| *** ho has joined #openstack-swift | 00:07 | |
| ho | good morning! | 00:10 |
|---|---|---|
| mattoliverau | ho: Morning | 00:16 |
| ho | mattoliverau: morning! | 00:17 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 00:23 |
| *** kei_yama_ has joined #openstack-swift | 00:28 | |
| *** kei_yama has quit IRC | 00:29 | |
| *** diazjf1 has quit IRC | 00:34 | |
| *** mwheckmann has quit IRC | 00:37 | |
| *** m_kazuhiro has joined #openstack-swift | 00:42 | |
| *** dmorita has joined #openstack-swift | 00:46 | |
| *** diazjf has joined #openstack-swift | 00:49 | |
| *** jerrygb has quit IRC | 00:55 | |
| *** hrou has joined #openstack-swift | 00:55 | |
| *** gyee has quit IRC | 00:57 | |
| torgomatic | well, I'm now pretty confident that clayg's ring change won't break on initial rebalance; I fuzz-tested about 11K random rings but didn't manage to crash it | 01:02 |
| notmyname | cool | 01:02 |
| torgomatic | now to work on making changes, then rebalancing | 01:02 |
| torgomatic | also, now to let my laptop quiet down | 01:02 |
| *** hrou has quit IRC | 01:04 | |
| *** thumpba has joined #openstack-swift | 01:06 | |
| *** mzhou has quit IRC | 01:08 | |
| *** jamielennox is now known as jamielennox|away | 01:10 | |
| *** thumpba has quit IRC | 01:12 | |
| openstackgerrit | Merged openstack/swift: go: fix problem with using single config file https://review.openstack.org/246500 | 01:13 |
| *** zhill has quit IRC | 01:50 | |
| *** diazjf has quit IRC | 01:55 | |
| mattoliverau | Lol, nice work torgomatic | 01:58 |
| *** mwheckmann has joined #openstack-swift | 02:05 | |
| *** mwheckmann has quit IRC | 02:17 | |
| *** mwheckmann has joined #openstack-swift | 02:17 | |
| *** jerrygb has joined #openstack-swift | 02:29 | |
| *** 7JTAA216T has joined #openstack-swift | 02:30 | |
| *** 7JTAA216T has quit IRC | 02:30 | |
| *** 21WAACTNP has joined #openstack-swift | 02:31 | |
| *** sanchitmalhotra has joined #openstack-swift | 02:43 | |
| jrichli | timburke: I just debugged the "old" version of what I had working. It used to work because seg_hash AND seg_resp.etag were both always None. | 02:53 |
| *** 21WAACTNP has quit IRC | 03:01 | |
| *** haomaiwang has joined #openstack-swift | 03:01 | |
| kota_ | notmyname: hello, i'd notify you that i will be absent from tommorow IRC meeting because of offline in my business flight. | 03:04 |
| kota_ | notmyname: sorry | 03:04 |
| *** kei_yama_ has quit IRC | 03:05 | |
| *** lpabon has quit IRC | 03:05 | |
| *** venkat_p has joined #openstack-swift | 03:06 | |
| *** kei_yama has joined #openstack-swift | 03:07 | |
| *** janonymous has quit IRC | 03:26 | |
| *** tsg has joined #openstack-swift | 03:30 | |
| kota_ | oh...tomorrow meeting addressed PyECLib updates I'd catch up the status, I won't be able to attend on time tho :( | 03:34 |
| *** janonymous has joined #openstack-swift | 03:36 | |
| *** mac_ified has joined #openstack-swift | 03:40 | |
| notmyname | kota_: thanks for letting me know | 03:42 |
| *** diazjf has joined #openstack-swift | 03:42 | |
| *** tsg has quit IRC | 03:43 | |
| *** haomaiwang has quit IRC | 04:01 | |
| *** haomaiwang has joined #openstack-swift | 04:01 | |
| *** km has joined #openstack-swift | 04:01 | |
| *** km is now known as Guest8285 | 04:02 | |
| *** Guest6043 has quit IRC | 04:03 | |
| *** kei_yama_ has joined #openstack-swift | 04:03 | |
| *** kei_yama has quit IRC | 04:04 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 04:11 | |
| *** sanchitmalhotra has quit IRC | 04:13 | |
| *** sanchitmalhotra has joined #openstack-swift | 04:20 | |
| *** elligottmc has quit IRC | 04:22 | |
| *** sanchitmalhotra1 has quit IRC | 04:22 | |
| *** ppai has joined #openstack-swift | 04:26 | |
| *** mwheckmann has quit IRC | 04:36 | |
| jrichli | timburke: sorry for the alert. I think that it might be some crypto changes of late causing the crpyto-meta for the etag to be missing. Still looking at verifying that. i still think its interesting those values used to be None, but oh well. | 04:40 |
| *** hrou has joined #openstack-swift | 04:43 | |
| *** jrichli has quit IRC | 04:43 | |
| *** trifon has joined #openstack-swift | 04:44 | |
| *** kei_yama_ has quit IRC | 04:46 | |
| *** Guest8285 has quit IRC | 04:47 | |
| *** km_ has joined #openstack-swift | 04:47 | |
| *** mac_ified has quit IRC | 04:49 | |
| *** trifon has quit IRC | 04:49 | |
| *** kei_yama has joined #openstack-swift | 04:52 | |
| *** eariasn has quit IRC | 04:52 | |
| *** eariasn has joined #openstack-swift | 04:53 | |
| *** jamielennox|away is now known as jamielennox | 04:54 | |
| *** eariasn has quit IRC | 04:57 | |
| *** janonymous has quit IRC | 04:59 | |
| *** haomaiwang has quit IRC | 05:01 | |
| *** klrmn2 has joined #openstack-swift | 05:01 | |
| *** klrmn1 has quit IRC | 05:01 | |
| *** haomaiwang has joined #openstack-swift | 05:01 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift: Add concurrent reads option to proxy https://review.openstack.org/117710 | 05:03 |
| *** tsg has joined #openstack-swift | 05:11 | |
| tsg | kota_: https://review.openstack.org/#/c/244287/ has the latest update in comments | 05:20 |
| tsg | kota_: s/update/status | 05:20 |
| *** jerrygb has quit IRC | 05:25 | |
| *** SkyRocknRoll has joined #openstack-swift | 05:33 | |
| kota_ | tsg: thx, I'll see that. | 05:35 |
| tsg | kota_: I added you to https://bugs.launchpad.net/precise-backports/+bug/1515710. Please see if you can ping someone on #ubuntu-devel to get a resolution accelerated. :-) I haven't had much luck during day time here last few days. Thank you | 05:38 |
| openstack | Launchpad bug 1515710 in vivid-backports "Please backport liberasurecode-1.1.0, python-pyeclib-1.1.1 to precise, trusty, vivid" [Undecided,New] | 05:38 |
| tsg | kota_: thanks | 05:38 |
| *** tsg has left #openstack-swift | 05:38 | |
| *** diazjf1 has joined #openstack-swift | 05:42 | |
| *** diazjf has quit IRC | 05:45 | |
| *** trifon has joined #openstack-swift | 05:56 | |
| *** hrou has quit IRC | 06:00 | |
| *** haomaiwang has quit IRC | 06:01 | |
| *** haomaiwang has joined #openstack-swift | 06:01 | |
| *** sanchitmalhotra1 has joined #openstack-swift | 06:04 | |
| *** SkyRocknRoll has quit IRC | 06:05 | |
| *** sanchitmalhotra has quit IRC | 06:07 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:20 | |
| *** rcernin has joined #openstack-swift | 06:37 | |
| *** diazjf1 has quit IRC | 06:37 | |
| *** haomaiwang has quit IRC | 06:42 | |
| *** ppai has quit IRC | 06:43 | |
| *** haomaiwang has joined #openstack-swift | 06:49 | |
| *** ppai has joined #openstack-swift | 06:54 | |
| *** zaitcev has quit IRC | 06:54 | |
| *** haomaiwang has quit IRC | 07:01 | |
| *** haomaiwang has joined #openstack-swift | 07:01 | |
| *** ppai has quit IRC | 07:01 | |
| *** pchng_ has joined #openstack-swift | 07:05 | |
| *** pchng has quit IRC | 07:09 | |
| *** ppai has joined #openstack-swift | 07:15 | |
| *** jerrygb has joined #openstack-swift | 07:25 | |
| *** jerrygb has quit IRC | 07:30 | |
| *** jmccarthy1 has joined #openstack-swift | 07:33 | |
| *** jmccarthy1 has quit IRC | 07:35 | |
| *** SkyRocknRoll has quit IRC | 07:36 | |
| *** eranrom has joined #openstack-swift | 07:43 | |
| *** eranrom has left #openstack-swift | 07:43 | |
| *** jmccarthy has quit IRC | 07:47 | |
| *** jmccarthy has joined #openstack-swift | 07:48 | |
| *** SkyRocknRoll has joined #openstack-swift | 07:49 | |
| *** breitz has quit IRC | 07:58 | |
| *** breitz1 has joined #openstack-swift | 07:58 | |
| *** haomaiwang has quit IRC | 08:01 | |
| *** haomaiwang has joined #openstack-swift | 08:01 | |
| *** arnox has joined #openstack-swift | 08:03 | |
| *** hseipp has joined #openstack-swift | 08:09 | |
| *** rledisez has joined #openstack-swift | 08:10 | |
| *** hseipp has quit IRC | 08:14 | |
| *** hseipp has joined #openstack-swift | 08:15 | |
| *** mahatic has joined #openstack-swift | 08:27 | |
| *** geaaru has joined #openstack-swift | 08:33 | |
| *** SkyRocknRoll has quit IRC | 08:43 | |
| *** jordanP has joined #openstack-swift | 08:49 | |
| *** ppai has quit IRC | 08:50 | |
| *** ho has quit IRC | 08:54 | |
| *** haomaiwang has quit IRC | 09:01 | |
| *** haomaiwang has joined #openstack-swift | 09:01 | |
| *** janonymous has joined #openstack-swift | 09:02 | |
| *** ppai has joined #openstack-swift | 09:04 | |
| *** mahatic has quit IRC | 09:05 | |
| *** mahatic has joined #openstack-swift | 09:08 | |
| *** dmorita has quit IRC | 09:19 | |
| *** m_kazuhiro has quit IRC | 09:19 | |
| *** joeljwright has joined #openstack-swift | 09:27 | |
| *** ChanServ sets mode: +v joeljwright | 09:27 | |
| *** janonymous has quit IRC | 09:36 | |
| *** SkyRocknRoll has joined #openstack-swift | 09:40 | |
| *** janonymous has joined #openstack-swift | 09:42 | |
| *** jistr has joined #openstack-swift | 09:42 | |
| *** SkyRocknRoll has quit IRC | 09:49 | |
| *** km_ has quit IRC | 09:51 | |
| *** kei_yama has quit IRC | 09:52 | |
| *** haomaiwang has quit IRC | 10:01 | |
| *** haomaiwang has joined #openstack-swift | 10:01 | |
| *** jmccarthy has quit IRC | 10:05 | |
| *** aix has joined #openstack-swift | 10:08 | |
| *** jmccarthy has joined #openstack-swift | 10:24 | |
| *** jerrygb has joined #openstack-swift | 10:26 | |
| *** jmccarthy has quit IRC | 10:26 | |
| *** jmccarthy has joined #openstack-swift | 10:27 | |
| *** daemontool has joined #openstack-swift | 10:28 | |
| *** daemontool has quit IRC | 10:30 | |
| *** daemontool has joined #openstack-swift | 10:30 | |
| *** jerrygb has quit IRC | 10:31 | |
| *** daemontool has quit IRC | 10:31 | |
| *** haomaiwang has quit IRC | 11:01 | |
| *** haomaiwang has joined #openstack-swift | 11:01 | |
| *** jeblair has quit IRC | 11:04 | |
| *** jeblair has joined #openstack-swift | 11:05 | |
| *** jmccarthy has quit IRC | 11:09 | |
| *** jmccarthy has joined #openstack-swift | 11:10 | |
| *** daemontool has joined #openstack-swift | 11:17 | |
| *** daemontool has quit IRC | 11:18 | |
| *** daemontool has joined #openstack-swift | 11:19 | |
| *** daemontool has quit IRC | 11:21 | |
| *** daemontool has joined #openstack-swift | 11:24 | |
| *** aix has quit IRC | 11:31 | |
| *** aix has joined #openstack-swift | 11:32 | |
| *** sanchitmalhotra1 has quit IRC | 11:34 | |
| *** SkyRocknRoll has joined #openstack-swift | 11:39 | |
| *** daemontool has quit IRC | 11:47 | |
| *** jmccarthy has quit IRC | 11:47 | |
| *** jmccarthy has joined #openstack-swift | 11:47 | |
| *** daemontool has joined #openstack-swift | 11:47 | |
| *** haomaiwang has quit IRC | 11:51 | |
| *** silor has joined #openstack-swift | 12:06 | |
| *** wolsen has quit IRC | 12:20 | |
| *** alejandrito has joined #openstack-swift | 12:23 | |
| *** eariasn has joined #openstack-swift | 12:27 | |
| *** hrou has joined #openstack-swift | 12:29 | |
| *** eariasn has quit IRC | 12:29 | |
| *** eariasn has joined #openstack-swift | 12:30 | |
| *** eariasn_ has joined #openstack-swift | 12:31 | |
| *** hrou has quit IRC | 12:34 | |
| *** eariasn has quit IRC | 12:34 | |
| venkat_p | Hi all | 12:44 |
| venkat_p | I am not able to delete the endpoint for swift using keystone | 12:44 |
| venkat_p | what can be the problem | 12:44 |
| venkat_p | This is devstack installation | 12:44 |
| venkat_p | I have not done anything after installation | 12:44 |
| *** eariasn_ has quit IRC | 12:49 | |
| *** eariasn has joined #openstack-swift | 12:49 | |
| *** eariasn has quit IRC | 12:53 | |
| *** NM has joined #openstack-swift | 12:54 | |
| *** joeljwright has quit IRC | 12:59 | |
| *** jerrygb has joined #openstack-swift | 13:06 | |
| *** jmccarthy has quit IRC | 13:09 | |
| *** jerrygb has quit IRC | 13:10 | |
| *** tab has joined #openstack-swift | 13:14 | |
| *** eariasn has joined #openstack-swift | 13:15 | |
| *** tab is now known as Guest99452 | 13:15 | |
| *** mac_ified has joined #openstack-swift | 13:16 | |
| *** eariasn_ has joined #openstack-swift | 13:16 | |
| *** joeljwright has joined #openstack-swift | 13:19 | |
| *** ChanServ sets mode: +v joeljwright | 13:19 | |
| *** eariasn has quit IRC | 13:19 | |
| *** joeljwright has quit IRC | 13:25 | |
| *** venkat_p has quit IRC | 13:25 | |
| *** joeljwright has joined #openstack-swift | 13:28 | |
| *** ChanServ sets mode: +v joeljwright | 13:28 | |
| *** jmccarthy has joined #openstack-swift | 13:30 | |
| *** mac_ified has quit IRC | 13:44 | |
| *** bill_az has joined #openstack-swift | 13:53 | |
| *** hrou has joined #openstack-swift | 13:59 | |
| *** bapalm has joined #openstack-swift | 14:08 | |
| *** ppai has quit IRC | 14:08 | |
| *** mwheckmann has joined #openstack-swift | 14:14 | |
| *** wbhuber has joined #openstack-swift | 14:19 | |
| *** daemontool_ has joined #openstack-swift | 14:25 | |
| openstackgerrit | Mike Fedosin proposed openstack/python-swiftclient: Fix swift client reauthentication https://review.openstack.org/246943 | 14:25 |
| *** daemontool_ has quit IRC | 14:26 | |
| *** daemontool_ has joined #openstack-swift | 14:27 | |
| *** daemontool has quit IRC | 14:27 | |
| *** daemontool_ has quit IRC | 14:28 | |
| *** daemontool_ has joined #openstack-swift | 14:29 | |
| *** hrou has left #openstack-swift | 14:30 | |
| *** dustins has joined #openstack-swift | 14:30 | |
| *** openstackgerrit has quit IRC | 14:31 | |
| *** openstackgerrit has joined #openstack-swift | 14:32 | |
| *** hrou has joined #openstack-swift | 14:33 | |
| *** bapalm has quit IRC | 14:43 | |
| *** bapalm has joined #openstack-swift | 14:43 | |
| *** lpabon has joined #openstack-swift | 14:46 | |
| *** pchng_ is now known as pchng | 14:51 | |
| *** daemontool_ has quit IRC | 14:51 | |
| *** petertr7_away is now known as petertr7 | 14:52 | |
| *** zaitcev has joined #openstack-swift | 14:55 | |
| *** ChanServ sets mode: +v zaitcev | 14:55 | |
| *** dustins has quit IRC | 14:57 | |
| *** dustins has joined #openstack-swift | 15:00 | |
| *** daemontool has joined #openstack-swift | 15:01 | |
| jordanP | guys, I'd like your opinion on https://review.openstack.org/#/c/246601/ | 15:01 |
| jordanP | "Add liberasurecode-dev as a swift dependency" | 15:01 |
| jordanP | if this gets merged, it won"t be possible to install swift-master on Ubuntu 12 | 15:02 |
| jordanP | because liberasurecode-dev is not available, as a package, on Ubuntu 12 | 15:02 |
| *** blmartin has joined #openstack-swift | 15:05 | |
| *** jrichli has joined #openstack-swift | 15:05 | |
| jordanP | I don't understand why, now, the system package liberasurecode-dev would be a requirement. | 15:15 |
| *** silor1 has joined #openstack-swift | 15:28 | |
| *** silor has quit IRC | 15:32 | |
| *** silor1 is now known as silor | 15:32 | |
| *** haomaiwang has joined #openstack-swift | 15:37 | |
| pchng | jordanP: I believe that change was done because later PyECLib versions don't include an EC implementation | 15:40 |
| pchng | jordanP: Perhaps related to this commit, but someone else might know better: https://github.com/openstack/swift/commit/8086a0e53406dc95856df2d3e015d3eaaf81380a | 15:41 |
| pchng | jordanP: So I think the change was done to "get ready" for the later versions of PyECLib, to include the requirements for it | 15:41 |
| jordanP | pchng, I am aware of this issue, but pining pyeclib to 1.0.7 worked so far | 15:41 |
| jordanP | so am not sure why we have to change this | 15:42 |
| pchng | jordanP: I believe you are right, things work fine, but I think the change to pin to PyECLib 1.0.7 was considered a temporary/interim one, at least that's what I can tell from that github commit message | 15:43 |
| jordanP | pchng, I understand. I still think that dropping support of Ubuntu 12 is not worth it and until we really can"t live with pyeclib 1.0.7, then we should stick to 1.0.7 but in the end it's not my call | 15:45 |
| jordanP | I'll wait and see then | 15:45 |
| notmyname | jordanP: tsg is working on getting that available https://bugs.launchpad.net/precise-backports/+bug/1515710 | 15:47 |
| openstack | Launchpad bug 1515710 in vivid-backports "Please backport liberasurecode-1.1.0, python-pyeclib-1.1.1 to precise, trusty, vivid" [Undecided,New] | 15:47 |
| notmyname | actually, that's one of the reasons the update to pyeclib hasn't landed yet | 15:48 |
| jordanP | notmyname, yup I've seen that bug report, I really hope his request will be accepted | 15:49 |
| notmyname | yeah, me too :-) | 15:49 |
| jordanP | I'll test liberasurecode 1.1.0-2 on Ubuntu 12 and leave a comment on the bug report | 15:50 |
| *** Yeo09 has joined #openstack-swift | 15:51 | |
| jordanP | notmyname, If you could leave a comment on https://review.openstack.org/#/c/246601 to slow things a bit that would help, cause at $my_job, I need to test swift on Ubuntu 12 and that patch will only make my job harder | 15:51 |
| jordanP | (I guess I shouldn't have picked devstack as my swift installer) | 15:52 |
| *** proteusguy__ has quit IRC | 15:53 | |
| Yeo09 | I have questions. How to create (PUT) and retrieve (GET) object files from (device, partition, node) | 15:54 |
| Yeo09 | tuple? (For example, when you store 'object_00001' object into 'container_1' | 15:54 |
| Yeo09 | container and 'account_1', describe which directories & files of that object | 15:54 |
| Yeo09 | are created in local filesystem.) | 15:54 |
| *** mzhou has joined #openstack-swift | 15:55 | |
| *** minwoob has joined #openstack-swift | 15:55 | |
| notmyname | jordanP: on the other hand, this whole process has taken so long I'm reticent to try to put the brakes on a particular aspect of it (except for things that are dependent on something else) | 15:56 |
| *** jerrygb has joined #openstack-swift | 15:57 | |
| notmyname | Yeo09: I'm not sure I understand your question | 15:57 |
| jordanP | agreed, but LP#1515710 was opened only 5 days ago. Maybe things will finally move | 15:57 |
| *** silor1 has joined #openstack-swift | 15:57 | |
| *** silor has quit IRC | 15:58 | |
| *** silor1 is now known as silor | 15:58 | |
| *** daemontool has quit IRC | 15:58 | |
| *** dustins has quit IRC | 15:58 | |
| *** dustins has joined #openstack-swift | 15:58 | |
| *** daemontool has joined #openstack-swift | 15:59 | |
| *** haomaiwang has quit IRC | 16:01 | |
| *** haomaiwang has joined #openstack-swift | 16:01 | |
| *** swiftchobo has joined #openstack-swift | 16:03 | |
| swiftchobo | please could you explain to me that the effect of configuration parameters defined in /etc/swift/***.conf. | 16:04 |
| pchng | swiftchobo: Configuration parameters are documented here: http://docs.openstack.org/developer/swift/deployment_guide.html | 16:06 |
| *** proteusguy__ has joined #openstack-swift | 16:06 | |
| pchng | swiftchobo: Is there a question about a specific configuration parameter you had? | 16:06 |
| *** rcernin has quit IRC | 16:09 | |
| swiftchobo | pchng : what are proxy-server.conf-write_affinity, read_affinity parameter | 16:09 |
| *** wuhg has quit IRC | 16:11 | |
| pchng | swiftchobo: I am not really familiar with those, but there is some documentation about it in this admin guide: http://docs.openstack.org/developer/swift/admin_guide.html#geographically-distributed-clusters | 16:12 |
| *** sayali has quit IRC | 16:13 | |
| *** jmccarthy has quit IRC | 16:14 | |
| swiftchobo | pchng : Okay. I will check it. thank you | 16:15 |
| *** jlhinson has joined #openstack-swift | 16:17 | |
| *** wolsen has joined #openstack-swift | 16:19 | |
| *** diazjf has joined #openstack-swift | 16:22 | |
| *** klrmn2 has quit IRC | 16:24 | |
| *** wbhuber_ has joined #openstack-swift | 16:24 | |
| *** sayali has joined #openstack-swift | 16:26 | |
| *** wbhuber has quit IRC | 16:26 | |
| *** jmccarthy has joined #openstack-swift | 16:27 | |
| *** trifon has quit IRC | 16:30 | |
| *** jmccarthy has quit IRC | 16:32 | |
| *** jmccarthy has joined #openstack-swift | 16:33 | |
| timburke | jrichli: interesting...it makes sense that if seg_resp.etag is None, then seg_hash would be as well. i'm assuming some recent crypto work started passing back an etag, which triggered the md5 check | 16:39 |
| timburke | in the error that gets logged, does the "headers had ____" etag match either of the encrypted etags? | 16:40 |
| jrichli | timburke: i'll have an answer in a big | 16:42 |
| jrichli | s/big/bit/ | 16:42 |
| *** jlhinson has quit IRC | 16:45 | |
| *** jlhinson has joined #openstack-swift | 16:48 | |
| jrichli | timburke: I am pretty sure its the one that is encrypted with the obj key. it just isnt decrypted because the crypto-meta is not present at decrypt time. that is what i am looking into now. | 16:49 |
| jrichli | but, there is no added passing back an etag or anything from the time when the items were "none" last time I tested. at least, nothing that was knowingly different. i guess a bug could have been fixed. | 16:51 |
| timburke | jrichli: cool, makes sense. another thought i had: you said yesterday that the slo PUT path isn't working yet, so...how is encryption coming into play here? is it that we had an existing slo, then replaced some segments with encrypted versions of themselves? | 16:51 |
| *** blmartin has quit IRC | 16:51 | |
| timburke | huh | 16:51 |
| timburke | weird | 16:51 |
| *** haomaiwang has quit IRC | 16:51 | |
| jrichli | the slo PUT path does work for the normal stuff on our feature branch | 16:51 |
| *** blmartin has joined #openstack-swift | 16:52 | |
| timburke | against encrypted segments? | 16:52 |
| jrichli | it just isnt working right now with my patch that replaces fake_footers with real footers support | 16:52 |
| *** chsc has joined #openstack-swift | 16:52 | |
| timburke | ah :) | 16:52 |
| *** petertr7 is now known as petertr7_away | 16:52 | |
| jrichli | yes, against encrypted segments - normal slo PUT works | 16:52 |
| timburke | love it | 16:52 |
| *** arnox has quit IRC | 16:54 | |
| notmyname | good morning (again) | 16:57 |
| *** tongli has joined #openstack-swift | 17:00 | |
| *** jlhinson_ has joined #openstack-swift | 17:00 | |
| *** jlhinson has quit IRC | 17:02 | |
| *** klrmn1 has joined #openstack-swift | 17:03 | |
| *** SkyRocknRoll_ has joined #openstack-swift | 17:05 | |
| *** diazjf has quit IRC | 17:06 | |
| *** diazjf has joined #openstack-swift | 17:08 | |
| *** petertr7_away is now known as petertr7 | 17:09 | |
| *** sileht has quit IRC | 17:10 | |
| *** sileht has joined #openstack-swift | 17:11 | |
| *** jistr has quit IRC | 17:11 | |
| notmyname | joeljwright: timburke: follow-up from last week's py26 conversation. I think we should drop py26 in swiftclient, and if a user is on a distro (ie red hat) that is supporting/pinning py26, the user needs to ask that distro for support | 17:13 |
| notmyname | joeljwright: timburke: I also think this should warrant a major version bump for swiftclinet | 17:13 |
| joeljwright | notmyname: agreed | 17:13 |
| joeljwright | I found enough ways to work round the problem | 17:13 |
| joeljwright | ius and softwarecollections both have a parallel installable py27 | 17:14 |
| timburke | seems fair. if we're planning a major version bump *anyway*, are there any breaking api changes we'd like to see? | 17:14 |
| *** diazjf has quit IRC | 17:16 | |
| joeljwright | notmyname: are we talking 2.6 -> 2.7? | 17:16 |
| joeljwright | with 2.7+ losing py26 support (nice parallel there) | 17:17 |
| timburke | joeljwright: 2.6 -> 3.0 | 17:17 |
| timburke | dun-dun-DUUUUN | 17:18 |
| joeljwright | :D | 17:18 |
| joeljwright | are there any swift api updates in the pipeline? | 17:18 |
| joeljwright | there have been chats about a v2 api bubbling under the surface for a while | 17:18 |
| joeljwright | but not sure if anyone's actually serious… | 17:19 |
| *** diazjf has joined #openstack-swift | 17:19 | |
| timburke | i think probably not terribly. notmyname is over at my desk now, though, tossing about the idea of changing the default auth version as part of this...i was thinking purely client api breaks | 17:19 |
| *** daemontool has quit IRC | 17:20 | |
| *** diazjf1 has joined #openstack-swift | 17:20 | |
| *** gyee has joined #openstack-swift | 17:20 | |
| joeljwright | there was a patch a while back that we couldn't implement because of potentially breaking clients | 17:21 |
| timburke | yeah...lemme look for that... | 17:21 |
| * joeljwright digging in history | 17:22 | |
| joeljwright | https://review.openstack.org/#/c/189815/ | 17:22 |
| timburke | yep, just found it too :) | 17:23 |
| *** diazjf has quit IRC | 17:23 | |
| notmyname | there would essentially be 2 releases if we did that. the last 2.x series and then a 3.0 | 17:23 |
| joeljwright | it never felt important enough at the time | 17:23 |
| joeljwright | I'd be happy enough to cut a 2.7 release and leave breaking changes for now | 17:24 |
| notmyname | generally, I'd think that it's not good to go do a lot of API changes right now just because we're doing a bump for py support | 17:24 |
| notmyname | yeah, what joeljwright said :-) | 17:24 |
| joeljwright | it's no great trauma to install python27 in parallel to the system python on rhel based distros anyway | 17:25 |
| notmyname | however, we had talked about raising the default auth version and that would be a major version bump too | 17:26 |
| joeljwright | yeah | 17:26 |
| notmyname | that's the only think I'd consider | 17:26 |
| notmyname | *thing | 17:27 |
| joeljwright | it needs to be done eventually | 17:27 |
| joeljwright | and now is as good a time as any :) | 17:27 |
| notmyname | maybe ;-) | 17:27 |
| joeljwright | now, how do we make pip flash red 'sorry I broke your client' messages on install...? | 17:28 |
| notmyname | for the auth version change? | 17:29 |
| joeljwright | yeah :) | 17:29 |
| joeljwright | let's face it, that's gonna bite more people | 17:29 |
| notmyname | oh yeah | 17:29 |
| notmyname | including us. since there's no way right now for a v3 auth to even work with our out-of-the box installs | 17:30 |
| notmyname | or rather it just makes it a lot harder (/me hasn't looked at the keystone vSAIO stuff yet) | 17:31 |
| joeljwright | and you're still putting this forward? you might want to check whether any of your support guys are dangerous first... | 17:31 |
| notmyname | lol | 17:31 |
| notmyname | it's the swift devs I'm more worried about ;-) | 17:31 |
| joeljwright | :) | 17:31 |
| notmyname | I think the default auth version bump should happen. and whenever it does happen, it should have a major version bump | 17:32 |
| notmyname | also, we probably shouldn't do more than one major version bump in a single openstack release cycle (that's an assumption) | 17:32 |
| *** jordanP has quit IRC | 17:33 | |
| notmyname | but changing the default will have significant impact on end users | 17:33 |
| joeljwright | so do we lose py26 now and go to 2.7.0 | 17:34 |
| joeljwright | and save the authv3 for the N-release? | 17:34 |
| joeljwright | or intermediate release | 17:34 |
| *** tsg has joined #openstack-swift | 17:34 | |
| joeljwright | tie the client breaking change to a significant event | 17:34 |
| notmyname | IIRC what we talked about in tokyo is making a right default based on envvars and options given, and then respecting the version if explicitly given | 17:35 |
| notmyname | joeljwright: I don't know | 17:35 |
| * joeljwright needs to look carefully at all those envvars | 17:35 | |
| timburke | notmyname: i have a fear related to the auth v3 idea...i feel like it'll just push more swift devs to curl instead of our own cli | 17:35 |
| notmyname | on the other hand, waiting another 6 months for an auth version update seems to be a long time | 17:35 |
| timburke | it's a hell of a lot of env vars... | 17:36 |
| notmyname | the OS_* ones | 17:36 |
| notmyname | timburke: I *like* curl ;-) | 17:36 |
| timburke | i like it too! but if we want a better client, we need people using the client, finding the stupid, and fixing it | 17:37 |
| *** jlhinson_ has quit IRC | 17:37 | |
| joeljwright | there must be combinations of OS_* options that would let us make the right choice in most cases | 17:39 |
| notmyname | making the right choice based on the options given might be a reasonable midpoint | 17:39 |
| notmyname | yeah | 17:39 |
| notmyname | and that might be possible to do in a release along with another major version bump | 17:41 |
| notmyname | or rather, would that require a major version bump at all? | 17:41 |
| joeljwright | is parsing the auth_url not also an option for a guess? | 17:41 |
| notmyname | I don't know. someone said that's a bad idea. or that we can't/shouldn't rely on that for version discovery | 17:42 |
| notmyname | but I don't remember who was saying that or why | 17:42 |
| *** jlhinson_ has joined #openstack-swift | 17:42 | |
| joeljwright | it could always be a maybe | 17:42 |
| notmyname | it's definitely a maybe | 17:42 |
| *** jmccarthy has quit IRC | 17:43 | |
| joeljwright | I think there will be plenty of people with string opinons later | 17:43 |
| notmyname | we're never short on that | 17:43 |
| joeljwright | I know I currently don't specify an auth version in the OS_* vars we give to clients | 17:43 |
| joeljwright | so I'm somewhat nervous | 17:43 |
| joeljwright | would like to hear other people's opinions on the matter | 17:45 |
| *** mac_ified has joined #openstack-swift | 17:45 | |
| notmyname | definitely | 17:46 |
| notmyname | I added it to a swiftclient topic for today's meeting | 17:46 |
| joeljwright | I wonder if the ksclient has any logic for this… | 17:46 |
| *** jlhinson_ has quit IRC | 17:46 | |
| *** jlhinson_ has joined #openstack-swift | 17:47 | |
| *** diazjf1 has quit IRC | 17:47 | |
| *** hseipp has quit IRC | 17:48 | |
| *** rledisez has quit IRC | 17:50 | |
| *** aix has quit IRC | 17:51 | |
| joeljwright | we must be able to query the auth_url for supported versions | 17:52 |
| notmyname | joeljwright: timburke: the side effect of even just the py26 stuff is doing another swiftclient release, so that's good :-) | 17:52 |
| notmyname | joeljwright: yeah, I think keystone supports that | 17:52 |
| *** zhill has joined #openstack-swift | 17:52 | |
| joeljwright | so let's pick a supported version starting at 3.0 :) | 17:53 |
| *** petertr7 is now known as petertr7_away | 17:53 | |
| joeljwright | I'll have a think about this before the meeting | 17:53 |
| joeljwright | unfortunately I now have to go collect my wife from the station and buy&cook dinner :S | 17:54 |
| *** diazjf has joined #openstack-swift | 17:54 | |
| *** jerrygb_ has joined #openstack-swift | 17:54 | |
| *** jerrygb_ has quit IRC | 17:54 | |
| *** diazjf has quit IRC | 17:55 | |
| *** jerrygb_ has joined #openstack-swift | 17:55 | |
| *** diazjf has joined #openstack-swift | 17:57 | |
| *** jerrygb has quit IRC | 17:57 | |
| *** jerrygb_ has quit IRC | 17:59 | |
| *** jerrygb_ has joined #openstack-swift | 18:01 | |
| *** jerrygb_ has quit IRC | 18:01 | |
| *** itlinux has joined #openstack-swift | 18:01 | |
| *** jerrygb_ has joined #openstack-swift | 18:01 | |
| *** joeljwright has quit IRC | 18:01 | |
| *** wbhuber has joined #openstack-swift | 18:04 | |
| *** daemontool has joined #openstack-swift | 18:05 | |
| *** wbhuber_ has quit IRC | 18:07 | |
| *** daemontool has quit IRC | 18:16 | |
| *** jerrygb_ has quit IRC | 18:17 | |
| *** diazjf has quit IRC | 18:18 | |
| *** mac_ified has quit IRC | 18:20 | |
| *** wbhuber_ has joined #openstack-swift | 18:42 | |
| *** trifon has joined #openstack-swift | 18:43 | |
| *** jerrygb has joined #openstack-swift | 18:44 | |
| *** wbhuber has quit IRC | 18:44 | |
| *** petertr7_away is now known as petertr7 | 18:45 | |
| *** geaaru has quit IRC | 18:51 | |
| *** badari has joined #openstack-swift | 18:56 | |
| *** petertr7 is now known as petertr7_away | 19:00 | |
| *** jlhinson_ has quit IRC | 19:00 | |
| *** wbhuber has joined #openstack-swift | 19:06 | |
| *** itlinux_ has joined #openstack-swift | 19:07 | |
| *** wbhuber_ has quit IRC | 19:08 | |
| *** diazjf has joined #openstack-swift | 19:08 | |
| *** itlinux has quit IRC | 19:08 | |
| *** petertr7_away is now known as petertr7 | 19:21 | |
| *** bapalm has quit IRC | 19:23 | |
| *** wbhuber_ has joined #openstack-swift | 19:25 | |
| *** bapalm has joined #openstack-swift | 19:26 | |
| *** wbhuber has quit IRC | 19:28 | |
| *** blmartin has quit IRC | 19:28 | |
| *** mahatic has quit IRC | 19:30 | |
| *** jlhinson_ has joined #openstack-swift | 19:41 | |
| *** afaranha has joined #openstack-swift | 19:47 | |
| *** tsg has quit IRC | 19:48 | |
| *** onder has quit IRC | 19:50 | |
| *** onder has joined #openstack-swift | 19:53 | |
| afaranha | Hey people, I just tried to run the devstack with swift and I got an error while installing the PyECLib: http://paste.openstack.org/show/479333/ | 19:57 |
| afaranha | can anyone see this problem? | 19:57 |
| *** nadeem has joined #openstack-swift | 19:57 | |
| afaranha | does anyone have any idea? | 19:58 |
| *** itlinux_ has quit IRC | 19:59 | |
| NM | afaranha: Can you do pip install PyECLib===1.0.7 ? | 19:59 |
| *** zhill has quit IRC | 19:59 | |
| afaranha | NM, yes, I can | 19:59 |
| afaranha | it install without problem, but when running stack.sh it raises error | 20:00 |
| afaranha | I just tried run "from pyeclib.ec_iface import ECDriver, ECDriverError, VALID_EC_TYPES" in a python and I got an error | 20:00 |
| afaranha | ah, When running pip install I get: "Failed to build PyECLib" | 20:01 |
| *** itlinux has joined #openstack-swift | 20:02 | |
| afaranha | NM, some news, running pip install with -v we got: Error: liberasurecode-1.0.7 build failed! Please install liberasurecode-1.0.7 manually and retry | 20:03 |
| *** silor1 has joined #openstack-swift | 20:06 | |
| *** diazjf has quit IRC | 20:08 | |
| *** jlhinson_ has quit IRC | 20:08 | |
| *** NM has quit IRC | 20:09 | |
| *** diazjf has joined #openstack-swift | 20:09 | |
| *** [1]eranrom has joined #openstack-swift | 20:09 | |
| *** petertr7 is now known as petertr7_away | 20:09 | |
| *** silor has quit IRC | 20:09 | |
| *** silor1 is now known as silor | 20:09 | |
| *** jlhinson_ has joined #openstack-swift | 20:09 | |
| *** itlinux has quit IRC | 20:10 | |
| *** wbhuber_ is now known as wbhuber | 20:11 | |
| *** itlinux has joined #openstack-swift | 20:11 | |
| [1]eranrom | Greetings everyone. I got a question on get_info https://github.com/openstack/swift/blob/master/swift/proxy/controllers/base.py Line 534 | 20:11 |
| *** zhill has joined #openstack-swift | 20:13 | |
| [1]eranrom | seems like we rely on GetorHead_base to populate the environment with the info, which is true only on another environment. I beleive that the env GetorHead_base operates on is different then the response env in line 534 | 20:13 |
| [1]eranrom | I feel I am missing something, as this implies that any first call to get_info just populates the cache but do not actually return anything usefull | 20:14 |
| [1]eranrom | s/usefull/useful | 20:14 |
| *** swiftchobo has quit IRC | 20:15 | |
| *** silor1 has joined #openstack-swift | 20:15 | |
| *** silor has quit IRC | 20:16 | |
| *** silor1 is now known as silor | 20:16 | |
| *** diazjf has quit IRC | 20:16 | |
| *** NM has joined #openstack-swift | 20:17 | |
| *** diazjf has joined #openstack-swift | 20:18 | |
| NM | Nice. Can you find that lib? | 20:18 |
| *** petertr7_away is now known as petertr7 | 20:24 | |
| *** diazjf1 has joined #openstack-swift | 20:27 | |
| *** diazjf has quit IRC | 20:30 | |
| *** diazjf has joined #openstack-swift | 20:30 | |
| *** diazjf2 has joined #openstack-swift | 20:31 | |
| *** itlinux has quit IRC | 20:33 | |
| *** diazjf1 has quit IRC | 20:34 | |
| *** diazjf has quit IRC | 20:34 | |
| *** tongli has quit IRC | 20:38 | |
| *** ho-mobile has joined #openstack-swift | 20:46 | |
| *** SkyRocknRoll has quit IRC | 20:48 | |
| *** ho-mobile_ has joined #openstack-swift | 20:48 | |
| *** wbhuber has quit IRC | 20:49 | |
| *** wbhuber has joined #openstack-swift | 20:50 | |
| *** ho-mobile has quit IRC | 20:51 | |
| *** ho-mobile_ is now known as ho-mobile | 20:51 | |
| *** itlinux has joined #openstack-swift | 20:51 | |
| *** vinsh_ has joined #openstack-swift | 20:51 | |
| torgomatic | [1]eranrom: yeah, that get_info stuff is pretty darn crazy IMO. should probably get reworked to populate stuff based on the response, not based on fishing around in getorhead_base | 20:52 |
| *** vinsh has quit IRC | 20:54 | |
| torgomatic | at a minimum, get_info() doesn't work with the 410 response from an account, for example | 20:54 |
| torgomatic | like, if you GET or HEAD an account that's recently deleted, the proxy gives you back a 410 instead of a 404 | 20:54 |
| torgomatic | but get_info() returns 404 because it's doing its work underneath that code, down in the base controller, instead of reading the response | 20:55 |
| torgomatic | a coworker of mine ran into that recently | 20:55 |
| torgomatic | I'm sure there are other such things that you also can't read via get_info() as well, but that's the only one I can think of at the moment. | 20:56 |
| mattoliverau | morning | 20:57 |
| torgomatic | hi | 20:57 |
| notmyname | meeting time | 21:00 |
| [1]eranrom | torgomatic: ok, thanks very much for this. | 21:00 |
| *** vinsh has joined #openstack-swift | 21:00 | |
| *** vinsh_ has quit IRC | 21:00 | |
| *** [1]eranrom has left #openstack-swift | 21:00 | |
| *** joeljwright has joined #openstack-swift | 21:02 | |
| *** ChanServ sets mode: +v joeljwright | 21:02 | |
| *** mzhou has quit IRC | 21:02 | |
| *** openstack has joined #openstack-swift | 21:05 | |
| *** itlinux has quit IRC | 21:05 | |
| *** trifon has quit IRC | 21:07 | |
| *** petertr7 is now known as petertr7_away | 21:08 | |
| *** petertr7_away is now known as petertr7 | 21:09 | |
| *** jistr has joined #openstack-swift | 21:10 | |
| *** jerrygb_ has joined #openstack-swift | 21:12 | |
| *** NM has quit IRC | 21:12 | |
| timburke | oh, joeljwright: mind taking a look at patch 226897 when you get a chance? i think it should address the underlying issue that patch 246943 is trying to work around | 21:13 |
| patchbot | timburke: https://review.openstack.org/#/c/226897/ - Make LengthWrappers resettable if their _readable ... | 21:13 |
| patchbot | timburke: https://review.openstack.org/#/c/246943/ - Fix swift client reauthentication | 21:13 |
| joeljwright | timburke: kk, will do | 21:13 |
| timburke | thanks! | 21:14 |
| *** jerrygb has quit IRC | 21:15 | |
| *** jerrygb_ has quit IRC | 21:22 | |
| *** diazjf2 has quit IRC | 21:22 | |
| *** jerrygb_ has joined #openstack-swift | 21:23 | |
| *** breitz1 is now known as breitz | 21:25 | |
| *** jerrygb__ has joined #openstack-swift | 21:26 | |
| *** jerrygb__ has quit IRC | 21:27 | |
| *** jerrygb has joined #openstack-swift | 21:28 | |
| *** jerrygb_ has quit IRC | 21:28 | |
| openstackgerrit | Alan Erwin proposed openstack/swift-specs: Updating spec to reflect a few changes. https://review.openstack.org/247185 | 21:28 |
| *** silor has quit IRC | 21:30 | |
| *** diazjf has joined #openstack-swift | 21:30 | |
| *** diazjf has joined #openstack-swift | 21:30 | |
| *** mzhou has joined #openstack-swift | 21:37 | |
| *** jistr has quit IRC | 21:41 | |
| *** blmartin has joined #openstack-swift | 21:44 | |
| *** dustins has quit IRC | 21:45 | |
| *** NM has joined #openstack-swift | 21:47 | |
| *** tsg has joined #openstack-swift | 21:56 | |
| *** ho-mobile has quit IRC | 21:57 | |
| notmyname | tsg: oh hi. just missed the meeting. any updates on pyeclib? just waiting for those distro packaging things? | 21:57 |
| gmmaha | notmyname: didnt want the hijack the meeting, but when my env has both ST_* and OS_* set & keystone enabled, swift client always default to ST_* and fails to auth with keystone. is this what discover feature will solve? | 21:59 |
| notmyname | gmmaha: not immediately | 21:59 |
| notmyname | gmmaha: the plan would be that if you had no ST_* options and no auth version specified, it would do auth version discoverability | 22:00 |
| gmmaha | aaah | 22:00 |
| *** jerrygb has quit IRC | 22:00 | |
| gmmaha | guess then the other is an enhancement at some point.. cool.. thanks | 22:00 |
| notmyname | of course, if you want to future-proof your app, always explicitly set the auth version | 22:01 |
| jrichli | will a lot of people be out next week due to Thanksgiving? I will be on holiday technically, but will prob still be doing swift and asking questions :-) | 22:01 |
| notmyname | oh yeah | 22:02 |
| notmyname | next week | 22:02 |
| notmyname | is US holiday | 22:02 |
| * torgomatic will be in a turkey coma and completely unresponsive | 22:02 | |
| acoles | yeah i was wondering if you guys would be here | 22:02 |
| gmmaha | notmyname: yeah. thats definitely something that might solve that | 22:02 |
| notmyname | shouldn't be any meeting | 22:02 |
| torgomatic | so, you know, about like normal | 22:02 |
| notmyname | torgomatic: on wednesday? | 22:02 |
| wbhuber | lol at torgomatic | 22:02 |
| minwoob | Quick question | 22:03 |
| acoles | notmyname: all week is holiday? | 22:03 |
| wbhuber | acoles: usually thurs and fri of next week | 22:03 |
| notmyname | I'm taking off wed-fri | 22:03 |
| minwoob | why do I not see the process/daemon that I created, when I do swift-init, or even a ps -ef ? | 22:03 |
| acoles | wbhuber: ah | 22:03 |
| notmyname | I'll be on an airplane or in a car during the meeting time next week | 22:03 |
| joeljwright | notmyname: I'll be away for the first half of December, so I guess the next meeting I can make is Dec 16th | 22:04 |
| minwoob | I've modified common/manager.py, bin, setup.cfg, and added the actual file in swift/container/ | 22:04 |
| joeljwright | but I'll keep an eye out for swiftclient docs info | 22:04 |
| jrichli | notmyname: anywhere new? | 22:04 |
| notmyname | joeljwright: yikes. ok. we definitely should be talking more than just in the meetings though :-) | 22:04 |
| acoles | joeljwright has THE best plans for december! | 22:04 |
| joeljwright | acoles: don't tell everyone! | 22:05 |
| notmyname | jrichli: I'll be in the exciting town of dallas ;-) | 22:05 |
| acoles | jrichli: which days are you on holiday? | 22:05 |
| notmyname | jrichli: majorca? ibiza? | 22:05 |
| notmyname | joeljwright: ^ | 22:05 |
| acoles | joeljwright: i was vague! | 22:05 |
| acoles | :) | 22:05 |
| jrichli | Big D! have fun | 22:05 |
| joeljwright | notmyname: Zermatt | 22:05 |
| notmyname | nice | 22:06 |
| joeljwright | taking a course to become a ski guide | 22:06 |
| notmyname | cool | 22:06 |
| tsg | notmyname: sorry missed the meeting due to a conflict. on the pkg updates, the backports (https://bugs.launchpad.net/trusty-backports/+bug/1515710) are still pending, although I finally heard from ubuntu-backporters - need to resubmit using the ubuntu tools, which I will do this afternoon. | 22:06 |
| openstack | Launchpad bug 1515710 in vivid-backports "Please backport liberasurecode-1.1.0, python-pyeclib-1.1.1 to precise, trusty, vivid" [Undecided,Invalid] | 22:06 |
| jrichli | acoles: again, technically - I am taking off work all week next week. but that just means I might be more productive ;-) | 22:06 |
| wbhuber | jrichli: that makes 2 of us | 22:06 |
| tsg | In the meantime, I submitted a new review request for devstack to get liberasurecode-dev in - you may have seen it | 22:06 |
| notmyname | jrichli: I think that's a common sentiment among swift devs ;-) | 22:06 |
| acoles | jrichli: no, think long term, take your vacation | 22:07 |
| mattoliverau | minwoob: does it need a configuration section in a config or it's own config? | 22:08 |
| jrichli | acoles: its ok, really. i will get enough of the relaxing stuff. the work parts i will do are just the fun parts of work - i wont have to do the things that feel more like work. | 22:08 |
| gmmaha | notmyname: if you dont mind, wanted to get your thoughts on one more thing. I ran the testr patch and functests on my SAIO setup and ran into a bunch of errors | 22:08 |
| gmmaha | http://paste.openstack.org/show/479352/ | 22:08 |
| minwoob | mattoliverau: I modified the container-server config file and added it in (just like the container-sharder). | 22:08 |
| gmmaha | i did verify that all the requirements are satisfied. Not sure what i might be missing | 22:08 |
| minwoob | mattoliverau: This is for the local table partitioner daemon that I'm trying to create, btw. | 22:09 |
| mattoliverau | minwoob: I guessed as much ;) | 22:09 |
| minwoob | lol | 22:09 |
| mattoliverau | minwoob: you said you added it to manager.py, when happens when you start it manually "swift-init <service> start -n" | 22:11 |
| notmyname | gmmaha: looking | 22:11 |
| gmmaha | notmyname: thanks.. No rush, i am most certain its a user error.. ;) | 22:11 |
| *** diazjf has quit IRC | 22:12 | |
| notmyname | gmmaha: you've validated your test.conf | 22:12 |
| notmyname | ? | 22:12 |
| minwoob | mattoliverau: dev@dev-VirtualBox:~/swift/bin$ sudo swift-init container-partitioner start -n | 22:13 |
| gmmaha | yes. havent changed it in a while.. its the one i have been using | 22:13 |
| minwoob | Starting container-partitioner...(/etc/swift/container-server/1.conf) | 22:13 |
| gmmaha | without the testr patch, functests passes just fine | 22:13 |
| minwoob | swift-container-partitioner does not exist | 22:13 |
| * gmmaha goes to look at it again | 22:13 | |
| minwoob | That's it. | 22:13 |
| notmyname | gmmaha: that was just an initial guess. not sure what's going on | 22:13 |
| minwoob | It's there in bin, though. | 22:13 |
| minwoob | Permissions are the same, etc. | 22:14 |
| gmmaha | notmyname: if it helps, this is my test.conf http://paste.openstack.org/show/479354/ | 22:15 |
| mattoliverau | minwoob: have you 'pip installed -e .' or 'sudo python setup.py develop' in case it isn't actually installed to the system properly | 22:15 |
| minwoob | mattoliverau: I issued setup.py again and now it works. | 22:17 |
| minwoob | Thanks a lot! | 22:17 |
| *** diazjf has joined #openstack-swift | 22:18 | |
| minwoob | I'll let you know how the rest of it goes. | 22:18 |
| mattoliverau | minwoob: awesome! please do! | 22:19 |
| *** acoles is now known as acoles_ | 22:31 | |
| peluse | can't seem to get the final steps of keystone on my SAIO working right. can interact with keystone OK, created all the relevant enteties I think and updated my proxy settings. Puking on me though and likely something silly. When anyone has a second: https://gist.github.com/anonymous/bfa221c146ffad435eb8 | 22:31 |
| *** petertr7 is now known as petertr7_away | 22:35 | |
| timburke | peluse: what are these headers? i thought v2 auth expected you to POST json...? | 22:36 |
| peluse | timburke: beats the hell out of me, I can't find devent docs showing exactly what is supposed to work so I've tried a few different things, that's just the last thing I tried | 22:36 |
| notmyname | peluse: heh https://twitter.com/alrs/status/667107893246726145 | 22:37 |
| timburke | give http://developer.openstack.org/api-ref-identity-v2.html a try | 22:37 |
| peluse | notmyname: no shit! | 22:38 |
| *** tsg has quit IRC | 22:38 | |
| timburke | peluse: i'm guessing you'll want something like {"auth":{"tenantName":"swift","passwordCredentials":{"username":"swift","password":"swift"}}} | 22:39 |
| peluse | timburke: yeah, was just looking at the output of 'get versions' and trying to figure out which is the simplest | 22:39 |
| *** jrichli has quit IRC | 22:40 | |
| *** diazjf has quit IRC | 22:42 | |
| *** diazjf has joined #openstack-swift | 22:43 | |
| peluse | bah nothing works, this is a pile of dogshit | 22:44 |
| notmyname | peluse: http://cdn.honestnetworker.com/dns/CRVCgHBUEAAj-Vg.mp4.gif | 22:49 |
| *** rjaiswal has joined #openstack-swift | 22:52 | |
| peluse | ok, still dogshit but timburke helped me get a token finally.... thanks tim! | 23:04 |
| timburke | happy to help | 23:04 |
| *** jerrygb has joined #openstack-swift | 23:05 | |
| *** tsg has joined #openstack-swift | 23:05 | |
| peluse | BTW if anyone is looking for a simple curl to get a token: curl -d '{"auth": {"tenantName": "swift", "passwordCredentials": {"username": "swift", "password": "swift"}}}' -H 'Content-type: application/json' http://localhost:5000/v2.0/tokens | 23:06 |
| *** km has joined #openstack-swift | 23:06 | |
| *** km is now known as Guest53794 | 23:07 | |
| *** arcimboldo has joined #openstack-swift | 23:08 | |
| arcimboldo | hi all, I'm trying to configure swift to use keystone, but I get EmptyCatalog on proxy-server | 23:09 |
| *** jerrygb has quit IRC | 23:09 | |
| *** minwoob has quit IRC | 23:11 | |
| *** jerrygb has joined #openstack-swift | 23:14 | |
| peluse | arcimboldo: I'm no expert but just now got my swift & keystone working. I only have like 3 min before I'm walking out the door but if you can explain a little more about what's happening maybe I can help | 23:15 |
| arcimboldo | thnx peluse, I think I've got it | 23:15 |
| arcimboldo | i was using the wrong sample fro proxy-server.conf | 23:15 |
| hrou | Ah no meeting next week ! Cool that means I'll just get to bug some of you here re: symlinks on this channel ;) | 23:16 |
| peluse | arcimboldo: great! | 23:16 |
| arcimboldo | I've spent like 2 hours before finding the correct configuration :( | 23:17 |
| notmyname | arcimboldo: is there something you can recommend changing in the sample config or the docs? if it's fresh on your mind, any help would be appreciated | 23:17 |
| notmyname | (even just filing a bug) | 23:18 |
| arcimboldo | well, I'm not sure it's an issue with the doc or the sample | 23:19 |
| *** jlhinson_ has quit IRC | 23:19 | |
| arcimboldo | for sure, the sample in /usr/share/doc/swift-proxy for the debian package is incorrect | 23:19 |
| arcimboldo | missing identity_uri, still using auth_host auth_port etc | 23:20 |
| arcimboldo | then, I was following the documentation in http://docs.openstack.org/developer/swift/admin_guide.html and then liberty, but I was using kilo :( | 23:20 |
| notmyname | arcimboldo: in which section? auth_token? | 23:20 |
| arcimboldo | [filter:authtoken] | 23:20 |
| arcimboldo | btw: it's never clear what you are supposed to do when using keystone v3 or v2 | 23:21 |
| notmyname | arcimboldo: how does https://github.com/openstack/swift/blob/master/etc/proxy-server.conf-sample#L284 look? | 23:21 |
| arcimboldo | looks like the one I'm using | 23:21 |
| notmyname | the good one or the bad one? | 23:21 |
| arcimboldo | but the debian package is not using that | 23:21 |
| arcimboldo | yes the good one | 23:21 |
| notmyname | oh | 23:21 |
| notmyname | zigo: :-( | 23:21 |
| notmyname | zigo: who'd you pass the debian swift packaging on to? | 23:22 |
| *** kei_yama has joined #openstack-swift | 23:25 | |
| arcimboldo | ok, now I can list, but can't create a container :( | 23:25 |
| arcimboldo | it gives me a 404 | 23:26 |
| notmyname | arcimboldo: what the list you did that worked? | 23:32 |
| arcimboldo | swift -V 3 list | 23:33 |
| arcimboldo | also openstack container list | 23:33 |
| notmyname | and you got an empty list back? | 23:33 |
| notmyname | but you can't create a container? | 23:33 |
| arcimboldo | yes | 23:33 |
| arcimboldo | yes | 23:33 |
| arcimboldo | Nov 18 23:34:18 swift1 proxy-server: 172.23.0.3 172.23.0.3 18/Nov/2015/23/34/18 POST /v1/AUTH_6a8c8c3ed987477b82f475742d695fef/test HTTP/1.0 400 - python-swiftclient-2.6.0 137d8ebd3d3c457c... - 65 - tx15e05f44603c4e6da4960-00564d0afa - 0.0264 - - 1447889658.110282898 1447889658.136686087 | 23:34 |
| notmyname | my first guess would be incorrect roles settings in keystone | 23:34 |
| *** chsc has quit IRC | 23:34 | |
| clayg | reading the logs I think the meeting went smoother without me today | 23:34 |
| arcimboldo | I am admin and _member_ in this tenant, and operator_roles = admin, Member, _member_, user in [filter:keystoneauth] | 23:35 |
| notmyname | arcimboldo: that sounds right | 23:37 |
| notmyname | arcimboldo: do you have account_autocreate set to true? https://github.com/openstack/swift/blob/master/etc/proxy-server.conf-sample#L131 | 23:37 |
| arcimboldo | uhm | 23:38 |
| arcimboldo | yes | 23:38 |
| arcimboldo | I wonder | 23:38 |
| arcimboldo | I might have created the rings wrong | 23:38 |
| arcimboldo | they *all* use port 6002 | 23:38 |
| notmyname | arcimboldo: ok. use `swift-recon --validate-servers` | 23:39 |
| notmyname | oh, yeah. that's a problem | 23:39 |
| arcimboldo | I've just deleted rings and data, will re-run my playbook | 23:40 |
| arcimboldo | ok, while it runs, I have a more theoretical question | 23:40 |
| notmyname | oh. not normally the first swift troubleshooting step I'd suggest ;-) | 23:40 |
| arcimboldo | it's just 2 vms | 23:41 |
| *** asettle is now known as asettle-bfast | 23:41 | |
| arcimboldo | so my question is: how swift distribute the data over the various servers? And what happens when a disk/server goes down? | 23:41 |
| arcimboldo | I have experience with ceph, which is, I think, pretty different | 23:41 |
| arcimboldo | (btw, now it created the container) | 23:42 |
| notmyname | arcimboldo: using a modified consistent hash ring. oh. similar to crush. and swift works around failures. a disk failure rebuilds the data somewhere else. a server failure doesn't | 23:42 |
| arcimboldo | so, when a disk fails, it automatically rebuilds the data somewhere else, as ceph does? | 23:42 |
| timburke | joeljwright: thanks for the review! don't forget to sleep, though :) | 23:43 |
| notmyname | arcimboldo: here's a starting point on the swift rings http://docs.openstack.org/developer/swift/overview_ring.html | 23:43 |
| arcimboldo | is it the *-sync daemons taht take care of syncing the data? | 23:43 |
| arcimboldo | and what happen if I have, let's say, only one replica and the server goes down? | 23:43 |
| arcimboldo | (let's say one replica for objects, but 3 replica for container and accounts) | 23:44 |
| notmyname | arcimboldo: yes. for an individual drive failure. for a server failure, it's treated as an availability issue, not durability. https://www.youtube.com/watch?v=_sUvfGKhaMo | 23:44 |
| *** joeljwright has quit IRC | 23:44 | |
| notmyname | if you have one replica and the server goes down, you won't be able to read it. but overwrites will work, and everything will be resolved (based on timestamps) when the server comes back up | 23:44 |
| arcimboldo | the idea I have in mind is to use swift for some sort of not-necessarly-always-online storage, for archive/cold storage | 23:45 |
| arcimboldo | and I would like to keep the cost down | 23:45 |
| arcimboldo | since we will need a lot of data | 23:45 |
| notmyname | where are your costs today? | 23:45 |
| arcimboldo | what do you mean? | 23:45 |
| notmyname | are your costs around keeping servers online? | 23:46 |
| notmyname | as opposed to drives or other hardware | 23:46 |
| arcimboldo | well, there is the energy question, of course, but this it's unavoidable unless we use tapes | 23:46 |
| arcimboldo | and there is the large amount of disks/servers you need to have a storage with 3 replica | 23:47 |
| arcimboldo | what I would like to have it's something that I guess it's not possible right now | 23:47 |
| notmyname | I know many people using swift for archival(-ish) storage. but swift is designed for high availability | 23:47 |
| *** mwheckmann has quit IRC | 23:47 | |
| notmyname | there's also erasure codes you could use to get better storage efficiency | 23:47 |
| arcimboldo | I would like to fill the storage pods and *power off* them | 23:47 |
| arcimboldo | when they are full | 23:47 |
| notmyname | no, swift isn't designed for turning off servers | 23:47 |
| *** diazjf has quit IRC | 23:47 | |
| arcimboldo | it would be enough to spin down the disks | 23:48 |
| notmyname | that won't work since the data distribution is all about spreading it across every server (drive) in the cluster | 23:48 |
| arcimboldo | but I can't find anything open source to do it | 23:48 |
| arcimboldo | I see | 23:48 |
| notmyname | that sounds similar to facebook's haystack. or maybe whatever backblaze uses | 23:48 |
| notmyname | (guessing on backblaze) | 23:48 |
| arcimboldo | that's right | 23:49 |
| arcimboldo | I think it makes sense | 23:49 |
| notmyname | what scale of storage are you talking about? | 23:49 |
| arcimboldo | I work for an university, and we have this use cases, where people produces huge amount of data that *might need in 2-3 years or maybe also next week* | 23:49 |
| notmyname | right :-) | 23:50 |
| arcimboldo | ~10PB | 23:50 |
| notmyname | yeah, swift is really really good at that (keeping stuff available for whenever you need it) | 23:50 |
| arcimboldo | but if it works, can very easily reach the 100PB over the years | 23:50 |
| notmyname | company I work for is called swiftstack and we have several university (and research/scientific) customers doing stuff very similar to what you describe | 23:50 |
| notmyname | arcimboldo: that's what everyone says ;-) | 23:51 |
| *** m_kazuhiro has joined #openstack-swift | 23:51 | |
| *** ho has joined #openstack-swift | 23:51 | |
| arcimboldo | notmyname, well, what can we do: they have microscopes and sequencing machines that produce 100x times the data they used to produce 10 years ago | 23:51 |
| notmyname | oh cool! | 23:52 |
| arcimboldo | and the researchers often do not have the experience to filter the raw data | 23:52 |
| arcimboldo | it's a brand new world for a lot of scientists | 23:52 |
| notmyname | any chance you're using illumina sequencers? I've seen those hacked to put the data directly in swift | 23:52 |
| notmyname | oh yeah | 23:52 |
| arcimboldo | yes we have them | 23:52 |
| arcimboldo | really? That's really interesting | 23:52 |
| arcimboldo | do you have any pointers? | 23:52 |
| notmyname | cool. we've been working with hudson alpha for putting stuff directly from the sequencers into swift | 23:53 |
| arcimboldo | (please do not quote xkcd: https://xkcd.com/138/) | 23:53 |
| notmyname | lol | 23:53 |
| notmyname | actually, get the book :-) https://swiftstack.com/books/ | 23:53 |
| notmyname | the genomics book talks all about that workflow | 23:53 |
| notmyname | fill out the marketing form (of course ;-) and we'll mail you one for free | 23:54 |
| notmyname | arcimboldo: where are you based? | 23:54 |
| arcimboldo | which form? | 23:55 |
| arcimboldo | zurich | 23:55 |
| notmyname | arcimboldo: sorry, this one http://info.swiftstack.com/object-storage-for-genomics-deploying-and-administering-openstack-swift-with-swiftstack-for-bioinformatics | 23:56 |
| arcimboldo | the "preorder"? | 23:56 |
| arcimboldo | ah ok | 23:56 |
| notmyname | yeah, I just asked our marketing people about that "preorder" word. doesn't make sense | 23:56 |
| *** eariasn_ has quit IRC | 23:56 | |
| arcimboldo | what would the marketing people say about the "added value" of swiftstack? :) | 23:57 |
| notmyname | "lots" | 23:57 |
| notmyname | ;-) | 23:57 |
| notmyname | (1) I try to keep marketing stuff out of this community channel (2) swiftstack's product is about making it easier for you to deploy, manage, and run swift | 23:58 |
| arcimboldo | I just wanted to know if it's "just" a deployment solution or you build something on top of swift | 23:59 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
