| *** mat128 has joined #openstack-swift | 00:01 | |
| *** catintheroof has quit IRC | 00:06 | |
| *** mingyu has quit IRC | 00:11 | |
| mattoliverau | timburke: when the scanner ran and finds ranges, it writes them and then replicates itself. So the other nodes should get it's extra objects.. well not in your patch, but pre it would. But I don't think cleaving replicates the cleaving contianer, so the scanner wont get updated. but I not sure where the others got each others info. | 00:13 |
|---|---|---|
| mattoliverau | Unless someone misplaced objects did something, because there is calls to _replicate in there. | 00:13 |
| timburke | mattoliverau: the trouble i was seeing was just the scanner sending its extra objects to the other dbs. it felt weird since we're expecting them to be pretty large themselves | 00:14 |
| *** mingyu has joined #openstack-swift | 00:14 | |
| mattoliverau | yeah | 00:14 |
| mattoliverau | its a good thought. | 00:15 |
| mattoliverau | if the others are in sharding state it should be ok to send, as only the shard_db table would be updated | 00:15 |
| mattoliverau | but in unsharded it could be a bad move. | 00:15 |
| mattoliverau | this is ofcourse only on large container. so how much do we cater for them. | 00:16 |
| timburke | also, i'm really in favor of getting those shard ranges out there as fast as we can, and my gut says that waiting for max_diff object rows might take a bit... | 00:16 |
| mattoliverau | I hope 99% of the time the containers are small and continuing the shard as required. But on _very_ large containers, we could be in a sharding state for a long time, so no usyncing could be a problem as they get more and more out of date? | 00:16 |
| mattoliverau | yeah, large containers so we don't want to slow it down any more, but if we are going to be sharding for say 1 week, do we still want to replicate objects so we don't have a chance of loosing anything thats missing?? | 00:18 |
| mattoliverau | pros and cons :) | 00:19 |
| mattoliverau | maybe it does make sense to always send the ranges over first, so things can get redirected. | 00:19 |
| timburke | yeah, it's that last bit that i'm torn about. how long can we risk it? | 00:20 |
| timburke | maybe we can have that initial exchange of metadata trigger the receiver to create the shard db? | 00:20 |
| timburke | moving from unsharded to sharding | 00:20 |
| mattoliverau | I think <sharding, sharded> -> unsharded means no moving data (only pivots).. but <sharding, sharded> -> sharding means only usync.. that would be ok. | 00:21 |
| mattoliverau | so a trigger might work. Send the shard ranges first if I'm dealing with an unsharded replica. the replica broker could trigger on recieving ranges if it's unsharded | 00:23 |
| mattoliverau | we can't go backwards, sharded to unsharded, and if I'm getting shards, then I have to cleave. | 00:24 |
| mattoliverau | so I do need to move in to sharding state. | 00:24 |
| timburke | i kinda like the idea of the replicator being able to change remote state, particularly as part of the metadata exchange. db state becomes like another piece of metadata (just tracked on the filesystem instead of the normal way) | 00:25 |
| mattoliverau | yeah | 00:25 |
| mattoliverau | yeah I like it too. and we'll still replicate objects, so we're doing our best at keeping containers durable.. and at least always attempting to move in the right direction. (keeping as insync as possible). | 00:28 |
| *** rcernin has quit IRC | 00:29 | |
| mattoliverau | That would mean the replicate RPC on the other replica would need some more smarts, or at least the broker. | 00:29 |
| mattoliverau | yeah the broker, that's ok :) | 00:30 |
| mattoliverau | sorry thinking out loud :P | 00:30 |
| timburke | that's perfect :-) thanks mattoliverau! | 00:30 |
| mattoliverau | nps, thank you | 00:32 |
| timburke | might mean a little more traffic as a result of sharded -> sharded movement, but that should be a rapidly-diminishing set of differences | 00:32 |
| timburke | ok, i oughta head home. and hope for a happier gate when i come in tomorrow ;-) | 00:33 |
| *** mat128 has quit IRC | 00:34 | |
| *** itlinux has joined #openstack-swift | 00:38 | |
| mattoliverau | timburke: o/ | 00:46 |
| *** tovin07_ has joined #openstack-swift | 00:51 | |
| *** wes_dillingham has quit IRC | 00:56 | |
| *** wes_dillingham has joined #openstack-swift | 01:10 | |
| *** Sukhdev has quit IRC | 01:31 | |
| *** mat128 has joined #openstack-swift | 01:36 | |
| *** itlinux has quit IRC | 01:39 | |
| *** gyee has quit IRC | 02:09 | |
| *** mat128 has quit IRC | 02:09 | |
| *** rcernin has joined #openstack-swift | 02:21 | |
| *** itlinux has joined #openstack-swift | 02:24 | |
| *** wes_dillingham has quit IRC | 02:27 | |
| *** spotz has quit IRC | 02:29 | |
| *** spotz has joined #openstack-swift | 02:37 | |
| *** mingyu has quit IRC | 02:44 | |
| *** mingyu has joined #openstack-swift | 02:47 | |
| *** itlinux has quit IRC | 02:50 | |
| *** gkadam has joined #openstack-swift | 03:26 | |
| *** links has joined #openstack-swift | 04:02 | |
| *** mat128 has joined #openstack-swift | 04:07 | |
| *** abhitechie has quit IRC | 04:19 | |
| *** tovin07_ has quit IRC | 04:27 | |
| *** tovin07_ has joined #openstack-swift | 04:28 | |
| *** mat128 has quit IRC | 04:40 | |
| *** Sukhdev has joined #openstack-swift | 04:43 | |
| *** abhitechie has joined #openstack-swift | 04:44 | |
| *** psachin has joined #openstack-swift | 04:56 | |
| *** ianychoi has quit IRC | 05:11 | |
| *** ianychoi has joined #openstack-swift | 05:12 | |
| *** rcernin has quit IRC | 05:15 | |
| *** chsc has joined #openstack-swift | 05:18 | |
| *** chsc has joined #openstack-swift | 05:18 | |
| *** pcaruana has joined #openstack-swift | 05:24 | |
| *** pcaruana has quit IRC | 05:29 | |
| *** chsc has quit IRC | 05:29 | |
| *** Sukhdev has quit IRC | 05:37 | |
| *** cbartz has joined #openstack-swift | 05:40 | |
| *** skudlik has joined #openstack-swift | 05:48 | |
| *** skudlik has left #openstack-swift | 05:48 | |
| *** ChubYann has quit IRC | 05:50 | |
| *** rcernin has joined #openstack-swift | 05:52 | |
| *** SkyRocknRoll has quit IRC | 06:08 | |
| *** silor has joined #openstack-swift | 06:18 | |
| *** kei_yama has quit IRC | 06:30 | |
| *** kei_yama has joined #openstack-swift | 06:35 | |
| *** mat128 has joined #openstack-swift | 06:38 | |
| *** pcaruana has joined #openstack-swift | 07:04 | |
| *** mat128 has quit IRC | 07:10 | |
| *** hseipp has joined #openstack-swift | 07:14 | |
| *** tovin07_ has quit IRC | 08:03 | |
| *** tovin07_ has joined #openstack-swift | 08:04 | |
| *** geaaru has joined #openstack-swift | 08:11 | |
| *** mabrams has quit IRC | 08:30 | |
| *** Dw_Sn has joined #openstack-swift | 08:47 | |
| Dw_Sn | Hello I have a question, I added the two new luns to my current storage node, did the rebalance and eveyrhing but the partitions on those two new luns are not the same... any idea why ? | 08:49 |
| *** gkadam has quit IRC | 08:51 | |
| *** gkadam has joined #openstack-swift | 08:55 | |
| *** gkadam has quit IRC | 09:00 | |
| *** gkadam has joined #openstack-swift | 09:01 | |
| *** kei_yama has quit IRC | 09:04 | |
| *** mat128 has joined #openstack-swift | 09:08 | |
| *** gkadam_ has joined #openstack-swift | 09:21 | |
| *** gkadam has quit IRC | 09:24 | |
| *** mingyu_ has joined #openstack-swift | 09:29 | |
| *** mingyu has quit IRC | 09:32 | |
| *** mingyu has joined #openstack-swift | 09:32 | |
| *** mingyu_ has quit IRC | 09:36 | |
| *** mingyu_ has joined #openstack-swift | 09:38 | |
| *** mat128 has quit IRC | 09:41 | |
| *** mingyu has quit IRC | 09:42 | |
| *** mingyu_ has quit IRC | 09:45 | |
| *** mingyu has joined #openstack-swift | 09:58 | |
| *** abhinavtechie has joined #openstack-swift | 10:02 | |
| *** abhitechie has quit IRC | 10:02 | |
| *** abhinavtechie has quit IRC | 10:18 | |
| *** abhinavtechie has joined #openstack-swift | 10:18 | |
| *** aagrawal has joined #openstack-swift | 10:20 | |
| *** aagrawal has quit IRC | 10:22 | |
| *** abhinavtechie has quit IRC | 10:22 | |
| *** abhitechie has joined #openstack-swift | 10:22 | |
| *** mingyu has quit IRC | 10:24 | |
| *** mingyu has joined #openstack-swift | 10:29 | |
| *** abhinavtechie has joined #openstack-swift | 10:30 | |
| *** abhitechie has quit IRC | 10:32 | |
| *** mingyu has quit IRC | 10:34 | |
| *** mingyu has joined #openstack-swift | 10:34 | |
| *** abhitechie has joined #openstack-swift | 10:35 | |
| *** abhinavtechie has quit IRC | 10:35 | |
| *** mingyu has quit IRC | 10:38 | |
| *** mingyu has joined #openstack-swift | 10:38 | |
| *** mingyu has quit IRC | 10:40 | |
| *** mingyu has joined #openstack-swift | 10:42 | |
| *** abhitechie has quit IRC | 10:42 | |
| *** abhitechie has joined #openstack-swift | 10:43 | |
| *** tovin07_ has quit IRC | 10:45 | |
| *** abhinavtechie has joined #openstack-swift | 10:46 | |
| *** mingyu has quit IRC | 10:46 | |
| *** mingyu has joined #openstack-swift | 10:46 | |
| *** abhitechie has quit IRC | 10:47 | |
| *** mingyu_ has joined #openstack-swift | 10:48 | |
| *** abhinavtechie has quit IRC | 10:49 | |
| *** aagrawal has joined #openstack-swift | 10:49 | |
| *** mingyu has quit IRC | 10:51 | |
| *** aagrawal has quit IRC | 10:51 | |
| *** aagrawal has joined #openstack-swift | 10:51 | |
| *** mjseger has quit IRC | 10:52 | |
| *** aagrawal has quit IRC | 10:59 | |
| *** abhitechie has joined #openstack-swift | 11:00 | |
| *** mingyu_ has quit IRC | 11:03 | |
| *** mingyu has joined #openstack-swift | 11:40 | |
| *** mat128 has joined #openstack-swift | 11:40 | |
| *** vint_bra has joined #openstack-swift | 11:56 | |
| *** mingyu has quit IRC | 11:57 | |
| *** vint_bra1 has joined #openstack-swift | 11:58 | |
| *** abhitechie has quit IRC | 12:00 | |
| *** vint_bra has quit IRC | 12:01 | |
| *** vint_bra1 has quit IRC | 12:04 | |
| *** mat128 has quit IRC | 12:12 | |
| *** mat128 has joined #openstack-swift | 12:18 | |
| *** MVenesio has joined #openstack-swift | 12:20 | |
| *** wes_dillingham has joined #openstack-swift | 12:24 | |
| *** mingyu has joined #openstack-swift | 12:27 | |
| *** 07IAA8DSW has joined #openstack-swift | 12:29 | |
| *** 5EXAACMRJ has joined #openstack-swift | 12:29 | |
| *** 07IAA8DSW has quit IRC | 12:33 | |
| *** 5EXAACMRJ has quit IRC | 12:34 | |
| *** hoonetorg has joined #openstack-swift | 12:35 | |
| *** MVenesio has quit IRC | 12:55 | |
| *** smira has joined #openstack-swift | 12:59 | |
| *** catintheroof has joined #openstack-swift | 13:03 | |
| *** smira has quit IRC | 13:04 | |
| *** cbartz has quit IRC | 13:14 | |
| *** mat128 has quit IRC | 13:22 | |
| *** mat128 has joined #openstack-swift | 13:22 | |
| *** vint_bra has joined #openstack-swift | 13:34 | |
| *** Dw_Sn has quit IRC | 13:34 | |
| *** psachin has quit IRC | 13:36 | |
| *** links has quit IRC | 13:39 | |
| *** Dinesh_Bhor has quit IRC | 13:48 | |
| *** vint_bra has quit IRC | 14:14 | |
| *** vint_bra has joined #openstack-swift | 14:18 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:50 | |
| *** saint_ has joined #openstack-swift | 14:56 | |
| *** rcernin has quit IRC | 15:01 | |
| *** SkyRocknRoll has quit IRC | 15:01 | |
| *** chsc has joined #openstack-swift | 15:15 | |
| *** chsc has joined #openstack-swift | 15:15 | |
| *** mingyu has quit IRC | 15:21 | |
| *** mingyu has joined #openstack-swift | 15:22 | |
| *** d0ugal has joined #openstack-swift | 15:31 | |
| *** chlong has quit IRC | 15:35 | |
| *** mingyu has quit IRC | 16:08 | |
| *** zaitcev has quit IRC | 16:09 | |
| *** mingyu has joined #openstack-swift | 16:11 | |
| *** saint_ has quit IRC | 16:18 | |
| clayg | 1) using SAN arrays to provision swift capacity is not common, I've seen it done and it didn't go that great 2) each individual storage device in the swift ring gets assigned unique replicas of unique partitions, i wouldn't expect them to be the same literal partition numbers - but their should be roughly the same # of partitions on the same sized devices | 16:18 |
| *** zaitcev has joined #openstack-swift | 16:22 | |
| *** ChanServ sets mode: +v zaitcev | 16:22 | |
| *** itlinux has joined #openstack-swift | 16:23 | |
| *** hseipp has quit IRC | 16:25 | |
| *** pcaruana has quit IRC | 16:26 | |
| *** SkyRocknRoll has joined #openstack-swift | 16:44 | |
| *** itlinux has quit IRC | 16:44 | |
| *** gyee has joined #openstack-swift | 16:45 | |
| *** lnxnut_ has joined #openstack-swift | 16:46 | |
| *** itlinux has joined #openstack-swift | 16:47 | |
| *** zaitcev has quit IRC | 16:52 | |
| *** zaitcev has joined #openstack-swift | 16:54 | |
| *** ChanServ sets mode: +v zaitcev | 16:54 | |
| *** shaner has quit IRC | 16:54 | |
| *** itlinux has quit IRC | 17:02 | |
| *** mingyu_ has joined #openstack-swift | 17:03 | |
| *** mingyu_ has quit IRC | 17:03 | |
| *** mingyu_ has joined #openstack-swift | 17:04 | |
| *** mingyu has quit IRC | 17:04 | |
| *** hseipp has joined #openstack-swift | 17:07 | |
| *** zaitcev has quit IRC | 17:11 | |
| timburke | good morning | 17:12 |
| *** hseipp has quit IRC | 17:13 | |
| *** mingyu_ has quit IRC | 17:21 | |
| *** mingyu has joined #openstack-swift | 17:21 | |
| *** MooingLe1ur is now known as MooingLemur | 17:26 | |
| *** NM has joined #openstack-swift | 17:35 | |
| *** mingyu_ has joined #openstack-swift | 17:40 | |
| *** dhellmann has quit IRC | 17:41 | |
| *** mingyu has quit IRC | 17:44 | |
| *** dhellmann has joined #openstack-swift | 17:46 | |
| *** geaaru has quit IRC | 17:57 | |
| *** NM has quit IRC | 18:04 | |
| *** hoonetorg has quit IRC | 18:06 | |
| *** NM has joined #openstack-swift | 18:15 | |
| *** vint_bra has quit IRC | 19:04 | |
| *** vint_bra has joined #openstack-swift | 19:08 | |
| *** catintheroof has quit IRC | 19:20 | |
| *** catintheroof has joined #openstack-swift | 19:20 | |
| *** catintheroof has quit IRC | 19:24 | |
| *** ChubYann has joined #openstack-swift | 19:34 | |
| *** mat128 has quit IRC | 19:46 | |
| *** silor has quit IRC | 19:55 | |
| *** wes_dillingham has quit IRC | 19:57 | |
| *** SkyRocknRoll has quit IRC | 19:57 | |
| *** mingyu_ has quit IRC | 19:58 | |
| *** mingyu has joined #openstack-swift | 19:58 | |
| *** stewie925 has quit IRC | 20:22 | |
| *** Sukhdev has joined #openstack-swift | 20:49 | |
| *** mingyu has quit IRC | 21:09 | |
| clarkb | timburke: re https://review.openstack.org/#/c/508589/1 looking at the old config it looks like the job wasn't voting? I take it that means you are ready to make it voting now or am I missing a case where our conversion of zuul configs broke (mostly want to make sure I'm not missing a fialure in the conversion process) | 21:13 |
| patchbot | patch 508589 - openstack-infra/project-config - Make legacy-swift-tox-xfs-tmp-func-ec voting | 21:13 |
| *** wes_dillingham has joined #openstack-swift | 21:13 | |
| timburke | clarkb: yeah, i'm wanting to turn it into a voting job now, since it had been running happily as a non-voting job for some 6 months | 21:14 |
| timburke | i maybe should have made that more clear in the commit message :-) | 21:14 |
| *** mwheckmann has quit IRC | 21:26 | |
| *** hoonetorg has joined #openstack-swift | 21:50 | |
| *** mat128 has joined #openstack-swift | 21:51 | |
| *** shaner has joined #openstack-swift | 21:58 | |
| *** NM has quit IRC | 22:17 | |
| *** mat128 has quit IRC | 22:21 | |
| *** d0ugal has quit IRC | 22:23 | |
| *** mat128 has joined #openstack-swift | 22:57 | |
| *** mat128 has quit IRC | 22:58 | |
| *** vint_bra has quit IRC | 23:33 | |
| *** chsc has quit IRC | 23:48 | |
| *** gyee has quit IRC | 23:53 | |
| *** Sukhdev has quit IRC | 23:59 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!