| mattoliverau | clayg: yeah, though I didn't leave until monday morning UK time. | 00:05 |
|---|---|---|
| *** yashmurty has joined #openstack-swift | 00:28 | |
| yashmurty | @kota_ are you attending the Tokyo Ops Meetup today? | 00:32 |
| yashmurty | What time will you be arriving? | 00:32 |
| *** tovin07_ has joined #openstack-swift | 00:51 | |
| *** gerardgine has quit IRC | 01:05 | |
| *** germs has joined #openstack-swift | 01:06 | |
| *** germs has quit IRC | 01:06 | |
| *** germs has joined #openstack-swift | 01:06 | |
| *** gyee has quit IRC | 01:08 | |
| openstackgerrit | Tim Burke proposed openstack/swift feature/deep: Pick up the new probetests check on feature/deep https://review.openstack.org/550291 | 01:09 |
| kei-ichi | Hi everyone.Hi everyone. Thank you for giving me a permission to proceed slogging task. | 01:10 |
| *** itlinux has joined #openstack-swift | 01:10 | |
| kei-ichi | Now launchpad page has created and I and yashmurty is going to start gerrit relevant work. | 01:10 |
| kei-ichi | I'll report progress of our task here in regular period. o/ | 01:10 |
| kei-ichi | (I also got home yesterday evening in JST same as mattoliverau ..) | 01:10 |
| *** germs has quit IRC | 01:11 | |
| mattoliverau | kei-ichi: welcome home and thanks for coming last week, it was great to meet you! | 01:12 |
| kei-ichi | Thank you mattoliverau ! | 01:14 |
| kei-ichi | That was first time for me to attend swift's meeting and I'd really appreciate everyone's warm welcome. | 01:14 |
| kei-ichi | I hope I can contribute to swift from Horizon/slogging side firstlly and finally I would like to make contribution for swift itself ! | 01:14 |
| notmyname | kei-ichi: it was good to meet you, and I'm glad you came. good luck with the slogging work! | 01:15 |
| mattoliverau | kei-ichi: great! | 01:15 |
| kei-ichi | Thaks notmyname ! | 01:16 |
| kei-ichi | I would like to enjoy this wonderful opportunity o/ | 01:16 |
| *** SkyRocknRoll has quit IRC | 01:19 | |
| *** gerardgine has joined #openstack-swift | 01:29 | |
| openstackgerrit | Tim Burke proposed openstack/swift feature/deep: Pick up the new probetests check on feature/deep https://review.openstack.org/550291 | 01:42 |
| *** itlinux has quit IRC | 01:42 | |
| *** gerardgine has quit IRC | 01:58 | |
| *** tovin07_ has quit IRC | 02:03 | |
| notmyname | timburke: you didn't want to get it via a merge from master? | 02:08 |
| timburke | notmyname: i did. see https://storyboard.openstack.org/#!/story/2001496 | 02:09 |
| timburke | feature/deep has it, but zuul doesn't know it's there yet | 02:09 |
| timburke | that's why i only did a whitespace change to .zuul.yaml | 02:11 |
| timburke | tdasilva: got any ideas on how to get swift logs for the probe test job? | 02:15 |
| timburke | probe test output ... isn't always terribly useful | 02:15 |
| *** tovin07_ has joined #openstack-swift | 02:17 | |
| *** itlinux has joined #openstack-swift | 02:21 | |
| *** germs has joined #openstack-swift | 02:25 | |
| *** germs has quit IRC | 02:29 | |
| *** itlinux has quit IRC | 02:36 | |
| *** yashmurty has quit IRC | 02:55 | |
| *** itlinux has joined #openstack-swift | 03:04 | |
| *** yashmurty has joined #openstack-swift | 03:04 | |
| *** itlinux has quit IRC | 03:07 | |
| *** yashmurty has quit IRC | 03:35 | |
| tdasilva | timburke: no ideas yet, but was thinking of the same problem....i thought i saw someone making a change about getting logs from one of the jobs...was it kota_?? or am i confused... | 03:47 |
| tdasilva | will look into that... | 03:47 |
| *** links has joined #openstack-swift | 03:59 | |
| *** yashmurty has joined #openstack-swift | 04:00 | |
| *** yashmurty has quit IRC | 04:01 | |
| *** yashmurt_ has joined #openstack-swift | 04:03 | |
| *** yashmurt_ has quit IRC | 04:21 | |
| *** yashmurty has joined #openstack-swift | 04:23 | |
| *** links has quit IRC | 04:25 | |
| *** germs has joined #openstack-swift | 04:26 | |
| *** germs has quit IRC | 04:26 | |
| *** germs has joined #openstack-swift | 04:26 | |
| *** yashmurty has quit IRC | 04:27 | |
| *** links has joined #openstack-swift | 04:27 | |
| *** germs has quit IRC | 04:31 | |
| *** yashmurty has joined #openstack-swift | 04:32 | |
| *** psachin has joined #openstack-swift | 04:40 | |
| *** silor has joined #openstack-swift | 05:12 | |
| *** silor has quit IRC | 05:36 | |
| *** Supun has joined #openstack-swift | 05:52 | |
| *** cshastri has joined #openstack-swift | 06:00 | |
| *** rcernin_ has joined #openstack-swift | 06:18 | |
| *** rcernin has quit IRC | 06:21 | |
| *** germs has joined #openstack-swift | 06:27 | |
| *** germs has joined #openstack-swift | 06:27 | |
| *** Supun has quit IRC | 06:30 | |
| *** germs has quit IRC | 06:31 | |
| *** cbartz has joined #openstack-swift | 06:40 | |
| *** links has quit IRC | 06:41 | |
| *** yashmurty has quit IRC | 06:43 | |
| *** links has joined #openstack-swift | 06:51 | |
| *** gkadam has joined #openstack-swift | 06:58 | |
| *** yashmurty has joined #openstack-swift | 06:59 | |
| *** yashmurty has quit IRC | 06:59 | |
| *** yashmurty has joined #openstack-swift | 07:00 | |
| *** rcernin has joined #openstack-swift | 07:05 | |
| *** rcernin_ has quit IRC | 07:06 | |
| *** mlanner has quit IRC | 07:09 | |
| *** threestrands_ has joined #openstack-swift | 07:09 | |
| *** threestrands_ has quit IRC | 07:09 | |
| *** threestrands_ has joined #openstack-swift | 07:09 | |
| *** threestrands has quit IRC | 07:10 | |
| *** threestrands_ has quit IRC | 07:13 | |
| *** yashmurty has quit IRC | 07:19 | |
| *** yashmurty has joined #openstack-swift | 07:28 | |
| *** yashmurty has quit IRC | 07:29 | |
| *** yashmurty has joined #openstack-swift | 07:29 | |
| *** hseipp has joined #openstack-swift | 07:41 | |
| *** armaan has joined #openstack-swift | 07:51 | |
| *** pcaruana has joined #openstack-swift | 07:54 | |
| *** pcaruana has quit IRC | 07:55 | |
| *** pcaruana has joined #openstack-swift | 07:55 | |
| kota_ | online at Ops-MeetUp Tokyo ¥o/ | 07:57 |
| * kota_ was tighten up to my office in this morning | 07:57 | |
| *** tesseract has joined #openstack-swift | 08:22 | |
| *** hseipp has quit IRC | 08:27 | |
| *** germs has joined #openstack-swift | 08:27 | |
| *** germs has quit IRC | 08:27 | |
| *** germs has joined #openstack-swift | 08:27 | |
| *** threestrands has joined #openstack-swift | 08:32 | |
| *** threestrands has quit IRC | 08:32 | |
| *** threestrands has joined #openstack-swift | 08:32 | |
| *** germs has quit IRC | 08:32 | |
| *** hseipp has joined #openstack-swift | 08:33 | |
| *** yashmurty has quit IRC | 08:36 | |
| *** threestrands has quit IRC | 08:36 | |
| *** cshastri has quit IRC | 08:42 | |
| *** rcernin has quit IRC | 08:42 | |
| *** cshastri has joined #openstack-swift | 08:43 | |
| *** geaaru has joined #openstack-swift | 08:43 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift master: DNM: Try out to pull execution logs of functests https://review.openstack.org/548226 | 08:44 |
| *** yashmurty has joined #openstack-swift | 08:45 | |
| kota_ | tdasilva: that is one I'm still trying to pull the logs | 08:46 |
| kota_ | tdasilva: but I don't think I complete the work with the patch, yet. | 08:46 |
| acoles | good morning | 08:46 |
| kota_ | I still needs to dig out around log setup with in-process. When i tried to grep the results from /var/log/syslog, the vm for testing denied because of permission. | 08:47 |
| kota_ | acoles: morning! | 08:47 |
| acoles | hi kota_ did you make it home ok? | 08:47 |
| kota_ | it was delayed tho, I'm fine here in Tokyo now. | 08:47 |
| kota_ | how were you? | 08:48 |
| acoles | I made it home on Saturday after delays, there was also snow in my street but not too bad | 08:49 |
| *** yashmurty has quit IRC | 08:49 | |
| kota_ | sounds not so bad | 08:51 |
| kota_ | that was crazy snow tho. | 08:52 |
| *** kei_yama has quit IRC | 08:56 | |
| kota_ | acoles: do you have any good idea to dump the swift log into tox log dir at in process testing? | 09:01 |
| acoles | kota_: no, sorry. I'm just looking at a new gate probe test and wanting to see logs too! | 09:02 |
| kota_ | im digging around log_address in conf file but not yet reach the answer | 09:03 |
| kota_ | acoles: ok. no worries | 09:04 |
| *** hseipp has quit IRC | 09:11 | |
| *** hseipp has joined #openstack-swift | 09:12 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Move auto_create_account_prefix option to DEFAULT section https://review.openstack.org/550409 | 09:22 |
| *** armaan has quit IRC | 09:24 | |
| *** armaan has joined #openstack-swift | 09:25 | |
| *** armaan has quit IRC | 09:38 | |
| *** armaan has joined #openstack-swift | 09:38 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add an update_meta method to ShardRange https://review.openstack.org/550414 | 09:45 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add an update_meta method to ShardRange https://review.openstack.org/550414 | 09:48 |
| *** tovin07_ has quit IRC | 09:59 | |
| *** kukacz has quit IRC | 10:17 | |
| *** kei-ichi has quit IRC | 10:26 | |
| *** kei-ichi has joined #openstack-swift | 10:27 | |
| *** germs has joined #openstack-swift | 10:28 | |
| *** germs has quit IRC | 10:28 | |
| *** germs has joined #openstack-swift | 10:28 | |
| *** germs has quit IRC | 10:33 | |
| *** kazsh has quit IRC | 10:40 | |
| *** kazsh has joined #openstack-swift | 10:45 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add an update_meta method to ShardRange https://review.openstack.org/550414 | 11:18 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add assertions for cleaved broker own shard range state https://review.openstack.org/550436 | 11:18 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add assertions for cleaved broker own shard range state https://review.openstack.org/550436 | 11:51 |
| openstackgerrit | yanwei proposed openstack/swift master: remove the if statement https://review.openstack.org/550457 | 11:58 |
| openstackgerrit | yanwei proposed openstack/swift master: 1,According to http://www.gevent.org/gevent.html Parameters: exception (type) – The type of exception to raise in the greenlet. The default is GreenletExit, which indicates a successful() completion of the greenlet. https://review.openstack.org/550457 | 12:00 |
| *** mattoliverau has quit IRC | 12:21 | |
| *** mattoliverau has joined #openstack-swift | 12:22 | |
| *** armaan has quit IRC | 12:26 | |
| *** armaan has joined #openstack-swift | 12:26 | |
| *** psachin has quit IRC | 12:27 | |
| *** germs has joined #openstack-swift | 12:29 | |
| *** germs has quit IRC | 12:29 | |
| *** germs has joined #openstack-swift | 12:29 | |
| *** germs has quit IRC | 12:34 | |
| *** cshastri has quit IRC | 12:53 | |
| *** armaan has quit IRC | 12:55 | |
| *** armaan has joined #openstack-swift | 12:56 | |
| *** cshastri has joined #openstack-swift | 13:01 | |
| *** germs has joined #openstack-swift | 14:08 | |
| *** germs has quit IRC | 14:08 | |
| *** germs has joined #openstack-swift | 14:08 | |
| *** germs has quit IRC | 14:09 | |
| *** germs has joined #openstack-swift | 14:09 | |
| *** germs has quit IRC | 14:09 | |
| *** germs has joined #openstack-swift | 14:09 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:15 | |
| *** armaan has quit IRC | 14:16 | |
| *** armaan has joined #openstack-swift | 14:20 | |
| *** armaan has quit IRC | 14:22 | |
| *** germs has quit IRC | 14:33 | |
| *** links has quit IRC | 14:37 | |
| *** mvk has quit IRC | 14:41 | |
| *** cshastri has quit IRC | 14:51 | |
| *** cbartz has quit IRC | 14:56 | |
| m3m0 | hello, I'm using a red hat :( openstack deployment and apparently the swift deployment is using "virtual drives" on my controllers, is there a way to get the current capacity of the disk and the "expected capacity"? swift-recon -d does not return anything useful | 15:00 |
| tdasilva | m3m0: by default on a tripleo (i.e., director) deployment, swift uses the OS disk on a controller node | 15:02 |
| m3m0 | tdasilva: mmm in that case, swift-recon -d should work the same no? | 15:04 |
| m3m0 | but complains about: Error. No drive info available. | 15:05 |
| tdasilva | m3m0: I don't have a tripleo environment available to me atm to test :/ | 15:10 |
| m3m0 | tdasilva: thanks a lot anyway :) and one last question, where can I find the replica configuration for my objects | 15:13 |
| tdasilva | in the rings | 15:13 |
| tdasilva | m3m0: what openstack version are you running? | 15:13 |
| m3m0 | tdasilva: mitaka | 15:14 |
| tdasilva | oh | 15:15 |
| m3m0 | or red hat openstack 9 | 15:15 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add assertions for cleaved broker own shard range state https://review.openstack.org/550436 | 15:16 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Store own shard range in the shard ranges table https://review.openstack.org/550246 | 15:16 |
| tdasilva | m3m0: deployment has worked better in newer releases, starting with newton | 15:17 |
| m3m0 | tdasilva: indeed, I was deploying queens with docker and it was easier to manage than this one :) | 15:18 |
| m3m0 | but deploying is one thing and managing is another one :) | 15:20 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add assertions for cleaved broker own shard range state https://review.openstack.org/550436 | 15:21 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Store own shard range in the shard ranges table https://review.openstack.org/550246 | 15:21 |
| *** mvk has joined #openstack-swift | 15:30 | |
| *** germs has joined #openstack-swift | 15:31 | |
| *** germs has quit IRC | 15:31 | |
| *** germs has joined #openstack-swift | 15:31 | |
| *** armaan has joined #openstack-swift | 15:32 | |
| *** germs has quit IRC | 15:32 | |
| *** germs has joined #openstack-swift | 15:35 | |
| *** germs has quit IRC | 15:35 | |
| *** germs has joined #openstack-swift | 15:35 | |
| *** pcaruana has quit IRC | 16:10 | |
| *** guimaluf has joined #openstack-swift | 16:16 | |
| tdasilva | what happened to that s3 doc we created in denver with the supported features? | 16:20 |
| tdasilva | was it ever published somewhere? | 16:20 |
| *** itlinux has joined #openstack-swift | 16:35 | |
| *** geaaru has quit IRC | 16:47 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Store own shard range in the shard ranges table https://review.openstack.org/550246 | 16:55 |
| *** geaaru has joined #openstack-swift | 17:09 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: DO NOT MERGE Allow a container PUT with shards body to autocreate container https://review.openstack.org/550535 | 17:10 |
| *** gkadam has quit IRC | 17:11 | |
| *** armaan has quit IRC | 17:15 | |
| *** armaan has joined #openstack-swift | 17:15 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add assertions for cleaved broker own shard range state https://review.openstack.org/550436 | 17:15 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Store own shard range in the shard ranges table https://review.openstack.org/550246 | 17:15 |
| timburke | good morning | 17:21 |
| notmyname | good morning | 17:23 |
| *** cshastri has joined #openstack-swift | 17:27 | |
| *** hseipp has quit IRC | 17:29 | |
| *** armaan has quit IRC | 17:47 | |
| *** armaan has joined #openstack-swift | 17:48 | |
| *** gyee has joined #openstack-swift | 17:48 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Add assertions to probe test https://review.openstack.org/550548 | 17:52 |
| openstackgerrit | Merged openstack/swift feature/deep: Add an update_meta method to ShardRange https://review.openstack.org/550414 | 17:53 |
| *** armaan has quit IRC | 17:53 | |
| *** armaan has joined #openstack-swift | 17:54 | |
| *** pcaruana has joined #openstack-swift | 17:57 | |
| *** kallenp has joined #openstack-swift | 18:06 | |
| *** cshastri has quit IRC | 18:10 | |
| *** kallenp_ has joined #openstack-swift | 18:11 | |
| kallenp_ | test only | 18:12 |
| *** kallenp_ has quit IRC | 18:13 | |
| *** SkyRocknRoll has quit IRC | 18:29 | |
| *** mvk has quit IRC | 18:31 | |
| *** geaaru has quit IRC | 18:42 | |
| *** tesseract has quit IRC | 18:53 | |
| *** kallenp has quit IRC | 19:08 | |
| *** kallenp has joined #openstack-swift | 19:10 | |
| *** kallenp has quit IRC | 19:10 | |
| *** kallenp has joined #openstack-swift | 19:13 | |
| *** mvk has joined #openstack-swift | 19:14 | |
| *** kallenp has quit IRC | 19:19 | |
| timburke | notmyname: i had a similar concern about https://review.openstack.org/#/c/548948/ -- see the response; since the change also overrides the fmt arg, it should be ok... though it still seems a little odd. otoh, do we ever override fmt elsewhere? how do (can?) operators *not* have "%s(server)s:" included in log messages today? | 19:20 |
| patchbot | patch 548948 - swift - Add template in proxy to create custom and anonymo... | 19:20 |
| *** bartden has joined #openstack-swift | 19:20 | |
| *** kallenp has joined #openstack-swift | 19:30 | |
| *** kallenp has quit IRC | 19:34 | |
| *** kallenp has joined #openstack-swift | 19:35 | |
| *** kallenp has quit IRC | 19:35 | |
| *** kallenp has joined #openstack-swift | 19:37 | |
| *** yashmurty has joined #openstack-swift | 19:37 | |
| *** kallenp has quit IRC | 19:39 | |
| *** silor has joined #openstack-swift | 19:41 | |
| *** yashmurty has quit IRC | 19:42 | |
| *** germs_ has joined #openstack-swift | 19:52 | |
| *** germs__ has joined #openstack-swift | 19:54 | |
| *** germs has quit IRC | 19:54 | |
| *** germs_ has quit IRC | 19:57 | |
| *** germs__ is now known as germs | 19:58 | |
| *** germs has quit IRC | 19:58 | |
| *** germs has joined #openstack-swift | 19:58 | |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Persist own shard stats after reporting to root https://review.openstack.org/550586 | 20:00 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Persist own shard stats after reporting to root https://review.openstack.org/550586 | 20:01 |
| *** silor1 has joined #openstack-swift | 20:09 | |
| *** gerardgine has joined #openstack-swift | 20:09 | |
| *** silor has quit IRC | 20:12 | |
| *** silor1 is now known as silor | 20:12 | |
| *** kallenp has joined #openstack-swift | 20:14 | |
| *** kallenp has quit IRC | 20:23 | |
| *** silor1 has joined #openstack-swift | 20:31 | |
| *** silor has quit IRC | 20:32 | |
| *** silor1 is now known as silor | 20:32 | |
| *** gerardgine has quit IRC | 20:35 | |
| *** hseipp has joined #openstack-swift | 20:36 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Functional tests for staticweb middleware https://review.openstack.org/503262 | 20:38 |
| *** d0ugal has quit IRC | 20:40 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Make storage_domain configurable in func tests https://review.openstack.org/549277 | 20:41 |
| *** kallenp has joined #openstack-swift | 20:41 | |
| *** bartden has left #openstack-swift | 20:42 | |
| openstackgerrit | Tim Burke proposed openstack/swift feature/deep: Pick up the new probetests check on feature/deep https://review.openstack.org/550291 | 20:44 |
| *** d0ugal has joined #openstack-swift | 20:53 | |
| kota_ | morning | 20:55 |
| notmyname | kota_: I hope you got home ok | 20:56 |
| kota_ | notmyname: yeah, it was delayed but fine for now | 20:57 |
| openstackgerrit | Tim Burke proposed openstack/swift master: py3: have Timestamp accept bytestrings, better validate input https://review.openstack.org/550602 | 20:57 |
| mattoliverau | Morning | 20:57 |
| openstackgerrit | Merged openstack/swift feature/deep: Add assertions for cleaved broker own shard range state https://review.openstack.org/550436 | 20:59 |
| notmyname | meeting time in #openstack-meeting | 20:59 |
| *** hseipp has quit IRC | 20:59 | |
| *** hseipp has joined #openstack-swift | 21:00 | |
| timburke | acoles: a thought on sharding and sync points -- once we cleave and unlink the retired db, could we update *all* sync points to the last row of the db we just unlinked? we no longer have them to send, and we've got an assumption that the rows are now durable... | 21:01 |
| *** hseipp has quit IRC | 21:06 | |
| *** hseipp has joined #openstack-swift | 21:06 | |
| *** hseipp has quit IRC | 21:07 | |
| *** pcaruana has quit IRC | 21:12 | |
| notmyname | rledisez: we were hoping to cover https://etherpad.openstack.org/p/Swift_Operational_Feedback with you. but there's the etherpad, so maybe we can talk more in an async fashion | 21:18 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Store own shard range in the shard ranges table https://review.openstack.org/550246 | 21:24 |
| openstackgerrit | Alistair Coles proposed openstack/swift feature/deep: Persist own shard stats after reporting to root https://review.openstack.org/550586 | 21:24 |
| *** silor has quit IRC | 21:28 | |
| notmyname | rledisez: ok, consistency engine work | 21:30 |
| rledisez | notmyname: ready :) | 21:30 |
| *** pdardeau has quit IRC | 21:31 | |
| notmyname | the major stuff we (swiftstack) will be working on include the patches you saw in the meeting. this gives some of the same tuning knobs for the DB replicator as the object replicator | 21:31 |
| notmyname | we also need to add concurrency-per-disk to both the object replicator and the DB replicator | 21:31 |
| notmyname | we want to reexamine some tuning defaults, too. this *may* involve changing the baseline for some values. eg moving from "total workers" to a "workers per Foo" metric | 21:32 |
| *** gerardgine has joined #openstack-swift | 21:32 | |
| notmyname | so that the same config file can be better suited to run on different hardware | 21:32 |
| notmyname | may be -per-disk or -per-core values | 21:33 |
| notmyname | I'm being vague because we haven't really started examining that in detail yet. it's just a meaningful short-term improvement we can make and get done sooner than "remove rsync" | 21:33 |
| rledisez | one related question: we always assume ssync is slow. and it's true. did someone profile that? do we know what is slow? | 21:34 |
| notmyname | the two big things I do not thing we'll be able to get done as a community in 2018 (unless something radically changes in our community) are creating a replacement replication protocol that removes rsync and rewriting storage node servers in golang | 21:34 |
| notmyname | rledisez: clayg has some great thoughts on that, but it mostly comes down to rsync replication being decent at bulk replication (aside from the lack of backpressure from us not being in the data path) and ssync being decent and replicating individual objects | 21:35 |
| notmyname | they're respectively terrible at the other | 21:35 |
| notmyname | rledisez: IIRC ssync is pretty slow because it's sending (and managing) individual objects instead of partitions, suffixes, or anything batch-related | 21:39 |
| rledisez | notmyname: ok, i see your point | 21:40 |
| notmyname | oh, and I just had a call this morning about a customer who has a pretty small amount of storage (15 servers) but they have billions and billions of tiny objects. they're wondering how to make it faster. "well, there's this thing rledisez is working on...." ;-) | 21:41 |
| rledisez | well, i won't say it's easy to use for now, but we deployed it on 10 servers (~2PB), and will move to 100 servers in a week | 21:42 |
| notmyname | nice! | 21:42 |
| notmyname | rledisez: on the consistency engine front, torgomatic is currently working on improving the object replicator to better organize the replication work jobs. that, along with the concurrency per disk, should help make some meaningful improvement in the replication engine | 21:43 |
| notmyname | when will alecuyer be back from holiday? could we set something up for next week? | 21:44 |
| torgomatic | yeah, ssync is a chatty protocol, and it also goes through the normal diskfile paths so it's calling fsync() left and right | 21:44 |
| rledisez | notmyname: he'll be back on monday, i'll drop him a mail right now to ask him if he's available fo smthing | 21:45 |
| notmyname | ok. maybe tdasilva can host a video chat next week | 21:45 |
| rledisez | notmyname: torgomatic: if it can help, we use this on our clusters for reconstruction: https://review.openstack.org/#/c/548664/ | 21:45 |
| patchbot | patch 548664 - swift - Split REVERT and SYNC processes in reconstructor | 21:45 |
| tdasilva | notmyname, rledisez let me know when and i can set it up | 21:45 |
| timburke | starred | 21:45 |
| torgomatic | rledisez: that sounds like a good idea. I'd briefly considered sorting the jobs and doing all REVERT first since those are cheaper and might save some other node from doing a full reconstruction | 21:46 |
| torgomatic | I never got to writing code for that, though... | 21:46 |
| rledisez | BTW, what's the thing with swift-tox-py27? is there any issue? all my patches fails on that one | 21:46 |
| rledisez | or am I a terrible dev? yeah, probably true also :D | 21:46 |
| notmyname | lol | 21:46 |
| notmyname | should be the same as `.unittests` locally | 21:46 |
| torgomatic | rledisez: if it's a test about the lockup detector, you can blame someone else. Otherwise... ;) | 21:47 |
| rledisez | i'll check my SAIO, because unittests are OK for me | 21:47 |
| rledisez | torgomatic: on that subject of revert first (it's dirty, sorry :)): https://review.openstack.org/#/c/389676/ | 21:48 |
| patchbot | patch 389676 - swift - WIP: reconstructor: do not sync recently moved parts | 21:48 |
| kallenp | Anyone, who can remotely help me with OpenStack Swift implementation ? | 21:49 |
| notmyname | kallenp: we're happy to help answer specific questions, but none of us here are equipped to do professional services or contract work. however, several of us work for companies who provide those paid products | 21:50 |
| rledisez | thx for all the info notmyname and torgomatic, i'll sleep on that | 21:55 |
| notmyname | rledisez: let me know what alecuyer says and we'll get a video chat set up | 21:56 |
| rledisez | sure | 21:56 |
| *** ChanServ sets mode: +v mattoliverau | 22:00 | |
| *** mlanner has joined #openstack-swift | 22:06 | |
| *** kallenp has quit IRC | 22:10 | |
| *** kallenp has joined #openstack-swift | 22:11 | |
| *** ggine has joined #openstack-swift | 22:11 | |
| *** openstackstatus has quit IRC | 22:13 | |
| *** openstack has joined #openstack-swift | 22:15 | |
| *** ChanServ sets mode: +o openstack | 22:15 | |
| *** ggine has quit IRC | 22:35 | |
| *** itlinux has quit IRC | 22:36 | |
| *** rcernin has joined #openstack-swift | 22:37 | |
| *** kallenp has quit IRC | 22:37 | |
| *** kallenp has joined #openstack-swift | 22:38 | |
| *** kallenp has joined #openstack-swift | 22:39 | |
| *** kallenp has quit IRC | 22:39 | |
| *** openstackstatus has quit IRC | 22:42 | |
| *** openstack has joined #openstack-swift | 22:44 | |
| *** ChanServ sets mode: +o openstack | 22:44 | |
| *** openstackstatus has joined #openstack-swift | 22:44 | |
| *** ChanServ sets mode: +v openstackstatus | 22:44 | |
| *** kallenp has joined #openstack-swift | 22:57 | |
| *** armaan has quit IRC | 22:58 | |
| *** armaan has joined #openstack-swift | 22:59 | |
| *** threestrands has joined #openstack-swift | 23:12 | |
| *** kallenp has left #openstack-swift | 23:15 | |
| *** kei_yama has joined #openstack-swift | 23:19 | |
| *** armaan has quit IRC | 23:33 | |
| notmyname | http://not.mn/reviews.html <-- dashboard updated. was broken for some reason | 23:33 |
| *** armaan has joined #openstack-swift | 23:34 | |
| *** threestrands_ has joined #openstack-swift | 23:34 | |
| *** threestrands_ has quit IRC | 23:34 | |
| *** threestrands_ has joined #openstack-swift | 23:34 | |
| *** threestrands_ has quit IRC | 23:35 | |
| *** threestrands_ has joined #openstack-swift | 23:36 | |
| *** threestrands has quit IRC | 23:36 | |
| *** gerardgine has joined #openstack-swift | 23:49 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!