| *** X019 has joined #openstack-swift | 00:08 | |
| *** Masahiro has joined #openstack-swift | 00:12 | |
| *** lcurtis has quit IRC | 00:14 | |
| *** fandi has joined #openstack-swift | 00:15 | |
| *** Masahiro has quit IRC | 00:17 | |
| *** miurahr has joined #openstack-swift | 00:17 | |
| *** daddyjoseph97 has quit IRC | 00:22 | |
| *** fandi has quit IRC | 00:28 | |
| *** annegent_ has joined #openstack-swift | 00:29 | |
| *** ho has quit IRC | 00:32 | |
| *** mvenesio_ has quit IRC | 00:34 | |
| *** annegent_ has quit IRC | 00:34 | |
| *** occup4nt is now known as occupant | 00:35 | |
| *** david-lyle is now known as david-lyle_afk | 00:35 | |
| *** fandi has joined #openstack-swift | 00:35 | |
| *** dmorita has joined #openstack-swift | 00:35 | |
| *** miurahr has quit IRC | 00:37 | |
| *** ho has joined #openstack-swift | 00:38 | |
| *** oomichi has joined #openstack-swift | 00:41 | |
| *** andrews has joined #openstack-swift | 00:48 | |
| *** Masahiro has joined #openstack-swift | 00:51 | |
| andrews | @here Don't you know, guys, if a possibility to delete a container with objects is a new feature in swift? I use python-swiftclient==2.3.1. All docs say the I can delete only empty containers. | 00:57 |
|---|---|---|
| *** addnull has joined #openstack-swift | 00:57 | |
| swifterdarrell | andrews: python-swiftclient just lists the objects in the container, deletes them, and then deletes the container, all on your behalf; no change was made to the Swift API re: deleting containers | 00:58 |
| *** panbalag has quit IRC | 01:05 | |
| andrews | swifterdarrell: before, when i tried to delete a container with children I got 409 error and it didn't allow me to do this until I kill all the objects. it is a bit confusing as i expect to get 409 again according to docs. Don't you know it is documented somewhere? | 01:05 |
| *** swifterdarrell has left #openstack-swift | 01:05 | |
| *** swifterdarrell has joined #openstack-swift | 01:05 | |
| *** ChanServ sets mode: +v swifterdarrell | 01:05 | |
| swifterdarrell | andrews: this was with the python-swiftclient vs. the "swift" CLI program? I dunno, maybe that did change... | 01:06 |
| swifterdarrell | andrews: but the Swift API itself has not changed in this respect, though | 01:06 |
| swifterdarrell | andrews: pretty sure Swift doesn't let you DELETE containers that still have objects. | 01:06 |
| *** lpabon has quit IRC | 01:09 | |
| andrews | swifterdarrell: I have noticed this working through the CLI. and I was pretty surprised being able to delete container with objects. here the code: https://github.com/openstack/python-swiftclient/blob/202bbab2d06b9ef751fba306effd09f3bae8f383/swiftclient/shell.py#L114 | 01:10 |
| andrews | swifterdarrell: it deleted all objects first and then container itself with no notification that container has objects, no expected 409 errors | 01:13 |
| *** shri has quit IRC | 01:13 | |
| swifterdarrell | andrews: my understanding is that the "swift" CLI has had that behavior for as long as I can remember | 01:16 |
| swifterdarrell | andrews: I think there was a /bug/ at one point where there was a race condition between the last few object deletes and the DELETE of the Container, preventing the "swift" CLI from accomplishing the user's goal (of deleting the container and all objects within it) | 01:16 |
| andrews | swifterdarrell: under "that behavior" you mean possibility to delete container with objects? | 01:22 |
| swifterdarrell | andrews: I believe "swift delete container_name" has always semantically meant "delete all objects in container_name and then delete container_name itself", and if that was broken and only deleted some objects before saying something about "409", that was a bug that my memory vaguely says we had for a while | 01:23 |
| *** X019 has quit IRC | 01:27 | |
| andrews | swifterdarrell: ok, if I try to delete a container with objects through the API call, can I expect error message? | 01:28 |
| *** annegent_ has joined #openstack-swift | 01:30 | |
| *** addnull has quit IRC | 01:34 | |
| *** annegent_ has quit IRC | 01:35 | |
| *** X019 has joined #openstack-swift | 01:42 | |
| *** panbalag has joined #openstack-swift | 01:47 | |
| swifterdarrell | andrews: definitely | 01:47 |
| *** ho has quit IRC | 01:49 | |
| *** ho has joined #openstack-swift | 01:49 | |
| *** haomaiwang has joined #openstack-swift | 01:57 | |
| *** panbalag has quit IRC | 01:57 | |
| andrews | swifterdarrell: thanks a lot for explanations, I just checked it, indeed, API call gives me 409 error, CLI allows me to delete container with objects - it is just different behavior for API and CLI ways. Thanks! | 02:07 |
| *** gyee has quit IRC | 02:09 | |
| *** lcurtis has joined #openstack-swift | 02:13 | |
| *** oomichi has quit IRC | 02:16 | |
| *** annegent_ has joined #openstack-swift | 02:29 | |
| *** andrews has quit IRC | 02:32 | |
| *** annegent_ has quit IRC | 02:34 | |
| *** rdaly2 has quit IRC | 02:48 | |
| *** addnull has joined #openstack-swift | 02:51 | |
| openstackgerrit | Hisashi Osanai proposed openstack/swift: Fix the GET's response code when there is a missing segment in LO https://review.openstack.org/136258 | 02:51 |
| *** masonhsiung has joined #openstack-swift | 02:54 | |
| *** addnull has quit IRC | 02:56 | |
| *** lcurtis has quit IRC | 03:01 | |
| *** Masahiro has quit IRC | 03:02 | |
| *** Masahiro has joined #openstack-swift | 03:02 | |
| *** annegent_ has joined #openstack-swift | 03:03 | |
| *** annegent_ has quit IRC | 03:05 | |
| *** lcurtis has joined #openstack-swift | 03:14 | |
| *** fandi has quit IRC | 03:25 | |
| *** lcurtis has quit IRC | 03:41 | |
| *** rdaly2 has joined #openstack-swift | 03:48 | |
| *** rdaly2 has quit IRC | 03:52 | |
| *** oomichi has joined #openstack-swift | 03:53 | |
| *** oomichi has quit IRC | 03:54 | |
| openstackgerrit | Yuan Zhou proposed openstack/swift: Add node/pair index patch back into feature/EC https://review.openstack.org/134065 | 03:57 |
| *** ho has quit IRC | 04:12 | |
| *** ho has joined #openstack-swift | 04:18 | |
| *** ppai has joined #openstack-swift | 04:21 | |
| *** haomaiwang has quit IRC | 04:22 | |
| *** haomaiwang has joined #openstack-swift | 04:23 | |
| *** miurahr has joined #openstack-swift | 04:27 | |
| *** miurahr has left #openstack-swift | 04:27 | |
| *** vishy has quit IRC | 04:35 | |
| *** sileht has quit IRC | 04:36 | |
| *** zacksh has quit IRC | 04:36 | |
| *** zacksh has joined #openstack-swift | 04:38 | |
| *** vishy has joined #openstack-swift | 04:39 | |
| *** sileht has joined #openstack-swift | 04:40 | |
| *** miurahr has joined #openstack-swift | 04:42 | |
| *** addnull has joined #openstack-swift | 04:47 | |
| *** rdaly2 has joined #openstack-swift | 04:50 | |
| *** rdaly2 has quit IRC | 04:54 | |
| *** miurahr has quit IRC | 04:56 | |
| *** zaitcev has quit IRC | 05:05 | |
| *** addnull has quit IRC | 05:13 | |
| *** addnull has joined #openstack-swift | 05:20 | |
| *** masonhsi_ has joined #openstack-swift | 05:24 | |
| *** masonhsiung has quit IRC | 05:27 | |
| *** addnull has quit IRC | 05:31 | |
| *** masonhsiung has joined #openstack-swift | 05:31 | |
| *** masonhsi_ has quit IRC | 05:35 | |
| *** oomichi has joined #openstack-swift | 05:38 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:47 | |
| *** rdaly2 has joined #openstack-swift | 05:50 | |
| *** rdaly2 has quit IRC | 05:55 | |
| *** oomichi has quit IRC | 06:02 | |
| *** masonhsiung has quit IRC | 06:05 | |
| *** masonhsiung has joined #openstack-swift | 06:05 | |
| *** nshaikh has joined #openstack-swift | 06:13 | |
| *** addnull has joined #openstack-swift | 06:15 | |
| *** rmcall has joined #openstack-swift | 06:19 | |
| *** rmcall has quit IRC | 06:24 | |
| *** rmcall has joined #openstack-swift | 06:25 | |
| mattoliverau | I'm, calling it a day, night swift land | 06:32 |
| *** kopparam has joined #openstack-swift | 06:42 | |
| *** kopparam has quit IRC | 07:00 | |
| *** rmcall has quit IRC | 07:01 | |
| *** kopparam has joined #openstack-swift | 07:03 | |
| *** kopparam has quit IRC | 07:03 | |
| *** ttrumm_ has joined #openstack-swift | 07:08 | |
| *** nshaikh has quit IRC | 07:10 | |
| *** pcaruana has joined #openstack-swift | 07:19 | |
| *** nshaikh has joined #openstack-swift | 07:20 | |
| *** jyoti-ranjan has joined #openstack-swift | 07:22 | |
| *** rdaly2 has joined #openstack-swift | 07:52 | |
| *** rdaly2 has quit IRC | 07:56 | |
| *** addnull has quit IRC | 08:03 | |
| *** kopparam has joined #openstack-swift | 08:04 | |
| ho | mattoliverau: what is "night swift land"? | 08:05 |
| mattoliverau | ho: just saying good evening/good night to the swift channel :) | 08:06 |
| ho | mattoliverau: thanks! | 08:07 |
| *** rledisez has joined #openstack-swift | 08:09 | |
| ondergetekende | Night swift land sounds like some realm in a fantasy series... ;-) | 08:10 |
| *** kopparam has quit IRC | 08:11 | |
| *** nellysmitt has joined #openstack-swift | 08:16 | |
| mattoliverau | Lol, yes yes it does :) | 08:29 |
| *** geaaru has joined #openstack-swift | 08:36 | |
| *** Krast has joined #openstack-swift | 08:42 | |
| *** k4n0 has joined #openstack-swift | 08:47 | |
| *** kopparam has joined #openstack-swift | 08:50 | |
| *** ppai has quit IRC | 08:54 | |
| *** jordanP has joined #openstack-swift | 08:56 | |
| *** kopparam has quit IRC | 08:57 | |
| *** kopparam has joined #openstack-swift | 08:59 | |
| *** fandi has joined #openstack-swift | 09:03 | |
| *** kopparam has quit IRC | 09:07 | |
| *** ppai has joined #openstack-swift | 09:08 | |
| *** kopparam has joined #openstack-swift | 09:09 | |
| *** kopparam_ has joined #openstack-swift | 09:28 | |
| *** kopparam has quit IRC | 09:29 | |
| *** addnull has joined #openstack-swift | 09:33 | |
| *** sandywalsh has joined #openstack-swift | 09:34 | |
| *** sandywalsh_ has quit IRC | 09:36 | |
| *** acoles_away is now known as acoles | 09:37 | |
| *** addnull has quit IRC | 09:42 | |
| *** nellysmitt has quit IRC | 09:51 | |
| *** rdaly2 has joined #openstack-swift | 09:53 | |
| *** rdaly2 has quit IRC | 09:58 | |
| *** kopparam_ has quit IRC | 10:00 | |
| *** jasondotstar has joined #openstack-swift | 10:05 | |
| *** ekarlso- has quit IRC | 10:07 | |
| *** ekarlso- has joined #openstack-swift | 10:07 | |
| *** nellysmitt has joined #openstack-swift | 10:25 | |
| *** fandi has quit IRC | 10:27 | |
| *** addnull has joined #openstack-swift | 10:50 | |
| *** kopparam has joined #openstack-swift | 11:00 | |
| openstackgerrit | Koert van der Veer proposed openstack/swift: Add option to insert a default reseller prefix https://review.openstack.org/142067 | 11:06 |
| *** haomaiwang has quit IRC | 11:14 | |
| *** kopparam has quit IRC | 11:16 | |
| *** kopparam has joined #openstack-swift | 11:16 | |
| *** dmorita has quit IRC | 11:17 | |
| *** Masahiro has quit IRC | 11:23 | |
| *** mahatic has joined #openstack-swift | 11:30 | |
| *** masonhsiung has quit IRC | 11:33 | |
| *** Masahiro has joined #openstack-swift | 11:39 | |
| *** infotection has quit IRC | 11:53 | |
| *** rdaly2 has joined #openstack-swift | 11:54 | |
| openstackgerrit | Koert van der Veer proposed openstack/swift: Add option to insert a default reseller prefix https://review.openstack.org/142067 | 11:57 |
| *** infotection has joined #openstack-swift | 11:59 | |
| *** rdaly2 has quit IRC | 11:59 | |
| *** kopparam has quit IRC | 12:00 | |
| *** jasondotstar is now known as jasondotstar|afk | 12:01 | |
| *** ppai has quit IRC | 12:03 | |
| *** lpabon has joined #openstack-swift | 12:05 | |
| *** kopparam has joined #openstack-swift | 12:05 | |
| *** Masahiro has quit IRC | 12:07 | |
| *** infotection has quit IRC | 12:09 | |
| *** infotection has joined #openstack-swift | 12:11 | |
| *** otoolee has quit IRC | 12:14 | |
| *** sileht has quit IRC | 12:14 | |
| *** ipolyzos has quit IRC | 12:14 | |
| *** kopparam has quit IRC | 12:15 | |
| *** sileht has joined #openstack-swift | 12:16 | |
| *** ipolyzos has joined #openstack-swift | 12:16 | |
| *** EmilienM is now known as EmilienM|afk | 12:17 | |
| *** ppai has joined #openstack-swift | 12:18 | |
| *** nexusz99 has quit IRC | 12:18 | |
| *** ppai has quit IRC | 12:18 | |
| *** ppai has joined #openstack-swift | 12:18 | |
| *** nottrobin__ has joined #openstack-swift | 12:19 | |
| *** nottrobin_ has quit IRC | 12:20 | |
| *** nottrobin__ is now known as nottrobin_ | 12:20 | |
| *** nottrobin_ is now known as nottrobin | 12:20 | |
| *** otoolee has joined #openstack-swift | 12:21 | |
| *** acoles has quit IRC | 12:22 | |
| *** ondergetekende has quit IRC | 12:22 | |
| *** kopparam has joined #openstack-swift | 12:25 | |
| *** zackmdavis_ has joined #openstack-swift | 12:26 | |
| *** ppai has quit IRC | 12:31 | |
| *** ho has quit IRC | 12:31 | |
| *** ujjain has quit IRC | 12:31 | |
| *** zackmdavis has quit IRC | 12:31 | |
| openstackgerrit | Daniel Wakefield proposed openstack/python-swiftclient: Verify MD5 of uploaded objects. https://review.openstack.org/129254 | 12:32 |
| *** ujjain has joined #openstack-swift | 12:33 | |
| *** miurahr has joined #openstack-swift | 12:33 | |
| *** lcurtis has joined #openstack-swift | 12:33 | |
| *** ondergetekende has joined #openstack-swift | 12:34 | |
| *** acoles_away has joined #openstack-swift | 12:34 | |
| *** acoles_away is now known as acoles | 12:35 | |
| *** ChanServ sets mode: +v acoles | 12:35 | |
| *** ppai has joined #openstack-swift | 12:36 | |
| *** miurahr has quit IRC | 12:38 | |
| *** addnull has quit IRC | 12:40 | |
| *** lcurtis has quit IRC | 12:44 | |
| *** panbalag has joined #openstack-swift | 12:50 | |
| *** miurahr has joined #openstack-swift | 12:50 | |
| *** panbalag has quit IRC | 12:51 | |
| *** miurahr has quit IRC | 12:51 | |
| *** tellesnobrega has quit IRC | 12:56 | |
| *** linuxgeek_ has joined #openstack-swift | 12:56 | |
| linuxgeek_ | hello, with swift as the glance back end it takes 30 minutes for a 30gb file to be uploaded to glance. is there any changes which can be made to better this? | 12:58 |
| linuxgeek_ | in the swift or glance conf files | 12:58 |
| *** rdaly2 has joined #openstack-swift | 12:58 | |
| *** Masahiro has joined #openstack-swift | 13:07 | |
| openstackgerrit | Yuan Zhou proposed openstack/swift: Add node/pair index patch back into feature/EC https://review.openstack.org/134065 | 13:08 |
| rledisez | ls -lh | 13:08 |
| *** tellesnobrega has joined #openstack-swift | 13:08 | |
| rledisez | oups, sorry :) | 13:09 |
| *** Masahiro has quit IRC | 13:12 | |
| *** kopparam has quit IRC | 13:16 | |
| *** kopparam has joined #openstack-swift | 13:19 | |
| mahatic | hi, I'm looking for tests for OPTIONS verb written in proxy. I can't locate it in the code. Can someone point me to it? | 13:19 |
| *** kopparam has quit IRC | 13:20 | |
| openstackgerrit | Doug Hellmann proposed openstack/swift: Remove unneeded condition https://review.openstack.org/141818 | 13:26 |
| *** EmilienM|afk is now known as EmilienM | 13:28 | |
| *** aix has joined #openstack-swift | 13:31 | |
| *** ppai has quit IRC | 13:41 | |
| *** ppai has joined #openstack-swift | 13:55 | |
| *** ppai has quit IRC | 14:02 | |
| *** nshaikh has quit IRC | 14:11 | |
| *** mvenesio has joined #openstack-swift | 14:17 | |
| openstackgerrit | Daniel Wakefield proposed openstack/python-swiftclient: Fix misplaced check for None in SwiftUploadObject. https://review.openstack.org/133107 | 14:18 |
| *** jwang__ has joined #openstack-swift | 14:20 | |
| *** sfineberg_ has joined #openstack-swift | 14:22 | |
| *** mtreinish_ has joined #openstack-swift | 14:22 | |
| *** acorwin_ has joined #openstack-swift | 14:23 | |
| *** otherjon_ has joined #openstack-swift | 14:24 | |
| *** rdaly2 has quit IRC | 14:24 | |
| *** jeblair_ has joined #openstack-swift | 14:25 | |
| *** HenryG_ has joined #openstack-swift | 14:28 | |
| *** portante_ has joined #openstack-swift | 14:30 | |
| *** k4n0 has quit IRC | 14:30 | |
| *** SkyRocknRoll has quit IRC | 14:30 | |
| *** HenryG has quit IRC | 14:30 | |
| *** mtreinish has quit IRC | 14:30 | |
| *** jeblair has quit IRC | 14:30 | |
| *** jwang_ has quit IRC | 14:30 | |
| *** sfineberg has quit IRC | 14:30 | |
| *** otherjon has quit IRC | 14:30 | |
| *** acorwin has quit IRC | 14:30 | |
| *** IRTermite has quit IRC | 14:30 | |
| *** portante has quit IRC | 14:30 | |
| *** otherjon_ is now known as otherjon | 14:30 | |
| *** mtreinish_ is now known as mtreinish | 14:30 | |
| *** k4n0 has joined #openstack-swift | 14:31 | |
| *** aix has quit IRC | 14:31 | |
| *** IRTermite has joined #openstack-swift | 14:31 | |
| *** aix has joined #openstack-swift | 14:32 | |
| *** SkyRocknRoll has joined #openstack-swift | 14:33 | |
| openstackgerrit | Merged openstack/swift: Adjust MAX_FILE_SIZE during test on 32 bit systems https://review.openstack.org/127030 | 14:37 |
| *** Masahiro has joined #openstack-swift | 14:40 | |
| *** tellesnobrega has quit IRC | 14:40 | |
| *** tellesnobrega has joined #openstack-swift | 14:40 | |
| *** lihkin has joined #openstack-swift | 14:44 | |
| *** Masahiro has quit IRC | 14:44 | |
| *** mvenesio has quit IRC | 14:49 | |
| *** ttrumm_ has quit IRC | 14:52 | |
| *** X019 has quit IRC | 14:55 | |
| *** dmsimard_away is now known as dmsimard | 15:03 | |
| *** lihkin has quit IRC | 15:05 | |
| *** tdasilva has joined #openstack-swift | 15:06 | |
| *** lihkin has joined #openstack-swift | 15:07 | |
| *** X019 has joined #openstack-swift | 15:08 | |
| *** jasondotstar|afk is now known as jasondotstar | 15:15 | |
| *** jyoti-ranjan has quit IRC | 15:17 | |
| *** SkyRocknRoll has quit IRC | 15:18 | |
| *** HenryG_ has quit IRC | 15:25 | |
| *** HenryG_ has joined #openstack-swift | 15:25 | |
| *** portante_ has quit IRC | 15:25 | |
| *** portante_ has joined #openstack-swift | 15:25 | |
| *** k4n0 has quit IRC | 15:25 | |
| *** k4n0 has joined #openstack-swift | 15:25 | |
| *** aix has quit IRC | 15:25 | |
| *** aix has joined #openstack-swift | 15:25 | |
| *** rdaly2 has joined #openstack-swift | 15:32 | |
| *** erlon has joined #openstack-swift | 15:32 | |
| *** pcaruana has quit IRC | 15:42 | |
| *** daddyjoseph97 has joined #openstack-swift | 15:44 | |
| *** miurahr has joined #openstack-swift | 15:48 | |
| *** rdaly2 has quit IRC | 15:51 | |
| *** rmcall has joined #openstack-swift | 16:08 | |
| imkarrer | Hey guys. I have a question about deploying multi-region swift. This question reguards the memcached servers. Should the list of memcached servers in proxy-server.conf and object-expirer.conf be Region specific? In other words, should Region1 only have memcached servers listed which reside in region1? | 16:19 |
| *** daddyjoseph97 has quit IRC | 16:19 | |
| remix_tj | imkarrer: in my plan there is one memcached on each region | 16:24 |
| *** jhfeng has joined #openstack-swift | 16:24 | |
| remix_tj | imkarrer: in case of client connecting to proxy on other region will need to reauth | 16:24 |
| openstackgerrit | paul luse proposed openstack/swift-specs: Updates to the reconstructor section https://review.openstack.org/142146 | 16:25 |
| imkarrer | remix_tj: Thanks, that makes since. I also think that since swift is eventually consistent that region specific memcached makes since. But the auth seals the deal for me. Thanks again. | 16:26 |
| *** kragniz is now known as kragwhale | 16:26 | |
| *** lihkin has quit IRC | 16:27 | |
| *** Masahiro has joined #openstack-swift | 16:29 | |
| *** SkyRocknRoll has joined #openstack-swift | 16:29 | |
| *** Masahiro has quit IRC | 16:34 | |
| *** EmilienM is now known as EmilienM|afk | 16:34 | |
| *** rmcall has quit IRC | 16:42 | |
| *** david-lyle_afk is now known as david-lyle | 16:44 | |
| *** ajiang has joined #openstack-swift | 16:48 | |
| *** kani has joined #openstack-swift | 16:48 | |
| kani | While uploading image to glance which has swift backend , the image is going to killed state - swift log shows the error as ChunkWriteTimeOut(10s). | 16:51 |
| kani | Can anyone tell what is this error and how to fix it? | 16:51 |
| *** ajiang38740 has joined #openstack-swift | 16:58 | |
| *** zackmdavis_ is now known as zackmdavis | 17:00 | |
| *** lpabon has quit IRC | 17:01 | |
| *** X019 has left #openstack-swift | 17:01 | |
| *** jhfeng has quit IRC | 17:04 | |
| kani | While uploading image to glance which has swift backend , the image is going to killed state - swift log shows the error as ChunkWriteTimeOut(10s). Can anyone tell what is this error and how to fix it? | 17:06 |
| *** nellysmitt has quit IRC | 17:10 | |
| notmyname | good morning | 17:13 |
| mahatic | good morning! | 17:14 |
| *** rmcall has joined #openstack-swift | 17:14 | |
| peluse | morning | 17:16 |
| acoles | morning | 17:19 |
| *** mvenesio has joined #openstack-swift | 17:19 | |
| mahatic | notmyname, I'm unable to locate tests for OPTIONS verb in proxy, can you point me to it? | 17:22 |
| notmyname | yup | 17:22 |
| notmyname | let me find them | 17:22 |
| mahatic | okay | 17:22 |
| torgomatic | mahatic: if nothing else, comment out the OPTIONS method, run tests, and see which ones fail | 17:23 |
| notmyname | heh | 17:23 |
| *** dmsimard is now known as dmsimard_away | 17:23 | |
| mahatic | torgomatic, oh. that's a way too :D thanks! | 17:23 |
| torgomatic | I do that all the time when I can't find tests for something. | 17:23 |
| notmyname | mahatic: test/unit/proxy/test_server.py:4855 | 17:23 |
| notmyname | mahatic: there are a few test methods in that file | 17:24 |
| mahatic | torgomatic, your timezone is PST too? | 17:24 |
| kani | notmyname: While uploading image to glance which has swift backend , the image is going to killed state - swift log shows the error as ChunkWriteTimeOut(10s). Could you help me understand what is this error and how to fix it? | 17:24 |
| torgomatic | mahatic: yes, I'm near San Francisco | 17:24 |
| mahatic | okay | 17:24 |
| *** dmsimard_away is now known as dmsimard | 17:25 | |
| notmyname | mahatic: when he's not working from home, he sits about 3 feet in front of me :-) | 17:25 |
| mahatic | notmyname, okay. I was digging through test_base.py all this time | 17:25 |
| mahatic | notmyname, ah yes. He works for swiftstack too i believe (I don't know how i remember, probably from a discussion here :D) | 17:26 |
| *** rledisez has quit IRC | 17:26 | |
| notmyname | kani: that error is when the proxy server times out taking to an object server. there are a few possible causes | 17:29 |
| torgomatic | kani: sounds like glance isn't sending data to swift in a timely fashion | 17:29 |
| torgomatic | notmyname: it's also in the proxy, FWIW; happens when the client fails to send data for $node_timeout seconds | 17:29 |
| notmyname | kani: importantly, though, what is "killed state". and what message are you getting from swift? ie what response code? swift works around a lot of that | 17:30 |
| kani | 503 Service Unavailable . this is in glance-api.log | 17:30 |
| notmyname | kani: yup. both ways | 17:31 |
| notmyname | kani: there's a few next steps | 17:31 |
| notmyname | kani: first, get the transaction id for that failed request. then grep the swift logs for that id so you have all the logs relating to that request | 17:32 |
| notmyname | kani: you'll see the timeout, but maybe some other stuff | 17:32 |
| *** mkerrin has quit IRC | 17:32 | |
| kani | this does not happen always | 17:32 |
| notmyname | kani: what does the full error message say? | 17:33 |
| kani | While trying to upload 4 images in parallel, I saw this error | 17:33 |
| notmyname | kani: that will tell you if it's a client read issue or a object server write issue | 17:33 |
| *** kbee has joined #openstack-swift | 17:33 | |
| kani | Got the error like this: proxy-server ERROR with Object server 127.0.0.1:6010/sda8 re: Trying to write to /v1/AUTH_8b0342f9b71f4f3db7272452ef31fa1c/myfiles/testCD.iso: ChunkWriteTimeout (10s) | 17:34 |
| notmyname | kani: ok. so the server listening on port 6010 is timing out. could be over-busy. maybe the drive is having issues (contention or failures). | 17:36 |
| notmyname | kani: interesting that you're only seeing one timeout error and getting a 503. are you using 3 replicas? | 17:36 |
| kani | yeah... 3 replicas | 17:37 |
| notmyname | kani: can you find all the logs with that trans id and pastebin them? | 17:38 |
| *** NM has joined #openstack-swift | 17:39 | |
| kani | notmyname: Yes.. I ll do paste | 17:40 |
| *** SkyRocknRoll has quit IRC | 17:40 | |
| *** kragwhale is now known as kragniz | 17:43 | |
| kbee | Hello notmyname , can you pls give your opinion on https://review.openstack.org/#/c/140972/ | 17:44 |
| notmyname | kbee: if you have multiple instances running on the same box and need to demux the logs, you can do that with different syslog facilities, right? logs still go to syslog, but you can configure syslog to put the logs where you need them. | 17:50 |
| notmyname | kbee: I'm not saying you don't have a valid use case (far from it!). but I think what you're looking for is already possible without adding new config options | 17:51 |
| *** lcurtis has joined #openstack-swift | 17:53 | |
| kbee | notmyname: the angle is for isolated drive-audit runs, its easier to mine 3-4 lines of data than go through heap of syslog. this sounds more prudent, and we can achieve with minimal code.. | 17:54 |
| kbee | notmyname: and most important usecase is : if there are simultaneous runs of swift drive audit, the first run logs to syslog and unmounts and second one does nothing since is already unmounted, mining through latest syslog message will be incorrect | 17:58 |
| kbee | notmyname: one should correctly find out the error state for a particular run of drive-audit. and that was the intention | 17:59 |
| notmyname | kbee: so your proposal is to log the first one to syslog and the second one to the console. then you look in the respective place for the log messages associated with whatever one you're checking. right? | 18:00 |
| notmyname | kbee: and you'd tell the second run to log to the console with a config option | 18:00 |
| kbee | notmyname: The proposal is to just add a console handler for logger. so when logging is done, it logs to syslog as well as console. | 18:01 |
| kbee | notmyname: if log_to_console is turned on(default: off), it logs to both syslog and console, else logs only to syslog as all these days | 18:02 |
| *** rmcall has quit IRC | 18:02 | |
| *** jordanP has quit IRC | 18:03 | |
| *** jeblair_ is now known as jeblair | 18:04 | |
| notmyname | kbee: and what do the configs look like? both log to console? or one logs to syslog and the other logs to syslog+console? | 18:04 |
| *** rmcall has joined #openstack-swift | 18:05 | |
| kbee | notmyname: just one extra config option - log_to_console (default: False). if set to true, it logs to console. | 18:05 |
| kbee | notmyname: By default, anyways it always gets logged to syslog | 18:06 |
| *** k4n0 has quit IRC | 18:06 | |
| *** rmcall has quit IRC | 18:06 | |
| kbee | notmyname: no change in that functionality. this is just extra ability to log to console depending on config | 18:06 |
| notmyname | kbee: right. I get that. so how are you running it? both log to console+syslog? or one logs to console and one logs to console+syslog? | 18:06 |
| *** fandi has joined #openstack-swift | 18:06 | |
| *** rmcall has joined #openstack-swift | 18:08 | |
| kbee | notmyname: sorry. i'm not getting it -> 'both log', do you mean logs or something else ? | 18:08 |
| *** shri has joined #openstack-swift | 18:08 | |
| notmyname | kbee: how are you running each instance of swift-drive-audit? do they have the same config file or different config files? | 18:08 |
| kbee | same config file | 18:09 |
| kbee | notmyname: same config file | 18:09 |
| kani | The following is the error from swift log. | 18:10 |
| kani | Jun 7 08:51:48 cs-swift proxy-server: ERROR with Object server 192.168.0.11:6000/d1 re: Trying to get final status of PUT to /v1/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84: Timeout (10s) | 18:10 |
| kani | Jun 7 08:51:48 cs-swift proxy-server: ERROR with Object server 192.168.0.10:6000/d1 re: Trying to get final status of PUT to /v1/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84: Timeout (10s) | 18:10 |
| kani | Jun 7 08:51:48 cs-swift proxy-server: ERROR with Object server 192.168.0.12:6000/d1 re: Trying to get final status of PUT to /v1/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84: Timeout (10s) | 18:10 |
| kani | Jun 7 08:51:48 cs-swift proxy-server: 192.168.0.4 192.168.0.4 07/Jun/2014/08/51/48 PUT /v1/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84 HTTP/1.0 503 - python-swiftclient-2.3.1.45.g892be34 25d50bf1b8814886... 2147483648 118 - tx2ac01b9ab3b24f71849b4-005392d259 - 75.0299 - - 1402131033.757571936 1402131108.787444115 | 18:10 |
| kani | Jun 7 08:52:13 cs-swift container-server: 192.168.0.10 - - [07/Jun/2014:08:52:13 +0000] "PUT /d1/990/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84" 201 - "PUT http://192.168.0.10:8080/d1/400/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84" "tx2ac01b9ab3b24f71849b4-005392d259" "object-server 4519" 0.0005 "-" 4388 | 18:11 |
| kani | Jun 7 08:52:13 cs-swift object-server: 192.168.0.10 - - [07/Jun/2014:08:52:13 +0000] "PUT /d1/400/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84" 201 - "PUT http://192.168.0.10:8080/v1/AUTH_4a6f81f6d3a94b469053da84fc16a8f9/glance/3116d48f-1bc6-43aa-8f45-61d31599ad84" "tx2ac01b9ab3b24f71849b4-005392d259" "proxy-server 4176" 99.5822 "-" 4519 | 18:11 |
| kani | Swift has 3 replicas. And I am trying to upload 4 images in parallel to glance | 18:12 |
| *** wer_ is now known as wer | 18:13 | |
| *** NM1 has joined #openstack-swift | 18:15 | |
| *** acoles is now known as acoles_away | 18:15 | |
| *** NM has quit IRC | 18:17 | |
| notmyname | kbee: beyond the logging, let's back up. you aren't having any issues with running multiple instances of swift-drive-audit at the same time? there isn't any sort of protection built in to the script, and all instances of it will read from the same log source and write the same output file. why are you running more than one at a time? | 18:17 |
| *** Masahiro has joined #openstack-swift | 18:17 | |
| kbee | notmyname: one can be run from admin to check, one can be run from mointoring tools such as icinga, other can be a cron.. thats the reason | 18:18 |
| kbee | notmyname: ^ the above may not happen all the time, but there is a window of opportunity | 18:20 |
| notmyname | kani: interesting. seems like 2 of those object server puts succeeded, but the response code didn't get back to the proxy server. so the issue is either that the proxy is overloaded and couldn't get to the response before the timer fired. or that the object server is overloaded. what else is going on in this cluster? what sort of hardware is it running on? | 18:20 |
| *** NM1 has quit IRC | 18:20 | |
| notmyname | kbee: ok, so you normally have it running one way. but sometimes you want to also run it from another tool, and you want to make sure you can cleanly demux the logs to find the right output. right? | 18:21 |
| *** Masahiro has quit IRC | 18:22 | |
| notmyname | kbee: still seems like a good way to corrupt your fstab file (ie it writes to fstab.new every time. not a tempfile) | 18:22 |
| *** NM has joined #openstack-swift | 18:22 | |
| kbee | notmyname: correct.. for every run, one should be able to get the output for that run predictably | 18:22 |
| notmyname | kbee: ok. that makes sense to me. I still think you could probably get that with some syslog config, but that's secondary to the other issue of multiple writers to fstab.new. that's not really going to work out well for you | 18:23 |
| *** NM has quit IRC | 18:23 | |
| kbee | notmyname: totally agree.. drive-audit needs to change for that. but all i had in the patch for ability to log to console in addition to syslog :) .. and thats useful in certian scenarios | 18:24 |
| notmyname | granted, there's also the new config option there so you don't unmount failed drives... | 18:24 |
| kbee | notmyname: true.. | 18:24 |
| *** rmcall has quit IRC | 18:25 | |
| notmyname | but in that case you'd be running with different config files anyway. ie one for the normal runner that unmounts. and a different one that is used when you run "by hand" and that doesn't unmount | 18:25 |
| notmyname | but in the second case, the reason for running it would be just as a checker. not to actually unmount. do you agree there? | 18:26 |
| notmyname | is that how you're using it? | 18:26 |
| *** rmcall has joined #openstack-swift | 18:27 | |
| *** rmcall has quit IRC | 18:29 | |
| mahatic | notmyname, OPTIONS is in base.py and ObjectController class (in obj.py) does not refer to OPTIONS. And tests are in test_server.py but not test_base.py? | 18:29 |
| mahatic | notmyname, I mentioned ObjectController because the tests are sending the request for OPTIONS to ObjectController | 18:30 |
| notmyname | mahatic: fun! makes perfect sense! ;-) | 18:30 |
| kbee | notmyname: agree.. , i am trying to use with unmount set to true. i am using the same config file. once i unmount from one session, i will not see those errors again since drive-audit does not read unmounted devices. | 18:31 |
| kani | SAIO is running on 3 vms on KVM | 18:31 |
| mahatic | notmyname, :D sorry, Let me do it in a paste.openstack maybe | 18:31 |
| *** gyee has joined #openstack-swift | 18:31 | |
| kani | notmyname: SAIO is running on 3 vms on KVM | 18:31 |
| notmyname | mahatic: that's because the existing OPTIONS also tests for CORS functionality | 18:31 |
| *** gordc has joined #openstack-swift | 18:31 | |
| notmyname | mahatic: so that's why it's using the ObjectController | 18:32 |
| notmyname | mahatic: as to test_server vs test_base. I don't know | 18:32 |
| kbee | notmyname: here's the code snippet -> https://gist.github.com/keshavab/f34bb63d93025bdff2b7 | 18:32 |
| mahatic | notmyname, true. But there is no OPTIONS in the ObjectController, but it's passing that request as well | 18:32 |
| mahatic | and it's passing* | 18:33 |
| notmyname | mahatic: will ObjectController has OPTIONS because it's inherited from Controller in base.py | 18:33 |
| mahatic | notmyname, oh I missed that | 18:33 |
| notmyname | mahatic: so I'm not sure if there's a good reason for the way it is. or a reason at all. maybe it should move to test_base.py. I don't know | 18:34 |
| mahatic | hmm | 18:34 |
| *** EmilienM|afk is now known as EmilienM | 18:35 | |
| kbee | notmyname: i was giving the prev example of multiple simultaneous invocations as just an usecase to explain why we need to have ability to log to console for predictable outputs | 18:35 |
| *** abhirc has joined #openstack-swift | 18:37 | |
| kbee | notmyname: so what do you say ? | 18:38 |
| *** rmcall has joined #openstack-swift | 18:38 | |
| notmyname | kani: I don't know the answer. but here's (basically) what's going on. either the proxy is overloaded and not getting to the backend requests fast enough. this would happen if you have a lot of concurrent requests. the solution is to upgrade the proxy or add more proxy servers. or the object servers are getting slow when doing the fsync to disk and not sending a response soon enough | 18:39 |
| notmyname | kani: because you're running storage nodes on VMs (never recommended), I suspect it's the latter. but again, that's just a guess. those are the areas to investigate though | 18:39 |
| notmyname | mahatic: for now, don't worry about moving the tests around. that is, that's a separate issue from "there is a test, and it is run" | 18:40 |
| notmyname | mahatic: you're looking at writing a test for your current patch? | 18:40 |
| mahatic | notmyname, yup, I'm not :) | 18:40 |
| mahatic | notmyname, yeah, test for current patch | 18:40 |
| mahatic | notmyname, and it comes off quite small (10-15 lines). I'm mentioning because you were saying "tests" the other day | 18:42 |
| kani | notmyname: Thanks for the response. | 18:44 |
| notmyname | mahatic: ya. as you pointed out, your patch will need tests. so I'm glad you're looking in to that. do you know what needs to be tested and why? | 18:45 |
| *** kani has quit IRC | 18:46 | |
| mahatic | notmyname, I guess - testing if the OPTIONS is indeed responding with the allowed methods (basically it's functionality). right? | 18:46 |
| notmyname | mahatic: right. this is a unit test to make sure the OPTIONS method is actually responding with what you expect it to respond with. that way we don't break it in the future. | 18:48 |
| *** NM has joined #openstack-swift | 18:48 | |
| *** rmcall has quit IRC | 18:48 | |
| *** HenryG_ has quit IRC | 18:48 | |
| notmyname | mahatic: since you're adding this to a storage node process, there won't be a functional test for it. clients can't talk to object servers, so there's no way to run the test | 18:48 |
| mahatic | notmyname, and correct me if i'm wrong - right now, I'm only writing test for the patch I've submitted. We're yet to decide where to put the OPTIONS (for all the nodes), consolidate it into one patch and add more tests | 18:49 |
| *** HenryG has joined #openstack-swift | 18:49 | |
| mahatic | notmyname, yeah | 18:49 |
| kbee | notmyname: So do you find a value in the patch to have ability to log to console additionally and optionally based on a config ? :) | 18:49 |
| notmyname | mahatic: correct. but I'd expect that the consolidation of OPTIONS will need to happen before your patch lands. | 18:49 |
| *** rmcall has joined #openstack-swift | 18:50 | |
| mahatic | notmyname, yeah of course (i'm assuming landing as in merging?) | 18:51 |
| *** NM has quit IRC | 18:51 | |
| notmyname | mahatic: right | 18:51 |
| notmyname | kbee: not sure yet. I'm not convinced either way yet. I think the end functionality might be possible with syslog configs locally with no need for changes in swift. (might). but also I'm concerned about multiple processes at the same time. drive-audit already works with swift-recon to report what's going on... | 18:51 |
| mahatic | notmyname, so what about where to put OPTIONS (for all the nodes to access), is that for a lil later? after the test? | 18:52 |
| *** aix has quit IRC | 18:53 | |
| notmyname | mahatic: well, I don't think the patch will land without both, so whichever you want to work on first is fine :-) | 18:54 |
| mahatic | notmyname, yeah, But I don't know where to put after yesterday's inputs from the others. | 18:55 |
| *** NM has joined #openstack-swift | 18:55 | |
| kbee | notmyname: hmmm...can you pls explain how its possible with local syslog configs.. | 18:55 |
| kbee | notmyname: if i want to isolate drive-audit-logs only ? | 18:56 |
| notmyname | mahatic: my first guess is a common base class for each storage node. that new class would live in swift.common. And for now all it would have is the common OPTIONS method. maybe later (much later) we could figure out if there's any other code consolidation that could be done | 18:58 |
| mahatic | notmyname, if I were to take a call, I would just go ahead and put it in swift.common coz that's the part i understand ;) | 18:58 |
| mahatic | notmyname, okay sure! | 18:58 |
| notmyname | mahatic: right. some new file there. maybe a BaseStorageServer() class. naming is hard | 18:59 |
| *** fandi has quit IRC | 18:59 | |
| mahatic | notmyname, yeah. okay | 19:00 |
| *** gordc has left #openstack-swift | 19:01 | |
| notmyname | kbee: I don't know. but maybe the ErrorMessagesToStderr setting for rsyslog. maybe syslod-ng has something too. combined with the syslog facility, you could redirect the output elsewhere. I've not tried any of that. only just did a quick google search | 19:02 |
| torgomatic | swift.common.uhh_stuff_and_things | 19:03 |
| notmyname | torgomatic: perfect! | 19:03 |
| mahatic | :D sure, I'll also put torgomatic as reviewer ;) | 19:04 |
| torgomatic | mahatic: be careful what you ask for; you may get it ;) | 19:04 |
| mahatic | :D sure | 19:06 |
| tdasilva | notmyname: hi, just saw this: https://www.brightbox.com/blog/2014/12/16/announcing-orbit-cloud-storage/ thought you might be interested | 19:08 |
| *** shri1 has joined #openstack-swift | 19:09 | |
| notmyname | tdasilva: cool! thanks | 19:09 |
| *** shri1 has quit IRC | 19:09 | |
| kbee | notmyname: hmm.. thnx.. | 19:10 |
| *** nellysmitt has joined #openstack-swift | 19:11 | |
| *** shri has quit IRC | 19:11 | |
| *** kbee has quit IRC | 19:15 | |
| *** nellysmitt has quit IRC | 19:16 | |
| peluse | anyone know what's up with recheck no bug apparently not doing anything? | 19:17 |
| *** acorwin_ is now known as acorwin | 19:18 | |
| *** lpabon has joined #openstack-swift | 19:20 | |
| *** zaitcev has joined #openstack-swift | 19:20 | |
| *** ChanServ sets mode: +v zaitcev | 19:20 | |
| notmyname | peluse: link? | 19:20 |
| peluse | notmyname, I see a bunch of them but this is the one I was trying https://review.openstack.org/#/c/132389/ | 19:21 |
| notmyname | peluse: looks like it reran (check the "toggle ci" button). but still has an error. I think that's a known thing that people are looking in to. maybe something with ceilometer causing the failure | 19:24 |
| notmyname | I'm looking into that issue. sortof. at least I was asked in -qa about it. to look in to it. | 19:24 |
| notmyname | there might be a workaround | 19:25 |
| peluse | notmyname, ahh, thanks. Man, it fails fast.... | 19:26 |
| linuxgeek_ | with swift as the glance back end it takes 30 minutes for a 30gb file to be uploaded to glance, is this expected? how do i track what is causing the delay? | 19:31 |
| *** mahatic has quit IRC | 19:34 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 19:36 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 19:37 |
| *** mahatic has joined #openstack-swift | 19:41 | |
| *** linuxgeek_ has left #openstack-swift | 19:48 | |
| *** rmcall has quit IRC | 19:58 | |
| *** rmcall has joined #openstack-swift | 19:59 | |
| *** exploreshaifali has joined #openstack-swift | 20:03 | |
| *** Masahiro has joined #openstack-swift | 20:06 | |
| *** Masahiro has quit IRC | 20:11 | |
| mahatic | notmyname, this is more python related question (didn't know how to form this for google search) - when a class A imports class B, the instance self will be of class A even for class B's functions? | 20:25 |
| mahatic | i.e. when I call class B's function in class A | 20:25 |
| mahatic | torgomatic,^^ | 20:27 |
| torgomatic | mahatic: you mean "class Stuff(Thing)"? | 20:27 |
| mahatic | torgomatic, :) yes | 20:27 |
| torgomatic | mahatic: yes, "self" will be the object on which the method is invoked, so you can have a derived-class self in a base-class method. | 20:28 |
| mahatic | class A(self) { B.stuff(self) } | 20:29 |
| torgomatic | mahatic: are you inheriting or composing? perhaps some pastebinned code would help. | 20:29 |
| mahatic | torgomatic, now class B will have class A "self" object, correct? | 20:29 |
| mahatic | torgomatic, inheriting | 20:30 |
| mahatic | okay, just a min | 20:30 |
| mahatic | torgomatic, http://paste.openstack.org/show/152005/ | 20:35 |
| mahatic | torgomatic, now class B will have the object "self" of A when invoked from A right? | 20:35 |
| zaitcev | mahatic: but the two in your example are not related | 20:36 |
| mahatic | let me know if the pastebin doesn't make sense :) | 20:36 |
| mahatic | zaitcev, sorry! let me correct | 20:36 |
| zaitcev | well, I guess you meant class A(B): | 20:38 |
| mahatic | http://paste.openstack.org/show/152007/ | 20:38 |
| mahatic | zaitcev, makes sense? | 20:38 |
| torgomatic | mahatic: http://paste.openstack.org/show/152008/ is more like what you're after, I think | 20:39 |
| torgomatic | and yes, in the invocation of Base.anothermethod(), "self" will be an instance of Derived | 20:39 |
| zaitcev | there are exmaples in db.py and backend.py that one could pilfer easily | 20:40 |
| mahatic | zaitcev, yes class A(B) | 20:41 |
| torgomatic | also, http://learnpythonthehardway.org/book/ex44.html is one of many examples online of Python inheritance, but it has some amusing words at the start that make it better than most. | 20:41 |
| mahatic | torgomatic, you're right. but OPTIONS not needed? when it is invoked in a node, and if that method is inherited, it need not be declared? | 20:42 |
| mahatic | torgomatic, thank you, will look at it. | 20:42 |
| torgomatic | mahatic: yes. if foo() is declared in the base class, and the base-class implementation is what you want the derived class to have too, then you don't need to define foo() in the derived class | 20:43 |
| mahatic | torgomatic, yes, right. I got confused with a similar code elsewhere. Thanks for the clarification! | 20:45 |
| mahatic | zaitcev, will look at them - db.py and backend.py - thanks | 20:49 |
| *** dmsimard is now known as dmsimard_away | 20:50 | |
| imkarrer | Good afternoon! I am curious. Which service is responsible for creating and setting permissions on the tmp directory on the devices holding the account and container DB's? The db_replicator depends tmp being available but I cannot find who is responsible for creating and setting permissions on the tmp directory | 20:56 |
| imkarrer | Is that something which needs to be created when swift is deployed? | 20:57 |
| notmyname | imkarrer: it's done in diskfile.py:1596 | 21:03 |
| notmyname | imkarrer: ie in the object server | 21:03 |
| imkarrer | notmyname: hmmm, its seems that tmp did not inherit the permissions of its parent directory. It does not appear that mkdirs assigns permissions after creating tmp | 21:06 |
| imkarrer | notmyname: it is probably more likely that the parent's permissions were not set properly, not allowing /tmp to be created. I need to further debug my issue. Thanks for the info! | 21:08 |
| notmyname | imkarrer: np | 21:08 |
| notmyname | imkarrer: it would be created by the user running the swift process. the default is to drop permissions to the "swift" user. so ya, maybe the parent dir wasn't set right | 21:09 |
| notmyname | does anyone (swift devs) consider keeping the sample config files in etc/ up to date to be hard or burdensome? | 21:10 |
| swifterdarrell | notmyname: yes, but I think it's totally worth it | 21:10 |
| notmyname | swifterdarrell: the current cross-project meeting is talking about some standard way to generate them. because recently they were removed from the source tree in other projects and operators complained | 21:11 |
| notmyname | ie no static file in the source tree. but also available to packagers | 21:11 |
| *** nellysmitt has joined #openstack-swift | 21:11 | |
| swifterdarrell | notmyname: doesn't seem hard or burdensome to maintain some wacky code to generate good examples files at all </scathing-sarcasm> | 21:12 |
| tdasilva | lol | 21:13 |
| mattoliverau | Morning | 21:14 |
| tdasilva | notmyname: wouldn't you have to generate a config to generate the config? or is it going to look in the code to find the default values for options? | 21:15 |
| *** tellesnobrega has quit IRC | 21:15 | |
| swifterdarrell | notmyname: perhaps it can achieve a similar level of quality as "git shortlog" for Changelog... | 21:15 |
| tdasilva | torgomatic: "Inheritance is the evil forest", pretty funny stuff | 21:16 |
| notmyname | tdasilva: I think it all comes from oslo config | 21:16 |
| *** nellysmitt has quit IRC | 21:17 | |
| *** tellesnobrega has joined #openstack-swift | 21:17 | |
| torgomatic | besides, sometimes the action of typing in the sample config shows me that my config option is stupidly named or has dumb semantics | 21:18 |
| mattoliverau | I like having the configs in /etc in most cases its better then the doc cause its something concrete with default values and explanations. Putting on my sysadmin hat, I found sample configs very useful. But that could just be me :) | 21:19 |
| *** mvenesio has quit IRC | 21:19 | |
| *** lpabon has quit IRC | 21:22 | |
| clayg | mattoliverau: it's definately not just you | 21:23 |
| clayg | notmyname: get out the sanity stick - THE BEATINGS WILL CONTINUE!!! | 21:23 |
| mattoliverau | Lol | 21:23 |
| notmyname | watching this conversation in the meeting is fascinating. I had no idea these are problems elsewhere in openstack | 21:25 |
| clayg | notmyname: conf.d with example configs from oslo messaging that get copied in when you do the oslo update dance? | 21:26 |
| notmyname | clayg: except that oslo is actually better about libraries now. so it's that an external library (real module!) getting updated can change config options in your app | 21:27 |
| clayg | i guess if you have a truly external depends that you configure pass-through the *only* correct answer is to say "see docs in this other project, because we *really* have no idea what's the union of options it's supports that work with our app" | 21:27 |
| notmyname | ya that | 21:27 |
| notmyname | and since ranges of versions for libraries are supported, then basically and statically generated file is always out of date unless installed versions match exactly | 21:28 |
| mattoliverau | I thought de stack generated the nova comfig but for other projects it grabs the sample config and then seds... But I guess that must've changed now | 21:28 |
| *** geaaru has quit IRC | 21:28 | |
| mattoliverau | *devstack | 21:28 |
| mattoliverau | Wow my typing is great this morning (I blame not enough coffee) | 21:28 |
| clayg | mattoliverau: i'm not so sure that's changed, we still have to merge default changes for keystone stuff into our /etc configs to "make devstack better" | 21:29 |
| * notmyname doesn't know if they are joking about an openstack-specs specs repo | 21:29 | |
| mattoliverau | clayg: how could that ever go wrong :p | 21:30 |
| mattoliverau | notmyname: can they call it openstack-specs^2 :p | 21:31 |
| clayg | this is how I imagine specs work for oslo -> https://www.youtube.com/watch?v=w3E1frdYzuo#t=16 | 21:31 |
| mattoliverau | Lol | 21:32 |
| *** NM has quit IRC | 21:41 | |
| *** dmsimard_away is now known as dmsimard | 21:55 | |
| *** Masahiro has joined #openstack-swift | 21:55 | |
| notmyname | oh my. they aren't joking https://review.openstack.org/#/q/status:open+project:openstack/openstack-specs+branch:master,n,z | 21:58 |
| *** Masahiro has quit IRC | 22:00 | |
| *** rmcall has quit IRC | 22:01 | |
| *** rmcall has joined #openstack-swift | 22:03 | |
| *** cdelatte has joined #openstack-swift | 22:06 | |
| openstackgerrit | Merged openstack/swift: Splice: Use `operator.or_` instead of a lambda-expression https://review.openstack.org/137307 | 22:11 |
| zaitcev | "we introduce a tiny but powerful library" | 22:11 |
| notmyname | zaitcev: ya, I liked that too :-) | 22:13 |
| *** bkopilov has quit IRC | 22:31 | |
| *** bkopilov has joined #openstack-swift | 22:36 | |
| *** rmcall has quit IRC | 23:00 | |
| *** miurahr has quit IRC | 23:07 | |
| *** rmcall has joined #openstack-swift | 23:07 | |
| *** rmcall has quit IRC | 23:10 | |
| *** nellysmitt has joined #openstack-swift | 23:12 | |
| *** rmcall has joined #openstack-swift | 23:14 | |
| *** nellysmitt has quit IRC | 23:17 | |
| *** rmcall has quit IRC | 23:18 | |
| *** rmcall has joined #openstack-swift | 23:20 | |
| *** dmsimard is now known as dmsimard_away | 23:23 | |
| *** ajiang38740 has quit IRC | 23:27 | |
| *** ajiang has quit IRC | 23:27 | |
| *** fandi has joined #openstack-swift | 23:38 | |
| *** fandi has quit IRC | 23:38 | |
| *** Masahiro has joined #openstack-swift | 23:44 | |
| *** Masahiro has quit IRC | 23:48 | |
| *** rmcall has quit IRC | 23:52 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!