| *** dja has joined #openstack-swift | 00:04 | |
| tdasilva | clayg: sorry, back... | 00:30 |
|---|---|---|
| tdasilva | clayg: I would imagine something in our tree would be better just because we can control it | 00:31 |
| tdasilva | but I just followed what I already knew, plust -infra folks were super helpful | 00:31 |
| clayg | awesome! | 00:32 |
| tdasilva | not sure yet about a test that breaks against package vs master, I'll take a look at libec master and see what's landed since last release | 00:32 |
| *** zhurong has joined #openstack-swift | 00:33 | |
| *** ediardo has quit IRC | 00:52 | |
| *** vint_bra has joined #openstack-swift | 01:07 | |
| *** vint_bra has quit IRC | 01:15 | |
| *** klrmn has quit IRC | 01:15 | |
| *** JimCheung has quit IRC | 01:33 | |
| *** ediardo has joined #openstack-swift | 01:34 | |
| *** m_kazuhiro has joined #openstack-swift | 01:35 | |
| m_kazuhiro | good morning | 01:36 |
| *** Sukhdev has joined #openstack-swift | 01:48 | |
| Sukhdev | Hey Swift experts, I am hitting no space left on device error - how do I get around it? Please see the error here - http://paste.openstack.org/show/605584/ | 01:48 |
| *** jamielennox is now known as jamielennox|away | 02:05 | |
| *** chosafine has joined #openstack-swift | 02:06 | |
| *** bkopilov has quit IRC | 02:13 | |
| *** jamielennox|away is now known as jamielennox | 02:19 | |
| *** Sukhdev has quit IRC | 02:21 | |
| *** vint_bra has joined #openstack-swift | 02:29 | |
| *** vint_bra has quit IRC | 02:31 | |
| *** vint_bra has joined #openstack-swift | 03:14 | |
| *** vint_bra has quit IRC | 03:18 | |
| *** chosafine has quit IRC | 03:38 | |
| *** bkopilov has joined #openstack-swift | 03:40 | |
| *** gkadam has joined #openstack-swift | 03:40 | |
| *** gkadam has quit IRC | 03:43 | |
| *** links has joined #openstack-swift | 03:45 | |
| notmyname | zaitcev: I'm not sure it's your cultural background... | 03:49 |
| *** gkadam has joined #openstack-swift | 03:49 | |
| zaitcev | notmyname: I would prefer a document that eschews the dramatic fictialization and just states what the challenges are, what we plan to do, and what is the expected outcome. | 03:50 |
| zaitcev | Although I don't wear suit and tie to work. | 03:51 |
| notmyname | zaitcev: yeah. I read it as a blog post written two years from now. from that perspective, it's still sortof dramatic, but makes more sense. as is, it's not similar to other vision statements I've been involved with in the past | 03:52 |
| notmyname | however, I understand the intent behind it, and I don't want to disregard the work that's gone into making it | 03:52 |
| zaitcev | yeah | 03:52 |
| notmyname | oftentimes, the actual vision statement doesn't actually matter. the process by which the community comes up with it and agrees to it is much more important to long-term success | 03:52 |
| *** Sukhdev has joined #openstack-swift | 03:53 | |
| zaitcev | But some of it is really strange from my limited perspective. For example, I haven't heard of git namespaces, but anyway, who cares? Our customers just want to be sure that the installer installs things correctly, and in our case it just gets RPMs from the media. | 03:53 |
| zaitcev | Is OpenStack's user the operator or integrator? | 03:54 |
| notmyname | that's what I'm currently grappling with. it seems to be focused on openstack internals as a goal. there's not anything like "there are 30 million cores runing X Y or Z" or "there are 13 EB stored" or anything about solving cloud infrastructure | 03:55 |
| notmyname | but openstack's user is in fact the operator | 03:55 |
| notmyname | the API consumers are our user's users | 03:56 |
| notmyname | *our users' users | 03:56 |
| notmyname | but yeah, I'm struggling with it to. I need to ruminate on it for a while | 03:57 |
| zaitcev | "The new bot that converts github pull requests to gerrit change-sets, instead of discarding them, imported several patches in each of the last three months." | 04:03 |
| zaitcev | oh god | 04:03 |
| zaitcev | " The TC membership now includes several women and representatives from APAC and European countries." . . . | 04:06 |
| notmyname | I'm glad that's being called out | 04:12 |
| *** m_kazuhiro_ has joined #openstack-swift | 04:18 | |
| *** m_kazuhiro has quit IRC | 04:19 | |
| mahatic | zaitcev: it's probably not all that irrelevant for the doc, after all it talks about vision and not just exclusively of technical goals | 04:36 |
| mahatic | re: calling out women rep etc | 04:37 |
| *** abhitechie has joined #openstack-swift | 04:38 | |
| *** ediardo has quit IRC | 04:42 | |
| *** psachin_ has joined #openstack-swift | 04:47 | |
| *** abhitechie has quit IRC | 04:53 | |
| *** ChubYann has quit IRC | 05:18 | |
| *** dja has quit IRC | 05:34 | |
| *** JimCheung has joined #openstack-swift | 06:00 | |
| *** JimCheung has quit IRC | 06:05 | |
| *** ppai has joined #openstack-swift | 06:05 | |
| clayg | oh no, how did no one help Sukhdev :'( | 06:08 |
| *** cshastri has joined #openstack-swift | 06:08 | |
| ppai | Hello. Does swift has an official metadata search feature yet ? | 06:09 |
| Sukhdev | clayg : Thanks for reaching out; after some poking around, I realized that I was running out of disk space. Deleting some of the images fixed my problem | 06:10 |
| clayg | Sukhdev: if you're buried in your own tears - you've probably figured out running out of space sucks - adding capacity is your best hope -> https://bugs.launchpad.net/swift/+bug/1359160 | 06:11 |
| openstack | Launchpad bug 1359160 in OpenStack Object Storage (swift) "Problem with disk overflowing (help my swift cluster is full)" [Undecided,Expired] | 06:11 |
| clayg | Sukhdev: oh nice! sometimes it's too late to delete | 06:11 |
| clayg | there is a fallocate_reserve option you simply must use! it'll prevent PUT when close to full to save room on disk to write tombstone (DELETE) - you don't need much, few hundered meg is fine | 06:12 |
| clayg | other problem tho is tha replication can backfill since rsync is out of data path - using ssync might be an option for some people (but you definately need seperate replication services) - afaik for replicated objects monitoring and disabling rsync modules is not uncommon -> https://docs.openstack.org/developer/swift/admin_guide.html?highlight=disable%20rsync#preventing-disk-full-scenarios | 06:14 |
| Sukhdev | clayg : I tried to see if there was a way to change the config of swift to increase the size of "/dev/loop0 /opt/stack/data/swift/drives/sdb1" from 2G to something higher, but, could not find any way | 06:15 |
| *** jaosorior_away is now known as jaosorior | 06:16 | |
| clayg | ppai: nope, not yet - cc timur rledisez did we decide at PTG we were going to make SwiftStack POC elastic search integration "official"? https://github.com/swiftstack/swift-metadata-sync | 06:19 |
| ppai | clayg, Great! Thanks | 06:20 |
| clayg | Sukhdev: umm... except in single vm development environments you shouldn't use loop device for swift storage - but I guess this is devstack deployment? | 06:21 |
| Sukhdev | clayg : yes | 06:21 |
| clayg | Sukhdev: SWIFT_LOOPBACK_DISK_SIZE | 06:21 |
| clayg | probably in your local.conf | 06:22 |
| clayg | i was just diggin in https://github.com/openstack-dev/devstack/blob/master/lib/swift | 06:22 |
| Sukhdev | clayg : yes, I saw that, but, it requires unstack/stack - did not want to go through that. It would be nice to have a knob in config file so that one could modify the config and restart swift | 06:23 |
| clayg | doesn't devstack do a single replica? you could probably just cp -a sdb1 sdb1.new then unmount sdb1 and mv sdb1.new o sdb1 ; set mount_check = false ; and restart | 06:27 |
| clayg | i'm not well... no you'd want a filesystem that supports xattrs - probably xfs - idk | 06:28 |
| clayg | you could add another device | 06:29 |
| clayg | the thing is just that I don't think devstack was really designed to support ongoing mainteance and capacity growth | 06:29 |
| Sukhdev | clayg : agreed | 06:31 |
| *** kirill_ has joined #openstack-swift | 06:51 | |
| *** pcaruana has joined #openstack-swift | 07:02 | |
| *** tesseract has joined #openstack-swift | 07:04 | |
| *** rcernin has joined #openstack-swift | 07:10 | |
| *** rcernin has quit IRC | 07:10 | |
| *** rcernin has joined #openstack-swift | 07:10 | |
| clayg | notmyname: we need to get some of rledisez's and pavel's bufixes on https://wiki.openstack.org/wiki/Swift/PriorityReviews | 07:14 |
| clayg | what was the one we were talking about in the meeting today? something like 507 on a&c can fill up your cluster with container db's? | 07:15 |
| acoles | good morning | 07:22 |
| acoles | clayg: ! | 07:23 |
| clayg | acoles: had a bad day; trying to make up for it | 07:25 |
| *** Sukhdev has quit IRC | 07:26 | |
| jaosorior | acoles: hey, could you take another look at this patch https://review.openstack.org/#/c/443104/4 ? | 07:28 |
| patchbot | patch 443104 - python-swiftclient - Use generic keystone client instead of versioned one | 07:28 |
| *** geaaru has joined #openstack-swift | 07:31 | |
| acoles | clayg: :( | 07:31 |
| acoles | jaosorior: I'll try to get back to it, I am keen to understand Joel's concern, but I think he may be on vacation | 07:32 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Do not sync suffixes when remote rejects reconstructor revert https://review.openstack.org/435152 | 07:32 |
| *** amoralej|off is now known as amoralej | 07:33 | |
| *** oshritf has joined #openstack-swift | 07:41 | |
| acoles | clayg: ^^ makes sense, I remember you talking about that needless REPLICATE | 07:44 |
| acoles | clayg: I'll try to look at patch today, got to run an errand now. back later | 07:45 |
| *** oshritf has quit IRC | 07:45 | |
| *** adriant has quit IRC | 07:50 | |
| *** oshritf has joined #openstack-swift | 07:51 | |
| *** oshritf has quit IRC | 07:58 | |
| *** hseipp has joined #openstack-swift | 07:59 | |
| *** cbartz has joined #openstack-swift | 08:09 | |
| kirill_ | hello, everybody! newbie here. how do I delete account and it's data if I deleted project first? | 08:38 |
| clayg | how does one get a reseller admin token from keystone i wonder | 08:41 |
| clayg | you could setup a noauth proxy and just DELETE to http://swift-storage.example.com/v1/AUTH_<project_id> | 08:42 |
| clayg | allow_account_management = true is recommended but not the default - can't trigger account reaper w/o it tho | 08:43 |
| *** jordanP has joined #openstack-swift | 08:44 | |
| clayg | kirill_: I found this https://ask.openstack.org/en/question/53466/is-there-a-reseller-admin-role-for-swift-under-keystone/, but devstack vm `openstack role list` says 403 | 08:49 |
| kirill_ | I cannot assign reselleradmin role if there is no project | 08:59 |
| clayg | kirill_: adapted some junk i had laying around, WOMM https://gist.github.com/clayg/1813aeb207a2411c823a5e5af32e169b | 09:01 |
| clayg | kirill_: you wouldn't assign it to *the* project you're trying to delete - you'd assign it to a super-duper-admin user (or some service that integrates with decomissions) | 09:02 |
| *** hseipp has quit IRC | 09:06 | |
| clayg | mahatic: it's quite possible I didn't fully understand what I thought I tested | 09:12 |
| clayg | mahatic: I agree with | 09:12 |
| clayg | ... Archana on the surface that X-Static-Large-Object is normally set by the mw | 09:12 |
| clayg | I *thougth* when slo was out of my pipeline i just uploaded with `swift --use-slo --segment-size` | 09:13 |
| mahatic | clayg: oic, lemme test that | 09:15 |
| mahatic | and I mixed up for which case header is used in my comment to Archana | 09:15 |
| mahatic | ! | 09:15 |
| clayg | no, i'm pretty sure I can upload a manifest with swift client and no slo in the pipeline and get X-Static-Large-Object in the object metadata | 09:19 |
| clayg | i bet finding out *why* that works is going to be a gas | 09:19 |
| clayg | "not it!" | 09:19 |
| mahatic | ah okay. right, and also emitting a warning in such a case sounds right to me | 09:21 |
| kirill_ | clayg: i've assigned a ResellerAdmin role to the admin user in the domain, however when I try to DELETE, I recieve 403 forbidden | 09:21 |
| clayg | wait, wat? no swiftclient does it https://github.com/openstack/python-swiftclient/blob/2710ff255b5d3954eebdf469587bf73bbfd48431/swiftclient/service.py#L1986 !!! | 09:22 |
| clayg | i had *just* convinced myself swiftclient couldn't possibly? https://github.com/openstack/swift/blob/bc9a7075d3accf3160043e8366745a1dd895b08d/swift/common/middleware/slo.py#L1190 | 09:23 |
| clayg | kirill_: oh, maybe allow_account_management then | 09:23 |
| *** mcarden has quit IRC | 09:23 | |
| clayg | kirill_: or srly, use the internal client script - it'll totally work (is this prod?) | 09:23 |
| *** mcarden has joined #openstack-swift | 09:24 | |
| *** foutatoro has joined #openstack-swift | 09:24 | |
| kirill_ | clayg: allow_account_management is true, This is prod unfortunately | 09:27 |
| mahatic | clayg: wow that's interesting. I wonder why we even need to set it in swiftclient in the first place, I mean was there some reasoning for it.. | 09:31 |
| mahatic | because clearly it's supposed to get set by the middleware | 09:32 |
| clayg | kirill_: i'm sorry i don't know enought about how folks *normally* do account deletion with keystone to tell you what to do | 09:48 |
| clayg | you need to make DELETE request to the account, normally that's restricted to allow_account_management = true + reseller admin privledge | 09:49 |
| *** SkyRocknRoll has joined #openstack-swift | 09:51 | |
| clayg | using internal client get a proxy-app that doesn't talk to auth really isn't such a crazy thing to do - but if I was operating your cluster I'd want to test/understand https://gist.github.com/clayg/1813aeb207a2411c823a5e5af32e169b in another environment before poking something like that at my accounts | 09:51 |
| clayg | alternatively you could use swift-get-nodes and track down the account db's | 09:51 |
| clayg | pretty sure you could just send a DELETE directly to the storage node api (/dev/part/AUTH_XYZ) for any replia and replicator would propogate the 410ness | 09:53 |
| clayg | still i'd probably test it in an integration environment | 09:53 |
| kirill_ | clayg: thank you very much! I'll try to experiment with keystone service account first, and I'm happy that there are alternatives, like your script and DELETE directly to the account server. | 09:59 |
| clayg | kirill_: awesome - sounds like you're all over it - I hope you figure out the keystone service account thing - if you can write up a gist or anything please throw it back this way - or even stick it on launchpad as a bug "wherever this should be written down I coudln't find it - but this wfm ... " | 10:02 |
| *** oshritf has joined #openstack-swift | 10:03 | |
| openstackgerrit | Gábor Antal proposed openstack/swift master: Use more specific asserts in test/unit/container https://review.openstack.org/342808 | 10:03 |
| *** m_kazuhiro_ has quit IRC | 10:21 | |
| kirill_ | clayg: I've managed to DELETE account trough proxy. I had to assign user a ResellerAdmin role in service project. And now it's 410) | 10:24 |
| clayg | kirill_: did you have to edit your proxy config at all? the keystone middleware sections? Did they already mention the "ResellerAdmin role" | 10:27 |
| kirill_ | clayg: I had to enable allow_account_management. ResellerAdmin was already implemented to assign account quotas. | 10:28 |
| *** silor has joined #openstack-swift | 10:28 | |
| clayg | so... in your domain you have a username "admin" - and he gets some roles in projects by default? But apparenlty in this case keystone mw was looking for ResellerAdmin role on he swift service project (the same one you've configured to validate tokens) | 10:29 |
| *** silor1 has joined #openstack-swift | 10:29 | |
| kirill_ | clayg: exectly! | 10:29 |
| clayg | and you use "openstack" cli to manage keystone - or the python-keystoneclient cli? | 10:30 |
| clayg | ... or maybe horizon or something!? | 10:31 |
| kirill_ | clayg: I am using "openstack" cli to manage keystone. | 10:32 |
| clayg | so basically openstack role add --user admin --project swift ResellerAdmin | 10:32 |
| *** silor has quit IRC | 10:32 | |
| *** silor1 is now known as silor | 10:32 | |
| *** gkadam is now known as gkadam-mtg | 10:33 | |
| kirill_ | clayg: yes, the account swift uses to connect to the keystone | 10:35 |
| clayg | so OS_USERNAME=admin OS_PROJECT_NAME=admin openstack role list got me going in devstack | 10:35 |
| kirill_ | maybe this is not the way it has to be done, but it workes | 10:36 |
| clayg | near as I can tell from the uuids - i have a ResellerAdmin role - but role assignment list doesn't give me a good indication any project/user's are assigned with it | 10:36 |
| kirill_ | use --names | 10:37 |
| *** SkyRocknRoll has quit IRC | 10:38 | |
| *** links has quit IRC | 10:39 | |
| kirill_ | and when you issue a token, you have to specify project, in my case it is service ptoject | 10:39 |
| *** mvk has quit IRC | 10:40 | |
| clayg | kirill_: hah! yeah that was last bit i was missing | 10:41 |
| clayg | oh tfw --names - thanks | 10:42 |
| kirill_ | np:) | 10:43 |
| clayg | oh you know what tho ... and I guess this makes sense ... if I `openstack role add --user admin --project admin ResellerAdmin` then I can use the token I get from the admin project | 10:44 |
| clayg | i'm not sure if it's easier to remmeber to specify the service project or not | 10:45 |
| clayg | apparently all that matters is the token has to associate any role with *name* to that identity? | 10:45 |
| kirill_ | great! I used to assign ResellerAdmin role to every project before. | 10:50 |
| kirill_ | now I assign role once | 10:50 |
| clayg | I'ma assign it to every user I think ;) | 10:53 |
| *** links has joined #openstack-swift | 10:54 | |
| clayg | srly, I might have missed what you said about quotas tho - you only assign admin user(s) the ResellerAdmin role yeah? but you were doing it for every project... oh and then getting token/catalog scoped to that project | 10:55 |
| foutatoro | hi all, I would like to know if there are opensource frameworks that allow to build frontend web storage application with swift ? | 10:55 |
| clayg | that makes sense - if you don't have a role associated from admin user to project you have to craft the storage url by hand instead of snarfing it from the catalog | 10:55 |
| clayg | kirill_: holy crap - i understand keystone | 10:55 |
| clayg | i can't ever go to sleep | 10:55 |
| *** bkopilov has quit IRC | 10:56 | |
| clayg | foutatoro: basically anything flat files or spa ;) | 10:57 |
| clayg | foutatoro: you *can* do some cool stuff in middlewares tho | 10:57 |
| clayg | foutatoro: I was turned on to serverless framework recently and all the s3 events got me thinking about how something like an opensource asw lambda (OpenWhisk?) could integrate with swift to cool effect | 10:59 |
| foutatoro | clayg: i mean using swif sdk ? | 10:59 |
| clayg | foutatoro: sorry, this is openstack-swift not apple-swift - i'm not sure what their channel is | 10:59 |
| clayg | #swift-lang maybe | 11:00 |
| foutatoro | clayg: yeah I know I'm using openstack swift | 11:00 |
| clayg | oh rofl | 11:00 |
| clayg | swift sdk.. swift's got a pretty thin api - folks tend to think of it having more like clients or bindings | 11:01 |
| clayg | you wouldn't have a "redis sdk" - redis is a powerful service - but the api is simple... | 11:02 |
| foutatoro | I'm looking for a tool like "IBM bluemix" to build nice frontend for end users intead of using openstack-dashboard | 11:02 |
| clayg | foutatoro: I don't think I understand what you're looking for - but you shouldn't let that discourage you - just elaborate on what you want! :D | 11:02 |
| clayg | *OH* ... no wait ... you want a good swift client? Or you need like a good swift client that you can hack/fork/put-your-branding-on | 11:03 |
| foutatoro | clayg: yes | 11:04 |
| clayg | IBM Bluemix includes OpenWhisk [1] !!! yeah that's what I'm talking about 1. https://en.wikipedia.org/wiki/Bluemix | 11:05 |
| clayg | do you work on or use bluemix? | 11:05 |
| foutatoro | no I don't use bluemix | 11:05 |
| clayg | oh... | 11:05 |
| clayg | foutatoro: so maybe poke around if any of these sound like what you want https://docs.openstack.org/developer/swift/associated_projects.html#associated-projects | 11:09 |
| clayg | I don't think there is a killer goto Opensource OpenStack Swift web/gui client right now | 11:10 |
| foutatoro | clayg: thanks so much for these information | 11:11 |
| kirill_ | clayg: yeah! after assignning admin a reseller role in project I could simlpy use python-swiftclient to modify quotas. and now it't like global permition. If only I knew it before:) | 11:18 |
| clayg | mahatic: oh, ok - it's legit - we only return "X-Static-Large-Object is a reserved header" if you 1) have slo in the pipeline and 2) *don't* send multipart-manifest=put | 11:19 |
| clayg | if you send multipart-manifest=put (like swiftclient does) it doesn't really hurt/matter if you *also* send X-Static-Large-Object because the mw is going to set it anyway | 11:20 |
| *** jordanP has quit IRC | 11:20 | |
| clayg | *however* allowing the "reserved" header to get persisted when the middleware is *not* there - is - well it's just awesome ... I can't even | 11:21 |
| clayg | kirill_: I'm glad you came in here to work this out this evening man. I was no help. But I learned something! Appreicate you greatly! | 11:22 |
| kirill_ | there is another issue tough. I have some more leftover swift account for which projects have been removed. for this one I alredy managed to delete I knew the account name (AUTH_<project id>). | 11:23 |
| clayg | hehhehe | 11:24 |
| kirill_ | is there a way to get account name from .db file? | 11:24 |
| clayg | swift-account-info totally | 11:25 |
| clayg | https://github.com/notmyname/slogging/blob/master/slogging/db_stats_collector.py | 11:25 |
| clayg | ^ example of trolling the nodes and ouptut to csv | 11:25 |
| clayg | one time someone pointed out most swift operators regularlly have container db's with 10's of M of objects and most of us rarely have more than 10's of K of accounts - so it's sort of annoying we don't just keep a global list/index of all accounts (swift-account-updater) | 11:27 |
| clayg | ... but nobody really got too pumped up about writing it | 11:27 |
| mahatic | clayg: :) thanks for digging that up, you should sleep! | 11:28 |
| kirill_ | clayg: thank you very much, clayg! swift-account-info will do the job for me. | 11:34 |
| clayg | kirill_: man you are on the ball! what cluster you running - I wanna tell people they should use it | 11:34 |
| *** mvk has joined #openstack-swift | 11:35 | |
| kirill_ | clayg: ahaha! please don't))) it is an local installation. several internal customers) | 11:37 |
| *** oshritf_ has joined #openstack-swift | 11:41 | |
| kirill_ | I wonder, can I issue DELETE command through swiftclient.service.SwiftService interface? | 11:44 |
| *** oshritf has quit IRC | 11:44 | |
| *** abhitechie has joined #openstack-swift | 11:53 | |
| *** SkyRocknRoll has joined #openstack-swift | 11:55 | |
| *** abhinavtechie has joined #openstack-swift | 12:13 | |
| *** abhitechie has quit IRC | 12:13 | |
| *** klamath has joined #openstack-swift | 12:19 | |
| *** klamath has quit IRC | 12:19 | |
| *** klamath has joined #openstack-swift | 12:19 | |
| *** oshritf__ has joined #openstack-swift | 12:32 | |
| *** oshritf_ has quit IRC | 12:36 | |
| *** zhurong has quit IRC | 12:42 | |
| *** jaosorior has quit IRC | 12:52 | |
| *** jaosorior has joined #openstack-swift | 12:53 | |
| *** psachin_ has quit IRC | 12:56 | |
| *** catintheroof has joined #openstack-swift | 13:05 | |
| *** vint_bra has joined #openstack-swift | 13:08 | |
| *** links has quit IRC | 13:09 | |
| *** vint_bra has quit IRC | 13:16 | |
| *** gkadam-mtg has quit IRC | 13:21 | |
| *** stradling has joined #openstack-swift | 13:21 | |
| *** kei_yama has quit IRC | 13:25 | |
| *** chlong has joined #openstack-swift | 13:31 | |
| *** JimCheung has joined #openstack-swift | 13:35 | |
| mathiasb | acoles jrichli mahatic: I don't suppose any of you have had a chance to try out the latest barbican patch yet? | 13:39 |
| *** JimCheung has quit IRC | 13:39 | |
| *** xinli has joined #openstack-swift | 13:43 | |
| jrichli | mathiasb: I'll look tonight. | 13:43 |
| *** jistr is now known as jistr|afk | 13:47 | |
| openstackgerrit | Pavel Kvasnička proposed openstack/swift master: Container drive error results double space usage on rest drives https://review.openstack.org/454174 | 13:52 |
| *** jaosorior is now known as jaosorior_away | 13:55 | |
| *** rickflare2 has joined #openstack-swift | 14:11 | |
| *** csmart has quit IRC | 14:11 | |
| *** rickflare has quit IRC | 14:11 | |
| *** csmart1 has joined #openstack-swift | 14:11 | |
| *** rickflare2 is now known as rickflare | 14:12 | |
| *** acoles has quit IRC | 14:16 | |
| dewanee_ | quick questions about .lock files | 14:16 |
| *** acoles has joined #openstack-swift | 14:16 | |
| *** ChanServ sets mode: +v acoles | 14:16 | |
| dewanee_ | who writes them? | 14:16 |
| dewanee_ | I have a .lock in each device with a really old creation date | 14:17 |
| dewanee_ | and sometimes I line like: | 14:17 |
| dewanee_ | swift-object-reconstructor: 10.129.31.142:6000/mpath85p1/13357 Unexpected response: ":ERROR: 0 '15 seconds: /srv/node/mpath85p1/.lock'" | 14:18 |
| *** bkopilov has joined #openstack-swift | 14:28 | |
| *** ppai has quit IRC | 14:40 | |
| *** jistr|afk is now known as jistr | 14:53 | |
| *** abhinavtechie has quit IRC | 14:55 | |
| *** cshastri has quit IRC | 14:56 | |
| *** SkyRocknRoll has quit IRC | 14:56 | |
| asettle | notmyname: ooohhh ahhhh https://review.openstack.org/#/c/453557/ | 15:02 |
| patchbot | patch 453557 - openstack-manuals - [config-ref] Updates swift config files | 15:02 |
| notmyname | good morning | 15:06 |
| notmyname | asettle: thanks. I saw it yesterday but didn't get a chance to look. it's on the list for today | 15:06 |
| asettle | No worries, I just updated it and now it's done all the right things :D | 15:06 |
| asettle | Let me know if that suits you. Then I'll delete the tooling afterwards. | 15:06 |
| *** vinsh_ has quit IRC | 15:11 | |
| *** cbartz has left #openstack-swift | 15:21 | |
| *** foutatoro has quit IRC | 15:22 | |
| *** pcaruana has quit IRC | 15:28 | |
| *** kirill_ has quit IRC | 15:33 | |
| *** mvk has quit IRC | 15:33 | |
| *** vint_bra has joined #openstack-swift | 15:37 | |
| *** oshritf__ has quit IRC | 15:41 | |
| *** zhurong has joined #openstack-swift | 15:47 | |
| *** zhurong has quit IRC | 15:49 | |
| *** _JZ_ has joined #openstack-swift | 15:52 | |
| *** vint_bra has quit IRC | 15:55 | |
| *** vint_bra has joined #openstack-swift | 16:00 | |
| *** jaosorior_away has quit IRC | 16:14 | |
| *** jaosorior has joined #openstack-swift | 16:15 | |
| *** vint_bra has quit IRC | 16:26 | |
| jrichli | my understanding was that both DLO and SLO were required middleware, and they are both auto-inserted in the pipeline. But I have not had time yet today to verify that - I was going to test using the pipeline in the bug. | 16:29 |
| *** JimCheung has joined #openstack-swift | 16:30 | |
| tdasilva | jrichli: slo is not IIRC | 16:30 |
| acoles | jrichli: what tdasilva said ^^ | 16:30 |
| jrichli | ah, ok | 16:31 |
| *** mvk has joined #openstack-swift | 16:37 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Do not sync suffixes when remote rejects reconstructor revert https://review.openstack.org/435152 | 16:37 |
| *** geaaru has quit IRC | 16:42 | |
| *** jaosorior is now known as jaosorior_away | 16:48 | |
| *** klrmn has joined #openstack-swift | 16:50 | |
| *** tesseract has quit IRC | 16:50 | |
| *** chsc has joined #openstack-swift | 16:51 | |
| *** chsc has joined #openstack-swift | 16:51 | |
| *** SkyRocknRoll has joined #openstack-swift | 16:57 | |
| *** ChubYann has joined #openstack-swift | 17:01 | |
| *** vint_bra has joined #openstack-swift | 17:36 | |
| *** chlong has quit IRC | 18:04 | |
| *** Sukhdev has joined #openstack-swift | 18:18 | |
| *** chlong has joined #openstack-swift | 18:18 | |
| *** xinli has quit IRC | 19:34 | |
| *** openstack has joined #openstack-swift | 19:46 | |
| clayg | dewanee_: they are from calls to lockdir | 19:47 |
| *** xinli has joined #openstack-swift | 20:07 | |
| *** Sukhdev has quit IRC | 20:31 | |
| *** chsc has quit IRC | 20:44 | |
| *** rcernin has quit IRC | 20:44 | |
| *** chsc has joined #openstack-swift | 20:47 | |
| *** chsc has joined #openstack-swift | 20:47 | |
| *** chlong has quit IRC | 20:57 | |
| *** openstackstatus has quit IRC | 20:59 | |
| *** openstackstatus has joined #openstack-swift | 21:02 | |
| *** ChanServ sets mode: +v openstackstatus | 21:02 | |
| *** stradling has quit IRC | 21:02 | |
| *** openstackgerrit has quit IRC | 21:07 | |
| *** chlong has joined #openstack-swift | 21:10 | |
| *** silor has quit IRC | 21:23 | |
| *** Sukhdev has joined #openstack-swift | 21:29 | |
| *** sputnik13 has quit IRC | 21:33 | |
| *** sputnik13 has joined #openstack-swift | 21:34 | |
| *** catintheroof has quit IRC | 21:38 | |
| *** xinli has quit IRC | 21:49 | |
| *** SkyRocknRoll has quit IRC | 21:56 | |
| *** openstack has joined #openstack-swift | 22:02 | |
| *** chlong has quit IRC | 22:05 | |
| *** vint_bra has quit IRC | 22:09 | |
| *** sputnik13 has quit IRC | 22:10 | |
| *** sputnik13 has joined #openstack-swift | 22:12 | |
| *** vint_bra has joined #openstack-swift | 22:12 | |
| *** klamath has quit IRC | 22:24 | |
| *** sputnik13 has quit IRC | 22:37 | |
| *** sputnik13 has joined #openstack-swift | 22:38 | |
| *** sputnik13 has quit IRC | 22:39 | |
| *** sputnik13 has joined #openstack-swift | 22:40 | |
| *** chsc has quit IRC | 22:45 | |
| mattoliverau | morning | 22:49 |
| *** chsc has joined #openstack-swift | 22:50 | |
| *** openstackgerrit has joined #openstack-swift | 23:02 | |
| openstackgerrit | Colleen Murphy proposed openstack/swift master: Fix install guide filter:authtoken config example https://review.openstack.org/454386 | 23:02 |
| *** vint_bra has quit IRC | 23:21 | |
| *** adriant has joined #openstack-swift | 23:37 | |
| *** kei_yama has joined #openstack-swift | 23:39 | |
| *** chsc has quit IRC | 23:45 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!