| *** diogogmt has joined #openstack-swift | 00:20 | |
| *** mathiasb has quit IRC | 00:24 | |
| *** zaitcev has joined #openstack-swift | 00:24 | |
| *** ChanServ sets mode: +v zaitcev | 00:24 | |
| kota_ | good morning | 00:28 |
|---|---|---|
| kota_ | notmyname: talking on liberasurecode warning, https://bugs.launchpad.net/swift/+bug/1628906/comments/10, you can ignore the warning | 00:30 |
| openstack | Launchpad bug 1628906 in OpenStack Object Storage (swift) "swift-object-reconstructor memory leak" [High,Confirmed] | 00:30 |
| kota_ | notmyname: it appears when backend ".so" library not found. And IIRC, it will run only once when initialize the pyeclib module to check the backends availability. | 00:32 |
| kota_ | notmyname: so that warning means, you can't use the backend anyway, however, if you don't want to use it, nothing you have to do on the message. | 00:33 |
| kota_ | notmyname: looking back at the bug comment, rledisez doesn't install isal (Intel ErasureCode engine, you know) and shss (NTT's backend) in his environment. I think you see a warning log just for shss, doesn't it? | 00:35 |
| *** mathiasb has joined #openstack-swift | 00:36 | |
| kota_ | notmyname: Speaking of EC work, please let me check the conflict status. | 00:38 |
| mattoliverau | kota_: o/ | 00:39 |
| *** oak has quit IRC | 00:39 | |
| kota_ | notmyname: in my feeling, i want to land mine first because it's already too old but in the community perspective, the acoles's one seems to have more benefit | 00:40 |
| kota_ | mattoliverau: o/ | 00:40 |
| *** adu has quit IRC | 00:41 | |
| kota_ | notmyname: on another perspective, iirc, mine vs acoles's doesn't seem to conflict a lot because his work should change on diskfile layer basically and my work is on proxy and index mapping so... | 00:43 |
| kota_ | notmyname: anyway, I'll take a look for his one and will update my opinion. | 00:44 |
| kota_ | mattoliverau: btw, does gerrit gate have been broken yet? | 00:44 |
| mattoliverau | kota_: well I've seen some stuff merge... so I hope that means it's working again | 00:45 |
| *** tqtran has quit IRC | 00:46 | |
| kota_ | mattolvierau: good to hear, thanks! | 00:46 |
| *** adu has joined #openstack-swift | 00:49 | |
| notmyname | kota_: thanks. yeah, I get that it's just a warning. but it gets printed on every call of `swift-recon` | 01:04 |
| kota_ | notmyname: hmm... yeah, I'm just writing the response meesage for the comment | 01:05 |
| kota_ | notmyname: could you tell me the *liberasurecode* version? | 01:05 |
| kota_ | i think timur was working on the suppressing that https://github.com/openstack/liberasurecode/commit/c7a94df0724af30b26e3856f9c14344fc9b73a09 | 01:06 |
| *** natarej_ has joined #openstack-swift | 01:07 | |
| kota_ | yeah, swift recon client import the POLICIES and storage_policy module import VALID_EC_TYPES which will trigger the log. | 01:08 |
| kota_ | if you're using older liberasurecode. | 01:08 |
| notmyname | kota_: https://gist.github.com/notmyname/8a75111177a66675d92085dbaa62614a | 01:08 |
| kota_ | notmyname: let me see | 01:08 |
| kota_ | notmyname: probably what you need to do is upgrading liberasurecode :0) | 01:09 |
| kota_ | notmyname: https://github.com/openstack/pyeclib/blob/master/pyeclib/ec_iface.py#L42-L46 | 01:09 |
| notmyname | ok, maybe because I grabbed the distro version (xenial) | 01:09 |
| notmyname | kk | 01:09 |
| notmyname | ok, I need to go eat dinner | 01:10 |
| notmyname | but let me know on the patches | 01:10 |
| kota_ | notmyname: enjoy | 01:10 |
| *** natarej has quit IRC | 01:10 | |
| *** klrmn has quit IRC | 01:16 | |
| *** cdelatte has quit IRC | 01:18 | |
| kota_ | notmyname: comment added at your gist. if you need more on something, please ping me. | 01:24 |
| *** dmorita has quit IRC | 01:27 | |
| *** dmorita has joined #openstack-swift | 01:28 | |
| *** dmorita has quit IRC | 01:32 | |
| *** adu has quit IRC | 01:33 | |
| *** adu has joined #openstack-swift | 01:37 | |
| openstackgerrit | Kazuhiro MIYAHARA proposed openstack/swift: WIP: Conditional Fast-POST https://review.openstack.org/381363 | 01:57 |
| *** chsc has quit IRC | 02:11 | |
| jrichli | I haven't gotten any email that allows me to cast my vote for TC. That is how this works, right? you get an email? | 02:11 |
| *** dmorita has joined #openstack-swift | 02:13 | |
| timburke | jrichli: i hadn't either. contact the election folks with one of your commit SHAs per http://lists.openstack.org/pipermail/openstack-dev/2016-October/105002.html -- tonyb got me sorted out inside of 10 mins (thanks again tonyb!) | 02:18 |
| tonyb | timburke: np | 02:19 |
| jrichli | thanks, will do. Hi tonyb :-) | 02:19 |
| tonyb | jrichli: o/ | 02:21 |
| jrichli | tonyb: do you need the official email, or can I send you the sha here? | 02:33 |
| tonyb | jrichli: Please send it to all the officals addresses are here: http://governance.openstack.org/election/#election-officials | 02:34 |
| jrichli | ok, thx | 02:34 |
| tonyb | jrichli: that way all the officals are on the same page | 02:34 |
| jrichli | gotcha | 02:34 |
| *** klrmn has joined #openstack-swift | 02:39 | |
| *** tqtran has joined #openstack-swift | 02:43 | |
| *** _JZ_ has joined #openstack-swift | 02:44 | |
| *** adu has quit IRC | 02:45 | |
| *** tqtran has quit IRC | 02:48 | |
| *** trananhkma has joined #openstack-swift | 02:55 | |
| *** dmorita has quit IRC | 02:58 | |
| *** dmorita has joined #openstack-swift | 02:59 | |
| notmyname | good evening | 03:01 |
| *** tqtran has joined #openstack-swift | 03:02 | |
| *** dmorita has quit IRC | 03:03 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/python-swiftclient: Updated from global requirements https://review.openstack.org/89250 | 03:04 |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 03:04 |
| *** dmorita has joined #openstack-swift | 03:05 | |
| notmyname | oh, yikes. is there a bug for https://review.openstack.org/#/c/381363/ | 03:05 |
| patchbot | patch 381363 - swift - WIP: Conditional Fast-POST | 03:05 |
| zaitcev | oh noes | 03:05 |
| zaitcev | OTL orz ... | 03:06 |
| *** tqtran has quit IRC | 03:06 | |
| openstackgerrit | Tuan Luong-Anh proposed openstack/swift: Fix a typo in documentation https://review.openstack.org/381377 | 03:07 |
| openstackgerrit | Pete Zaitcev proposed openstack/swift: Add a configurable URL base to staticweb https://review.openstack.org/372809 | 03:09 |
| *** links has joined #openstack-swift | 03:10 | |
| *** _JZ_ has quit IRC | 03:13 | |
| *** chsc has joined #openstack-swift | 03:25 | |
| *** dmorita has quit IRC | 03:27 | |
| openstackgerrit | Merged openstack/swift: Fix typo: remove redundant 'that' https://review.openstack.org/380938 | 03:27 |
| *** chsc has quit IRC | 03:30 | |
| *** dmorita has joined #openstack-swift | 03:31 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift: Updated from global requirements https://review.openstack.org/88736 | 03:33 |
| mattoliverau | notmyname: o/ | 03:38 |
| notmyname | yo | 03:38 |
| openstackgerrit | John Dickinson proposed openstack/swift: adding reno sphinx tree https://review.openstack.org/381382 | 03:49 |
| *** natarej__ has joined #openstack-swift | 04:06 | |
| *** mingyu has joined #openstack-swift | 04:08 | |
| *** natarej_ has quit IRC | 04:09 | |
| *** mingyu has quit IRC | 04:14 | |
| openstackgerrit | Kazuhiro MIYAHARA proposed openstack/swift: [Don't merge] Use __init__() of Daemon class in sub classes https://review.openstack.org/365614 | 04:38 |
| mahatic_ | jrichli: timburke :thanks for the info. I was waiting for TC email too | 04:43 |
| *** klrmn has quit IRC | 04:44 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: EC Fragment Duplication - Foundational Global EC Cluster Support https://review.openstack.org/219165 | 04:51 |
| *** zaitcev has quit IRC | 05:03 | |
| *** dmorita has quit IRC | 05:07 | |
| *** nadeem has joined #openstack-swift | 05:09 | |
| *** dmorita has joined #openstack-swift | 05:23 | |
| *** dmorita has quit IRC | 05:24 | |
| *** dmorita has joined #openstack-swift | 05:25 | |
| *** dmorita has quit IRC | 05:30 | |
| *** takashi has joined #openstack-swift | 05:47 | |
| *** mingyu has joined #openstack-swift | 06:02 | |
| *** tqtran has joined #openstack-swift | 06:05 | |
| *** tqtran has quit IRC | 06:10 | |
| *** Jeffrey4l__ has quit IRC | 06:14 | |
| *** Jeffrey4l__ has joined #openstack-swift | 06:14 | |
| *** Jeffrey4l__ has quit IRC | 06:14 | |
| *** Jeffrey4l_ has joined #openstack-swift | 06:15 | |
| *** Jeffrey4l_ is now known as Jeffrey4l | 06:15 | |
| *** winggundamth has quit IRC | 06:17 | |
| *** nadeem has quit IRC | 06:28 | |
| *** nadeem has joined #openstack-swift | 06:29 | |
| *** tqtran has joined #openstack-swift | 06:31 | |
| *** tqtran has quit IRC | 06:35 | |
| *** tqtran has joined #openstack-swift | 06:38 | |
| *** rcernin has joined #openstack-swift | 06:43 | |
| *** takashi has quit IRC | 06:44 | |
| *** ppai has joined #openstack-swift | 06:46 | |
| *** admin6_ has joined #openstack-swift | 06:50 | |
| *** tesseract- has joined #openstack-swift | 06:52 | |
| *** hseipp has joined #openstack-swift | 06:54 | |
| openstackgerrit | Hanxi Liu proposed openstack/swift: Add links to more detailed overview in overview_architecture https://review.openstack.org/381446 | 06:56 |
| openstackgerrit | Hanxi Liu proposed openstack/swift: Add links to more detailed overview in overview_architecture https://review.openstack.org/381446 | 06:58 |
| openstackgerrit | Hanxi Liu proposed openstack/swift: Add links to more detailed overview in overview_architecture https://review.openstack.org/381446 | 07:01 |
| *** rledisez has joined #openstack-swift | 07:19 | |
| openstackgerrit | Kota Tsuyuzaki proposed openstack/swift: EC Fragment Duplication - Foundational Global EC Cluster Support https://review.openstack.org/219165 | 07:24 |
| *** geaaru has joined #openstack-swift | 07:31 | |
| *** takashi has joined #openstack-swift | 07:55 | |
| *** jistr has joined #openstack-swift | 07:57 | |
| *** admin6_ has quit IRC | 07:59 | |
| *** amoralej|off is now known as amoralej | 08:03 | |
| openstackgerrit | Hanxi Liu proposed openstack/swift: Add links to more detailed overview in overview_architecture https://review.openstack.org/381446 | 08:03 |
| *** lhx_ has joined #openstack-swift | 08:03 | |
| *** takashi has quit IRC | 08:05 | |
| *** jordanP has joined #openstack-swift | 08:06 | |
| *** tqtran has quit IRC | 08:09 | |
| *** takashi has joined #openstack-swift | 08:11 | |
| *** takashi_ has joined #openstack-swift | 08:31 | |
| *** takashi has quit IRC | 08:32 | |
| *** takashi_ is now known as takashi | 08:37 | |
| *** lhx_ has quit IRC | 08:44 | |
| *** acoles_ is now known as acoles | 08:45 | |
| acoles | good morning | 09:01 |
| kota_ | acoles: morning :-) | 09:02 |
| acoles | so should we have received TC voting mail? anyone know the email subject so i can search for it | 09:04 |
| acoles | hi kota_ | 09:04 |
| kota_ | acoles: Poll: OpenStack Technical Committe (TC) Election - October 2016 | 09:06 |
| kota_ | acoles: the subject which you are looking for. | 09:06 |
| acoles | kota_: thanks | 09:08 |
| acoles | looks like I didn't receive that | 09:08 |
| kota_ | acoles: that's too bad :/ | 09:09 |
| *** nadeem has quit IRC | 09:11 | |
| kota_ | notmyname: since the result of quick look at the durable eliminate patch and my global ec work, the order is not significant for the technical reason because the conflict area is just test/unit/proxy/test_server.py | 09:14 |
| kota_ | notmyname: the order for landing. | 09:15 |
| kota_ | notmyname: it seems to still conflict with mine but it looks easy to rebase rather than ever. | 09:15 |
| * kota_ has working on rebasing, rebasing and rebasing anytime on the change at test/unit/porxy/test_server.py | 09:16 | |
| kota_ | since the experience, I'm realizing, i want to cleanup the proxy/test_server.py code because it's so long lines in *a* file. | 09:17 |
| kota_ | but it's also not ugent, anyway, though. | 09:17 |
| kota_ | notmyname: since i can rebase easily, i think, it's ok for vice-versa | 09:19 |
| * kota_ is leaving his office | 09:22 | |
| kota_ | see you swifters. | 09:22 |
| mahatic_ | acoles: good morning | 09:25 |
| mahatic_ | acoles: If you're expecting the poll email it to your gmail, you should ask them to resend it. They suspect there's a problem between CIVS and gmail | 09:26 |
| acoles | mahatic_: thanks, it should be coming to me at hpe though. I requested another. | 09:48 |
| mahatic_ | acoles: ah okay | 09:49 |
| *** jistr is now known as jistr|mtg | 09:59 | |
| *** zigo has quit IRC | 10:01 | |
| *** zigo has joined #openstack-swift | 10:04 | |
| *** zigo is now known as Guest84780 | 10:05 | |
| cnf | why would swift be constantly using about 800 IOPS while havig no traffic? | 10:07 |
| cnf | and every 7 minutes it drops to 0 for a little while | 10:08 |
| ahale | maybe auditor or replicator processes ? | 10:14 |
| *** Guest84780 has quit IRC | 10:14 | |
| *** zigo_ has joined #openstack-swift | 10:16 | |
| cnf | it's a single node wih replication set to 1 | 10:16 |
| *** SkyRocknRoll has joined #openstack-swift | 10:20 | |
| cnf | notmyname any idea? | 10:25 |
| *** abhitechie has joined #openstack-swift | 10:29 | |
| *** SkyRocknRoll has quit IRC | 10:54 | |
| *** rcernin has quit IRC | 10:57 | |
| *** rcernin has joined #openstack-swift | 10:58 | |
| *** jistr|mtg is now known as jistr | 11:00 | |
| *** amoralej is now known as amoralej|lunch | 11:13 | |
| *** silor has joined #openstack-swift | 11:21 | |
| *** zigo_ is now known as zigo | 11:23 | |
| *** dmorita has joined #openstack-swift | 11:25 | |
| *** silor has quit IRC | 11:28 | |
| *** silor has joined #openstack-swift | 11:28 | |
| *** dmorita has quit IRC | 11:29 | |
| *** admin6 has quit IRC | 11:30 | |
| *** admin6 has joined #openstack-swift | 11:31 | |
| *** silor1 has joined #openstack-swift | 11:31 | |
| *** silor has quit IRC | 11:32 | |
| *** silor1 is now known as silor | 11:32 | |
| *** ppai has quit IRC | 11:35 | |
| *** mingyu has quit IRC | 11:40 | |
| *** lhx_ has joined #openstack-swift | 11:49 | |
| *** ppai has joined #openstack-swift | 11:51 | |
| *** jlwhite has quit IRC | 12:14 | |
| *** jlwhite has joined #openstack-swift | 12:15 | |
| *** furlongm has quit IRC | 12:15 | |
| *** furlongm has joined #openstack-swift | 12:16 | |
| *** mingyu has joined #openstack-swift | 12:18 | |
| cnf | hmm, this is so strange | 12:20 |
| cnf | constant 800IOPS or about 100MB/s reads | 12:21 |
| *** takashi has quit IRC | 12:26 | |
| cnf | it seems to be /usr/bin/swift-object-auditor /etc/swift/object-server.conf | 12:30 |
| *** ppai has quit IRC | 12:32 | |
| *** cdelatte has joined #openstack-swift | 12:33 | |
| *** _JZ_ has joined #openstack-swift | 12:35 | |
| *** amoralej|lunch is now known as amoralej | 12:37 | |
| *** cbartz has joined #openstack-swift | 12:41 | |
| *** klamath has joined #openstack-swift | 13:03 | |
| *** klamath has quit IRC | 13:05 | |
| *** klamath has joined #openstack-swift | 13:05 | |
| *** klamath has quit IRC | 13:07 | |
| *** klamath has joined #openstack-swift | 13:10 | |
| *** kei_yama has quit IRC | 13:16 | |
| *** mingyu has quit IRC | 13:20 | |
| *** oak has joined #openstack-swift | 13:21 | |
| jrichli | acoles: I didn't get the mail either. I followed the instructions in http://lists.openstack.org/pipermail/openstack-dev/2016-October/105002.html | 13:31 |
| jrichli | they sent a mail to the same address that worked this time | 13:32 |
| *** daemontool has joined #openstack-swift | 13:33 | |
| acoles | jrichli: yep I did same | 13:33 |
| acoles | mathiasb: hi, I'm looking at your patch 364878 and trying to understand how to use openstackclient to interact with barbican, as shown in the commit message, but I can't see support for a 'secret' subcommand in the CLI. How does that work? | 13:33 |
| patchbot | https://review.openstack.org/#/c/364878/ - swift - WIP: Storing encryption root secret in Barbican | 13:33 |
| acoles | jrichli: have you any success with this ^^ ? | 13:34 |
| jrichli | acoles: sadly, I haven't attempted the barbican install myself yet. | 13:34 |
| acoles | jrichli: ok, nm. I have barbican running, just need to get my secret setup. resorting to curl now. | 13:36 |
| acoles | Anyone else still missing the 'f' shortcut in gerrit? It's been a while now but I'm still longing for it to come back :'( | 13:42 |
| tdasilva | yes! | 13:42 |
| tdasilva | acoles: if you run for the TC promising to bring back the 'f' shortcut you will get my vote ;) | 13:44 |
| acoles | tdasilva: LOL. The one sure ticket for victory... "Make the f key great again" | 13:45 |
| tdasilva | lol | 13:45 |
| jrichli | :-) | 13:46 |
| *** links has quit IRC | 13:49 | |
| *** diogogmt has quit IRC | 13:59 | |
| *** tmoreira has quit IRC | 14:01 | |
| -openstackstatus- NOTICE: The Gerrit service on review.openstack.org is being restarted to address performance degradation and should return momentarily | 14:01 | |
| *** abhitechie has quit IRC | 14:03 | |
| *** daemontool_ has joined #openstack-swift | 14:06 | |
| *** daemontool has quit IRC | 14:10 | |
| *** natarej__ has quit IRC | 14:21 | |
| *** natarej__ has joined #openstack-swift | 14:22 | |
| *** vint_bra has joined #openstack-swift | 14:24 | |
| *** zaitcev has joined #openstack-swift | 14:40 | |
| *** ChanServ sets mode: +v zaitcev | 14:40 | |
| *** mingyu has joined #openstack-swift | 14:50 | |
| *** admin6_ has joined #openstack-swift | 14:50 | |
| *** sgundur has joined #openstack-swift | 14:53 | |
| stevemar | acoles: key management support via barbican is available as a plug-in to openstackclient. you need to install openstackclient, then install barbicanclient and the 'secret' command will show up. cc jrichli | 14:55 |
| *** tmoreira has joined #openstack-swift | 14:56 | |
| acoles | stevemar: OIC. thanks! | 14:56 |
| jrichli | stevemar: thanks | 14:57 |
| mathiasb | acoles: maybe it's a version issue? i am running openstack client 2.3.0 | 14:58 |
| mathiasb | if you have the barbican client you can use that as well, "barbican secret store" | 14:58 |
| *** sgundur has quit IRC | 15:01 | |
| acoles | mathiasb: what stevemar said seems to work, I just didn't have barbicanclient installed, now I do and I get the secret subcommands | 15:02 |
| mathiasb | acoles: great, somehow I missed that comment, thanks stevemar! | 15:03 |
| acoles | mathiasb: btw i have openstack 3.2.0 | 15:04 |
| *** oak has quit IRC | 15:04 | |
| acoles | so now I need to set up barbican endpoints in keystone | 15:04 |
| *** klrmn has joined #openstack-swift | 15:06 | |
| stevemar | np swifterinos :) | 15:08 |
| *** rcernin has quit IRC | 15:10 | |
| *** daemontool_ has quit IRC | 15:19 | |
| *** diogogmt has joined #openstack-swift | 15:21 | |
| notmyname | good morning | 15:23 |
| *** dmorita has joined #openstack-swift | 15:26 | |
| *** daemontool has joined #openstack-swift | 15:27 | |
| *** dmorita has quit IRC | 15:31 | |
| *** chsc has joined #openstack-swift | 15:41 | |
| cnf | ohai notmyname | 15:43 |
| cnf | my workday is pretty much done | 15:43 |
| cnf | well, the office one, anyway | 15:43 |
| *** chsc has quit IRC | 15:51 | |
| openstackgerrit | Hanxi Liu proposed openstack/swift: Remove the file named MANIFEST.in https://review.openstack.org/381947 | 15:56 |
| notmyname | reminder that the cross-project sessions etherpad is at https://etherpad.openstack.org/p/ocata-cross-project-sessions. take a look now that there's a tentative schedule proposed | 15:57 |
| notmyname | also, please continue to fill out https://etherpad.openstack.org/p/ocata_swift_summit_topics for the swift sessions | 15:57 |
| *** arch-nemesis has joined #openstack-swift | 16:00 | |
| *** diogogmt_ has joined #openstack-swift | 16:07 | |
| *** diogogmt has quit IRC | 16:08 | |
| *** diogogmt_ is now known as diogogmt | 16:08 | |
| *** silor has quit IRC | 16:09 | |
| *** tongli has joined #openstack-swift | 16:11 | |
| *** rledisez has quit IRC | 16:17 | |
| *** links has joined #openstack-swift | 16:17 | |
| *** links has quit IRC | 16:18 | |
| *** sgundur has joined #openstack-swift | 16:21 | |
| *** cdelatte has quit IRC | 16:23 | |
| *** admin6_ has quit IRC | 16:26 | |
| *** openstackgerrit has quit IRC | 16:26 | |
| *** openstackgerrit has joined #openstack-swift | 16:27 | |
| *** openstackgerrit has quit IRC | 16:28 | |
| *** tesseract- has quit IRC | 16:29 | |
| *** openstackgerrit has joined #openstack-swift | 16:29 | |
| *** openstackgerrit has quit IRC | 16:30 | |
| *** chsc has joined #openstack-swift | 16:30 | |
| *** dmorita has joined #openstack-swift | 16:30 | |
| *** openstackgerrit has joined #openstack-swift | 16:30 | |
| *** daemontool has quit IRC | 16:35 | |
| *** cbartz has left #openstack-swift | 16:39 | |
| *** hseipp has quit IRC | 16:41 | |
| *** mingyu has quit IRC | 16:41 | |
| *** nikivi has joined #openstack-swift | 16:46 | |
| openstackgerrit | John Dickinson proposed openstack/swift: adding reno sphinx tree https://review.openstack.org/381382 | 16:51 |
| acoles | notmyname: once we move to reno will we cease to have CHANGELOG file? | 17:00 |
| *** chsc has quit IRC | 17:00 | |
| notmyname | acoles: I don't want to | 17:02 |
| * timburke wonders if it'd be worth adding an [extras] section to osc's setup.cfg... | 17:03 | |
| timburke | good morning | 17:03 |
| notmyname | acoles: with 2.10.0, I wrote the CHANGELOG entry and then translated it to yaml to put in the reno file | 17:03 |
| *** jordanP has quit IRC | 17:04 | |
| *** nadeem has joined #openstack-swift | 17:04 | |
| *** nikivi has quit IRC | 17:05 | |
| acoles | notmyname: and do you intend to continue that way? I ask because today I was trying to read the release notes for another project and ended up having to google my way to where they are published by openstack i.e. they weren't in the repo in a form I could read :/ | 17:05 |
| *** mingyu has joined #openstack-swift | 17:05 | |
| notmyname | I really don't like the idea of either (1) to read release notes, go download a dependency and run a script to get something printed to a screen or (2) go look at a webpage someplace else that may or may not be currently available | 17:06 |
| notmyname | so yeah, I really like the idea of human-written notes in plain text that live with the code in the repo | 17:06 |
| notmyname | for now, I expect to continue to maintain the CHANGELOG file, but I'll admit I'm fighting against gravity here, so it partly depends on how tired I get | 17:07 |
| *** mingyu has quit IRC | 17:07 | |
| *** mingyu has joined #openstack-swift | 17:08 | |
| timburke | notmyname: look at it this way: you fight against gravity and *win* every morning! | 17:08 |
| cnf | hey notmyname, is it normal for a mostly-kinda default install of swift to generate 800 iops constantly (about 100MB/s) | 17:09 |
| cnf | it was the auditor, but it was just doing it for hours on end, with 2 files in the storage ^^; | 17:10 |
| notmyname | that does seem high | 17:10 |
| cnf | yah | 17:10 |
| *** mingyu has quit IRC | 17:10 | |
| notmyname | cnf: swift will continually look at the drives in the cluster to check that the data is good and in the right place. this is done by the auditor and the replicator | 17:10 |
| cnf | right | 17:11 |
| notmyname | (there are several of each) | 17:11 |
| cnf | I figured that bit out | 17:11 |
| notmyname | and they can be tuned to not be as active | 17:11 |
| cnf | i also disabled them on my single node install :P | 17:11 |
| cnf | but 800iops seems incredibly expensive, even on enterprise HW | 17:11 |
| cnf | notmyname: disabeling them on a single node instance should not matter much, right? | 17:12 |
| notmyname | yeah, it matters, if you've got multiple drives and are using more than one replica | 17:13 |
| cnf | no, 1 replica | 17:13 |
| cnf | notmyname: i know it's not a typical setup :P | 17:15 |
| cnf | anyway, if you keep it on, 800iops on a default install does seem mighty high | 17:15 |
| cnf | is it supposed to just keep using that, constantly? | 17:15 |
| notmyname | with one replica, you can't recover from any failures, so there's no reason to run auditors. replication would only need to be run if you change the cluster capacity (to move a bit of data around) | 17:15 |
| cnf | right | 17:16 |
| cnf | i'm counting on the ZFS it is stored on to keep the data sane | 17:16 |
| cnf | I didn't find how to not start the auditor though | 17:16 |
| notmyname | remove/rename the auditor sections in {account,container,object}-server.conf | 17:17 |
| cnf | ah, ok | 17:17 |
| clayg | notmyname: you could contribute a plugin to make reno support plaintext output so you could generate them from the yaml | 17:17 |
| notmyname | and/or `swift-init account-auditor stop` | 17:17 |
| *** sgundur has quit IRC | 17:18 | |
| cnf | notmyname: so if I set up swift properly, 800iops constantly on every swift node is just part of normal upkeep? | 17:18 |
| notmyname | clayg: it currently takes the yaml and outputs plain test. or you mean a plugin to take plain text as input? that might be better | 17:18 |
| cnf | also, when read/writes happen, does that get throttled back at all? | 17:18 |
| notmyname | cnf: well, no single spinning drive can even do half of 800iops, so your number is dependent on how many drives you have in the server. it's normal to have some background activity as a baseline. the amount depends on several things, including your ring part power, how much data is in the cluster, and any failures currently in the cluster | 17:20 |
| *** lhx_ has quit IRC | 17:20 | |
| cnf | well, I did a create 8 1 1 | 17:21 |
| *** sgundur has joined #openstack-swift | 17:21 | |
| *** mingyu has joined #openstack-swift | 17:21 | |
| cnf | and yeah, "some" I was expecting, but I am trying to figure out why my system was getting hammered with 800iops | 17:22 |
| notmyname | part power of 8 is nice and small. glad you didn't go for something like 18 ;-) | 17:22 |
| clayg | i thought it generated rst or something? if it already spits out plaintext change logs why can you just check them in when you run it? i'm confused. | 17:23 |
| clayg | See - you have solved nothing - we're all going to reno experts in like 2 hours | 17:23 |
| cnf | notmyname: i was reading about it, and some page i read said smaller than 8 had an averse effect, and larger made things heaver on small installs | 17:24 |
| cnf | so i just went wirh 8 :P | 17:24 |
| cnf | hmm, I realize swift on a single node is kinda overkill | 17:25 |
| cnf | but I can't find any other halfdecent object store to run | 17:25 |
| notmyname | clayg: `reno report` generates plain text. but the sphinx doc tree is necessary for the CI job to generate the URL structure at http://docs.openstack.org/releasenotes/swift/* so that it can be linked from https://releases.openstack.org/newton/index.html | 17:27 |
| *** lhx_ has joined #openstack-swift | 17:28 | |
| *** sgundur has quit IRC | 17:36 | |
| *** sgundur has joined #openstack-swift | 17:36 | |
| *** geaaru has quit IRC | 17:42 | |
| *** nadeem has quit IRC | 17:42 | |
| *** nadeem has joined #openstack-swift | 17:42 | |
| *** lhx_ has quit IRC | 17:48 | |
| *** bjolo_ has joined #openstack-swift | 17:53 | |
| *** hseipp has joined #openstack-swift | 17:54 | |
| *** tqtran has joined #openstack-swift | 17:55 | |
| *** chsc has joined #openstack-swift | 17:58 | |
| *** chsc has joined #openstack-swift | 17:58 | |
| *** rcernin has joined #openstack-swift | 17:58 | |
| *** mingyu has quit IRC | 18:08 | |
| *** hseipp has quit IRC | 18:13 | |
| *** acoles is now known as acoles_ | 18:20 | |
| *** amoralej is now known as amoralej|off | 18:29 | |
| *** Suyi has joined #openstack-swift | 18:35 | |
| ahale | ahh, thats why i didnt get a TC vote email, I don't qualify to vote :( | 18:39 |
| notmyname | what are the criteria? I thought it was "be a foundation member" | 18:42 |
| ahale | " Foundation individual members that are also committers for one of the official project teams repositories over the Liberty-Mitaka timeframe" | 18:42 |
| notmyname | ah | 18:42 |
| *** silor has joined #openstack-swift | 18:43 | |
| *** nikivi has joined #openstack-swift | 18:45 | |
| nadeem | I have voted....lets hope more people vote for the change :) | 18:46 |
| ahale | make openstack great again ? | 18:46 |
| nadeem | I would rather destroy the wall rather than creating new ones (as far as new ideas are concerned) | 18:47 |
| ahale | a vote against golang is a vote against walls | 18:48 |
| *** silor has quit IRC | 18:48 | |
| *** sgundur has quit IRC | 19:05 | |
| *** sgundur has joined #openstack-swift | 19:08 | |
| openstackgerrit | Shashirekha Gundur proposed openstack/swift: Invalidate cached tokens api https://review.openstack.org/370319 | 19:15 |
| *** cdelatte has joined #openstack-swift | 19:34 | |
| openstackgerrit | Thiago da Silva proposed openstack/swift: Fix swiftdir option and usage of storage policy aliases https://review.openstack.org/344693 | 19:34 |
| *** nikivi has quit IRC | 19:36 | |
| cnf | notmyname: do you know if I can run swift and keystone behind the same reverse proxy on the same port? | 19:45 |
| notmyname | not on the same port | 19:46 |
| cnf | hmm, damn | 19:46 |
| cnf | they use the same paths? | 19:46 |
| notmyname | they're likely using unique paths, but they don't share a common URL router, so only one process gets the request to read off the socket. if you send a keystone request to the swift proecess, swift will return an error. and vice versa. | 19:48 |
| notmyname | you could write your own, but they'd still need to be listening on different sockets on the backend | 19:48 |
| cnf | well, yeah, but they are behind a reverse proxy for ssl termination anyway | 19:48 |
| *** bjolo_ has quit IRC | 19:49 | |
| *** bjolo_ has joined #openstack-swift | 19:49 | |
| *** sgundur has quit IRC | 19:50 | |
| *** sgundur has joined #openstack-swift | 19:51 | |
| cnf | notmyname: sorry for trying to do things with swift it's not meant for :P | 19:51 |
| notmyname | nicest user ever. "oops I'm actually trying to use your software. is that ok?" YES! totally ok :-) | 19:51 |
| cnf | ^^; | 19:52 |
| cnf | I need to read up more on how auth / keystone works, and how it interacts with swift | 19:53 |
| clarkb | I know horizon and keystone run on the same port with apache using different url prefixes for them | 19:54 |
| clarkb | so at least for keystone it is happy to do that | 19:54 |
| *** sgundur has quit IRC | 19:55 | |
| cnf | hmm | 19:55 |
| clarkb | er with devstack specifically | 19:55 |
| cnf | clarkb: do you know any docs for that? | 19:55 |
| *** sgundur has joined #openstack-swift | 19:55 | |
| clarkb | I am not aware of any specific docs no, but you could see how devstack sets that up. (both are running as wsgi too which may affect stuff) | 19:56 |
| cnf | ok | 19:56 |
| cnf | hmm, they all seem to use /v1 and /v2 etc though | 19:58 |
| notmyname | yeah, the "write your own" could also be spelled "use apache,etc and configure the paths yourself" | 19:58 |
| cnf | http://developer.openstack.org/api-ref/identity/v3/ that is keystone, right? | 19:59 |
| *** sgundur has quit IRC | 20:00 | |
| *** _JZ_ has quit IRC | 20:00 | |
| clarkb | yes | 20:00 |
| cnf | hmm | 20:01 |
| cnf | for now, sending /v3 to keystone would work, I guess | 20:01 |
| *** bjolo_ has quit IRC | 20:01 | |
| cnf | but if swift upgrades api version to 3, i'd be boned :P | 20:01 |
| clarkb | I think devstack uses eg /identity/v3 and /dashboard | 20:01 |
| cnf | hmm | 20:02 |
| cnf | interesting | 20:02 |
| cnf | so you can configure the base path somewhere? | 20:02 |
| *** sanchitmalhotra has quit IRC | 20:02 | |
| clarkb | ya I think there are different config options to set the root url | 20:02 |
| clarkb | and if not you can probably hack it with mod rewrite | 20:03 |
| cnf | hmm, admin_endpoint and public_endpoint, it seems | 20:03 |
| cnf | nice | 20:03 |
| *** mingyu has joined #openstack-swift | 20:04 | |
| cnf | ok, i'll play with it some, tomorrow | 20:07 |
| cnf | and then try figure out how the user id / token thing works | 20:07 |
| *** sgundur has joined #openstack-swift | 20:07 | |
| pdardeau | cnf: you might find swift-on-file interesting https://github.com/openstack/swiftonfile | 20:10 |
| cnf | pdardeau: hmm, thanks | 20:12 |
| cnf | pdardeau: i'm assuming you lose the accounting / etc functionality if you use it? | 20:12 |
| cnf | (which I don't use anyway) | 20:12 |
| pdardeau | cnf: i have no idea. i haven't used it myself. | 20:13 |
| cnf | right | 20:13 |
| cnf | i will have a look at this, once I figure out how to run swift and swift3 properly | 20:13 |
| cnf | thanks a lot | 20:14 |
| cnf | "It is also recommended that these Filesystems provide data durability as Swift-On-File should not use Swift's replication mechanisms." | 20:14 |
| cnf | makes sense | 20:14 |
| pdardeau | cnf: tdasilva can probably answer questions about it | 20:15 |
| cnf | ok | 20:16 |
| tdasilva | hello | 20:16 |
| cnf | ohai :P | 20:16 |
| cnf | do you use swiftonfile tdasilva ? | 20:19 |
| tdasilva | cnf: I don't have any deployment clusters, just a dev. | 20:20 |
| cnf | right | 20:20 |
| tdasilva | s/deployment/deployed | 20:20 |
| tdasilva | cnf: what's your question? | 20:20 |
| cnf | nothing atm, pdardeau just linked it 10 minutes ago | 20:21 |
| cnf | i'm doing some unconventional things with swift, atm | 20:21 |
| notmyname | tdasilva: cnf is running swift on zfs in a small cluster | 20:21 |
| pdardeau | tdasilva: cnf expressed interest in running swift on top of zfs with replica=1 | 20:21 |
| pdardeau | tdasilva: reminded me of swift-on-file | 20:21 |
| cnf | yeah | 20:22 |
| cnf | i mostly care about the swift (and S3) api, and not so much about the multi node features | 20:22 |
| cnf | well, for this current project at home, at $dayjob is a different thing | 20:22 |
| tdasilva | cnf: cool! are you trying to do something like a home nas support the swift api? | 20:23 |
| cnf | tdasilva: yep | 20:23 |
| cnf | and swift is the most workable answer I found to getting an object store API going | 20:24 |
| tdasilva | cnf: neat...the cool play with swiftonfile for that use case, is that a lot of 'home' use cases might still use like a file access | 20:25 |
| tdasilva | so you could write the data over file, and then with swift access be able to acess that data over web | 20:25 |
| cnf | indeed! | 20:25 |
| cnf | my main goal is to share storage to family etc for backups and the like | 20:25 |
| tdasilva | cnf: nice, i've thought about doing something like that, a small swift cluster at home, and then having something like rackspace/ovh's public cloud for remote backup | 20:27 |
| cnf | right, really looks interesting | 20:28 |
| cnf | but first on my list is figuring out the auth | 20:28 |
| cnf | tdasilva: backblaze b2 is really cheap as wel | 20:28 |
| tdasilva | cnf: for home, why don't you start with something simple like swauth? | 20:28 |
| cnf | yeah, been looking at swauth | 20:28 |
| cnf | but "home" is also a test env for me | 20:28 |
| cnf | if I do get keystone working, I can point openstack test installs to it | 20:28 |
| cnf | it's a "nice to have" and not a must work, though | 20:29 |
| tqtran | acoles_: ping, still around? | 20:29 |
| cnf | the "NAS" is an 8 core intel avaton with 32G ECC and 16T of usable storage btw | 20:30 |
| cnf | and I have an intel NUC as a play ESXi node as well | 20:30 |
| tdasilva | heh | 20:30 |
| cnf | so I have some room to play :P | 20:30 |
| tdasilva | cnf: if you want to give SoF a quick spin, https://github.com/thiagodasilva/ansible-saio sets everything up for you | 20:31 |
| tdasilva | cnf: i'd like to get keystone setup on that project one day too... | 20:31 |
| cnf | :P | 20:32 |
| cnf | I got keystone working in docker atm | 20:32 |
| cnf | and swift mostly | 20:32 |
| *** dfflanders has joined #openstack-swift | 20:32 | |
| tdasilva | cnf: cool, have you pushed any images to docker hub? | 20:33 |
| cnf | not yet, atm they are very very highly customised to my situation | 20:33 |
| cnf | so they need quite some cleanup | 20:34 |
| cnf | I can always send the files I have though | 20:34 |
| *** Jeffrey4l_ has joined #openstack-swift | 20:34 | |
| cnf | it's a multi domain setup with ldap | 20:34 |
| cnf | still not entirely certain how to integrate ldap cleanly into keystone | 20:35 |
| *** Jeffrey4l has quit IRC | 20:35 | |
| cnf | tdasilva: I do plan on pushing some stuff once I get it cleaned up a bit, though | 20:36 |
| tdasilva | cnf: cool, i'll keep an eye out, last time i looked on docker hub there wasn't anything there | 20:36 |
| cnf | yeah, indeed | 20:36 |
| tdasilva | cnf: i keep hearing about the kolla project, but haven't had a chance to play with it tho | 20:37 |
| cnf | kolla project? | 20:37 |
| cnf | oh, https://github.com/openstack/kolla ? | 20:37 |
| cnf | hmm | 20:37 |
| mattoliverau | Morning | 20:38 |
| cnf | hmm, no automated builds, darn | 20:39 |
| cnf | hey mattoliverau | 20:39 |
| cnf | hmm, I also need to figure out how to get openstack stats into influxdb, and logs into graylog | 20:41 |
| cnf | but later, need to get it working, first :P | 20:41 |
| tdasilva | cnf: yeah sorry, that's the project | 20:44 |
| cnf | had not seen that before | 20:44 |
| cnf | i'll poke around their stuff | 20:44 |
| *** sgundur has quit IRC | 20:46 | |
| *** sgundur has joined #openstack-swift | 20:46 | |
| cnf | it's a good project, I think | 20:48 |
| *** sgundur has quit IRC | 20:50 | |
| *** sgundur has joined #openstack-swift | 20:52 | |
| *** gyee has joined #openstack-swift | 20:55 | |
| *** diogogmt_ has joined #openstack-swift | 20:59 | |
| *** diogogmt has quit IRC | 21:01 | |
| *** diogogmt_ is now known as diogogmt | 21:01 | |
| *** sgundur has quit IRC | 21:11 | |
| *** sgundur has joined #openstack-swift | 21:11 | |
| *** sgundur has quit IRC | 21:20 | |
| *** sgundur has joined #openstack-swift | 21:23 | |
| *** dfflanders has quit IRC | 21:25 | |
| *** dfflanders has joined #openstack-swift | 21:28 | |
| *** cdelatte has quit IRC | 21:28 | |
| *** diogogmt has quit IRC | 21:44 | |
| *** sgundur has quit IRC | 21:44 | |
| *** mingyu has quit IRC | 21:44 | |
| *** mingyu has joined #openstack-swift | 21:47 | |
| *** nadeem has quit IRC | 22:01 | |
| *** mingyu has quit IRC | 22:03 | |
| tqtran | hello all, have a question. when we are doing things such as verifying account information in keystoneauth, are we making external calls to keystone? if so, could you point me to where the code is? | 22:14 |
| *** mingyu has joined #openstack-swift | 22:16 | |
| *** vint_bra has quit IRC | 22:16 | |
| *** mingyu has quit IRC | 22:17 | |
| mattoliverau | tqtran: that happens in the authtoken and keystone middleware. The first is keystones middleware and does the talkin to keystone and dumps the information into the enviromnent, then the keystone middleware does it's thing | 22:18 |
| mattoliverau | that might be a bit confusing when reading back.. keystone's provided middleware is authtoken, and it talks to keystone and dumps account and group impormation into the wsgi environment, the keystoneauth middleware is ours and reads the env and makes decisions. | 22:22 |
| mattoliverau | tqtran: ^ | 22:22 |
| timur_ | I don't know if anyone else is interested in this, but I'd like to attempt to ressurect https://review.openstack.org/#/c/196755/8. This looks like a pretty valuable thing to do with Swift. I do wonder if the database crawling approach (like suggested for metadata search) would subsume it (I can see arguments both ways), but it keeps popping up as something that Swift operators want to do | 22:29 |
| patchbot | patch 196755 - swift - Add notification middleware | 22:29 |
| *** timur_ is now known as timur | 22:30 | |
| tqtran | mattoliverau: keystone provided middleware is authtoken, is that part of swift's code? | 22:30 |
| tqtran | right, i figured that the information was populated prior to keystoneauth since everything is already stored in environment. i was trying to figure out where/how that happens. | 22:31 |
| mattoliverau | tqtran: its not in the swiftc code base, it comes from keystonemiddleware, a part of keystone. You can: pip install keystonemiddleware (and it probably packaged by your distro) or checkout the code locally or in github. | 22:33 |
| tqtran | i see, thanks for the pointer. so i guess if swift wanted to make a call directly into keystone, is that something that swift would have to implement or is it possible to somehow reuse that the keystonemiddleware to do its bidding? | 22:35 |
| jrichli | timur: bkeller` has been working on a notifications patch. It isn't ready to upload yet. It would be good to share ideas. | 22:35 |
| tqtran | i guess what im trying to say is, if swift needed to do something like - a name to id lookup - what would be the best approach? | 22:36 |
| timur | jrichli: oh, good to know. I guess maybe we can all talk about it at the summit | 22:37 |
| jrichli | timur: good plan! we will probably have a WIP patch uploaded by then, or an ideas page on it at the least. | 22:38 |
| timburke | tqtran: am i right to think this has to do with https://review.openstack.org/#/c/356715/ ? is the idea to let clients specify a name in the acl, then lookup the current id to actually store it? | 22:42 |
| patchbot | patch 356715 - swift - Supporting Account ACL in keystoneauth | 22:42 |
| *** klamath has quit IRC | 22:43 | |
| tqtran | timburke: yes, basing it off of what was discussed a few years ago | 22:45 |
| tqtran | https://wiki.openstack.org/wiki/Swift/ContainerACLWithKeystoneV3 | 22:45 |
| *** rcernin has quit IRC | 22:50 | |
| timburke | tqtran: but when reading it back, clients will still be presented with ids, yeah? i'm kinda inclined to say that clients should need to specify ids, always. especially if it's a new api -- keystone users have never been able to set account acls before, right? or at least, not with any useful effect | 22:54 |
| *** diogogmt has joined #openstack-swift | 23:06 | |
| tqtran | timburke: my current patch actually only support id. my intention is for my next patch to also allow the account owner to specify user name/domain. we then map this name/domain to an id which we store in the acl. | 23:07 |
| tqtran | this way, if name changes, it doesnt matter at all. | 23:08 |
| tqtran | but it does additional burden since we have to hit keystone for verification | 23:08 |
| *** diogogmt has quit IRC | 23:09 | |
| *** jamielennox is now known as jamielennox|away | 23:12 | |
| *** jamielennox|away is now known as jamielennox | 23:13 | |
| timburke | ...and if they specify, say, 30 name/domain pairs, we make 30 subrequests? | 23:16 |
| tqtran | i guess so. its a convenience feature, maybe we can have it as an option in the conf file to disable/enable it, disabled by default. realistically, do we normally have 30 people in the acl? | 23:22 |
| notmyname | oh my I think I'm finally finished with defcore updates | 23:23 |
| tqtran | or.. maybe it can happen in the background in a separate thread? iono =/ | 23:23 |
| *** kei_yama has joined #openstack-swift | 23:27 | |
| *** arch-nemesis has quit IRC | 23:28 | |
| *** chsc has quit IRC | 23:36 | |
| *** Suyi has quit IRC | 23:36 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!