| *** rcernin_ has joined #openstack-swift | 00:05 | |
| *** rcernin has quit IRC | 00:09 | |
| *** timburke_ has joined #openstack-swift | 00:25 | |
| *** ChanServ sets mode: +v timburke_ | 00:25 | |
| *** timburk__ has joined #openstack-swift | 00:29 | |
| *** ChanServ sets mode: +v timburk__ | 00:29 | |
| *** timburke_ has quit IRC | 00:30 | |
| *** gyee has quit IRC | 00:33 | |
| openstackgerrit | Tim Burke proposed openstack/swift stable/rocky: Fix stable gate https://review.opendev.org/693102 | 00:42 |
|---|---|---|
| openstackgerrit | Merged openstack/swift stable/stein: Fix stable gate https://review.opendev.org/693007 | 00:51 |
| *** timburk__ has quit IRC | 00:57 | |
| *** timburke_ has joined #openstack-swift | 01:02 | |
| *** ChanServ sets mode: +v timburke_ | 01:02 | |
| *** timburk__ has joined #openstack-swift | 01:07 | |
| *** ChanServ sets mode: +v timburk__ | 01:07 | |
| *** timburke_ has quit IRC | 01:07 | |
| kota_ | i got the reachability to irc! | 01:14 |
| *** timburk__ has quit IRC | 01:33 | |
| *** nanzha has joined #openstack-swift | 01:37 | |
| *** timburke_ has joined #openstack-swift | 01:40 | |
| *** ChanServ sets mode: +v timburke_ | 01:40 | |
| *** openstackstatus has joined #openstack-swift | 01:44 | |
| *** ChanServ sets mode: +v openstackstatus | 01:44 | |
| *** mwheckmann has joined #openstack-swift | 01:48 | |
| viks___ | My swift-recon shows the following. | 01:50 |
| viks___ | ``` | 01:50 |
| viks___ | Distribution Graph: | 01:50 |
| viks___ | 1% 3 ********************************************************************* | 01:50 |
| viks___ | 8% 3 ********************************************************************* | 01:50 |
| viks___ | 16% 1 *********************** | 01:50 |
| viks___ | 17% 1 *********************** | 01:50 |
| viks___ | 21% 1 *********************** | 01:50 |
| viks___ | 23% 1 *********************** | 01:50 |
| viks___ | 24% 1 *********************** | 01:50 |
| viks___ | 25% 1 *********************** | 01:50 |
| viks___ | 26% 1 *********************** | 01:50 |
| viks___ | 28% 1 *********************** | 01:50 |
| viks___ | 31% 2 ********************************************** | 01:50 |
| viks___ | Just wanted to confirm if this is normal.. because 2 disks are around 90%. Rest all well below. Or should all disks fill rate is almost around same range? | 01:52 |
| timburke_ | seongsoocho: hi! | 02:05 |
| seongsoocho | hi~ | 02:05 |
| timburke_ | https://github.com/swiftstack/vagrant-swift-all-in-one | 02:06 |
| clayg | viks___: if your rings have good balance, all disks should be about the same amount full | 02:08 |
| clayg | viks___: varience in the 2-5% is normaly 7-10 is quite high but maybe not indicating a problem | 02:09 |
| clayg | most likely you've changed rings (added node?) and are waiting to rebalance - you probably need to tune your replicators (workers & concurrency) | 02:09 |
| viks___ | clayg: Here is my object ring details: | 02:17 |
| viks___ | ``` | 02:17 |
| viks___ | object.builder, build version 92, id 1e06239cf36344a1934df49cd2c980fb | 02:17 |
| viks___ | 262144 partitions, 3.000000 replicas, 1 regions, 2 zones, 15 devices, 0.00 balance, 0.00 dispersion | 02:17 |
| viks___ | The minimum number of hours before a partition can be reassigned is 1 (0:00:00 remaining) | 02:17 |
| viks___ | The overload factor is 0.00% (0.000000) | 02:17 |
| viks___ | Ring file object.ring.gz not found, probably it hasn't been written yet | 02:17 |
| viks___ | Devices: id region zone ip address:port replication ip:port name weight partitions balance flags meta | 02:17 |
| viks___ | 8 1 1 10.255.252.24:6200 172.16.252.24:6200 disk3 100.00 52498 -0.00 | 02:17 |
| viks___ | 0 1 1 10.255.252.24:6200 172.16.252.24:6200 disk4 100.00 52499 0.00 | 02:17 |
| viks___ | 1 1 1 10.255.252.24:6200 172.16.252.24:6200 disk5 99.00 51974 0.00 | 02:17 |
| viks___ | 2 1 1 10.255.252.24:6200 172.16.252.24:6200 disk6 100.00 52500 0.00 | 02:17 |
| clayg | viks___: balance and dispersionlook good! | 02:19 |
| timburke_ | clayg: basically, have the get_more_nodes call include enough context that we could pre-populate https://github.com/openstack/swift/blob/2.23.0/swift/common/ring/ring.py#L411-L416 | 02:20 |
| clayg | viks___: there's a "find" command in https://docs.openstack.org/swift/latest/admin_guide.html#checking-handoff-partition-distribution you should run on your nodes | 02:23 |
| *** tkajinam has joined #openstack-swift | 02:24 | |
| clayg | all the disks in your ring say that should have 52K parts (that's kind of a lot btw) | 02:26 |
| clayg | so if you have one disk that is 99% full - but it has 100K parts the problem is just that it has a bunch of handoffs - it's replicator should be trying to shed them to the correct primaries | 02:26 |
| viks___ | clayg: In my case disk7 is having 92% ... rest all below that as seen from the distribution graph... do not have 99% filled disk.. | 02:30 |
| clayg | how many partition directories are on disk7? | 02:31 |
| timburke_ | seongsoocho: https://bugs.launchpad.net/swift/+bug/1835324 is pretty small, should be easy to reproduce the issue | 02:32 |
| openstack | Launchpad bug 1835324 in OpenStack Object Storage (swift) "Internal client's delete_object always logs 499 for 404s" [Low,Confirmed] | 02:32 |
| timburke_ | https://bugs.launchpad.net/swift/+bug/1634382 is one clayg's wanted for a while | 02:32 |
| openstack | Launchpad bug 1634382 in OpenStack Object Storage (swift) "return x-backend-storage-policy-index header from object server" [Wishlist,Confirmed] - Assigned to Christian Hugo (christianhugo) | 02:32 |
| seongsoocho | timburke_: ok. I will try it. | 02:33 |
| timburke_ | https://bugs.launchpad.net/swift/+bug/1693149 is surely a good idea, and the Ring code (as opposed to RingBuilder) already has it, so https://github.com/openstack/swift/commit/c6cebb6 would give you a good idea of what would be involved | 02:34 |
| openstack | Launchpad bug 1693149 in OpenStack Object Storage (swift) "RingBuilder.save isn't atomic" [Wishlist,Confirmed] | 02:34 |
| *** alexlecuyer-ch has joined #openstack-swift | 02:35 | |
| viks___ | clayg: | 02:37 |
| viks___ | ``` | 02:37 |
| viks___ | sudo find -L /srv/node/disk7 -maxdepth 3 -type d -wholename "*objects/*" | wc -l | 02:37 |
| viks___ | 39451 | 02:37 |
| viks___ | ``` | 02:37 |
| timburke_ | so, thinking about server reloads, we've got https://github.com/eventlet/eventlet/pull/590 (so our HUP currently kills the connection) and https://review.opendev.org/#/c/691423/ should let us have seamless reloads with the same socket rather than needing to close & re-open | 02:39 |
| patchbot | patch 691423 - swift - Seamlessly reload servers with SIGUSR1 - 8 patch sets | 02:39 |
| clayg | viks___: that's odd, that seems like significantly LESS partitions than the ring has assigned to it?! but you say it's nearly full!? | 02:43 |
| clayg | how many parts are on the other less full disks? Maybe I'm missing something 🤔 | 02:43 |
| openstackgerrit | Tim Burke proposed openstack/swift stable/stein: Authors/changelog for 2.21.1 https://review.opendev.org/690699 | 02:46 |
| *** diablo_rojo has joined #openstack-swift | 02:47 | |
| viks___ | clayg: should it match exactly the same number which is in the ring ? | 02:48 |
| clayg | viks___: when the cluster is settled and stable - yes! | 02:48 |
| viks___ | the below is for 18% filled disk: | 02:49 |
| viks___ | ``` | 02:49 |
| viks___ | sudo find -L /srv/node/disk6 -maxdepth 3 -type d -wholename "*objects/*" | wc -l | 02:49 |
| viks___ | 39918 | 02:49 |
| viks___ | ``` | 02:49 |
| clayg | !!! So you have one disk with 39K parts thats 92% full - and other disk with 39K parts that's only 17% full!? | 02:49 |
| openstack | clayg: Error: "!!" is not a valid command. | 02:49 |
| clayg | viks___: are these disks the same GB? | 02:50 |
| viks___ | yes... | 02:50 |
| *** alexlecuyer-ch has quit IRC | 02:50 | |
| *** alexlecuyer-ch has joined #openstack-swift | 02:50 | |
| viks___ | i'm checking on vagrant vm's.. so all the disk sizes are 2GB | 02:52 |
| clayg | vagrant? vms? 2GB disks... with 39K parts on them!? viks___ i'm getting confused 🤣 | 02:53 |
| viks___ | just for testing purpose i have created this setup.... i mean i'm using part_power 18... is that the problem? i mean i should'nt use that at all? can you clarify? | 02:57 |
| clayg | testing is great! | 03:09 |
| clayg | so i guess that makes sense - part power 18 is too big to get even distribution unless you have a millon or so objects in your "cluster" | 03:11 |
| clayg | and it sounds like you'd fill up your disks before then | 03:11 |
| clayg | part power 10 would probably be better for 15 disks | 03:11 |
| clayg | 👍 | 03:11 |
| viks___ | clayg: Oh ok...Thanks a lot.. Just to understand, how to migrate from 18 part power to 10 without fresh resetup.. is it doable? | 03:16 |
| zaitcev | No | 03:22 |
| clayg | 👋 @ zaitcev 😁 | 03:25 |
| clayg | zaitcev: rledisez says he thinks we can still make put+post work | 03:26 |
| zaitcev | clayg: does anyone need it? we managed to complete py3 with changes to eventlet. | 03:27 |
| clayg | I thought I'd given up - but then timburke_ reminded me we can't get rid of eventlet unless we fix EC PUT to not need the 100-continue behavior and then rledisez convineced me that we could probably make it work | 03:27 |
| viks___ | zaitcev: clayg.. Ok... Thanks.. | 03:27 |
| clayg | viks___: 😢 sorry - it took us a long to get a story for part power *increase* - going the other direction has less usecases in production | 03:28 |
| viks___ | clayg: Ok | 03:29 |
| zaitcev | The remaining hurdle was actually... hmm... I forgot now. I remember you brought up some issue with a behavior change, but I cannot remember what it was. | 03:30 |
| zaitcev | Ah | 03:30 |
| zaitcev | It was the lack of -HUP | 03:31 |
| zaitcev | Before, if HUP came while PUT was proceeding, it completed. But now, PUT completes, but POST does not, and the temp file disappears. On-disk consistency is not violated, but users see 500. | 03:32 |
| zaitcev | Did Romain resolve it? | 03:32 |
| *** psachin has joined #openstack-swift | 03:33 | |
| clayg | well, he had some idea about it otmp somewhere then link somewhere then move it later - i didn't quite follow - but he convinced me it would work | 03:33 |
| zaitcev | No, that was his suggestion for another issue: getting into the same inode group as the directory. | 03:33 |
| zaitcev | So theoretically he proposed a path, but I didn't implement it, only wrote down. | 03:34 |
| *** psachin has joined #openstack-swift | 03:35 | |
| clayg | zaitcev: sounds like you understood him better than me - I thought it was a NEW idea | 03:36 |
| *** alexlecuyer-ch has quit IRC | 03:36 | |
| zaitcev | There may be a misunderstanding. Did you go to Shanghai? | 03:38 |
| zaitcev | or we can hash it out on IRC | 03:38 |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: Rename "tests" directory to be "test" like in the swift repo https://review.opendev.org/687773 | 03:40 |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: v1auth: support endpoint_data_for() api https://review.opendev.org/687774 | 03:40 |
| timburke_ | zaitcev: yeah, clayg and rledisez are both in Shanghai. along with kota_, alecuyer, cschwede and i | 03:43 |
| *** alexlecuyer-ch has joined #openstack-swift | 03:44 | |
| *** alexlecuyer-ch has joined #openstack-swift | 03:45 | |
| *** tkajinam has quit IRC | 03:45 | |
| timburke_ | sounded like the idea was to have the PUT use O_TMPFILE to get the file in the right inode group, flush, and rather than keeping the fd around, go link it into the tmp directory. hand the client back the tmpfile name, so when the POST comes along we can go open it up again, add metadata, link it into place in the hashdir | 03:46 |
| zaitcev | So who is getting the HUP - proxy or object server? | 03:48 |
| clayg | seongsoocho: https://github.com/swiftstack/vagrant-swift-all-in-one/issues/87 | 03:50 |
| clayg | seongsoocho: I filed that, feel free to add detials or put your diff up as a PR! | 03:50 |
| seongsoocho | clayg: OK thanks! | 03:50 |
| alexlecuyer-ch | https://etherpad.openstack.org/p/swift-ptg-shanghai | 03:52 |
| timburke_ | zaitcev: i think we've convinced ourselves that the troublesome case is the object-server -- proxy *should* just keep processing until we've sent the reply back to the client | 03:54 |
| timburke_ | clayg: so i think that status field in the account_stat table actually *does* still matter... https://github.com/openstack/swift/blob/master/swift/account/backend.py#L269-L277 | 03:58 |
| *** diablo_rojo__ has joined #openstack-swift | 04:03 | |
| *** diablo_rojo has quit IRC | 04:04 | |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: Rename "tests" directory to be "test" like in the swift repo https://review.opendev.org/687773 | 04:04 |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: v1auth: support endpoint_data_for() api https://review.opendev.org/687774 | 04:04 |
| *** mwheckmann has quit IRC | 04:07 | |
| *** diablo_rojo__ has quit IRC | 04:12 | |
| openstackgerrit | Romain LE DISEZ proposed openstack/swift master: proxy: stop sending chunks to objects with a Queue https://review.opendev.org/693116 | 04:14 |
| *** alexlecuyer-ch has quit IRC | 04:22 | |
| *** timburke_ has quit IRC | 04:25 | |
| *** timburke_ has joined #openstack-swift | 04:27 | |
| *** ChanServ sets mode: +v timburke_ | 04:27 | |
| *** timburke_ has quit IRC | 04:32 | |
| *** timburke_ has joined #openstack-swift | 05:06 | |
| *** ChanServ sets mode: +v timburke_ | 05:06 | |
| *** alexlecuyer-ch has joined #openstack-swift | 05:10 | |
| *** alexlecuyer-ch has quit IRC | 05:12 | |
| *** alexlecuyer-ch has joined #openstack-swift | 05:19 | |
| alexlecuyer-ch | a few notes about how we may use larger (12+TB) on object servers https://etherpad.openstack.org/p/swift-ptg-shanghai-large-drives | 05:34 |
| *** irclogbot_0 has quit IRC | 05:34 | |
| *** irclogbot_1 has joined #openstack-swift | 05:38 | |
| *** tkajinam has joined #openstack-swift | 05:39 | |
| *** diablo_rojo has joined #openstack-swift | 05:40 | |
| *** diablo_rojo has quit IRC | 05:52 | |
| *** diablo_rojo has joined #openstack-swift | 05:53 | |
| *** timburke_ has quit IRC | 05:54 | |
| *** alexlecuyer-ch has quit IRC | 05:54 | |
| *** pcaruana has joined #openstack-swift | 05:59 | |
| *** timburke_ has joined #openstack-swift | 06:02 | |
| *** ChanServ sets mode: +v timburke_ | 06:02 | |
| mattoliverau | hey, it's cool to see more of you on IRC. I assume the PTG side of things has finally started ;) | 06:17 |
| *** nanzha has quit IRC | 06:24 | |
| mattoliverau | alecuyer: interesting. Wish I was there. | 06:25 |
| *** nanzha has joined #openstack-swift | 06:26 | |
| *** timburke_ has quit IRC | 06:26 | |
| *** timburke_ has joined #openstack-swift | 06:27 | |
| *** ChanServ sets mode: +v timburke_ | 06:27 | |
| *** timburke_ has quit IRC | 06:28 | |
| *** timburke_ has joined #openstack-swift | 06:28 | |
| *** ChanServ sets mode: +v timburke_ | 06:28 | |
| *** timburk__ has joined #openstack-swift | 06:29 | |
| *** ChanServ sets mode: +v timburk__ | 06:29 | |
| *** timburke_ has quit IRC | 06:33 | |
| *** alexlecuyer-ch has joined #openstack-swift | 06:49 | |
| *** alexlecuyer-ch has quit IRC | 06:51 | |
| *** alexlecuyer-ch has joined #openstack-swift | 06:51 | |
| *** nanzha has quit IRC | 06:58 | |
| *** nanzha has joined #openstack-swift | 07:00 | |
| *** tkajinam has quit IRC | 07:08 | |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift feature/losf: Erase object header if we fail to register it to the index-server https://review.opendev.org/693127 | 07:11 |
| *** diablo_rojo__ has joined #openstack-swift | 07:18 | |
| *** fungi has quit IRC | 07:20 | |
| *** diablo_rojo__ has quit IRC | 07:20 | |
| *** diablo_rojo__ has joined #openstack-swift | 07:20 | |
| *** diablo_rojo has quit IRC | 07:20 | |
| *** diablo_rojo__ is now known as diablo_rojo | 07:21 | |
| *** fungi has joined #openstack-swift | 07:24 | |
| clayg | mattoliverau: yes, we miss you!!! | 07:26 |
| openstackgerrit | Romain LE DISEZ proposed openstack/swift master: proxy: stop sending chunks to objects with a Queue https://review.opendev.org/693116 | 07:33 |
| *** diablo_rojo__ has joined #openstack-swift | 07:34 | |
| *** diablo_rojo__ has quit IRC | 07:35 | |
| *** diablo_rojo__ has joined #openstack-swift | 07:35 | |
| *** diablo_rojo has quit IRC | 07:35 | |
| *** diablo_rojoooooo has joined #openstack-swift | 07:37 | |
| *** diablo_rojo__ has quit IRC | 07:38 | |
| openstackgerrit | SeongSoo Cho proposed openstack/swift master: Fix bug internal client's delete_object logs 499 https://review.opendev.org/693128 | 07:40 |
| *** diablo_rojo__ has joined #openstack-swift | 07:41 | |
| *** diablo_rojoooooo has quit IRC | 07:45 | |
| *** diablo_rojo__ has quit IRC | 07:47 | |
| *** rdejoux has joined #openstack-swift | 08:04 | |
| *** rcernin_ has quit IRC | 08:06 | |
| *** tesseract has joined #openstack-swift | 08:08 | |
| openstackgerrit | SeongSoo Cho proposed openstack/swift master: Fix bug internal client's delete_object logs 499 https://review.opendev.org/693128 | 08:18 |
| *** nanzha has quit IRC | 08:35 | |
| *** nanzha has joined #openstack-swift | 08:36 | |
| openstackgerrit | SeongSoo Cho proposed openstack/swift master: Fix bug internal client's delete_object logs 499 https://review.opendev.org/693128 | 08:52 |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: v1auth: support endpoint_data_for() api https://review.opendev.org/687774 | 08:55 |
| *** timburk__ has quit IRC | 08:55 | |
| openstackgerrit | Alexandre Lécuyer proposed openstack/swift feature/losf: Erase object header if we fail to register it to the index-server https://review.opendev.org/693127 | 08:57 |
| *** alexlecuyer-ch has quit IRC | 08:57 | |
| *** mikecmpbll has joined #openstack-swift | 09:13 | |
| zigo | In the doc, I can read: | 09:46 |
| zigo | "The auto_shard option is currently NOT recommended for production systems and shoud be set to false" | 09:46 |
| zigo | Is this still truth? | 09:46 |
| *** timburke_ has joined #openstack-swift | 09:46 | |
| *** ChanServ sets mode: +v timburke_ | 09:46 | |
| zigo | timburke: ^ | 09:47 |
| openstackgerrit | Tim Burke proposed openstack/swift master: Fix bug internal client's delete_object logs 499 https://review.opendev.org/693128 | 09:49 |
| timburke_ | zigo: eh, what? sorry, not on my bouncer at the moment | 09:52 |
| *** timburke_ has quit IRC | 09:55 | |
| *** timburke_ has joined #openstack-swift | 09:56 | |
| *** ChanServ sets mode: +v timburke_ | 09:56 | |
| zigo | timburke_: I was wondering if "The auto_shard option is currently NOT recommended for production systems and shoud be set to false" in the doc is still valid, or if auto_shard is ok in production these days. | 09:56 |
| *** nanzha has quit IRC | 09:57 | |
| *** nanzha has joined #openstack-swift | 09:57 | |
| *** timburke_ has quit IRC | 09:57 | |
| *** timburke_ has joined #openstack-swift | 09:58 | |
| *** ChanServ sets mode: +v timburke_ | 09:58 | |
| zigo | timburke_: I was wondering if "The auto_shard option is currently NOT recommended for production systems and shoud be set to false" in the doc is still valid, or if auto_shard is ok in production these days. | 09:58 |
| timburke_ | i don't know of anyone currently running it in pod with that turned on. i know there are some concerns about what might happen during rebalances | 10:01 |
| timburke_ | the auto-sharding relies on replica index to choose a leader, so during a rebalance there's a chance that you'd get a second node thinking it ought to be the leader... | 10:01 |
| timburke_ | i thinki like the plan outlined in https://review.opendev.org/#/c/667579/ though! | 10:02 |
| patchbot | patch 667579 - swift - sharding: first attempt at _elect_leader - 5 patch sets | 10:02 |
| timburke_ | sorry, gotta go -- dinner time | 10:03 |
| zigo | Have a good evening! | 10:03 |
| *** timburke_ has quit IRC | 10:04 | |
| *** ianychoi has quit IRC | 10:13 | |
| *** timburke_ has joined #openstack-swift | 10:46 | |
| *** ChanServ sets mode: +v timburke_ | 10:46 | |
| *** timburke_ has quit IRC | 10:51 | |
| *** rcernin_ has joined #openstack-swift | 10:55 | |
| *** rcernin_ has quit IRC | 11:31 | |
| *** timburke_ has joined #openstack-swift | 11:50 | |
| *** ChanServ sets mode: +v timburke_ | 11:50 | |
| *** timburke_ has quit IRC | 11:54 | |
| openstackgerrit | Merged openstack/swift master: Fix bug internal client's delete_object logs 499 https://review.opendev.org/693128 | 12:14 |
| *** tomha has joined #openstack-swift | 12:29 | |
| *** timburke_ has joined #openstack-swift | 12:31 | |
| *** ChanServ sets mode: +v timburke_ | 12:31 | |
| *** timburke_ has quit IRC | 12:35 | |
| *** mikecmpbll has quit IRC | 12:39 | |
| *** mikecmpbll has joined #openstack-swift | 12:52 | |
| *** timburke_ has joined #openstack-swift | 13:01 | |
| *** ChanServ sets mode: +v timburke_ | 13:01 | |
| *** timburke_ has quit IRC | 13:06 | |
| *** timburke_ has joined #openstack-swift | 14:34 | |
| *** ChanServ sets mode: +v timburke_ | 14:34 | |
| *** alexlecuyer-ch has joined #openstack-swift | 14:41 | |
| *** timburk__ has joined #openstack-swift | 14:45 | |
| *** ChanServ sets mode: +v timburk__ | 14:45 | |
| *** alexlecuyer-ch has quit IRC | 14:45 | |
| *** timburke_ has quit IRC | 14:45 | |
| *** renich has joined #openstack-swift | 14:48 | |
| *** alexlecuyer-ch has joined #openstack-swift | 14:48 | |
| *** alexlecuyer-ch has quit IRC | 14:48 | |
| *** mikecmpbll has quit IRC | 14:56 | |
| *** renich_ has joined #openstack-swift | 15:02 | |
| *** renich has quit IRC | 15:03 | |
| *** timburk__ has quit IRC | 15:16 | |
| *** timburke_ has joined #openstack-swift | 15:26 | |
| *** ChanServ sets mode: +v timburke_ | 15:26 | |
| *** timburke_ has quit IRC | 15:30 | |
| *** gyee has joined #openstack-swift | 15:45 | |
| *** timburke_ has joined #openstack-swift | 16:06 | |
| *** ChanServ sets mode: +v timburke_ | 16:06 | |
| *** timburke_ has quit IRC | 16:11 | |
| *** nanzha has quit IRC | 16:40 | |
| *** psachin has quit IRC | 17:06 | |
| *** renich has joined #openstack-swift | 17:14 | |
| *** renich_ has quit IRC | 17:15 | |
| *** rdejoux has quit IRC | 17:38 | |
| openstackgerrit | Gilles Biannic proposed openstack/swift master: probe: Add test for syncing a delete when the remote 404s https://review.opendev.org/691463 | 17:43 |
| openstackgerrit | Gilles Biannic proposed openstack/swift master: Stop retrying every deletes in container-sync https://review.opendev.org/691729 | 17:43 |
| *** renich has quit IRC | 18:24 | |
| DHE | is there an ideal version of python for swift? if I'm installing a new system just for swift and had full flexibility in which version I'm using, what's suggested? 2.7? 3.6? | 19:09 |
| *** viks___ has quit IRC | 19:27 | |
| *** nanzha has joined #openstack-swift | 20:07 | |
| *** nanzha has quit IRC | 20:12 | |
| openstackgerrit | Merged openstack/swift stable/rocky: Fix stable gate https://review.opendev.org/693102 | 20:16 |
| *** timburke_ has joined #openstack-swift | 21:08 | |
| *** ChanServ sets mode: +v timburke_ | 21:08 | |
| *** pcaruana has quit IRC | 21:27 | |
| *** renich has joined #openstack-swift | 22:15 | |
| *** tesseract has quit IRC | 22:28 | |
| *** rcernin has joined #openstack-swift | 22:53 | |
| *** threestrands has joined #openstack-swift | 23:07 | |
| *** renich has quit IRC | 23:16 | |
| *** timburke_ has quit IRC | 23:54 | |
| *** timburke_ has joined #openstack-swift | 23:54 | |
| *** ChanServ sets mode: +v timburke_ | 23:54 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!