| *** tonanhngo has quit IRC | 00:03 | |
| *** tonanhngo has joined #openstack-swift | 00:04 | |
| *** tonanhngo has quit IRC | 00:08 | |
| *** gyee has quit IRC | 00:30 | |
| *** alenavolk has joined #openstack-swift | 00:30 | |
| *** jamielennox is now known as jamielennox|away | 00:44 | |
| *** tovin07_ has joined #openstack-swift | 00:48 | |
| *** neonpastor has quit IRC | 00:50 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Prefer auth_uri in functests configs https://review.openstack.org/469984 | 00:51 |
|---|---|---|
| *** alenavolk has quit IRC | 00:58 | |
| *** zhengyin has joined #openstack-swift | 01:00 | |
| *** klamath has quit IRC | 01:09 | |
| *** klamath has joined #openstack-swift | 01:10 | |
| *** klamath has quit IRC | 01:28 | |
| *** klamath has joined #openstack-swift | 01:29 | |
| kota_ | good morning | 01:49 |
| *** jamielennox|away is now known as jamielennox | 02:00 | |
| *** alenavolk has joined #openstack-swift | 02:12 | |
| *** alenavolk has quit IRC | 02:14 | |
| *** alenavolk has joined #openstack-swift | 02:16 | |
| *** alenavolk has joined #openstack-swift | 02:17 | |
| *** links has joined #openstack-swift | 02:22 | |
| *** alenavolk has quit IRC | 02:28 | |
| *** alenavolk has joined #openstack-swift | 02:29 | |
| *** alenavolk has quit IRC | 02:29 | |
| *** alenavolk has joined #openstack-swift | 02:29 | |
| *** alenavolk has joined #openstack-swift | 02:30 | |
| *** alenavolk has quit IRC | 02:30 | |
| *** klamath has quit IRC | 02:44 | |
| *** jamielennox is now known as jamielennox|away | 03:30 | |
| *** psachin has joined #openstack-swift | 03:45 | |
| *** oshritf has joined #openstack-swift | 04:02 | |
| *** oshritf has quit IRC | 04:07 | |
| openstackgerrit | Lingxian Kong proposed openstack/swift master: [WIP] Write-affinity aware object deletion https://review.openstack.org/470158 | 04:36 |
| *** chsc has joined #openstack-swift | 04:36 | |
| *** chsc has joined #openstack-swift | 04:36 | |
| *** jaosorior has joined #openstack-swift | 04:41 | |
| *** chsc has quit IRC | 04:42 | |
| notmyname | tdasilva: clayg: ^^ FYI on that patch | 05:03 |
| *** tovin07_ has quit IRC | 05:11 | |
| *** zhengyin has quit IRC | 05:12 | |
| *** zhengyin has joined #openstack-swift | 05:13 | |
| *** rcernin has joined #openstack-swift | 05:26 | |
| *** jeffl has joined #openstack-swift | 05:26 | |
| *** zhengyin has quit IRC | 05:33 | |
| *** zhengyin has joined #openstack-swift | 05:34 | |
| *** ChubYann has quit IRC | 05:49 | |
| *** cshastri has joined #openstack-swift | 05:50 | |
| *** sanchitmalhotra has quit IRC | 05:59 | |
| *** kei_yama has quit IRC | 06:16 | |
| *** skudlik has joined #openstack-swift | 06:26 | |
| *** kei_yama has joined #openstack-swift | 06:30 | |
| *** aselius has quit IRC | 06:32 | |
| *** hseipp has joined #openstack-swift | 06:42 | |
| jeffl | \quit | 06:51 |
| *** jeffl has quit IRC | 06:51 | |
| *** klrmn has quit IRC | 07:02 | |
| *** tesseract has joined #openstack-swift | 07:04 | |
| *** aselius has joined #openstack-swift | 07:34 | |
| *** sanchitmalhotra has joined #openstack-swift | 07:41 | |
| *** hseipp has quit IRC | 08:13 | |
| *** cbartz has joined #openstack-swift | 08:15 | |
| *** jamielennox|away is now known as jamielennox | 08:21 | |
| *** joeljwright has joined #openstack-swift | 08:22 | |
| *** ChanServ sets mode: +v joeljwright | 08:22 | |
| acoles | mattoliverau: hi! so sounds like the change at https://review.openstack.org/#/c/450491/13/swift/container/sharder.py@a471 is wrong - we *should* be applying the current meta timestamp to the new pivot range object because we are copying the bytes and object count across to it? | 08:46 |
| patchbot | patch 450491 - swift - fixup! Add container sharding to Swift containers | 08:46 |
| *** jamielennox is now known as jamielennox|away | 08:49 | |
| *** oshritf has joined #openstack-swift | 08:54 | |
| *** oshritf has quit IRC | 09:03 | |
| *** oshritf has joined #openstack-swift | 09:06 | |
| *** oshritf has quit IRC | 09:16 | |
| mattoliverau | Umm, no shouldn't we be setting a new timestamp because we are grabbing the latest versions of the stats. Tho I agree maybe we should be smarter so we don't always have to send an update | 09:21 |
| *** bikmak has joined #openstack-swift | 09:23 | |
| *** oshritf has joined #openstack-swift | 09:25 | |
| *** hseipp has joined #openstack-swift | 09:41 | |
| *** aselius has quit IRC | 09:44 | |
| acoles | mattoliverau: ok thanks, so looks like that patch is good to squash | 09:46 |
| *** kei_yama has quit IRC | 09:49 | |
| openstackgerrit | Merged openstack/swift master: Remove threads_per_disk from object-server.conf manpages https://review.openstack.org/470013 | 09:53 |
| *** jaosorior is now known as jaosorior_lunch | 10:01 | |
| *** silor has joined #openstack-swift | 10:04 | |
| *** oshritf has quit IRC | 10:10 | |
| *** mat128 has joined #openstack-swift | 10:45 | |
| *** mingyu has joined #openstack-swift | 10:46 | |
| openstackgerrit | Christopher Bartz proposed openstack/swift master: Exclude containers for account quota https://review.openstack.org/414232 | 10:48 |
| *** mingyu has quit IRC | 10:57 | |
| *** mingyu_ has joined #openstack-swift | 10:58 | |
| *** mingyu_ has quit IRC | 11:03 | |
| *** mingyu_ has joined #openstack-swift | 11:08 | |
| *** zhengyin has quit IRC | 11:14 | |
| *** mingyu_ has quit IRC | 11:25 | |
| *** Dw_Sn has joined #openstack-swift | 12:00 | |
| Dw_Sn | Hello, I have a n00b question, I want to make my buckets private and generate a signed URL for the user with expire date "QSA / Query String Authentication", I am exposing S3 API and using keystone for auth. what I need to look for ? | 12:01 |
| *** catintheroof has joined #openstack-swift | 12:08 | |
| Dw_Sn | is tempurl the answer ? and would it work with no issues along side keystone ? | 12:16 |
| acoles | Dw_Sn: sounds like tempurl is what you are looking for, it should work fine alongside keystone auth, make sure tempurl middleware is before (to left of) keystoneauth in the proxy server pipeline conf as per the sample configs | 12:40 |
| cbartz | Dw_Sn: tempurl should work along side keystone. Make sure to set delay_auth_decision = true in the proxy-server.conf . See here https://docs.openstack.org/developer/swift/api/temporary_url_middleware.html and here https://docs.openstack.org/developer/swift/overview_auth.html#configuring-swift-to-use-keystone | 12:41 |
| *** jaosorior_lunch is now known as jaosorior | 13:03 | |
| *** mdrabe has joined #openstack-swift | 13:10 | |
| tdasilva | good morning | 13:12 |
| *** lucasxu has joined #openstack-swift | 13:20 | |
| *** Dinesh_Bhor has quit IRC | 13:24 | |
| *** psachin has quit IRC | 13:26 | |
| *** links has quit IRC | 13:35 | |
| acoles | tdasilva: o/ | 13:36 |
| tdasilva | acoles: hello! 0/ | 13:36 |
| Dw_Sn | thanks guys :) | 13:46 |
| *** skudlik has quit IRC | 13:49 | |
| *** MVenesio has joined #openstack-swift | 13:51 | |
| Dw_Sn | one final question, in tempurl sig, the key value is the Secret from EC2 ? | 13:53 |
| *** klamath has joined #openstack-swift | 14:01 | |
| *** klamath has joined #openstack-swift | 14:01 | |
| Dw_Sn | nevermind found it :) I need to generate a new key :) | 14:07 |
| *** mingyu has joined #openstack-swift | 14:14 | |
| *** oshritf has joined #openstack-swift | 14:19 | |
| Dw_Sn | I am sorry again guys, I got a bit lost, I added a key to X-Container-Meta-Temp-URL-Key , do I need to set a full acocunt or cluster key using swift post -m "Temp-URL-Key:MYKEY" ? | 14:20 |
| *** mingyu has quit IRC | 14:34 | |
| *** lucasxu has quit IRC | 14:34 | |
| *** lucasxu has joined #openstack-swift | 14:36 | |
| *** oshritf has quit IRC | 14:41 | |
| Dw_Sn | I added the Temp-URL-Key and still getting Username/Password Authentication failed.. | 14:44 |
| *** catintheroof has quit IRC | 14:45 | |
| acoles | Dw_Sn: it shouldn't be necessary to have an account temp-url-key as well as container | 14:49 |
| Dw_Sn | acoles: I am following the links sent and now I have only Temp-URL-Key but getting auth failed | 14:49 |
| Dw_Sn | I can see Meta Temp-Url-Key: XXXXXXXXXXXXXXX when I do swift stat | 14:50 |
| acoles | Dw_Sn: to confirm container key is set, 'swift stat <container name>' will show the temp url key | 14:50 |
| acoles | Dw_Sn: ok you got it :) | 14:50 |
| Dw_Sn | acoles: mmm okay I will add it to the container only and see | 14:52 |
| Dw_Sn | acoles: maybe I need to have it on both ? | 14:53 |
| Dw_Sn | acoles: so I set Temp-URL-Key and container level one " same key" works perfectly now :) | 14:56 |
| *** cshastri has quit IRC | 14:58 | |
| *** ^andrea^ has joined #openstack-swift | 15:02 | |
| Dw_Sn | thanks acoles cbartz | 15:05 |
| cbartz | yw | 15:05 |
| *** rcernin has quit IRC | 15:05 | |
| Dw_Sn | one final question :) the temp_url_sig and other one, can't be modified to look like amazon AWS s3 one :) ? | 15:07 |
| *** bikmak has quit IRC | 15:10 | |
| *** klrmn has joined #openstack-swift | 15:11 | |
| acoles | Dw_Sn: glad it worked, but I don't think you should need both account and container level keys set to access an object in the same container as the container key is set | 15:14 |
| acoles | Dw_Sn: re s3, no idea sorry | 15:14 |
| timburke | Dw_Sn: swift3 supports s3-style pre-signed requests | 15:17 |
| *** gyee has joined #openstack-swift | 15:27 | |
| *** lucasxu has quit IRC | 15:38 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Prefer auth_uri in functests configs https://review.openstack.org/469984 | 15:45 |
| *** klamath has quit IRC | 15:45 | |
| ^andrea^ | can I install swift, (on CentOS 7) without installing the whole openstack environment? | 15:56 |
| ^andrea^ | Hi (first of all) :-D | 15:56 |
| *** cbartz has left #openstack-swift | 15:58 | |
| etienneme | Hi you probably want keystone too for the auth :) | 15:58 |
| ^andrea^ | OK, so I can use only the bits I need from openstack, without installing EVERYTHING... | 15:59 |
| ^andrea^ | I'll investigate further, thanks | 16:00 |
| *** tesseract has quit IRC | 16:00 | |
| etienneme | Yep | 16:00 |
| *** aselius has joined #openstack-swift | 16:01 | |
| *** eckesicle has joined #openstack-swift | 16:02 | |
| Dw_Sn | timburke: sorry, in which sense ? using same values as S3 ? | 16:02 |
| jrichli | ^andrea^ : You can setup swift standalone. And if you only want to setup a swift cluster for testing, you can use https://github.com/swiftstack/vagrant-swift-all-in-one | 16:02 |
| Dw_Sn | acoles: don't know, I tired to have either one of them, didn't work at all :/ | 16:02 |
| Dw_Sn | timburke: file.pdf?AWSAccessKeyId=XXXXXXXXXXXXXX&Expires=1523458993&Signature=XXXXXXXXXXXXXXXXX what will be AWSAccessKeyID here from swift ? | 16:04 |
| clarkb | clayg: your comments yesterday about how logn it takes to vagrant up a devstack got me looking at the swift functional job. And related to ^andrea^'s question above I thinik taht test job runs all of the default devstack services when ti really only needs keystone and swift. So we might be able to trim that down a bit at least for your use case | 16:05 |
| acoles | ^andrea^: you can use swift standalone and many people do, with either keystone or another auth service. Along with jrichli 's suggestion there is a 'swift-all-in-one' guide (for single node experimentation) https://docs.openstack.org/developer/swift/development_saio.html | 16:05 |
| *** lisek has joined #openstack-swift | 16:08 | |
| lisek | Hello. Does anybody know why object replicator doesn't use hashes.pkl for handoff partitions? | 16:08 |
| *** klrmn has quit IRC | 16:09 | |
| ^andrea^ | thank you all, that should get me started! ;-) | 16:09 |
| notmyname | good morning | 16:12 |
| ^andrea^ | and, since I'm here, let me say what I'm trying to achieve and if it makes sense | 16:12 |
| ^andrea^ | basically I/we supports lots of products/sites for a company | 16:13 |
| ^andrea^ | and every site has its own single NFS VM | 16:13 |
| ^andrea^ | so I want to get rid of all the single NFS servers, and replace it with something like swift/ceph/glusterfs... | 16:13 |
| ^andrea^ | *replace them | 16:14 |
| notmyname | ^andrea^: nice | 16:14 |
| ^andrea^ | object storage seems the way to go, which would rule out glusterfs | 16:14 |
| notmyname | ^andrea^: your biggest challenge will be to get the existing users to switch to use an object storage API instead of a filesystem interface | 16:15 |
| ^andrea^ | ceph seems good too, but there doesn't seem to be any Windows client... | 16:15 |
| ^andrea^ | notmyname: good point | 16:16 |
| *** pcaruana has quit IRC | 16:16 | |
| ^andrea^ | I assume glusterfs would be better with that... | 16:17 |
| *** Dw_Sn has quit IRC | 16:17 | |
| notmyname | I think object storage is an obviously better way to manage a bunch of storage (especially when compared to a bunch of NFS servers). you get better scalability, lower cost, higher throughput. all good things | 16:18 |
| ^andrea^ | but the applications/sites will need to be changed "more"... | 16:19 |
| notmyname | but I'd be crazy not to acknowledging the realities of how people are using the storage | 16:19 |
| ^andrea^ | yep | 16:19 |
| notmyname | so here's my high-level summary of your options | 16:19 |
| notmyname | (1) swift. really really great object storage. very stable and mature. does object storage exceptionally well and has a ton of great features. it's object storage only, but there are a few ways of getting a filesystem interface to it (including one proprietary that my employer is making) | 16:20 |
| notmyname | (2) gluster. distributed filesystem. pretty good if you just need a filesystem for your data (of the options I'm listing, I'm least familiar with gluster) | 16:21 |
| *** chsc has joined #openstack-swift | 16:21 | |
| *** chsc has joined #openstack-swift | 16:21 | |
| ^andrea^ | initially I've also asked the infrastructure guys (I'm the linux guy) for an NFS share coming directly from the storage, but we don't have a NAS header for that... (this was my option 0) | 16:22 |
| notmyname | (3) ceph. another good object storage system, although it was built to solve a different problem than swift. ceph is most commonly used as a block-storage provider, and they've recently made their filesystem features "production ready". | 16:22 |
| notmyname | one thing to note swift ceph is that the data in object, block, and file interfaces is *not* shared (ie you can't store as object and read as filesystem) | 16:23 |
| notmyname | (4) various proprietary object storage systems. don't do this ;-). they're expensive and all have copied both the swift and S3 API | 16:23 |
| ^andrea^ | how about Windows compatibility? | 16:25 |
| ^andrea^ | let's say the windows guys could use this storage for a 2TB samba share... | 16:25 |
| ^andrea^ | ceph says they don't have a Windows client | 16:25 |
| ^andrea^ | is Swift better at that? | 16:25 |
| notmyname | windows compat as in running on a windows server or as in the client? | 16:26 |
| notmyname | all of these options are built to run on linux, but the clients that are making the API calls can live anywhere | 16:26 |
| ^andrea^ | like maybe having a Windows VM using this HA storage cluster | 16:26 |
| notmyname | there are both open source and proprietary clients on all platforms | 16:27 |
| ^andrea^ | "mounting" it somehow, and offering as Windows shares to other systems... | 16:27 |
| ^andrea^ | OK | 16:27 |
| ^andrea^ | I took enough of your time | 16:27 |
| notmyname | to completely unfairly trivialize the work needed to have a windows client, all it needs to do is make HTTP requests to the storage system | 16:27 |
| ^andrea^ | thank you, you gave me a very good overview, now I only need to go and study it all :-D | 16:28 |
| notmyname | but if you're specifically looking for samba clients, if the storage system supports the SMB protocol, then "any" windows SMB client (ie windows itself) would work, right? | 16:28 |
| clarkb | clayg: left comment on that change, I think there is a second place you need to deal with cafile (possibly can be consolidated with the current one?) | 16:28 |
| notmyname | ^andrea^: please feel free to stay around and ask questions :-) | 16:28 |
| clayg | clarkb: maybe | 16:28 |
| *** hseipp has quit IRC | 16:29 | |
| *** mdrabe has quit IRC | 16:30 | |
| clayg | lisek: RE: hashes.pkl on handoffs -> http://eavesdrop.openstack.org/irclogs/%23openstack-dev/%23openstack-dev.2017-06-02.log.html#t2017-06-02T16:24:53 | 16:30 |
| *** jaosorior has quit IRC | 16:30 | |
| ^andrea^ | that's OK, forget about windows for now, I'm over complicating it... :-) | 16:31 |
| ^andrea^ | but my primary concern is replacing the NFS single point of failure/unrebootable monsters! | 16:32 |
| notmyname | ^andrea^: yep. a noble goal :-) | 16:32 |
| notmyname | ^andrea^: what kind of data is on the NFS mounts? is it like home directories? or is it more like shared storage? or is the data even shared between users? | 16:33 |
| clayg | those tracebacks are still coming from swiftclient - i looked at the swift_test_client.Connection and it uses httplib.HTTP[s]Connection's directly... but now I'm seeing it still uses swiftclient to talk to auth - so you're right | 16:33 |
| clayg | clarkb: I'm not really *helping* you here you know? | 16:34 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Prefer auth_uri in functests configs https://review.openstack.org/469984 | 16:41 |
| *** mdrabe has joined #openstack-swift | 16:44 | |
| *** silor1 has joined #openstack-swift | 16:44 | |
| *** silor has quit IRC | 16:46 | |
| *** silor1 is now known as silor | 16:46 | |
| *** mingyu has joined #openstack-swift | 16:50 | |
| ^andrea^ | notmyname: generally they are for shared storage amongst application servers | 16:52 |
| ^andrea^ | but they could be used for anything really... | 16:52 |
| ^andrea^ | thanks again for your time | 16:54 |
| *** mingyu has quit IRC | 16:54 | |
| *** lucasxu has joined #openstack-swift | 17:00 | |
| *** klrmn has joined #openstack-swift | 17:04 | |
| *** joeljwright has quit IRC | 17:15 | |
| notmyname | weird (linux) container question: I set up an lxd container, install some stuff (ie swift), shutdown the contaienr, snapshot it, start a second container from the snapshot, and the 2nd does *not* have .pyc files and the version of swift that's installed is reported differently | 17:19 |
| clarkb | clayg: progress! http://logs.openstack.org/74/372374/29/check/gate-swift-dsvm-functional-ubuntu-xenial-nv/f7fc35b/ it ran tests and some of them passed, I think there is another place where the cert needs to be passed though. ile "test/functional/__init__.py", line 1005, in get_url_token | 17:31 |
| *** mingyu has joined #openstack-swift | 17:35 | |
| *** mingyu has quit IRC | 17:40 | |
| clayg | yeah that's quite a few more failures than I was seeing locally when I was testing directly w/o the venv | 17:54 |
| clayg | clarkb: if I split up the auth_uri config fix form the cacert plumbing - can you add another depends-on to the super test everything change thingy? | 17:55 |
| clarkb | clayg: yup | 17:56 |
| clarkb | clayg: I'm also working on making your functional job run a bit quicker but not running the rest of the cloud (which should make your local reproductions simpler) | 17:56 |
| notmyname | cool | 17:57 |
| *** geaaru has joined #openstack-swift | 18:04 | |
| notmyname | timburke: for https://bugs.launchpad.net/bugs/1695273 sounds like it should be checking that the first element is in swift.valid_api_version | 18:06 |
| openstack | Launchpad bug 1695273 in OpenStack Object Storage (swift) "Subrequests from ratelimit for /auth/v1.0 considered as proxy-server errors" [Undecided,New] | 18:06 |
| *** vint_bra has joined #openstack-swift | 18:12 | |
| timburke | notmyname: i'd need to look at where we recommend placement wrt cname_lookup and domain_remap before feeling comfortable with that change | 18:18 |
| timburke | ...aaaand we don't. make recommendations, that is. :-/ | 18:20 |
| clayg | clarkb: I feel like I already have my local config only setup swift & keystone (and tempest I guess) - it's not that quick | 18:32 |
| *** catintheroof has joined #openstack-swift | 18:38 | |
| *** mingyu has joined #openstack-swift | 18:41 | |
| *** lucasxu has quit IRC | 18:52 | |
| *** catintheroof has quit IRC | 19:02 | |
| *** lucasxu has joined #openstack-swift | 19:04 | |
| clarkb | hrm I would expect it to be much quicker if just those services | 19:04 |
| *** klrmn has quit IRC | 19:08 | |
| *** klrmn has joined #openstack-swift | 19:08 | |
| *** gyee has quit IRC | 19:14 | |
| *** gyee has joined #openstack-swift | 19:17 | |
| clayg | yeah - idk, maybe I keep my laptop busy doing other things - the game seems to build nodes in ~20mins | 19:18 |
| clayg | *gate | 19:19 |
| *** mingyu has quit IRC | 19:19 | |
| clarkb | ya but that 20 minutes is all the services too. ANyways think I sorted out the problem with my initial attempt at trimming it down for the functional tests. Now to see if the run | 19:19 |
| *** mingyu has joined #openstack-swift | 19:20 | |
| *** eckesicle has quit IRC | 19:21 | |
| *** ChubYann has joined #openstack-swift | 19:43 | |
| *** mingyu has quit IRC | 19:52 | |
| *** mingyu has joined #openstack-swift | 19:53 | |
| *** wasmum has joined #openstack-swift | 19:55 | |
| *** mingyu has quit IRC | 19:57 | |
| clarkb | clayg: ok https://review.openstack.org/470383 and https://review.openstack.org/470421 should result in what we got at http://logs.openstack.org/83/470383/2/check/gate-swift-dsvm-functional-ubuntu-xenial-nv/4edd279/console.html (10 minute devstack time) | 20:00 |
| patchbot | patch 470383 - openstack-infra/devstack-gate - List swift and keystone in feature.yaml | 20:00 |
| patchbot | patch 470421 - openstack-infra/project-config - Only run swift and keystone in swift func jobs | 20:00 |
| clarkb | notmyname: ^ you maybe interested in those cahngse just to make sure I don't berak anything based on assuming only keystone and swift are necessary there | 20:01 |
| *** wasmum has quit IRC | 20:03 | |
| *** silor has quit IRC | 20:11 | |
| *** rcernin has joined #openstack-swift | 20:12 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift master: Return 404 on a GET if tombstone is newer https://review.openstack.org/371150 | 20:13 |
| *** wasmum has joined #openstack-swift | 20:14 | |
| *** silor has joined #openstack-swift | 20:18 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Prefer auth_uri in functests configs https://review.openstack.org/469984 | 20:20 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Add support for cacert plumbing via cafile in the test.conf https://review.openstack.org/470427 | 20:20 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Prefer auth_uri in functests configs https://review.openstack.org/469984 | 20:23 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Add support for cacert plumbing via cafile in the test.conf https://review.openstack.org/470427 | 20:24 |
| notmyname | clarkb: oh I'm interested :-) | 20:28 |
| notmyname | been following your conversations with clayg | 20:28 |
| clarkb | clayg: gerrit says 470427 is on top of patchset 6 of 469984 not patchset 7 (I think you can just hit the reabse button in gerrit to fix) | 20:30 |
| clayg | i only updated the commit message - does it have to be "fixed" ? | 20:31 |
| clayg | ... i mean I'll hit the button - i love buttons | 20:31 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Add support for cacert plumbing via cafile in the test.conf https://review.openstack.org/470427 | 20:32 |
| clarkb | gerrit wo't let you merge if you dep on an old patchset | 20:32 |
| clayg | BOOM | 20:32 |
| clarkb | bceause only the current patchset can merge in gerrit | 20:32 |
| clayg | oh, ok - guess that'd be a bummer if it had a +A on it then... | 20:32 |
| *** lucasxu has quit IRC | 20:45 | |
| clayg | clarkb: i keep loosing the change with all the depends - will you let me know when it fails? I'll try to classify the ~20 failures and see if which thing to fix next gets us the biggest bang for our buck | 20:46 |
| *** lucasxu has joined #openstack-swift | 20:47 | |
| clarkb | clayg: can do, there is a test backlog so it hasn't even started the functest job yet | 20:47 |
| *** lucasxu has quit IRC | 20:48 | |
| *** lucasxu has joined #openstack-swift | 20:48 | |
| *** lucasxu has quit IRC | 20:49 | |
| clayg | tdasilva: timburke: jrichli: you folks are all generally nicer/more-helpful than I am and I'm kinda petered out - I know at somepoint clarkb is going to need us to take on lp bug #1667833 - can I guilt anyone else into stealing time from other stuff to work on swift SSL functests in the gate? | 20:49 |
| openstack | Launchpad bug 1667833 in OpenStack Object Storage (swift) "func tests must not assert accept-ranges on 304" [Undecided,New] https://launchpad.net/bugs/1667833 | 20:49 |
| *** mat128 has quit IRC | 21:10 | |
| openstackgerrit | Pavel Kvasnička proposed openstack/swift master: WIP: Objects are deleted from additional handoffs https://review.openstack.org/470440 | 21:12 |
| clayg | PavelK!!!! you're... not in here.... | 21:12 |
| clayg | How you gunna be *working on swift* and your not online. Also this can *not* be normal hours for you? | 21:13 |
| mattoliverau | PavelK is the Phantom of the Swift channel | 21:27 |
| *** geaaru has quit IRC | 21:27 | |
| clarkb | clayg: http://logs.openstack.org/74/372374/30/check/gate-swift-dsvm-functional-ubuntu-xenial-nv/a811a96/ | 21:32 |
| clarkb | Failed: 20 | 21:32 |
| clayg | nice! | 21:32 |
| clarkb | http://logs.openstack.org/74/372374/30/check/gate-swift-dsvm-functional-ubuntu-xenial-nv/a811a96/console.html#_2017-06-02_21_26_00_776751 seems like maybe apache doesn't likethe inputs? | 21:37 |
| *** MVenesio has quit IRC | 21:39 | |
| *** mdrabe has quit IRC | 21:47 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements https://review.openstack.org/88736 | 22:09 |
| *** zacksh_ has joined #openstack-swift | 22:31 | |
| *** klamath has joined #openstack-swift | 22:33 | |
| *** sgundur| has joined #openstack-swift | 22:35 | |
| *** tonanhngo has joined #openstack-swift | 22:38 | |
| *** rcernin has quit IRC | 22:39 | |
| *** ejat is now known as fenris | 22:39 | |
| *** fenris is now known as ejat | 22:39 | |
| *** sgundur has quit IRC | 22:40 | |
| *** rickflare has quit IRC | 22:40 | |
| *** rickflare2 has joined #openstack-swift | 22:40 | |
| *** zacksh has quit IRC | 22:40 | |
| *** portante has quit IRC | 22:42 | |
| *** mtreinish has quit IRC | 22:42 | |
| *** portante has joined #openstack-swift | 22:44 | |
| *** mvk has quit IRC | 22:47 | |
| *** mtreinish has joined #openstack-swift | 22:47 | |
| *** klamath has quit IRC | 22:49 | |
| *** mingyu has joined #openstack-swift | 22:54 | |
| *** mingyu has quit IRC | 22:57 | |
| *** mvk has joined #openstack-swift | 22:59 | |
| *** ianychoi has quit IRC | 22:59 | |
| *** ianychoi has joined #openstack-swift | 23:01 | |
| clayg | umm... so my /opt/stack/logs/s-*.log I think is *not* the place to go see logs on devstack | 23:10 |
| clarkb | clayg: correct, with everything running under systemd now the logs go to journald | 23:11 |
| clarkb | clayg: you can get them via journalctl -u devstack@s-proxy for example | 23:12 |
| clayg | is there a journalctl list-all-units? | 23:12 |
| clarkb | or journalctl -u 'devstack@*' for all the services | 23:12 |
| clarkb | clayg: there is with systemctl | 23:12 |
| clarkb | systemctl list-units | 23:12 |
| portante | clayg: so swift has stopped logging to syslog when run under systemd? | 23:13 |
| clarkb | portante: sort of, if syslog is installed and enabled as a service most distros automatically have journald pipe into it | 23:14 |
| *** zacksh_ is now known as zacksh | 23:14 | |
| clarkb | and I think swift may directly syslog? beacuse it doesn't oslo.logging | 23:14 |
| portante | clarkb: it uses python logging, too, iirc | 23:15 |
| portante | and that is configured to send to syslog | 23:15 |
| clayg | ok, i guess I didn't know/understand what the -u was about - `journalctl devstack@s-*` was my first guess :\ | 23:15 |
| portante | so who ever is listening on /dev/log (or the like) will get log messages | 23:15 |
| clarkb | clayg: the -u restricts the journalctl output to the listed unit(s) which also accepts a globbing string | 23:16 |
| clarkb | fwiw the logs devstack collects definitely end up in journald Jun 02 21:15:39.033955 ubuntu-xenial-citycloud-la1-9113089 systemd[1]: Started Devstack devstack@s-object.service. | 23:18 |
| clarkb | and should also end up in syslog if rsyslog is installed and enabled on ubuntu and rhel iirc | 23:18 |
| portante | is journald listening to /dev/log? | 23:19 |
| clarkb | I don't know. | 23:20 |
| portante | so on RHEL journald listens to /dev/log, and handles all syslog messages incorporating them into the journald | 23:21 |
| portante | then rsyslog uses the imjournal module to pull them out of the journald into traditional /var/log files | 23:21 |
| portante | not sure how ubuntu configures that | 23:22 |
| portante | fedora uses to do that, but now has dropped rsyslog by default | 23:22 |
| clayg | `journalctl -u devstack@s-* -f` is my new fav | 23:24 |
| *** vint_bra has quit IRC | 23:45 | |
| *** chsc has quit IRC | 23:51 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!