| sweeper | *shrug* we use newrelic for monitoring, unless someone else has a plugin, I would have to write it. afaik NR is fairly popular so I guess more integration points are always nice? :P | 00:00 |
|---|---|---|
| *** mwheckmann has quit IRC | 00:00 | |
| *** minwoob has quit IRC | 00:02 | |
| *** torgomatic has joined #openstack-swift | 00:05 | |
| *** torgomatic has quit IRC | 00:09 | |
| *** torgomatic has joined #openstack-swift | 00:10 | |
| *** ChanServ sets mode: +v torgomatic | 00:10 | |
| *** nadeem has quit IRC | 00:10 | |
| *** haomaiwang has quit IRC | 00:30 | |
| *** ppai has joined #openstack-swift | 00:33 | |
| *** potato_farmer has joined #openstack-swift | 00:47 | |
| potato_farmer | Hi. Does anyone have any online resource links that explain the deployment of swift on multiple nodes? Over the last few days I have found many different instructions from many different sources and vendors. What would be the best way to deploy on Centos 6.6? | 00:48 |
| ppai | centos 7: http://docs.openstack.org/juno/install-guide/install/yum/content/ch_swift.html | 00:50 |
| *** mwheckmann has joined #openstack-swift | 00:51 | |
| *** esker has joined #openstack-swift | 00:54 | |
| potato_farmer | ppai, I will be installing on centos 6 (so perhaps icehouse instead of juno). I am installing on vanilla systems with nothing but the OS installed. Following that document does not get me to a place to run the required commands. | 00:56 |
| potato_farmer | ppai, there are prereqs, but I don't know what they are | 00:56 |
| ppai | yeah, you have to figure out what packages to yum install | 00:56 |
| potato_farmer | For example, do I need to install an identity service first? Or what packages contains the keystone commands I need. etc... | 00:56 |
| ppai | I'd suggest you start with no auth system and then later may be install keystone | 00:57 |
| ppai | http://docs.openstack.org/icehouse/install-guide/install/yum/content/installing-and-configuring-the-proxy-node.html | 00:58 |
| ppai | you'll need different set of packages for proxy and storage nodes | 00:59 |
| ppai | you'll need probably need .repo file to be put in /etc/yum.repos.d/ from https://repos.fedorapeople.org/repos/openstack/ | 01:02 |
| potato_farmer | ppai, thanks for the help, I'll give that a shot | 01:02 |
| *** potato_farmer has quit IRC | 01:03 | |
| *** gyee has quit IRC | 01:07 | |
| mattoliverau | Is anyone against me +A-ing some of the specs to get them landed? Specifically the ones we discussed at summit? Summit based updates can, if and in the prgress of being updated, can be added as a patch to landed specs. This way the spec.o.o site will container what we are actaully working on, not gerrit. | 01:19 |
| mattoliverau | s/container/contain/ | 01:20 |
| mattoliverau | everytime I start typing contain, it always comes out container.. #swiftdevproblems | 01:20 |
| *** ppai has quit IRC | 01:28 | |
| *** jrichli has joined #openstack-swift | 01:36 | |
| *** jkugel has joined #openstack-swift | 01:40 | |
| *** theanalyst has quit IRC | 01:55 | |
| *** theanalyst has joined #openstack-swift | 01:58 | |
| *** chenhuayi has joined #openstack-swift | 01:58 | |
| *** haomaiwang has joined #openstack-swift | 02:24 | |
| *** mwheckmann has quit IRC | 02:26 | |
| * mattoliverau goes to find some lunch | 02:29 | |
| *** jkugel has quit IRC | 02:34 | |
| *** jkugel has joined #openstack-swift | 02:35 | |
| *** bkopilov has quit IRC | 02:36 | |
| MooingLemur | peluse: ping, I'm guessing you aren't back as planned | 02:50 |
| MooingLemur | :P | 02:50 |
| *** zhonghua-lee has joined #openstack-swift | 02:52 | |
| mattoliverau | MooingLemur: haven't seen him yet | 02:53 |
| *** jrichli has quit IRC | 02:54 | |
| *** chenhuayi has quit IRC | 03:06 | |
| *** chenhuayi has joined #openstack-swift | 03:07 | |
| *** zaitcev has quit IRC | 03:46 | |
| *** ppai has joined #openstack-swift | 03:50 | |
| *** chenhuayi has quit IRC | 03:55 | |
| *** haomaiwang has quit IRC | 03:57 | |
| *** zhonghua-lee has quit IRC | 04:00 | |
| *** haomaiwang has joined #openstack-swift | 04:01 | |
| *** jaison has joined #openstack-swift | 04:05 | |
| *** bkopilov has joined #openstack-swift | 04:11 | |
| *** chenhuayi has joined #openstack-swift | 04:35 | |
| *** haomaiwang has quit IRC | 04:39 | |
| *** chenhuayi has quit IRC | 04:42 | |
| *** chenhuayi has joined #openstack-swift | 04:43 | |
| *** openstackgerrit has quit IRC | 04:50 | |
| *** openstackgerrit has joined #openstack-swift | 04:50 | |
| *** zul has quit IRC | 04:57 | |
| *** zul has joined #openstack-swift | 04:57 | |
| *** cloudm2 has quit IRC | 05:02 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:14 | |
| *** zynisch_o7 has joined #openstack-swift | 05:16 | |
| *** nadeem has joined #openstack-swift | 05:17 | |
| *** jaison has quit IRC | 05:18 | |
| *** nadeem has quit IRC | 05:20 | |
| *** fifieldt has joined #openstack-swift | 05:25 | |
| *** harlowja_ has quit IRC | 05:40 | |
| *** setmason has quit IRC | 05:44 | |
| *** zynisch_o7 has quit IRC | 05:53 | |
| *** haomaiwa_ has joined #openstack-swift | 05:56 | |
| *** mmcardle has joined #openstack-swift | 06:03 | |
| *** krykowski has joined #openstack-swift | 06:06 | |
| *** mmcardle has quit IRC | 06:08 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Imported Translations from Transifex https://review.openstack.org/186282 | 06:08 |
| *** setmason has joined #openstack-swift | 06:12 | |
| *** setmason_ has joined #openstack-swift | 06:14 | |
| *** setmason has quit IRC | 06:17 | |
| *** setmason_ is now known as setmason | 06:17 | |
| mattoliverau | cschwede: You have many +1's on SPEC patch 180914 Want to make note of the comments given and then we should land it. | 06:20 |
| patchbot | mattoliverau: https://review.openstack.org/#/c/180914/ | 06:20 |
| openstackgerrit | Merged openstack/swift-specs: Symlinks in Swift https://review.openstack.org/173609 | 06:20 |
| openstackgerrit | Merged openstack/swift-specs: Increasing ring partition power https://review.openstack.org/152935 | 06:24 |
| *** bhan has joined #openstack-swift | 06:36 | |
| *** setmason has quit IRC | 06:41 | |
| *** setmason has joined #openstack-swift | 06:42 | |
| bhan | hi, In a openstack environment, i want to use aws s3 instead of configuring swift. Is it possible? | 06:43 |
| *** yuan has joined #openstack-swift | 07:07 | |
| *** setmason has quit IRC | 07:10 | |
| *** chenhuayi has quit IRC | 07:17 | |
| *** zhonghua-lee has joined #openstack-swift | 07:27 | |
| *** jordanP has joined #openstack-swift | 07:31 | |
| *** silor has joined #openstack-swift | 07:31 | |
| *** silor has quit IRC | 07:40 | |
| *** geaaru has joined #openstack-swift | 07:45 | |
| cschwede | mattoliverau: morning! yes, you’re right - i’m going to update all my specs this week. talking about patch 180914 - i’m wondering if i should add more details (based on the comments) or less and make it more general… or add an example/to discuss section based on the comments? wdyt? | 07:49 |
| patchbot | cschwede: https://review.openstack.org/#/c/180914/ | 07:49 |
| *** SkyRocknRoll has quit IRC | 07:54 | |
| *** jistr has joined #openstack-swift | 07:54 | |
| mattoliverau | The way I see it, we've talked about it at summit and its something we will do. So it definitely can be landed. Seeing as its up, you could update it based on comments and discussions at summit. Anything new on the other hand let's put in the next review. That way we have something to show on specs.o.o | 08:02 |
| *** chenhuayi has joined #openstack-swift | 08:03 | |
| ho | bhan: unfortunately it's possible if you have network connectivity. but i recommend you to use swift with no reason :-) | 08:07 |
| *** chenhuayi has quit IRC | 08:11 | |
| *** chenhuayi has joined #openstack-swift | 08:11 | |
| *** leopoldj has joined #openstack-swift | 08:31 | |
| *** haomaiwa_ has quit IRC | 08:39 | |
| *** 7YUAAETE8 has joined #openstack-swift | 08:45 | |
| hugokuo | for adding X-Account-Sysmeta-Global-Write-Ratelimit to an account by direct_client. Should I use gen_header() ? | 08:55 |
| hugokuo | done | 09:10 |
| *** SkyRocknRoll has joined #openstack-swift | 09:16 | |
| *** Triveni has joined #openstack-swift | 09:18 | |
| *** chenhuayi has quit IRC | 09:18 | |
| *** chenhuayi has joined #openstack-swift | 09:18 | |
| *** 7YUAAETE8 has quit IRC | 09:27 | |
| *** hseipp has joined #openstack-swift | 09:37 | |
| *** hseipp has quit IRC | 09:37 | |
| *** hseipp has joined #openstack-swift | 09:37 | |
| *** chenhuayi has quit IRC | 09:53 | |
| *** zhonghua-lee has quit IRC | 10:22 | |
| *** early has quit IRC | 10:46 | |
| *** Triveni has quit IRC | 10:49 | |
| *** early has joined #openstack-swift | 10:53 | |
| *** mmcardle has joined #openstack-swift | 10:59 | |
| *** aix has joined #openstack-swift | 11:13 | |
| *** zhonghua-lee has joined #openstack-swift | 11:22 | |
| *** baffle_ is now known as baffle | 11:24 | |
| *** haomaiwang has joined #openstack-swift | 11:28 | |
| *** ho has quit IRC | 11:28 | |
| *** mwheckmann has joined #openstack-swift | 11:37 | |
| *** openstackgerrit has quit IRC | 11:39 | |
| *** openstackgerrit has joined #openstack-swift | 11:39 | |
| *** haomaiwang has quit IRC | 11:40 | |
| *** mwheckmann has quit IRC | 11:40 | |
| openstackgerrit | Tim Burke proposed openstack/python-swiftclient: Compare each chunk of large objects when downloading https://review.openstack.org/172212 | 11:42 |
| *** aix has quit IRC | 11:44 | |
| *** mmcardle has quit IRC | 11:45 | |
| *** early has quit IRC | 11:46 | |
| *** mmcardle has joined #openstack-swift | 11:47 | |
| *** early has joined #openstack-swift | 11:49 | |
| *** zhonghua-lee has quit IRC | 11:56 | |
| *** mmcardle has quit IRC | 11:58 | |
| *** aix has joined #openstack-swift | 11:59 | |
| *** jkugel has quit IRC | 12:13 | |
| *** proteusguy has quit IRC | 12:23 | |
| *** zhonghua-lee has joined #openstack-swift | 12:24 | |
| *** proteusguy has joined #openstack-swift | 12:26 | |
| *** bhan has quit IRC | 12:27 | |
| *** bkopilov has quit IRC | 12:38 | |
| tdasilva | good morning | 12:46 |
| ppai | morning tdasilva | 12:46 |
| *** ppai has quit IRC | 12:52 | |
| *** proteusguy has quit IRC | 12:53 | |
| *** zhonghua-lee has quit IRC | 12:57 | |
| *** krykowski has quit IRC | 12:59 | |
| *** krykowski has joined #openstack-swift | 12:59 | |
| *** mmcardle has joined #openstack-swift | 13:01 | |
| *** jkugel has joined #openstack-swift | 13:03 | |
| *** jkugel1 has joined #openstack-swift | 13:04 | |
| *** krykowski has quit IRC | 13:06 | |
| *** krykowski has joined #openstack-swift | 13:07 | |
| *** jkugel has quit IRC | 13:07 | |
| *** km has quit IRC | 13:08 | |
| *** fifieldt has quit IRC | 13:11 | |
| *** kei_yama has quit IRC | 13:11 | |
| *** esker has quit IRC | 13:12 | |
| *** mmcardle has quit IRC | 13:14 | |
| *** SkyRocknRoll has quit IRC | 13:14 | |
| *** jamielennox is now known as jamielennox|away | 13:29 | |
| *** krykowski has quit IRC | 13:37 | |
| *** acampbel11 has joined #openstack-swift | 13:38 | |
| *** acampbell has joined #openstack-swift | 13:38 | |
| *** setmason has joined #openstack-swift | 13:47 | |
| openstackgerrit | paul luse proposed openstack/swift: Add policy support to dispersion tools https://review.openstack.org/185459 | 13:49 |
| *** mmcardle has joined #openstack-swift | 13:56 | |
| *** admin6 has joined #openstack-swift | 13:58 | |
| *** wbhuber has joined #openstack-swift | 14:02 | |
| *** krykowski has joined #openstack-swift | 14:02 | |
| *** janonymous has joined #openstack-swift | 14:02 | |
| *** proteusguy has joined #openstack-swift | 14:03 | |
| openstackgerrit | Merged openstack/swift: Remove simplejson from swift-recon https://review.openstack.org/186169 | 14:09 |
| openstackgerrit | Merged openstack/swift: Imported Translations from Transifex https://review.openstack.org/186282 | 14:11 |
| *** krykowski has quit IRC | 14:13 | |
| *** blmartin has joined #openstack-swift | 14:19 | |
| *** breitz has quit IRC | 14:20 | |
| *** breitz has joined #openstack-swift | 14:21 | |
| *** jrichli has joined #openstack-swift | 14:22 | |
| admin6 | Hi all, in order to plan a deployment I’d like to know how I can get an average estimation of the SSD disk size I need to store the metadatas (Account and containers databases). | 14:30 |
| *** leopoldj has quit IRC | 14:30 | |
| admin6 | Nowadays, I’m storing 285TB of data, for about 375 Millions of files, with an average file siez of 800KB, and growing… | 14:32 |
| *** odsail has joined #openstack-swift | 14:38 | |
| *** mcnully has quit IRC | 14:40 | |
| openstackgerrit | Merged openstack/swift: Add process name checking into swift-init https://review.openstack.org/116203 | 14:43 |
| tdasilva | admin6: last week at the summit, HP had this presentation about their public cloud https://www.youtube.com/watch?v=wJirzmJQ7Y4 which containtained a graph showing that their account/container servers consumed about 1% of the overall capacity | 14:44 |
| tdasilva | admin6: maybe that can serve as a guide | 14:44 |
| admin6 | tdasilva: Thanks. I’ve already seen this presentation. However 1% seems to be a high value. Would that mean a database of 3TB in my current case? | 14:47 |
| *** minwoob has joined #openstack-swift | 14:47 | |
| *** mahatic has joined #openstack-swift | 14:48 | |
| *** esker has joined #openstack-swift | 14:48 | |
| tdasilva | admin6: not _one_ database as you would most likely be sharding your objects across multiples containers, but I think that number gives you a general idea of where you mind endup with overall consumption | 14:50 |
| cschwede | admin6: not a single 3TB db, but many small DBs (at least if there are not millions of objects in a single container). It really depends on the amount of accounts, containers and length of the object names - and some more | 14:50 |
| cschwede | admin6: you could summarize the disk usage of your account/container subdirectories on your current disks to get the currently used disk space | 14:52 |
| cschwede | admin6: for example using something like this: „find -L /srv/ -name containers -type d -exec du -sh {} \;“ | 14:54 |
| admin6 | cschwede: thanks, but I’m not using swift at the moment, my files are currently stored on nightmares RAID6 NAS… | 14:57 |
| cschwede | admin6: ah sorry, sure, you want to migrate. ok, can you estimate the average length of your object names? | 14:58 |
| *** mahatic_ has joined #openstack-swift | 14:59 | |
| *** mahatic has quit IRC | 15:01 | |
| *** krykowski has joined #openstack-swift | 15:01 | |
| admin6 | cschwede: gp7/2014/8/20/18/d_456266/original/73382688.jpg is a good exemple of my object name length. about 50 chars. | 15:03 |
| cschwede | admin6: how many objects do you expect on average per container? | 15:04 |
| admin6 | cschwede: as many as possible. In fact , I’ll probably limit the containers at about 1 Millions of object, just because I’ve read somewhere that it is nod a good idea to go too far over that value | 15:06 |
| cschwede | admin6: that’s true, especially if containers are stored on spinning disks. it becomes a lesser problem on ssds, but it’s still better to avoid containers with millions of objects in it | 15:07 |
| *** bkopilov has joined #openstack-swift | 15:11 | |
| *** acoles_away is now known as acoles | 15:14 | |
| *** fthiagogv has joined #openstack-swift | 15:16 | |
| acoles | notmyname: large screen usually | 15:17 |
| admin6 | cschwede: I store an average of 400.000 new files per day, so I’ll probably create a new container for each day. | 15:17 |
| *** acampbel11 has quit IRC | 15:18 | |
| *** acampbell has quit IRC | 15:18 | |
| openstackgerrit | Merged openstack/swift: Remove confusable query string on post as copy https://review.openstack.org/178927 | 15:20 |
| *** acoles is now known as acoles_away | 15:20 | |
| cschwede | admin6: i would expect at least 80MiB per container db file with a size of 400.000 objects, if you only add objects and don’t do many delete/adds on them | 15:25 |
| admin6 | cschwede: Ok, thanks. That give me a much more clear idea of my needs. :-) | 15:26 |
| cschwede | admin6: so, that would make roughly 170 TiB with 3 replicas, ignoring any other overhead (inode size, …) for 280 million objects. but that’s the lower boundary | 15:26 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 15:28 |
| *** josed has joined #openstack-swift | 15:31 | |
| admin6 | cschwede: I suppose you means 170 GiB (not TiB) with 3 replicas. | 15:33 |
| cschwede | admin6: oh yes, sorry! 170GiB. | 15:33 |
| *** mmcardle has quit IRC | 15:35 | |
| *** fthiagogv has quit IRC | 15:37 | |
| *** fthiagogv has joined #openstack-swift | 15:37 | |
| *** mmcardle has joined #openstack-swift | 15:39 | |
| *** gyee has joined #openstack-swift | 15:43 | |
| *** mwheckmann has joined #openstack-swift | 15:45 | |
| *** silor has joined #openstack-swift | 15:47 | |
| briancline | morning folks | 15:52 |
| *** mwheckmann has quit IRC | 15:53 | |
| briancline | would someone mind reviewing this? https://review.openstack.org/#/c/185221/ | 15:53 |
| *** chlong has quit IRC | 15:54 | |
| *** nadeem has joined #openstack-swift | 15:54 | |
| *** jistr has quit IRC | 15:56 | |
| *** mwheckmann has joined #openstack-swift | 15:57 | |
| *** krykowski_ has joined #openstack-swift | 16:02 | |
| *** nadeem has quit IRC | 16:03 | |
| *** krykowski has quit IRC | 16:05 | |
| *** chlong has joined #openstack-swift | 16:07 | |
| notmyname | good morning | 16:08 |
| *** nadeem has joined #openstack-swift | 16:09 | |
| notmyname | briancline: I starred it | 16:09 |
| briancline | notmyname: thank you sir | 16:11 |
| *** jordanP has quit IRC | 16:14 | |
| *** Triveni has joined #openstack-swift | 16:17 | |
| *** mmcardle has quit IRC | 16:22 | |
| *** janonymous has quit IRC | 16:27 | |
| *** Triveni has quit IRC | 16:27 | |
| *** mahatic_ has quit IRC | 16:29 | |
| *** haomaiwang has joined #openstack-swift | 16:29 | |
| *** admin6 has left #openstack-swift | 16:35 | |
| *** mahatic_ has joined #openstack-swift | 16:35 | |
| *** mahatic_ has quit IRC | 16:45 | |
| *** Triveni has joined #openstack-swift | 16:45 | |
| *** sileht has quit IRC | 16:47 | |
| *** Triveni has quit IRC | 16:52 | |
| notmyname | interesting. looks like Ironic is going to change their release model away from the milestones and going to something that looks more like Swift's pattern | 16:55 |
| notmyname | mailing list message subject is "[openstack-dev] [Ironic] [TC] Discussion: changing Ironic's release model" | 16:55 |
| *** gsilvis has quit IRC | 16:56 | |
| *** gsilvis has joined #openstack-swift | 16:58 | |
| *** acoles_away is now known as acoles | 17:01 | |
| *** aix has quit IRC | 17:04 | |
| *** gyee has quit IRC | 17:04 | |
| *** harlowja has joined #openstack-swift | 17:05 | |
| *** krykowski_ has quit IRC | 17:08 | |
| *** krykowski has joined #openstack-swift | 17:08 | |
| acoles | jrichli: clayg: did anyone remember the outcome of the whiteboard vote on what to call "massmeta"? clayg had an idea - was it x-object-sys-post-meta? or were there less hyphens? | 17:13 |
| *** zhill_ has joined #openstack-swift | 17:15 | |
| *** krykowski has quit IRC | 17:15 | |
| *** proteusguy has quit IRC | 17:19 | |
| *** hseipp has left #openstack-swift | 17:22 | |
| *** proteusguy has joined #openstack-swift | 17:23 | |
| *** sileht has joined #openstack-swift | 17:27 | |
| *** zaitcev has joined #openstack-swift | 17:30 | |
| *** ChanServ sets mode: +v zaitcev | 17:30 | |
| *** gsilvis has quit IRC | 17:36 | |
| *** gsilvis has joined #openstack-swift | 17:43 | |
| jrichli | acoles: I don't think I was there for the whiteboard vote :-/ | 17:49 |
| acoles | jrichli: when we were sat in corner discussing crypto - you were there ;) | 17:50 |
| acoles | or someone who looked very much like you | 17:50 |
| cschwede | briancline: pls have a look at patch 138342 from Hisashi as well - seems to be some somewhat similar | 17:50 |
| patchbot | cschwede: https://review.openstack.org/#/c/138342/ | 17:50 |
| *** gsilvis has quit IRC | 17:51 | |
| *** gsilvis has joined #openstack-swift | 17:53 | |
| *** cdelatte has quit IRC | 17:58 | |
| *** cdelatte has joined #openstack-swift | 17:58 | |
| *** acampbell has joined #openstack-swift | 17:59 | |
| *** acampbell has quit IRC | 18:00 | |
| *** acampbell has joined #openstack-swift | 18:00 | |
| *** acampbel11 has joined #openstack-swift | 18:00 | |
| *** gyee has joined #openstack-swift | 18:17 | |
| *** josed has quit IRC | 18:30 | |
| *** geaaru has quit IRC | 18:32 | |
| *** B4rker has joined #openstack-swift | 18:39 | |
| *** B4rker has quit IRC | 18:42 | |
| *** B4rker has joined #openstack-swift | 18:47 | |
| *** silor1 has joined #openstack-swift | 18:54 | |
| *** fthiagogv has quit IRC | 18:54 | |
| *** silor has quit IRC | 18:56 | |
| *** B4rker has quit IRC | 19:00 | |
| *** zhonghua-lee has joined #openstack-swift | 19:00 | |
| *** SkyRocknRoll has joined #openstack-swift | 19:10 | |
| *** acoles is now known as acoles_away | 19:11 | |
| *** tongli has joined #openstack-swift | 19:14 | |
| *** zhonghua-lee has quit IRC | 19:16 | |
| *** csmart has quit IRC | 19:20 | |
| *** mwheckmann has quit IRC | 19:35 | |
| *** kutija has joined #openstack-swift | 19:43 | |
| *** B4rker has joined #openstack-swift | 19:51 | |
| *** odsail has quit IRC | 19:53 | |
| *** silor1 has quit IRC | 19:54 | |
| *** charlesw has joined #openstack-swift | 19:54 | |
| *** B4rker has quit IRC | 19:58 | |
| *** csmart has joined #openstack-swift | 20:00 | |
| charlesw | Hi guys, we have a cluster across two data centers. The connection in btw is slow. PUT obj is therefore very slow, probably caused by cross-DC container update operations. We are considering decreasing the conn_timeout setting in object-server.conf in [object-server] section to some small values so that the remote container update will timeout quickly. Is there any negative impact? | 20:04 |
| *** kutija has quit IRC | 20:05 | |
| notmyname | charlesw: have you tried the write_affinity setting yet? | 20:07 |
| charlesw | yes | 20:07 |
| *** lyrrad has joined #openstack-swift | 20:08 | |
| notmyname | and that didn't work for you? | 20:08 |
| notmyname | (that's entirely possible. write_affinity isn't good in all cases) | 20:08 |
| charlesw | obj writes are local but container updates are still cross-DC | 20:08 |
| notmyname | yeah | 20:08 |
| notmyname | yet another reason for my in-progress async container updates patch... | 20:09 |
| charlesw | Link to the patch? | 20:09 |
| notmyname | I haven't pushed it yet because I still have failing tests and no new tests written :-) | 20:10 |
| notmyname | so reducing the conn_timeout would probably help. you'll probably end up with every object write resulting in an async pending file. and those are more expensive to resolve, but they aren't done in the write patch | 20:11 |
| notmyname | *path | 20:11 |
| notmyname | so you'll have to adjust your expectations of "normal" and "bad" levels of async pendings. probably will have to pay closer attention to the container servers to ensure they are healthy | 20:13 |
| notmyname | your container listings will take longer to be up to date with the objects | 20:13 |
| *** odsail has joined #openstack-swift | 20:13 | |
| notmyname | but, in general, it should all still work with very few, if any, end-user noticable things | 20:13 |
| *** kutija has joined #openstack-swift | 20:24 | |
| *** odsail has quit IRC | 20:24 | |
| charlesw | great, appreciate your feedback. Will report back about results and look forward to your patch!! | 20:29 |
| notmyname | me too :-) | 20:29 |
| *** esker has quit IRC | 20:30 | |
| *** tongli has quit IRC | 20:32 | |
| *** lyrrad has quit IRC | 20:44 | |
| *** mwheckmann has joined #openstack-swift | 20:45 | |
| *** blmartin has quit IRC | 20:45 | |
| briancline | cschwede: thanks, I had not seen that yet | 21:02 |
| *** silor has joined #openstack-swift | 21:07 | |
| *** mwheckmann has quit IRC | 21:14 | |
| *** charlesw_ has joined #openstack-swift | 21:23 | |
| *** charlesw has quit IRC | 21:25 | |
| *** charlesw_ is now known as charlesw | 21:25 | |
| clayg | i like object-sysmeta, object-meta, object-post-sysmeta [and object-post-meta] | 21:29 |
| clayg | acoles_away: jrichli: ^ | 21:29 |
| jrichli | clayg, acoles_away : I prefer the later 2, because the first two might be confusing seeing that there is already sysmeta and object meta | 21:32 |
| jrichli | clayg: I don't suppose you have been able to look into some of the functest errors, like the issue when encrypter exceeds max length on user-meta? | 21:33 |
| clayg | jrichli: yeah i mean we already have the first two, but we need a third - I think object-post-sysmeta will have synergy with a future object-post-meta (which I expect to need once we have object metadata that is updateable with a life-cycle more like sysmeta or containers) | 21:33 |
| clayg | jrichli: I have NOT! :'( | 21:33 |
| jrichli | clayg: np. I haven't been able to work on *my* stuff at all this week! | 21:34 |
| clayg | jrichli: can you do a gist or something - what's the test name - maybe i could at leave validate it's reproducable on my setup | 21:34 |
| *** jamielennox|away is now known as jamielennox | 21:34 | |
| clayg | just running it may give me an idea | 21:34 |
| clayg | bbiab | 21:34 |
| jrichli | clayg: sure, i can give you some details to make it quicker to do. can i send you mail tonight? | 21:35 |
| *** openstackgerrit has quit IRC | 21:36 | |
| *** openstackgerrit has joined #openstack-swift | 21:36 | |
| jrichli | clayg: testMetadataLengthLimits (test.functional.tests.TestFile) : I get 503: 'Service Unavailable' | 21:47 |
| *** silor has quit IRC | 21:47 | |
| jrichli | clayg: also, if you take out the special cases for Meta-Temp-Url-Key or Meta-Access-Control-Allow-Origin headers, | 21:51 |
| jrichli | then functional/tests.py:TestContainerTempurl.test_GET fails with ResponseError: 401 | 21:51 |
| *** jkugel1 has quit IRC | 21:51 | |
| *** acampbell has quit IRC | 21:56 | |
| *** acampbel11 has quit IRC | 21:56 | |
| *** charlesw has quit IRC | 22:08 | |
| mattoliverau | Morning | 22:12 |
| openstackgerrit | Michael Barton proposed openstack/swift: go: object server SYNC report newer files https://review.openstack.org/184617 | 22:18 |
| *** jrichli has quit IRC | 22:35 | |
| *** barra204 has joined #openstack-swift | 22:43 | |
| *** nadeem has quit IRC | 22:45 | |
| *** chlong has quit IRC | 22:52 | |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Remove simplejson from tests https://review.openstack.org/186590 | 22:56 |
| *** zhill_ has quit IRC | 22:56 | |
| *** km has joined #openstack-swift | 22:59 | |
| *** barra204 has quit IRC | 23:01 | |
| *** lyrrad has joined #openstack-swift | 23:01 | |
| openstackgerrit | Samuel Merritt proposed openstack/swift: Remove simplejson from tests https://review.openstack.org/186590 | 23:06 |
| *** zhill_ has joined #openstack-swift | 23:12 | |
| *** barra204 has joined #openstack-swift | 23:15 | |
| *** setmason has quit IRC | 23:18 | |
| *** proteusguy has quit IRC | 23:26 | |
| *** blmartin has joined #openstack-swift | 23:26 | |
| *** kei_yama has joined #openstack-swift | 23:29 | |
| MooingLemur | clayg: As it turns out, I lost a bunch of data in one container but lost nothing in another. Both on the same storage policy | 23:30 |
| MooingLemur | I'm confused. | 23:31 |
| MooingLemur | but here's the most confusing part, and probably a key to what's broken | 23:32 |
| MooingLemur | https://bpaste.net/show/5064e2d08c38 | 23:32 |
| MooingLemur | one partition, this one only has one object in it. Each pass, one fragment is missing, it seems to be reverting, but the file never reaches completion.. By the time the replicator comes around again, a different fragment is missing | 23:33 |
| MooingLemur | s/replicator/reconstructor/ | 23:35 |
| *** proteusguy has joined #openstack-swift | 23:43 | |
| MooingLemur | or anyone else that might be willing to help me figure out what's going on :) | 23:50 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!