| *** zul has quit IRC | 10:56 | |
| *** CrackerJackMack has quit IRC | 10:56 | |
| *** zacksh has quit IRC | 10:56 | |
| *** jroll has quit IRC | 10:56 | |
| *** zacksh has joined #openstack-swift | 10:58 | |
| *** bkopilov has quit IRC | 10:59 | |
| *** bkopilov has joined #openstack-swift | 11:00 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** sudorandom has joined #openstack-swift | 11:01 | |
| *** CrackerJackMack has joined #openstack-swift | 11:01 | |
| *** haomaiwang has joined #openstack-swift | 11:02 | |
| *** jroll has joined #openstack-swift | 11:02 | |
| *** jamielennox|away has joined #openstack-swift | 11:04 | |
| *** jamielennox|away is now known as jamielennox | 11:04 | |
| *** donagh has joined #openstack-swift | 11:05 | |
| openstackgerrit | Alistair Coles proposed openstack/python-swiftclient: Make tox -e pep8 check for lines being too long https://review.openstack.org/216192 | 11:05 |
| *** abhirc has joined #openstack-swift | 11:18 | |
| *** lpabon has joined #openstack-swift | 11:19 | |
| *** iepupp has quit IRC | 11:28 | |
| *** nakagawamsa has quit IRC | 11:42 | |
| *** haomaiwang has quit IRC | 11:55 | |
| *** ppai has quit IRC | 12:02 | |
| *** ppai has joined #openstack-swift | 12:12 | |
| *** abhirc has quit IRC | 12:22 | |
| *** lpabon_ has joined #openstack-swift | 12:24 | |
| *** lpabon has quit IRC | 12:24 | |
| *** ppai has quit IRC | 12:26 | |
| *** daemontool__ has joined #openstack-swift | 12:27 | |
| *** daemontool_ has quit IRC | 12:29 | |
| *** jrichli has joined #openstack-swift | 12:30 | |
| *** jistr has quit IRC | 12:31 | |
| *** jistr has joined #openstack-swift | 12:33 | |
| *** petertr7_away is now known as petertr7 | 12:37 | |
| *** zul has joined #openstack-swift | 12:39 | |
| *** dustins has joined #openstack-swift | 12:42 | |
| openstackgerrit | Alistair Coles proposed openstack/python-swiftclient: Stop Connection class modifying os_options parameter https://review.openstack.org/216240 | 12:43 |
| *** chlong has joined #openstack-swift | 12:47 | |
| *** mahatic has joined #openstack-swift | 12:49 | |
| *** chlong has quit IRC | 12:54 | |
| *** annegentle has joined #openstack-swift | 13:05 | |
| *** lpabon_ has quit IRC | 13:08 | |
| *** annegentle has quit IRC | 13:08 | |
| *** annegentle has joined #openstack-swift | 13:08 | |
| *** dustins has quit IRC | 13:10 | |
| *** dustins has joined #openstack-swift | 13:11 | |
| *** hrou has joined #openstack-swift | 13:11 | |
| *** chlong has joined #openstack-swift | 13:13 | |
| *** dustins has quit IRC | 13:18 | |
| *** ccavanna_ has quit IRC | 13:18 | |
| *** dustins has joined #openstack-swift | 13:19 | |
| *** haomaiwa_ has joined #openstack-swift | 13:21 | |
| *** jroll has quit IRC | 13:23 | |
| *** dustins has quit IRC | 13:25 | |
| *** jith_ has quit IRC | 13:25 | |
| *** jamielennox has quit IRC | 13:26 | |
| *** donagh has quit IRC | 13:26 | |
| *** SkyRocknRoll has joined #openstack-swift | 13:27 | |
| *** bapalm_ has joined #openstack-swift | 13:28 | |
| *** dustins has joined #openstack-swift | 13:29 | |
| *** wbhuber has joined #openstack-swift | 13:31 | |
| *** links has joined #openstack-swift | 13:31 | |
| *** jordanP has quit IRC | 13:34 | |
| *** petertr7 is now known as petertr7_away | 13:36 | |
| *** jroll has joined #openstack-swift | 13:37 | |
| *** donagh has joined #openstack-swift | 13:40 | |
| *** jamielennox|away has joined #openstack-swift | 13:43 | |
| *** jamielennox|away is now known as jamielennox | 13:43 | |
| mahatic | acoles: get_container and head_container (in swift client - client.py) do not accept headers, do you know if I could add them? or is there a process for it? (I'm aware swift-client has a diff core) | 13:46 |
| acoles | mahatic: you can add them, just propose a patch to swifclient. | 13:47 |
| mahatic | acoles: okay, thanks | 13:47 |
| acoles | mahatic: swiftclient cores are superset of swift cores | 13:47 |
| mahatic | acoles: ah okay | 13:47 |
| *** petertr7_away is now known as petertr7 | 13:57 | |
| *** haomaiwa_ has quit IRC | 14:01 | |
| *** haomaiwang has joined #openstack-swift | 14:02 | |
| *** sanchitmalhotra1 has quit IRC | 14:05 | |
| *** ccavanna has joined #openstack-swift | 14:07 | |
| *** lpabon has joined #openstack-swift | 14:07 | |
| *** jamielennox has quit IRC | 14:08 | |
| *** jordanP has joined #openstack-swift | 14:10 | |
| *** jamielennox|away has joined #openstack-swift | 14:10 | |
| *** jamielennox|away is now known as jamielennox | 14:11 | |
| *** marzif has quit IRC | 14:11 | |
| *** lpabon has quit IRC | 14:11 | |
| *** marzif has joined #openstack-swift | 14:12 | |
| *** petertr7 is now known as petertr7_away | 14:20 | |
| *** kairo has joined #openstack-swift | 14:25 | |
| *** petertr7_away is now known as petertr7 | 14:25 | |
| *** jlhinson has joined #openstack-swift | 14:33 | |
| *** jlhinson has quit IRC | 14:35 | |
| *** alejandrito has joined #openstack-swift | 14:37 | |
| *** bapalm_ has quit IRC | 14:41 | |
| *** pgbridge has quit IRC | 14:44 | |
| *** bapalm- has joined #openstack-swift | 14:45 | |
| *** pgbridge has joined #openstack-swift | 14:49 | |
| *** annegentle has quit IRC | 14:52 | |
| *** haomaiwang has quit IRC | 15:01 | |
| *** haomaiwa_ has joined #openstack-swift | 15:02 | |
| *** minwoob has joined #openstack-swift | 15:02 | |
| *** flwang has joined #openstack-swift | 15:04 | |
| flwang | cschwede: ping | 15:04 |
| cschwede | flwang: pong | 15:08 |
| flwang | cschwede: hi, there is Feilong from zaqar team. just want to chat about this https://review.openstack.org/#/c/196755/ | 15:09 |
| flwang | the notification middleware of swift | 15:09 |
| flwang | cschwede: are you still working on that? do you need any help from our side? | 15:09 |
| cschwede | flwang: yes, still working on it. going to send another update this week | 15:10 |
| flwang | cschwede: awesome, and we will also keep an eye on that and provide feedback from our side. thanks for working on this | 15:11 |
| cschwede | flwang: thanks, i appreciate it :) | 15:13 |
| *** zaitcev has joined #openstack-swift | 15:21 | |
| *** ChanServ sets mode: +v zaitcev | 15:21 | |
| *** haomaiwa_ has quit IRC | 15:22 | |
| *** mfalatic has joined #openstack-swift | 15:24 | |
| *** annegentle has joined #openstack-swift | 15:25 | |
| *** lyrrad has quit IRC | 15:35 | |
| *** nadeem has joined #openstack-swift | 15:35 | |
| *** jistr has quit IRC | 15:35 | |
| *** david-ly_ is now known as david-lyle | 15:38 | |
| *** jistr has joined #openstack-swift | 15:38 | |
| *** SkyRocknRoll has quit IRC | 15:48 | |
| *** alejandrito has quit IRC | 15:53 | |
| *** gyee has joined #openstack-swift | 15:54 | |
| *** petertr7 is now known as petertr7_away | 15:56 | |
| *** jlhinson has joined #openstack-swift | 15:57 | |
| *** links has quit IRC | 16:08 | |
| *** thumpba has joined #openstack-swift | 16:12 | |
| *** jistr has quit IRC | 16:13 | |
| *** marzif has quit IRC | 16:18 | |
| minwoob | All right -- so I know that EC policy is counter-productive when object size is less than the chunk size, but does that apply to objects larger than the chunk size, but still small in size? | 16:19 |
| minwoob | Basically, is the lower bound that is acceptable for EC actually higher than the specified chunk size? | 16:20 |
| minwoob | And is there an actual (formulaically definable) limit, or none? | 16:23 |
| *** jordanP has quit IRC | 16:24 | |
| minwoob | (This is for anyone who wants to participate in coming up with an auto-tiering strategy based on size). | 16:24 |
| *** BAKfr has quit IRC | 16:25 | |
| *** BAKfr has joined #openstack-swift | 16:26 | |
| *** lpabon has joined #openstack-swift | 16:27 | |
| *** rledisez has quit IRC | 16:28 | |
| egon | minwoob: how are you using "chunk size" in this context? I mean, how does it relate to block length and message length? | 16:29 |
| minwoob | egon: I mean the size of a segment before it gets encoded. | 16:29 |
| minwoob | egon: If you apply EC to such an object, you get more storage overhead than if you were to have used replication. | 16:30 |
| minwoob | egon: At the very least, it requires padding to the object with the current implementation. | 16:30 |
| egon | Yes, it's the same problem as file sizes below a filesystem block size. | 16:31 |
| peluse | minwoob, I think you mean segment size there wrt your very first comment | 16:31 |
| *** lpabon has quit IRC | 16:31 | |
| notmyname | good morning | 16:31 |
| peluse | notmyname, morning! | 16:31 |
| egon | notmyname: morning! | 16:31 |
| minwoob | notmyname: Good morning! | 16:31 |
| peluse | minwoob, terminology in the EC implementation is pretty specific and if you use the wrong word (liek we did many times early on) it can get very confusing | 16:31 |
| minwoob | peluse: Yes, exactly. | 16:31 |
| minwoob | peluse: Yeah, what I meant was segment size in that comment above. | 16:32 |
| peluse | minwoob, we don't pad to segment size though and it is configurable. We're running some tests this week to help characterize segment size vs object size wrt performance/efficiency | 16:34 |
| wbhuber | peluse: with the same definition files that Bill G. sent to us awhile ago? | 16:34 |
| wbhuber | peluse: do your tests include saturating the network while the workloads are running to see how the policies handle storing objects? | 16:37 |
| peluse | wbhuber, yup | 16:37 |
| peluse | wbhuber, the yes was to your first question, not sure I understand the 2nd | 16:38 |
| wbhuber | peluse: for example full cluster | 16:40 |
| wbhuber | with the disks at full capacity (prob same effects as replica though) | 16:40 |
| peluse | wbhuber, all: basically what we're doing w/EC this week in Phx i doing baseline stress testing of various sizes on a 16 node clsuter (2 proxy 1 LB 14 SN). Making sure things work, testing some of the outstanding patches, testing some eror scenarios (as many as we can get to) | 16:41 |
| peluse | ...and then leaving the working cluster for our benchmarking team to run repl vs Ec tests (which they tried to do before but ran into too many issues) | 16:41 |
| peluse | so if anyone has any suggestions (like wbhuber just did) please let us know! | 16:42 |
| peluse | wbhuber, doubt we'll do a cluster full test this week. focusing on (1) heavy load w/o failure (2) heavy load w/node failures (3) reconstructor scenarios (4) some parm tuing (like segment size) | 16:43 |
| wbhuber | peluse: noticed that in your cluster config, what do you expect from an LB? handling network bottlenecks? | 16:44 |
| peluse | wbhuber, we expect it to balance traffic over our 2 proxies :) | 16:44 |
| egon | we're not playing with EC yet, and are running an older version of swift, but we have 8 proxy nodes. | 16:45 |
| *** marcusvrn_ has joined #openstack-swift | 16:45 | |
| egon | and same here, we expect to balance traffic | 16:45 |
| *** barker has joined #openstack-swift | 16:53 | |
| wbhuber | peluse: are those nodes straight bare metal (i.e. not VMs)? | 17:01 |
| *** barker has quit IRC | 17:01 | |
| *** kevinc_ has joined #openstack-swift | 17:05 | |
| *** marzif has joined #openstack-swift | 17:12 | |
| *** barker has joined #openstack-swift | 17:12 | |
| *** mahatic has quit IRC | 17:13 | |
| *** marzif has quit IRC | 17:17 | |
| *** petertr7_away is now known as petertr7 | 17:20 | |
| *** aix has quit IRC | 17:20 | |
| *** bi_fa_fu has joined #openstack-swift | 17:23 | |
| *** jrichli has quit IRC | 17:24 | |
| *** barker has quit IRC | 17:24 | |
| *** barker has joined #openstack-swift | 17:26 | |
| *** barker has quit IRC | 17:33 | |
| peluse | wbhuber, yes, no VMs anywhere | 17:41 |
| peluse | ctennis will post pics of our config here shortly | 17:43 |
| *** geaaru has quit IRC | 17:44 | |
| *** bkopilov has quit IRC | 17:44 | |
| *** bkopilov has joined #openstack-swift | 17:45 | |
| ctennis | http://i.imgur.com/FS8o5Uh.png | 17:48 |
| clayg | good morning | 17:55 |
| peluse | clayg, yo! | 17:55 |
| ctennis | and http://i.imgur.com/Vt7wuRb.png | 17:57 |
| *** bi_fa_fu has quit IRC | 17:58 | |
| clayg | ctennis: both of those setups look great! | 17:58 |
| clayg | ctennis: can you plug them both *together*!? | 17:59 |
| egon | ctennis: is that the intel reference deployment? | 17:59 |
| ctennis | yes | 17:59 |
| peluse | clayg, yeah its going well. we are currently running patch 199043 on all nodes | 17:59 |
| patchbot | peluse: https://review.openstack.org/#/c/199043/ | 17:59 |
| clayg | peluse: yeah that's a nice fix - we really need to get that one in :\ | 18:00 |
| peluse | clayg, they are 2 different clusters totally. we are using the xeon one right now | 18:00 |
| clayg | peluse: what about the etag_buckets change - that one is pretty good too! | 18:00 |
| egon | ctennis: the backend is straight swift, not ceph, or something? | 18:00 |
| peluse | clayg, I will be ready to +2 it with a little bit of back history :) Its super tiny but I don't totally understand it | 18:00 |
| ctennis | yeah this is straight swift | 18:01 |
| torgomatic | minwoob: what do you mean by padding? there's some EC headers, but just because you use a 1 MiB EC chunk size doesn't mean all your stuff gets padded to 1 MiB | 18:01 |
| peluse | clayg, we are going to run with the etag buckets patch after we get done running a bunch of baseline stress testing - will fire that up probably after lunch | 18:02 |
| clayg | peluse: ah ok - cool! | 18:02 |
| peluse | clayg, because that's when we'll start inducing some failures and need it... | 18:03 |
| clayg | peluse: hehe | 18:06 |
| peluse | BTW cluster running almost nonstop since yesterday morning, various workloads, no exceptions/errors to report so far... | 18:08 |
| *** bkopilov has quit IRC | 18:08 | |
| clayg | why does swiftclient still get to have a whitelist of H checks? https://review.openstack.org/#/c/216192 | 18:09 |
| clayg | and why doesn't it have Flake8 checks (select = F) ??? | 18:10 |
| *** corvus is now known as jeblair | 18:13 | |
| acoles | clayg: i have no idea what is going on with those flake tests. i was just 'how did this merge?'!! | 18:15 |
| *** bapalm- is now known as bapalm_ | 18:19 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Use correct Storage-Policy header for REPLICATE requests https://review.openstack.org/215857 | 18:20 |
| timburke | clayg: it shouldn't, the whitelist doesn't work right. go review https://review.openstack.org/#/c/202909/ if you want swiftclient to have a sane pep8 check :P | 18:22 |
| openstackgerrit | Alistair Coles proposed openstack/swift: Enable object server to return non-durable data https://review.openstack.org/215276 | 18:22 |
| acoles | timburke: darn i didnt spot that patch | 18:24 |
| openstackgerrit | Clay Gerrard proposed openstack/swift: Make handoffs_first a more useful "mode" https://review.openstack.org/215867 | 18:25 |
| acoles | peluse: clayg: i lost track where patch 212187 is at - shoudl i be reviewing what is there or is there a change in the works? | 18:27 |
| patchbot | acoles: https://review.openstack.org/#/c/212187/ | 18:27 |
| clayg | acoles: I think peluse is busy this week and it's not quite ready for merge | 18:28 |
| clayg | acoles: there's a couple of fix ups in patch 215360 | 18:29 |
| patchbot | clayg: https://review.openstack.org/#/c/215360/ | 18:29 |
| clayg | acoles: maybe sanity that diff for me and I'll squash it down into the other for a proper review | 18:29 |
| clayg | acoles: what's the status on patch 215276 - ready for review? | 18:29 |
| patchbot | clayg: https://review.openstack.org/#/c/215276/ | 18:29 |
| *** jrichli has joined #openstack-swift | 18:30 | |
| clayg | it looks like it's two or three changes down the line from master - so there's probably gunna be some rebases - if you're happy with the changes and I break it during your overnight I'll try to keep it rebased for you | 18:31 |
| clayg | acoles: if you're still working on it anyway I'll probably just leave it busted for you | 18:31 |
| acoles | clayg: patch 215276 is ready for review. its not much use without the corresponding proxy changes but you can at least tell me it sucks :P | 18:32 |
| patchbot | acoles: https://review.openstack.org/#/c/215276/ | 18:32 |
| clayg | acoles: full-ack | 18:33 |
| acoles | clayg: i'm done for today and happy to rebase. i'll look at 215360 tomorrow, sorry if idint notice that | 18:33 |
| clayg | acoles: can you star patch 215317 too - it's a cheap one | 18:33 |
| patchbot | clayg: https://review.openstack.org/#/c/215317/ | 18:33 |
| minwoob | torgomatic: I see. If that's the case, then yeah, padding is not an issue. | 18:33 |
| acoles | s/if idint/i didnt/ | 18:33 |
| clayg | acoles: no it's ok, don't worry about that - i'm pretty sure *I* like it better at least - i'm going to squash it over patch 212187 and change etag_buckets to timestamp_buckets anyway | 18:34 |
| patchbot | clayg: https://review.openstack.org/#/c/212187/ | 18:34 |
| acoles | clayg: sure | 18:34 |
| acoles | clayg: sure re 215317 but tomorrow sorry | 18:35 |
| acoles | clayg: yeah i'm pretty convinced we should be comparing timestamps across the frag set | 18:35 |
| clayg | mattoliverau: the commit message on patch 117710 sounds great! | 18:36 |
| patchbot | clayg: https://review.openstack.org/#/c/117710/ | 18:36 |
| acoles | clayg: even if etags match the metadata could be different on different time'd frags | 18:36 |
| clayg | acoles: ok i'll do it then - have a good night | 18:36 |
| clayg | acoles: thanks for working late for the handoff! get out of there! | 18:37 |
| openstackgerrit | Michael Barton proposed openstack/swift: go: O_TMPFILE https://review.openstack.org/216375 | 18:39 |
| joeljwright | Is there any chance of some people casting their eyes over patch 130339 and patch 171692 please | 18:43 |
| patchbot | joeljwright: https://review.openstack.org/#/c/130339/ | 18:43 |
| joeljwright | just for patchbot patch 171692 | 18:43 |
| patchbot | joeljwright: https://review.openstack.org/#/c/171692/ | 18:43 |
| joeljwright | they've been sitting around for some time | 18:43 |
| joeljwright | would like to get them done and merged | 18:44 |
| joeljwright | but I'm aware that not many people have looked too hard at them | 18:44 |
| clayg | redbo: wohoooo O_TMPFILE! | 18:45 |
| acoles | joeljwright: ack will try to find some time | 18:46 |
| clayg | redbo: can I trick anyone at the rax into looking at matts' patch for concurrent gets? patch 117710 | 18:46 |
| patchbot | clayg: https://review.openstack.org/#/c/117710/ | 18:46 |
| joeljwright | acoles: thanks | 18:46 |
| acoles | good night | 18:47 |
| redbo | I don't know how tricky you can be! | 18:48 |
| *** DericHorn-HP has joined #openstack-swift | 18:50 | |
| tdasilva | redbo, clayg: or maybe patch 162243 ??? | 18:53 |
| patchbot | tdasilva: https://review.openstack.org/#/c/162243/ | 18:53 |
| tdasilva | :D | 18:53 |
| *** bkopilov has joined #openstack-swift | 18:59 | |
| *** acoles is now known as acoles_ | 19:01 | |
| *** bkopilov has quit IRC | 19:09 | |
| *** bkopilov has joined #openstack-swift | 19:11 | |
| *** annegent_ has joined #openstack-swift | 19:12 | |
| clayg | redbo: fine don't review it - see if I care | 19:16 |
| * clayg tries reverse trickery | 19:16 | |
| *** annegentle has quit IRC | 19:16 | |
| redbo | I bet you don't even know how to review it | 19:16 |
| clayg | redbo: I've reviewed it once before - it was a bunch of work - i had to setup dfg's slowdown (https://github.com/dpgoetz/slowdown) | 19:21 |
| clayg | redbo: but that's how I found the first round of bugs - maybe some of them are fixed - idk | 19:21 |
| clayg | I sorta don't want to look - if it's still broken I'll just have to -1 it - and *then* what did I accomplish? | 19:22 |
| * clayg thinks he's not selling this review work to redbo very well :\ | 19:22 | |
| *** petertr7 is now known as petertr7_away | 19:25 | |
| *** petertr7_away is now known as petertr7 | 19:29 | |
| redbo | I mean we could just blindly approve it, then we'll find all the problems 6 months or a year from now when we actually deploy the current code. | 19:35 |
| clayg | redbo: ok, well this was fun - nm | 19:35 |
| clayg | :) | 19:35 |
| *** zul has quit IRC | 19:49 | |
| *** kutija has joined #openstack-swift | 20:03 | |
| *** bkopilov has quit IRC | 20:03 | |
| *** annegent_ has quit IRC | 20:07 | |
| *** bkopilov has joined #openstack-swift | 20:07 | |
| *** annegentle has joined #openstack-swift | 20:07 | |
| *** silor has quit IRC | 20:15 | |
| *** DericHorn-HP has quit IRC | 20:20 | |
| openstackgerrit | Merged openstack/swift: Fix the missing SLO state on fast-post https://review.openstack.org/182564 | 20:22 |
| *** jlhinson_ has joined #openstack-swift | 20:22 | |
| *** jlhinson has quit IRC | 20:25 | |
| *** akle has quit IRC | 20:31 | |
| *** jerrygb has joined #openstack-swift | 20:33 | |
| *** jerrygb is now known as test0rz | 20:34 | |
| *** test0rz is now known as asd112z | 20:34 | |
| *** petertr7 is now known as petertr7_away | 20:42 | |
| *** dustins has quit IRC | 20:47 | |
| *** ccavanna has quit IRC | 20:59 | |
| *** then3rd has joined #openstack-swift | 21:01 | |
| then3rd | It appears that telemetry meters storage.objects.{incoming,outgoing}.bytes. My question: is there any way to measure bandwidth per object in each container? | 21:02 |
| clayg | then3rd: post-facto log processing | 21:03 |
| clayg | then3rd: like, you don't *really* want a metric name per object do you? ... maybe you do | 21:03 |
| then3rd | I would imagine that many metrics could become quite cumbersome. | 21:04 |
| clayg | ok, just checking | 21:04 |
| then3rd | I'm sort of spitballing ideas right now. Being so new to swift/openstack, I'm not sure where to start first. What logging mechanisms can gather what I need for processing? | 21:06 |
| *** nadeem has quit IRC | 21:07 | |
| clayg | proxy access logs - most folks keep 'em around "forever" (normally compressed and put back into swift) - you can pull them out and split them up (or index them) - it's a stable form => http://docs.openstack.org/developer/swift/logs.html#proxy-logs | 21:08 |
| clayg | the "request_path" part will have the /account/container/object - the bytes_[sent|recvd] has the 'bandwidth" stuff you were looking for | 21:09 |
| openstackgerrit | Tim Burke proposed openstack/swift: Fix use of delimiter in account listings https://review.openstack.org/216427 | 21:09 |
| clayg | then3rd: still - anything in swift "per object" is a large amount of keys :\ | 21:10 |
| *** jodah has joined #openstack-swift | 21:11 | |
| *** rook][ has joined #openstack-swift | 21:13 | |
| then3rd | clayg: Awesome, I belive that gives me something to run with. I'll probably be back with more questions sooner or later. :) Thanks! | 21:13 |
| clayg | then3rd: absolutely! come on by anytime! | 21:14 |
| clayg | timburke: I think maybe the fix from patch 28069 circa 2013 was just never extended to accounts? Like maybe, it didn't regress - it just never worked? re patch 216427 | 21:20 |
| patchbot | clayg: https://review.openstack.org/#/c/28069/ | 21:20 |
| *** hrou has quit IRC | 21:20 | |
| *** nadeem has joined #openstack-swift | 21:21 | |
| *** trifon has joined #openstack-swift | 21:23 | |
| timburke | clayg: oh yeah, you might be right...i'll fix up the commit message | 21:33 |
| openstackgerrit | Tim Burke proposed openstack/swift: Fix use of delimiter in account listings https://review.openstack.org/216427 | 21:34 |
| *** rook][ has quit IRC | 21:40 | |
| *** abhirc has joined #openstack-swift | 21:47 | |
| *** abhirc has quit IRC | 21:47 | |
| *** abhirc has joined #openstack-swift | 21:48 | |
| notmyname | then3rd: https://github.com/notmyname/slogging is something you could look at for processing swift logs | 21:57 |
| *** chlong has quit IRC | 22:00 | |
| *** pgbridge has quit IRC | 22:03 | |
| *** ujjain has quit IRC | 22:04 | |
| *** ujjain has joined #openstack-swift | 22:07 | |
| *** ujjain has joined #openstack-swift | 22:07 | |
| *** joeljwright has quit IRC | 22:10 | |
| mattoliverau | Morning | 22:13 |
| *** ccavanna has joined #openstack-swift | 22:13 | |
| *** marcusvrn_ has quit IRC | 22:15 | |
| *** chlong has joined #openstack-swift | 22:15 | |
| *** ho has joined #openstack-swift | 22:16 | |
| notmyname | good morning mattoliverau | 22:16 |
| briancline | +1 for slogging | 22:17 |
| briancline | it has aged well ;-) | 22:17 |
| notmyname | briancline: good to hear :-) | 22:17 |
| notmyname | briancline: are you using it too? | 22:18 |
| *** jrichli has quit IRC | 22:18 | |
| briancline | notmyname: yep, we are | 22:18 |
| notmyname | cool | 22:18 |
| ho | good morning! | 22:23 |
| *** hrou has joined #openstack-swift | 22:24 | |
| *** kairo_ has joined #openstack-swift | 22:25 | |
| *** asd112z has quit IRC | 22:25 | |
| *** kairo has quit IRC | 22:27 | |
| mattoliverau | ho: morning | 22:33 |
| ho | mattoliverau: morning! and thanks for the review. it took 8 month :-) | 22:35 |
| *** jlhinson_ has quit IRC | 22:35 | |
| *** nadeem has quit IRC | 22:36 | |
| mattoliverau | Lol | 22:37 |
| *** chlong has quit IRC | 22:40 | |
| *** thumpba has quit IRC | 22:42 | |
| ho | mattoliverau: hehe. thanks again! | 22:43 |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Make the object auditor's run-once mode run once. https://review.openstack.org/216449 | 22:43 |
| *** minwoob has quit IRC | 22:43 | |
| *** hrou has quit IRC | 22:44 | |
| *** dmorita has joined #openstack-swift | 22:48 | |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Disable pep8 check E127 https://review.openstack.org/216450 | 22:48 |
| *** chlong has joined #openstack-swift | 22:52 | |
| openstackgerrit | Hisashi Osanai proposed openstack/swift: Fix typo of a comment in replicator https://review.openstack.org/216453 | 22:54 |
| ho | I read this: http://www.openkinetic.org/index.php?title=Kinetic_Open_Storage_Group . interesting! and swiftstack and redhat are members for it. do you have any info? (what is the relation with swift?) | 23:01 |
| *** tongli has joined #openstack-swift | 23:02 | |
| notmyname | ho: (at swiftstack) we've worked on making swift work with kinetic drives | 23:11 |
| ho | notmyname: cooool!!!!! | 23:11 |
| ho | notmyname: Does swiftstack have the drives in your lab? | 23:13 |
| notmyname | ho: yeah | 23:17 |
| *** annegentle has quit IRC | 23:19 | |
| ho | notmyname: thanks! it's really cool! | 23:19 |
| *** kei_yama has joined #openstack-swift | 23:25 | |
| *** lyrrad has joined #openstack-swift | 23:27 | |
| *** km has joined #openstack-swift | 23:34 | |
| *** EmilienM has quit IRC | 23:34 | |
| *** EmilienM has joined #openstack-swift | 23:37 | |
| *** flwang has quit IRC | 23:39 | |
| *** morganfainberg is now known as morgan | 23:42 | |
| clayg | torgomatic: nice merge -> https://review.openstack.org/#/c/216453/ | 23:46 |
| clayg | ^ cc: anyone whos ever asked when is it ok for single core +A | 23:46 |
| clayg | also ho - thanks! | 23:46 |
| notmyname | ho: nice! keystone in vSAIO :-) | 23:51 |
| ho | notmyname: I won this race :-) | 23:52 |
| ho | notmyname: I would like to do more update on it later. | 23:53 |
| *** tongli has quit IRC | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
