| *** dtroyer is now known as dtroyer_zzz | 00:04 | |
| *** renier_ has joined #openstack-dev | 00:15 | |
| *** renier has quit IRC | 00:15 | |
| *** renier_ is now known as renier | 00:15 | |
| *** davidkranz has quit IRC | 00:16 | |
| *** davidkranz_ has joined #openstack-dev | 00:16 | |
| *** dachary has quit IRC | 00:18 | |
| *** mnewby has joined #openstack-dev | 00:19 | |
| *** matiu has quit IRC | 00:21 | |
| *** zhuadl has joined #openstack-dev | 00:22 | |
| *** spiffxp has quit IRC | 00:29 | |
| *** halfss has quit IRC | 00:31 | |
| *** yuanz has joined #openstack-dev | 00:33 | |
| *** gyee has joined #openstack-dev | 00:34 | |
| *** flaviamissi has quit IRC | 00:39 | |
| *** asalkeld has quit IRC | 00:39 | |
| *** asalkeld has joined #openstack-dev | 00:41 | |
| *** zhuadl has quit IRC | 00:43 | |
| *** s0mik has quit IRC | 00:46 | |
| *** matwood has joined #openstack-dev | 00:47 | |
| *** matwood has quit IRC | 00:47 | |
| *** anniec has joined #openstack-dev | 00:48 | |
| *** anniec_ has joined #openstack-dev | 00:49 | |
| *** edygarcia has joined #openstack-dev | 00:51 | |
| *** anniec has quit IRC | 00:52 | |
| *** anniec_ is now known as anniec | 00:52 | |
| *** novas0x2a|laptop has quit IRC | 00:55 | |
| *** roge has quit IRC | 00:56 | |
| *** dachary has joined #openstack-dev | 01:06 | |
| *** winston-d has joined #openstack-dev | 01:12 | |
| *** edygarcia has quit IRC | 01:17 | |
| *** chrisfer has quit IRC | 01:19 | |
| *** lzyeval has joined #openstack-dev | 01:21 | |
| *** mikal has quit IRC | 01:22 | |
| *** thingee has quit IRC | 01:22 | |
| *** nati_uen_ has joined #openstack-dev | 01:26 | |
| *** matiu has joined #openstack-dev | 01:27 | |
| *** matiu has quit IRC | 01:27 | |
| *** matiu has joined #openstack-dev | 01:27 | |
| *** littleidea has quit IRC | 01:28 | |
| *** nati_ueno has quit IRC | 01:29 | |
| *** edygarcia has joined #openstack-dev | 01:32 | |
| *** Aaton_off is now known as Aaton | 01:36 | |
| *** ncode has joined #openstack-dev | 01:38 | |
| *** gongys has joined #openstack-dev | 01:44 | |
| *** Ryan_Lane has quit IRC | 01:45 | |
| *** eglynn_ has joined #openstack-dev | 01:47 | |
| *** eglynn has quit IRC | 01:50 | |
| *** yuanz has left #openstack-dev | 01:53 | |
| *** Aaton is now known as Aaton_off | 01:56 | |
| *** ncode has quit IRC | 01:57 | |
| *** gyee has quit IRC | 01:57 | |
| *** PotHix has quit IRC | 02:01 | |
| *** jdurgin has quit IRC | 02:02 | |
| *** mikal has joined #openstack-dev | 02:02 | |
| *** anniec has quit IRC | 02:04 | |
| *** steveb___ has joined #openstack-dev | 02:06 | |
| *** steveb_ has quit IRC | 02:06 | |
| *** blamar_ has joined #openstack-dev | 02:09 | |
| *** blamar has quit IRC | 02:11 | |
| *** blamar_ is now known as blamar | 02:11 | |
| *** edygarcia has quit IRC | 02:13 | |
| *** tgall_foo has joined #openstack-dev | 02:16 | |
| *** tgall_foo has joined #openstack-dev | 02:16 | |
| *** steveb___ has quit IRC | 02:21 | |
| *** edygarcia has joined #openstack-dev | 02:21 | |
| *** steveb_ has joined #openstack-dev | 02:24 | |
| *** blamar has quit IRC | 02:27 | |
| *** dtroyer_zzz is now known as dtroyer | 02:30 | |
| *** danwent_ has quit IRC | 02:30 | |
| *** galstrom has joined #openstack-dev | 02:41 | |
| *** anderstj has joined #openstack-dev | 02:47 | |
| *** danwent has joined #openstack-dev | 03:06 | |
| *** ecarlin has quit IRC | 03:07 | |
| *** zhuadl has joined #openstack-dev | 03:08 | |
| *** jcooley has quit IRC | 03:14 | |
| *** nati_uen_ has quit IRC | 03:17 | |
| *** markmcclain has quit IRC | 03:18 | |
| *** gongys has quit IRC | 03:34 | |
| *** mnaser has joined #openstack-dev | 03:40 | |
| *** mnaser has joined #openstack-dev | 03:40 | |
| *** bencherian has quit IRC | 03:43 | |
| *** littleidea has joined #openstack-dev | 03:47 | |
| *** adjohn has quit IRC | 03:48 | |
| *** galstrom has quit IRC | 03:55 | |
| *** mnaser has quit IRC | 04:00 | |
| *** zhuadl has quit IRC | 04:03 | |
| *** samkottler has quit IRC | 04:08 | |
| *** jsgotangco has joined #openstack-dev | 04:13 | |
| *** mue_ has joined #openstack-dev | 04:17 | |
| *** mattstep has quit IRC | 04:18 | |
| *** steveb_ has quit IRC | 04:19 | |
| *** jsgotangco is now known as greeneggsnospam | 04:24 | |
| *** utlemming has quit IRC | 04:28 | |
| *** thingee has joined #openstack-dev | 04:30 | |
| *** utlemming has joined #openstack-dev | 04:30 | |
| *** s0mik has joined #openstack-dev | 04:31 | |
| *** mnewby has quit IRC | 04:33 | |
| *** mjfork has quit IRC | 04:36 | |
| *** tgall_foo has quit IRC | 04:36 | |
| *** matwood has joined #openstack-dev | 04:42 | |
| *** dtroyer is now known as dtroyer_zzz | 04:43 | |
| *** anderstj has quit IRC | 04:47 | |
| *** nati_ueno has joined #openstack-dev | 04:47 | |
| *** nati_ueno has quit IRC | 04:52 | |
| *** s0mik has quit IRC | 04:54 | |
| *** anderstj has joined #openstack-dev | 04:56 | |
| *** s0mik has joined #openstack-dev | 04:56 | |
| *** Mandell has joined #openstack-dev | 04:57 | |
| *** edygarcia has quit IRC | 04:57 | |
| *** utlemming has quit IRC | 05:11 | |
| *** zhuadl has joined #openstack-dev | 05:17 | |
| *** utlemming has joined #openstack-dev | 05:18 | |
| *** s0mik has quit IRC | 05:21 | |
| *** zaitcev has quit IRC | 05:34 | |
| *** adjohn has joined #openstack-dev | 05:48 | |
| *** adjohn has quit IRC | 05:50 | |
| *** anderstj has quit IRC | 05:52 | |
| *** anderstj has joined #openstack-dev | 05:53 | |
| *** anderstj has quit IRC | 05:57 | |
| *** asalkeld has quit IRC | 06:04 | |
| *** mrunge has joined #openstack-dev | 06:07 | |
| *** greeneggsnospam has quit IRC | 06:07 | |
| *** littleidea has quit IRC | 06:13 | |
| *** lzyeval has quit IRC | 06:16 | |
| *** mattstep has joined #openstack-dev | 06:36 | |
| *** dolphm has joined #openstack-dev | 06:38 | |
| *** eglynn_ has quit IRC | 06:43 | |
| *** armaan has joined #openstack-dev | 06:44 | |
| *** armaan has left #openstack-dev | 06:44 | |
| *** mindpixel has joined #openstack-dev | 06:45 | |
| *** dolphm has quit IRC | 06:46 | |
| *** armaan1 has joined #openstack-dev | 06:51 | |
| *** armaan1 has left #openstack-dev | 06:51 | |
| *** bencherian has joined #openstack-dev | 06:52 | |
| *** Mandell has quit IRC | 06:56 | |
| *** reidrac has joined #openstack-dev | 07:01 | |
| *** gongys has joined #openstack-dev | 07:02 | |
| *** EmilienM has joined #openstack-dev | 07:11 | |
| *** bencherian has quit IRC | 07:17 | |
| *** davidha has joined #openstack-dev | 07:17 | |
| *** thingee is now known as thingee_zz | 07:26 | |
| *** armaan has joined #openstack-dev | 07:34 | |
| *** armaan has quit IRC | 07:38 | |
| *** sulochan has joined #openstack-dev | 07:40 | |
| *** sulochan_ has joined #openstack-dev | 07:41 | |
| *** sulochan has quit IRC | 07:44 | |
| *** sulochan_ is now known as sulochan | 07:44 | |
| *** pquerna has quit IRC | 07:45 | |
| *** pquerna has joined #openstack-dev | 07:46 | |
| *** eglynn_ has joined #openstack-dev | 07:57 | |
| *** rods` has quit IRC | 08:02 | |
| *** pixelbeat has joined #openstack-dev | 08:07 | |
| *** rods has joined #openstack-dev | 08:09 | |
| *** rods has quit IRC | 08:14 | |
| *** winston-d has quit IRC | 08:18 | |
| *** rods has joined #openstack-dev | 08:22 | |
| *** darraghb has joined #openstack-dev | 08:30 | |
| *** sulochan has quit IRC | 08:37 | |
| *** asakhnov has joined #openstack-dev | 08:37 | |
| *** danpb has joined #openstack-dev | 08:40 | |
| *** asalkeld has joined #openstack-dev | 08:51 | |
| *** maploin has joined #openstack-dev | 09:00 | |
| *** maploin has quit IRC | 09:00 | |
| *** maploin has joined #openstack-dev | 09:00 | |
| *** sulochan has joined #openstack-dev | 09:22 | |
| *** heyho has joined #openstack-dev | 09:51 | |
| *** danwent has quit IRC | 09:57 | |
| *** zhuadl is now known as Edward_Zhang | 09:57 | |
| *** gongys has quit IRC | 10:06 | |
| *** armaan has joined #openstack-dev | 10:14 | |
| *** armaan has left #openstack-dev | 10:15 | |
| *** Edward_Zhang has quit IRC | 10:17 | |
| *** fattarsi has quit IRC | 10:20 | |
| *** rods_ has joined #openstack-dev | 10:36 | |
| *** amotoki has joined #openstack-dev | 10:38 | |
| *** flaviamissi has joined #openstack-dev | 11:05 | |
| *** halfss has joined #openstack-dev | 11:27 | |
| *** halfss has quit IRC | 11:30 | |
| *** otter1 has quit IRC | 11:30 | |
| *** markvoelker has joined #openstack-dev | 11:39 | |
| *** mattstep has quit IRC | 11:54 | |
| *** rods_ has quit IRC | 11:55 | |
| *** rods_ has joined #openstack-dev | 11:56 | |
| *** mattstep has joined #openstack-dev | 11:57 | |
| *** galstrom has joined #openstack-dev | 11:57 | |
| *** lts has joined #openstack-dev | 12:04 | |
| *** galstrom has quit IRC | 12:06 | |
| *** ayoung has joined #openstack-dev | 12:08 | |
| *** mrunge has quit IRC | 12:12 | |
| *** salgado has joined #openstack-dev | 12:12 | |
| *** salgado has joined #openstack-dev | 12:12 | |
| *** mjfork has joined #openstack-dev | 12:18 | |
| *** GheRivero has quit IRC | 12:20 | |
| *** amotoki has quit IRC | 12:22 | |
| *** reidrac has left #openstack-dev | 12:27 | |
| *** derekh has joined #openstack-dev | 12:32 | |
| *** tongli has joined #openstack-dev | 12:34 | |
| tongli | hi, someone in early June indicated that nova-manage is getting deprecated. any thing to replace it? | 12:35 |
|---|---|---|
| *** markmc has joined #openstack-dev | 12:36 | |
| tongli | anyone out there knows that what will be there and when to replace nova-manage or it is just a rumor? | 12:42 |
| *** maploin has quit IRC | 12:45 | |
| *** eafonichev has quit IRC | 12:46 | |
| *** asakhnov has quit IRC | 12:47 | |
| *** maploin has joined #openstack-dev | 12:47 | |
| *** maploin has quit IRC | 12:47 | |
| *** maploin has joined #openstack-dev | 12:47 | |
| *** tgall_foo has joined #openstack-dev | 12:55 | |
| *** tgall_foo has joined #openstack-dev | 12:55 | |
| *** dachary has quit IRC | 13:01 | |
| *** sandywalsh has joined #openstack-dev | 13:02 | |
| *** roge has joined #openstack-dev | 13:03 | |
| *** kbringard has joined #openstack-dev | 13:04 | |
| sandywalsh | vishy: ping? | 13:12 |
| sandywalsh | jaypipes: JAY!!! | 13:12 |
| *** dprince has joined #openstack-dev | 13:15 | |
| *** halfss has joined #openstack-dev | 13:21 | |
| *** dtroyer_zzz is now known as dtroyer | 13:36 | |
| *** nati_ueno has joined #openstack-dev | 13:57 | |
| *** dtroyer is now known as dtroyer_zzz | 14:00 | |
| jaypipes | sandywalsh: SANDY!!! | 14:00 |
| sandywalsh | jaypipes: that was all. :) | 14:00 |
| jaypipes | sandywalsh: :) ok | 14:01 |
| *** renier has quit IRC | 14:07 | |
| *** renier has joined #openstack-dev | 14:07 | |
| *** galstrom has joined #openstack-dev | 14:08 | |
| *** ewindisch has joined #openstack-dev | 14:09 | |
| *** dhellmann has joined #openstack-dev | 14:09 | |
| *** ncode has joined #openstack-dev | 14:12 | |
| *** littleidea has joined #openstack-dev | 14:13 | |
| *** e1mer has quit IRC | 14:14 | |
| *** dtroyer_zzz is now known as dtroyer | 14:15 | |
| *** edygarcia has joined #openstack-dev | 14:16 | |
| *** armaan has joined #openstack-dev | 14:26 | |
| *** armaan has left #openstack-dev | 14:26 | |
| *** dachary has joined #openstack-dev | 14:27 | |
| *** rnirmal has joined #openstack-dev | 14:30 | |
| *** armaan has joined #openstack-dev | 14:31 | |
| *** nati_uen_ has joined #openstack-dev | 14:35 | |
| *** nati_ueno has quit IRC | 14:35 | |
| *** armaan has quit IRC | 14:35 | |
| *** thingee_zz is now known as thingee | 14:43 | |
| *** ncode has quit IRC | 14:44 | |
| *** littleidea_ has joined #openstack-dev | 14:44 | |
| *** galstrom has left #openstack-dev | 14:46 | |
| *** littleidea has quit IRC | 14:48 | |
| *** littleidea_ is now known as littleidea | 14:48 | |
| *** datsun180b has joined #openstack-dev | 14:52 | |
| *** markmc has quit IRC | 14:54 | |
| jaypipes | eglynn_: got a sec to chat about https://bugs.launchpad.net/nova/+bug/1026600? | 14:55 |
| uvirtbot | Launchpad bug 1026600 in nova "Volume creation 201 API response does not include a Location header" [Low,In progress] | 14:55 |
| eglynn_ | jaypipes: sure | 14:55 |
| jaypipes | eglynn_: did you notice the failure int he tempest gate job> | 14:56 |
| jaypipes | ? | 14:56 |
| eglynn_ | jaypipes: BTW I just proposed a Tempest patch to take account of the change: https://review.openstack.org/10058 | 14:56 |
| eglynn_ | jaypipes: allows both old and new statuses as a tmp transition | 14:56 |
| jaypipes | eglynn_: ah, gotcha... | 14:57 |
| jaypipes | eglynn_: just haven't caught up to that email yet ;) | 14:57 |
| jaypipes | eglynn_: I will mark that bug as also affecting tempest then. | 14:57 |
| eglynn_ | jaypipes: np, I literally just pulled the trigger on the tempest patch ... thanks! | 14:58 |
| jaypipes | eglynn_: :) ty! | 14:58 |
| *** utlemming has quit IRC | 15:01 | |
| *** derekh has quit IRC | 15:01 | |
| *** utlemming has joined #openstack-dev | 15:03 | |
| *** flaviamissi has quit IRC | 15:03 | |
| *** derekh has joined #openstack-dev | 15:05 | |
| jaypipes | eglynn_: so changing the return code for the volume create call is indeed a change to the Compute API. Not sure we can just go around changing that if the spec calls for a 200... | 15:07 |
| eglynn_ | jaypipes: a-ha, I didn't realize the spec explicitly calls out that a 200 should be returned | 15:08 |
| *** markmc has joined #openstack-dev | 15:08 | |
| eglynn_ | jaypipes: (I was working on the basis of the normal RESTful idiom ...) | 15:09 |
| jaypipes | eglynn_: udnerstood. lemme grab the spec link. | 15:09 |
| eglynn_ | jaypipes: cool, ta | 15:09 |
| *** heyho has quit IRC | 15:09 | |
| jaypipes | eglynn_: and .. of course, I can't find it... | 15:10 |
| jaypipes | annegentle: are any of the volume commands doc'd in the Compute API spec? | 15:10 |
| eglynn_ | jaypipes: by the spec, do you mean this: https://github.com/openstack/compute-api/blob/master/openstack-compute-api-2/src/os-compute-2.wadl | 15:11 |
| *** sstent has quit IRC | 15:12 | |
| jaypipes | eglynn_: well that doesn't contain any volume commands either... | 15:13 |
| jaypipes | eglynn_: but no, was referrring to the human readable dosc | 15:13 |
| *** sstent has joined #openstack-dev | 15:13 | |
| eglynn_ | jaypipes: yep, just seacrhed thru' it now | 15:13 |
| *** primeministerp has quit IRC | 15:14 | |
| *** flaviamissi has joined #openstack-dev | 15:16 | |
| *** bencherian has joined #openstack-dev | 15:16 | |
| *** dprince has quit IRC | 15:18 | |
| *** _val_ has quit IRC | 15:18 | |
| *** ecarlin has joined #openstack-dev | 15:18 | |
| *** sacharya has joined #openstack-dev | 15:19 | |
| *** mattstep has quit IRC | 15:21 | |
| *** primeministerp has joined #openstack-dev | 15:22 | |
| *** derekh has quit IRC | 15:22 | |
| *** eglynn_ has quit IRC | 15:22 | |
| *** mindpixel has quit IRC | 15:24 | |
| *** eglynn_ has joined #openstack-dev | 15:25 | |
| *** mattstep has joined #openstack-dev | 15:27 | |
| *** zaitcev has joined #openstack-dev | 15:28 | |
| *** flaviamissi has quit IRC | 15:32 | |
| *** halfss has quit IRC | 15:32 | |
| jgriffith | eglynn_: I'm interested in your 201 status change patch as well... but I have the same concern raised by jaypipes | 15:33 |
| jgriffith | eglynn_: I think this would warrant broader discussion (ie the mailing lists) before changes are made | 15:33 |
| eglynn_ | jgriffith: cool, that makes sense ... I | 15:33 |
| eglynn_ | griffith: I'll fire off a mail to the ML proposing the change | 15:34 |
| eglynn_ | jgriffith ^^^ | 15:34 |
| jgriffith | eglynn_: Thanks... FWIW I agree with the logic completely! | 15:35 |
| jgriffith | If it were my cloud I'd change it :) | 15:35 |
| *** Mandell has joined #openstack-dev | 15:39 | |
| *** mnewby has joined #openstack-dev | 15:41 | |
| *** littleidea_ has joined #openstack-dev | 15:42 | |
| *** littleidea_ has joined #openstack-dev | 15:43 | |
| *** littleidea has quit IRC | 15:46 | |
| *** littleidea_ is now known as littleidea | 15:46 | |
| *** s0mik has joined #openstack-dev | 15:48 | |
| *** utlemming has quit IRC | 15:51 | |
| *** utlemming has joined #openstack-dev | 15:51 | |
| *** maploin has quit IRC | 15:52 | |
| dansmith | jaypipes: I'm secretly trying to see how many reviewers I can rack up and how many iterations I can push this silly change through :P | 15:53 |
| jaypipes | dansmith: sorry man :( | 15:54 |
| jaypipes | dansmith: but technically I was already on the review ;) | 15:54 |
| dansmith | heh, no, I'm just starting to fell ridiculous :) | 15:54 |
| dansmith | I know, didn't even help my score! :) | 15:54 |
| jaypipes | dansmith: oh, man, you are NOWHERE NEAR the record. | 15:54 |
| dansmith | s/fell/feel/ | 15:54 |
| dansmith | hah, dang, gotta work harder then | 15:54 |
| jaypipes | :) | 15:55 |
| *** utlemming has quit IRC | 15:55 | |
| jaypipes | I think the record is somewhere in the 40s. | 15:55 |
| dansmith | eesh | 15:55 |
| *** utlemming has joined #openstack-dev | 15:55 | |
| kbringard | bcwaldon: you about? | 16:00 |
| *** epim has joined #openstack-dev | 16:01 | |
| dansmith | jaypipes: er, scratch that suggestion of mine.. wrap_instance_fault is above checks_instance_lock (obviously) | 16:03 |
| *** cp16net is now known as cp16net|away | 16:04 | |
| *** cp16net|away is now known as cp16net | 16:04 | |
| *** markmc has quit IRC | 16:05 | |
| *** anderstj has joined #openstack-dev | 16:05 | |
| *** reed has joined #openstack-dev | 16:06 | |
| *** blamar has joined #openstack-dev | 16:06 | |
| *** nati_uen_ has quit IRC | 16:10 | |
| jaypipes | dansmith: hmm, so I'm not really following you on this logic... if you remove entirely the decoration of functions that don't revert states, the function will just be executed as-is, and that means the original behaviour will apply (returning False on lock or error) | 16:11 |
| *** nati_ueno has joined #openstack-dev | 16:11 | |
| jaypipes | dansmith: all I am suggesting is removing the decorator when you are passing no params. | 16:11 |
| jaypipes | i.e. on inject_file and agent_update | 16:12 |
| jaypipes | dansmith: but now I see that you changed the return of check_instance_locked()... | 16:13 |
| dansmith | right | 16:13 |
| dansmith | (per a prior review suggestion) | 16:13 |
| jaypipes | dansmith: mine? | 16:13 |
| dansmith | dunno, there have been so many.. I can go look if you want | 16:13 |
| dansmith | I don't really like the look of it with no params either, but was trying to keep it symmetric | 16:14 |
| jaypipes | dansmith: OK, how about a compromise... | 16:14 |
| *** danwent has joined #openstack-dev | 16:15 | |
| dansmith | (it was my suggestion and you agreed) | 16:15 |
| jaypipes | dansmith: how about adding the states param to the check_instance_lock decorator instead, and getting rid of the revert_task_states decorator entirely? | 16:15 |
| dansmith | because then you don't revert the state if the operation fails for some other reason than being locked | 16:16 |
| *** spiffxp has joined #openstack-dev | 16:16 | |
| * ttx waves from PDX airport | 16:16 | |
| dansmith | I was originally gunning for the locked case, but... | 16:16 |
| jaypipes | dansmith: gah, right.. :( | 16:16 |
| dansmith | ttx: ...and you didn't stop by to say hi? :) | 16:16 |
| ttx | dansmith: weren't you at OSCON ? | 16:16 |
| dansmith | ttx: nope... I guess you get a pass :) | 16:17 |
| *** markmc has joined #openstack-dev | 16:17 | |
| ttx | dansmith: I gave a talk there. Shortest path to the free all-inclusive pass :) | 16:18 |
| dansmith | ttx: hehe, yeah :) | 16:18 |
| jaypipes | dansmith: would you mind if I pulled your branch and gave it a once over? | 16:18 |
| dansmith | jaypipes: will it hurt? er, I mean, of course not, go ahead :) | 16:18 |
| jaypipes | dansmith: ok, cool. | 16:19 |
| dansmith | jaypipes: so, one other thing I had thought of for this, | 16:20 |
| *** flaviamissi has joined #openstack-dev | 16:20 | |
| dansmith | is to use an attr decorator to attach the revert states to a function, | 16:20 |
| dansmith | and then I could revert them separately in the lock check and instance failure decorators, | 16:20 |
| jaypipes | hmm... | 16:20 |
| jaypipes | possible. | 16:20 |
| dansmith | which seemed to be really nice, but I didn't see any @attr prior art | 16:20 |
| jaypipes | lemme play with it for a bit... | 16:21 |
| dansmith | dang, am I generating work for other people again? | 16:21 |
| jaypipes | hehe | 16:22 |
| *** nati_ueno has quit IRC | 16:22 | |
| *** spiffxp has quit IRC | 16:25 | |
| jk0 | bcwaldon: https://review.openstack.org/#/c/10030/ | 16:25 |
| *** spiffxp has joined #openstack-dev | 16:26 | |
| *** cp16net is now known as cp16net|away | 16:28 | |
| *** thovden has joined #openstack-dev | 16:29 | |
| Slower | hmm, so I'm looking at the nova compute code, it appears all authentication is done in the client and then calls to the engine over amqp are considered authenticated? that right? | 16:29 |
| *** anniec has joined #openstack-dev | 16:30 | |
| *** anderstj has quit IRC | 16:30 | |
| Slower | auth + policy checks | 16:30 |
| Slower | so I guess the amqp msg bus is considered secure then? | 16:31 |
| *** darraghb has quit IRC | 16:33 | |
| sandywalsh | vishy: ping? | 16:33 |
| *** jaypipes has quit IRC | 16:34 | |
| *** salgado is now known as salgado-lunch | 16:37 | |
| *** cp16net|away is now known as cp16net | 16:40 | |
| *** sandywalsh has quit IRC | 16:41 | |
| *** rods_ has quit IRC | 16:47 | |
| *** Mandell has quit IRC | 16:47 | |
| *** bencherian has quit IRC | 16:48 | |
| *** cp16net is now known as cp16net|away | 16:48 | |
| *** cp16net|away is now known as cp16net | 16:48 | |
| *** danpb has quit IRC | 16:49 | |
| *** Mandell has joined #openstack-dev | 16:49 | |
| *** jaypipes has joined #openstack-dev | 16:51 | |
| Slower | no one knows? | 16:52 |
| *** nati_ueno has joined #openstack-dev | 16:53 | |
| *** spiffxp has quit IRC | 16:55 | |
| *** danwent_ has joined #openstack-dev | 16:57 | |
| *** armaan has joined #openstack-dev | 16:59 | |
| *** armaan has left #openstack-dev | 16:59 | |
| *** rods_ has joined #openstack-dev | 16:59 | |
| *** danwent has quit IRC | 17:01 | |
| *** danwent_ is now known as danwent | 17:01 | |
| *** jdurgin has joined #openstack-dev | 17:03 | |
| *** jaypipes has quit IRC | 17:03 | |
| *** jaypipes has joined #openstack-dev | 17:05 | |
| *** ecarlin has quit IRC | 17:06 | |
| *** spiffxp has joined #openstack-dev | 17:06 | |
| *** dachary has quit IRC | 17:09 | |
| *** markmc has quit IRC | 17:11 | |
| *** anniec has quit IRC | 17:13 | |
| *** sulochan has quit IRC | 17:14 | |
| ewindisch | ttx: around? | 17:16 |
| *** jtran has joined #openstack-dev | 17:16 | |
| *** eglynn_ has quit IRC | 17:17 | |
| *** dachary has joined #openstack-dev | 17:19 | |
| *** rods_ has quit IRC | 17:21 | |
| *** rods_ has joined #openstack-dev | 17:21 | |
| *** s0mik has quit IRC | 17:22 | |
| *** s0mik has joined #openstack-dev | 17:23 | |
| *** cp16net is now known as cp16net|away | 17:33 | |
| *** issackelly has joined #openstack-dev | 17:39 | |
| *** littleidea has quit IRC | 17:40 | |
| *** anderstj has joined #openstack-dev | 17:41 | |
| *** littleidea has joined #openstack-dev | 17:46 | |
| *** anniec has joined #openstack-dev | 17:47 | |
| *** anderstj has quit IRC | 17:47 | |
| *** bencherian has joined #openstack-dev | 17:49 | |
| *** lts has quit IRC | 17:49 | |
| zykes- | Anyone here that know the paging stuff in glance ? | 17:51 |
| *** Ryan_Lane has joined #openstack-dev | 17:54 | |
| *** dachary has quit IRC | 17:55 | |
| *** anniec_ has joined #openstack-dev | 17:56 | |
| *** dachary has joined #openstack-dev | 17:57 | |
| *** anniec has quit IRC | 17:59 | |
| *** ewindisch has quit IRC | 17:59 | |
| notmyname | zaitcev: I added a comment to the review, but I'm happy to talk about it further here if you'd like | 18:01 |
| zaitcev | notmyname: thanks | 18:01 |
| notmyname | zaitcev: it showed up (for us at swiftstack) on replication. if replication is started before any device is added to the ring, the ring would never be reloaded | 18:01 |
| *** ecarlin has joined #openstack-dev | 18:02 | |
| *** anniec_ has quit IRC | 18:02 | |
| notmyname | zaitcev: the current patch adds the getter so that accessing .devs will check to see if the ring should be updated from disk | 18:03 |
| zaitcev | notmyname: I remember that, it made complete sense. It's only the re-defining devs[] that attracted my attention, because I use Ring class, so now every time I do for dev in r.devs: there's some funny stuff going on. | 18:04 |
| notmyname | zaitcev: that should still work. it will stat the file every so often now (how often is based on the config). I understand your concern though :-) | 18:05 |
| zaitcev | This practice in general prevents people like me from reading anything casually. You think it's a list, but it's not. It's a hell of a machinery behind, potentially. | 18:05 |
| zaitcev | very powerful technique, too powerful | 18:05 |
| notmyname | ya, it's actually one of my complaints with webob (all the magic that is done when you access .body or .content_type and similar | 18:06 |
| *** salgado-lunch is now known as salgado | 18:07 | |
| notmyname | zaitcev: I don't think the patch to the violates any established contract in the ring interface, though. I suppose it has the potential to add some latency to your ring operations | 18:07 |
| notmyname | zaitcev: do you have an issue with the current patch, or with the technique in general and how it could potentially be used? | 18:07 |
| zaitcev | notmyname: Sure. But I see a slippery slope, so I thought I'd ask about this one. | 18:08 |
| notmyname | zaitcev: please review +/- 1 as you see fit (also, thanks for reviewing swift patches) | 18:10 |
| Slower | is the amqp bus trusted in openstack? All the auth/policy checks seem to be done in teh api side, that right? | 18:10 |
| *** markmcclain has joined #openstack-dev | 18:11 | |
| Slower | I'm working on heat and wanted to know if that is truly the right model | 18:12 |
| *** rbasak has quit IRC | 18:16 | |
| *** ewindisch has joined #openstack-dev | 18:16 | |
| *** PotHix has joined #openstack-dev | 18:17 | |
| *** cp16net|away is now known as cp16net | 18:18 | |
| vishy | sandywalsh: pong | 18:20 |
| *** cp16net is now known as cp16net|away | 18:21 | |
| vishy | dansmith: so there was a point when we elevated all context at the manager layer so we didn't accidentally get rejected by a check context in the db layer | 18:21 |
| dansmith | vishy: so, is reboot out of sync with start/stop/etc? | 18:22 |
| vishy | dansmith: the idea when it was done was to eventually go through and find out exactly where it was needed, but since db access is going to be removed eventually and we have policy.json for access checks now, that never happened. | 18:22 |
| dansmith | ah, okay | 18:22 |
| dansmith | it's confusing because the test_lock() test right now will only work for reboot because of that elevation | 18:22 |
| *** anniec has joined #openstack-dev | 18:27 | |
| Slower | vishy: you sound like you know the auth stuff.. :) | 18:28 |
| Slower | vishy: you mind telling me if I'm correct.. all teh auth is done in the api and the amqp msgs to the engine are considered authenticated and secure? | 18:28 |
| jgriffith | jeblair: Wish I knew what were missing on that gate change | 18:29 |
| vishy | Slower: pretty much | 18:29 |
| vishy | jgriffith: ping | 18:30 |
| Slower | vishy: pretty much? :) | 18:30 |
| jgriffith | vishy: Hey... | 18:30 |
| Slower | there are exceptions? | 18:30 |
| vishy | Slower: yes, I mean there is some amount of checking at different layers, for example when nova talks to glance it uses the token from the original request | 18:30 |
| Slower | vishy: right | 18:30 |
| vishy | Slower: and the project is checked in the db layer in a few places | 18:31 |
| vishy | jgriffith: how is devstack gating on cinder coming | 18:31 |
| vishy | jgriffith: i noticed the devstack review expired | 18:31 |
| jgriffith | vishy: Can't figure out why the volume tests are failing | 18:31 |
| Slower | vishy: have you looked at heat-api at all? | 18:31 |
| jgriffith | vishy: Looking at it again now | 18:31 |
| vishy | Slower: I have not | 18:31 |
| Slower | vishy: so basically we are starting instaces from templates | 18:31 |
| vishy | jgriffith: I would like to get that in today, so let me know if i can help | 18:32 |
| Slower | vishy: and it's set up as another openstack api/engine | 18:32 |
| Slower | vishy: so currently we are passing in the creds to the engine and using those creds to call into openstack components | 18:32 |
| jgriffith | vishy: jeblair and I have both been scratching our heads so if you have a chance to take a quick look that would be great | 18:32 |
| Slower | vishy: but we are also doing HA stuff so we were thinking of just using a 'heat' user and having heat have its own policy in the api layer | 18:33 |
| vishy | jgriffith: you have a link to the failure somewhere? | 18:33 |
| jeblair | jgriffith: i think "[sudo] password for stack:" probably has something to do with it.... | 18:33 |
| Slower | heat user would be an admin and able to do whatever it needs to | 18:33 |
| jgriffith | vishy: Yeah, just a sec | 18:33 |
| jeblair | vishy: https://jenkins.openstack.org/job/gate-tempest-devstack-vm/4317 | 18:33 |
| vishy | Slower: that seems like a suboptimal solution to me | 18:33 |
| Slower | not sure what the best design would be.. | 18:33 |
| Slower | vishy: what would you recommend? | 18:33 |
| jgriffith | vishy: https://jenkins.openstack.org/job/gate-tempest-devstack-vm/4317/consoleFull | 18:33 |
| jeblair | here's the log with the sudo prompt: https://jenkins.openstack.org/job/gate-tempest-devstack-vm/4317/artifact/logs/screen-c-vol.txt | 18:33 |
| vishy | jeblair, jgriffith: perhaps cinder-rootwrap needs to be setup? | 18:33 |
| Slower | vishy: the current implementation stores the credentials of the user in the database and reauthenticates.. | 18:34 |
| vishy | Slower: getting proper credential passing seems like the right way | 18:34 |
| jgriffith | vishy: jeblair makes sense... | 18:34 |
| Slower | and heat itself does not have any policy | 18:34 |
| vishy | Slower: I guess the problem is that you are mimicking the amazon api, so you aren't getting creds from the user? | 18:34 |
| Slower | vishy: yeah we have to do token auth | 18:34 |
| Slower | vishy: and then get teh info from there.. which works.. | 18:35 |
| vishy | Slower: does it use the amazon secret/access? | 18:35 |
| Slower | vishy: but then we have to store all the auth info so we can reauth | 18:35 |
| Slower | vishy: yes | 18:35 |
| Slower | well we do both but yes | 18:35 |
| * vishy is unfamiliar with the api | 18:35 | |
| Slower | vishy: mostly it is intended to do secret/access, yes | 18:35 |
| vishy | Slower, can you use the keystone extension to validate ec2 tokens? | 18:35 |
| Slower | that is what we are doing yes | 18:35 |
| Slower | using keystone v2 api | 18:36 |
| Slower | and then we get back the user info as well | 18:36 |
| vishy | Slower: so it needs to store the token because it may automatically run commands for the user later? | 18:36 |
| Slower | vishy: yes | 18:36 |
| vishy | Slower: well that is a little nasty isn't it | 18:36 |
| Slower | yes ;) | 18:36 |
| vishy | Slower: ok so I think there are two approaches here | 18:36 |
| Slower | it's a tricky one | 18:37 |
| jeblair | jgriffith: i don't see any cinder-rootwrap code in devstack; perhaps it needs to be added there? | 18:37 |
| vishy | Slower: short term it seems like you will have to store credentials for the user | 18:37 |
| *** winston-d has joined #openstack-dev | 18:37 | |
| vishy | long term, we really need to figure out the right way for users to delegate auth to a service | 18:37 |
| jgriffith | jeblair: Yeah, trying to figure out how that works now. | 18:37 |
| vishy | which probably requires some changes to keystone etc. | 18:38 |
| Slower | vishy: right.. | 18:38 |
| vishy | but we have to deal with this more and more as we split services out | 18:38 |
| jgriffith | jeblair: Also not implemented in Cinder | 18:38 |
| vishy | passing the same token from nova -> cinder -> glance is probably incorrect | 18:38 |
| Slower | vishy: the other odd part in this is that we are passing the full creds to the engine via rpc.. | 18:38 |
| cloudfly | vishy i am probably going to put up for a review a patch to add the keystone api url to metadata api. you see anything potentially terrible about that? | 18:38 |
| vishy | jgriffith: wait, what we don't have a root wrap in cinder? | 18:38 |
| jgriffith | vishy: Nope, nothing in cinder/etc/ | 18:39 |
| vishy | cloudfly: I think it is ok under its own endpoint, although smoser is arguing vehemently against it | 18:39 |
| Slower | vishy: the tokens expire.. | 18:39 |
| Slower | vishy: and we could end up doing an HA operation days later.. | 18:39 |
| vishy | Slower: I would think you need to store access and secret | 18:39 |
| cloudfly | honestly he's being pedantic. | 18:39 |
| smoser | i do like to argue. | 18:39 |
| vishy | Slower: short term | 18:39 |
| * smoser reads | 18:39 | |
| Slower | vishy: yes | 18:40 |
| cloudfly | smoser i mean that in the literal sense not as a derrogatory statement | 18:40 |
| Slower | vishy: so you would suggest that instead of doing the heat-engine-has-its-own-user idea? | 18:40 |
| vishy | Slower: /me thinks | 18:40 |
| smoser | i'm generlaly against things. | 18:40 |
| smoser | i agree. | 18:40 |
| vishy | Slower: the problem with that is it totally messes up billing and quotas | 18:41 |
| Slower | smoser: haha :) | 18:41 |
| cloudfly | hah | 18:41 |
| smoser | but i really dont understand why there is a need for it in the metadata service | 18:41 |
| Slower | vishy: ah, right | 18:41 |
| cloudfly | it's the only pre-known address a system can query to find out where the catalog service is | 18:41 |
| cloudfly | well usually pre-known | 18:41 |
| smoser | you can write 25 lines of python that launches an instance with whatever user-data you like, and finds the api's keystone api and puts it in there. | 18:41 |
| jeblair | jgriffith: there's a bin/cinder-rootwrap | 18:41 |
| smoser | its just not necessary. | 18:41 |
| *** jcooley has joined #openstack-dev | 18:41 | |
| cloudfly | it's useful | 18:41 |
| smoser | and it becomes a wart when "keystone v2" | 18:41 |
| cloudfly | very useful | 18:41 |
| smoser | (or 3 as the case might be) | 18:42 |
| vishy | jeblair, jgriffith, ttx: we probably need to cherry pick the root wrap modifications into cinder | 18:42 |
| *** bencherian has quit IRC | 18:42 | |
| smoser | i'd like to avoid the wart | 18:42 |
| cloudfly | also how do you recommend finding the keystone api? | 18:42 |
| smoser | passed in via user-data | 18:42 |
| cloudfly | as far as i can tell you have to bake it in or find it out and pass it as user-data | 18:42 |
| cloudfly | and it | 18:43 |
| smoser | its easy. you clearly know the keystone api when you launch the instance. | 18:43 |
| vishy | (but we can do it after we get the gate working) | 18:43 |
| cloudfly | 's not advertised on horizon or anything | 18:43 |
| smoser | you get it, add it to user-data, and start an instance | 18:43 |
| cloudfly | that's silly | 18:43 |
| cloudfly | it | 18:43 |
| Slower | vishy: thanks for the chat, much appreciated :). | 18:43 |
| cloudfly | 's entirely unnecessary | 18:43 |
| smoser | no. its the same way you pass *any* information to an instance. | 18:43 |
| jeblair | jgriffith, vishy: makes sense. i think ttx is on an intercontenental flight right now, and i have no idea what i'm talking about, so i think you're up, jgriffith. :) | 18:43 |
| Slower | vishy: I'll do some more thinking on this then | 18:43 |
| cloudfly | it's not just any information though | 18:43 |
| smoser | thats what makes it cleaner. | 18:43 |
| *** shardy has joined #openstack-dev | 18:43 | |
| cloudfly | it's the catalog service | 18:43 |
| smoser | it is just any information | 18:43 |
| cloudfly | i disagree | 18:44 |
| smoser | theres nothing special about it | 18:44 |
| cloudfly | there is | 18:44 |
| cloudfly | it's the catalog service | 18:44 |
| smoser | and if you choose to go the route that I suggest, then your launcher works on other clouds also. | 18:44 |
| vishy | smoser, cloudfly: here is my argument for putting it in: we are trying to make the openstack api ubiquitous so that users can use any openstack cloud (big goal i know) | 18:44 |
| cloudfly | it provides a one stop shop access point to openstack apis | 18:44 |
| smoser | even authenticating via my cloud from your cloud | 18:44 |
| smoser | :) | 18:44 |
| jgriffith | jeblair: vishy: Well I'm in the process of adding etc/cinder/rootwrap... then was just going to try and fubmle through devstack using nova as the example | 18:44 |
| vishy | smoser, cloudfly: if we don't throw some required things in like this, we are going to end up with everyone doing it a little bit different. | 18:44 |
| smoser | or they dont do it at all. | 18:45 |
| vishy | jgriffith: my guess is the rootwrap conf doesn't exist in cinder | 18:45 |
| *** sandywalsh has joined #openstack-dev | 18:45 | |
| smoser | and you solve it outside, and it works everywhere. | 18:45 |
| vishy | jgriffith: i would think we just need to modify shudders properly | 18:45 |
| jgriffith | vishy: Nope, I just added it and rootwrap.d files for volume | 18:45 |
| cloudfly | it isn't able to be solved outside for many use cases | 18:45 |
| smoser | cloudfly, like? | 18:46 |
| jgriffith | vishy: Sadly I don't know what you mean by shudders? | 18:46 |
| vishy | jgriffith: (you cherry picked the code) ? | 18:46 |
| smoser | if you're launching an instance, you know the api end point. if you know the thing that your lauching needs that information, give it to it. | 18:46 |
| cloudfly | my current example is i am generating a public image that authenticates against keystone | 18:46 |
| vishy | * sudoers (damn autocorrect) | 18:46 |
| jgriffith | vishy: yes, out of nova | 18:46 |
| jgriffith | vishy: lol | 18:46 |
| cloudfly | right now i have to bake in the keystone API server for the cloud it is in | 18:46 |
| vishy | jgriffith: ok if you want to go that way | 18:46 |
| cloudfly | if the metadata api had just that one URL the image would be cross cloud portable | 18:46 |
| vishy | jgriffith: you will need to modify cinder-rootwrap as well and add the flags | 18:46 |
| jgriffith | vishy: K... if there's a preferred alternative I'm happy to dig and learn it | 18:47 |
| cloudfly | requiring every user who launches that image as an instance to find the keystone uri which isn't even easy to find in horizon and pass it as userdata is just too much | 18:47 |
| jgriffith | vishy: Either that or I change it again later | 18:47 |
| cloudfly | especially if it's their first time on the cloud | 18:47 |
| *** armaan has joined #openstack-dev | 18:47 | |
| vishy | smoser, cloudfly: I'm for putting it in, but I'm happy to do a vote on the mailing list if someone wants to throw together a quick email with pros and cons | 18:48 |
| *** armaan has left #openstack-dev | 18:48 | |
| smoser | cloudfly, but your argument can be used to add *anything* to the the metadata service. | 18:48 |
| cloudfly | i mean i agree it's a deviation from traditional metadata constraints and that's important to clarify | 18:48 |
| cloudfly | but i think there is value in this specifically since it is an openstack api access datum | 18:48 |
| smoser | i agree its sort of different due to that. | 18:49 |
| smoser | but its also "just data" | 18:49 |
| cloudfly | it's all just data | 18:49 |
| smoser | well, there are certain things the cloud knows that the user at launch time cannot know. | 18:49 |
| smoser | this is not that. | 18:49 |
| smoser | ie, i dont know at launch time the instance-id. | 18:49 |
| cloudfly | i'd say its exactly that | 18:49 |
| smoser | or IP address. | 18:49 |
| *** zaneb has joined #openstack-dev | 18:49 | |
| smoser | i do know at launch time the keystone api endpoint | 18:49 |
| smoser | because i'm using it | 18:50 |
| smoser | and i can't laucnh an instance without knowing it. | 18:50 |
| smoser | i'll stop arguing (as i said, i do like to argue). | 18:50 |
| *** dachary has quit IRC | 18:50 | |
| *** dachary has joined #openstack-dev | 18:51 | |
| smoser | but before putting it in, we need to have some definition written of what qualifies to go in, and what does not. | 18:51 |
| smoser | and why this data is special | 18:51 |
| cloudfly | users are acting via horizon | 18:51 |
| cloudfly | horizon at best gives you a downloadable script that you can edit in a text editor | 18:51 |
| smoser | users can act via nother web service. | 18:51 |
| cloudfly | if you know to look for it | 18:51 |
| cloudfly | and find the keystone URI | 18:51 |
| smoser | or horizon could just tag all instances with keystone uri | 18:51 |
| cloudfly | heck most users wouldn't know or care what the uri is | 18:51 |
| cloudfly | and shouldn't. | 18:51 |
| *** otter1 has joined #openstack-dev | 18:51 | |
| smoser | you're arguing "web portal/client" is broken and i dont want to change it | 18:52 |
| smoser | :) | 18:52 |
| cloudfly | hrmm | 18:52 |
| cloudfly | well i guess it's a question | 18:52 |
| cloudfly | if i am a cloud operator | 18:52 |
| cloudfly | nah | 18:52 |
| cloudfly | it all comes back to this being openstack api data | 18:53 |
| cloudfly | and important data | 18:53 |
| cloudfly | do we want it available to anything that would use the metadata values? | 18:53 |
| cloudfly | i mean what good is an instance-id if you can't query any apis other than metadata? | 18:53 |
| smoser | i dont necessarily agree that it is "important" data. | 18:53 |
| cloudfly | without user supplied and non standard datasets? | 18:53 |
| smoser | i think its important for your use case. | 18:53 |
| cloudfly | it's the only way to query other apis? | 18:53 |
| smoser | but for my use case, it is not. | 18:53 |
| smoser | but you can't query other apis without credentials | 18:54 |
| cloudfly | i think it's important as a uniform path for querying other api servers | 18:54 |
| smoser | we should put credentials in there too! | 18:54 |
| cloudfly | credential passing is a whole nother thing | 18:54 |
| cloudfly | and i feel that becomes a keystone thing | 18:54 |
| jgriffith | vishy: jeblair: https://review.openstack.org/#/c/10071 | 18:54 |
| vishy | smoser, cloudfly: there is also the potential future where the same cloud could support multiple different keystone endpoints | 18:55 |
| cloudfly | federation indeed | 18:55 |
| *** troytoman-away is now known as troytoman | 18:55 | |
| *** dachary has quit IRC | 18:55 | |
| cloudfly | in which case what I really want is a catalog service on metadata | 18:55 |
| vishy | jgriffith: still need to update cinder-rootwrap | 18:56 |
| *** dachary has joined #openstack-dev | 18:56 | |
| smoser | i dont know if that is an argument for or against, vishy . honestly. you're arguing "we might have to change this data". which was what i used to argue against it. | 18:56 |
| jgriffith | vishy: crap... forgot to add that file | 18:56 |
| smoser | so, i will stop arguing. and wont cry loudly. i'm willing to accept that this is somehow useful information. | 18:56 |
| jgriffith | vishy: Oh wait, no I didn't | 18:56 |
| smoser | but i will argue that we need *some* policy on changes to the metadata service | 18:57 |
| vishy | jgriffith: bin/cinder-rootwrap | 18:57 |
| smoser | as it is an API. | 18:57 |
| jgriffith | vishy: yeah, but wasn't sure what needed "changed" | 18:57 |
| jgriffith | vishy: Thought it was ok as is | 18:57 |
| vishy | jgriffith: the code changed since the split, those new files won't do anything without the new code | 18:57 |
| smoser | cloudfly, so please come up with some policy (clearly one that fits your desired datum) and propose that on the mailing list | 18:57 |
| vishy | jgriffith: that is what I was trying to say. it used to use python code | 18:57 |
| *** edygarcia_ has joined #openstack-dev | 18:57 | |
| cloudfly | smoser i'll try to and i expect you'll keep me from doing anything stupid =P | 18:58 |
| jgriffith | vishy: ah crap... | 18:58 |
| vishy | jgriffith: you will need to change the cinder/rootwrap directory as well | 18:58 |
| smoser | cloudfly, as long as you promise that i can find some security vulnerability exposed because of the data you're giving me :) | 18:58 |
| vishy | jgriffith: I was trying to say, you probably will have an easier time just updating sudoers for now and letting ttx cherry pick his changes in | 18:59 |
| vishy | jgriffith: it actually changed a lot | 18:59 |
| jgriffith | vishy: Yeah, I just wasn't sure what any of this looked like so I thought this would work | 18:59 |
| jgriffith | vishy: Apprantly not | 18:59 |
| cloudfly | smoser i was half tempted last night to suggest adding posix ids to keystone. then i regained my sanity. | 18:59 |
| smoser | cloudfly, very honestly, the best definition I can come up with of what belongs in there versus what does not, is "what data is not known by the creator" | 19:00 |
| vishy | jgriffith: the old style root wrap is in cinder. It should still work, but I'm guessing the root wrap command isn't added to sudoers | 19:00 |
| cloudfly | this certainly isn't that | 19:00 |
| smoser | and that clearly does not fit your datum | 19:00 |
| smoser | :) | 19:00 |
| smoser | so.. you find a good definition that does. | 19:00 |
| cloudfly | but it is what openstack access do we want to provide a uniform access to? | 19:00 |
| cloudfly | err openstack data | 19:00 |
| smoser | "that which the user doens't already know" :) | 19:01 |
| smoser | anyway. | 19:01 |
| smoser | i really need to work on something else. but i do look forward to your thoughts on it. | 19:01 |
| cloudfly | okay will do | 19:01 |
| *** edygarcia has quit IRC | 19:01 | |
| *** edygarcia_ is now known as edygarcia | 19:01 | |
| smoser | i just dont want a scenario of "look, lets through some data in there!" | 19:01 |
| vishy | jgriffith: so option 1 is look at devstack/stable/essex and look how it set up shudders before | 19:01 |
| jgriffith | vishy: Sorry, thought I could just add it in devstacks ROOTWRAP_SUDOER_CMD and be done with it | 19:01 |
| smoser | amazon has been very controlled on what they put there. | 19:01 |
| vishy | * sudoers | 19:01 |
| cloudfly | fair | 19:01 |
| *** armaan1 has joined #openstack-dev | 19:03 | |
| vishy | jgriffith: actually | 19:03 |
| *** armaan1 has left #openstack-dev | 19:03 | |
| vishy | jgriffith: i think you can leave cinder as is and just duplicate the code for ROOTWRAP_SUDOER_CMD and add one for cinder as well | 19:03 |
| jgriffith | vishy: That's what I was "hoping" | 19:04 |
| jgriffith | vishy: but TBH wasn't sure it would work, so I guess I can just "try" it and find out | 19:04 |
| *** PotHix is now known as SeuMadruga | 19:05 | |
| *** SeuMadruga is now known as PotHix | 19:06 | |
| vishy | jgriffith: something like this | 19:10 |
| vishy | http://paste.openstack.org/show/19515/ | 19:10 |
| vishy | (might want to wrap it in an if enabled service somewhere, but that "should" work | 19:10 |
| jgriffith | vishy: I've got a change, let me past it... just a sec | 19:11 |
| jgriffith | http://paste.openstack.org/show/19516/ | 19:11 |
| jgriffith | vishy: Ok, looks like we have the same thing for the most part | 19:13 |
| *** s0mik has quit IRC | 19:13 | |
| *** flaviamissi has quit IRC | 19:13 | |
| *** markmcclain has quit IRC | 19:17 | |
| jgriffith | vishy: jeblair: https://review.openstack.org/10073 | 19:19 |
| jgriffith | shit... forgot to wrap it | 19:20 |
| vishy | jgriffith: i think you should duplicate the file creation code also | 19:23 |
| vishy | jgriffith: putting the cinder root wrap command in a file called nova-rootwrap is non-intuitive | 19:23 |
| jgriffith | vishy: Yep | 19:24 |
| vishy | jgriffith: (commented inline) | 19:24 |
| jgriffith | vishy: Thanks | 19:24 |
| *** danwent has quit IRC | 19:27 | |
| *** dachary has quit IRC | 19:27 | |
| vishy | comstud: looks like padraig responded to your concern here: https://review.openstack.org/#/c/9944/4 Mind if i send it in? | 19:28 |
| *** jog0 has joined #openstack-dev | 19:29 | |
| *** jog0 has quit IRC | 19:32 | |
| *** mdomsch has joined #openstack-dev | 19:33 | |
| comstud | vishy: looking | 19:35 |
| *** littleidea has quit IRC | 19:36 | |
| comstud | vishy: pushed | 19:37 |
| vishy | comstud: thx | 19:37 |
| comstud | np | 19:37 |
| *** jog0 has joined #openstack-dev | 19:39 | |
| *** dachary has joined #openstack-dev | 19:39 | |
| jgriffith | vishy: Forgot... https://review.openstack.org/#/c/10073/3 | 19:40 |
| jgriffith | vishy: It'll still need the other cinder change though | 19:40 |
| *** danwent has joined #openstack-dev | 19:40 | |
| *** pixelbeat has quit IRC | 19:41 | |
| *** littleidea has joined #openstack-dev | 19:41 | |
| *** danwent_ has joined #openstack-dev | 19:41 | |
| *** s0mik has joined #openstack-dev | 19:42 | |
| jgriffith | Suppose this could've been moved into lib/cinder | 19:44 |
| *** danwent has quit IRC | 19:45 | |
| *** danwent_ is now known as danwent | 19:45 | |
| tongli | can anyone tell me how to use the code review site to respond to a comment (not inline comment)? | 19:49 |
| cloudfly | i didn't know that could be done. | 19:49 |
| dansmith | tongli: click the review button | 19:50 |
| *** sandywalsh has quit IRC | 19:51 | |
| tongli | ah, I thought that was just for the people who wants to review my code. | 19:51 |
| *** atdxs has joined #openstack-dev | 19:51 | |
| tongli | let me try that. | 19:51 |
| *** eglynn_ has joined #openstack-dev | 19:51 | |
| *** littleidea has quit IRC | 19:54 | |
| ewindisch | You can leave a +0 review with comment | 19:54 |
| ewindisch | (really there is no plus or minus on zero… but I guess a '0 review' wouldn't have made as much sense) | 19:55 |
| *** littleidea has joined #openstack-dev | 19:56 | |
| jgriffith | jeblair: Sorry... I went ahead and moved the changes into lib/cinder | 19:57 |
| jeblair | jgriffith: no prob, the change is on my dashboard now. :) is there a corresponding cinder change(s)? | 19:57 |
| jgriffith | jeblair: It turns out as vishy said this will work without those cinder changes | 19:58 |
| jgriffith | the bin file is the only one that should be needed | 19:58 |
| jgriffith | jeblair: *should* is the key | 19:58 |
| vishy | jgriffith, jeblair: that is my theory :) | 19:58 |
| jgriffith | vishy: I'm sure that's the case, just had to convince myself and look at the code :) | 19:59 |
| *** danwent has quit IRC | 20:02 | |
| *** Shrews has quit IRC | 20:04 | |
| *** eglynn_ has quit IRC | 20:10 | |
| tongli | can some helping soul approve this patch, https://review.openstack.org/#/c/10038/, this is basically 6 character change. | 20:12 |
| *** jog0 has quit IRC | 20:13 | |
| *** novas0x2a|laptop has joined #openstack-dev | 20:16 | |
| *** armaan has joined #openstack-dev | 20:18 | |
| *** bencherian has joined #openstack-dev | 20:19 | |
| *** armaan has quit IRC | 20:22 | |
| *** tongli has quit IRC | 20:24 | |
| jgriffith | dtroyer: ping | 20:27 |
| dtroyer | jgriffith: yo | 20:27 |
| *** bencherian has quit IRC | 20:27 | |
| jgriffith | dtroyer: I'm confused... it is in the configure_cinder function isn't it? | 20:28 |
| *** eglynn_ has joined #openstack-dev | 20:28 | |
| dtroyer | jgriffith: yes, that is where all of the configuration bits are handled | 20:28 |
| jgriffith | dtroyer: But your comment on the review said it "belongs" in config_cinder, but I did put the changes in configure_cinder I thought? | 20:29 |
| dtroyer | um, I was looking at patch set 3, we must have done that in parallel, let me re-look | 20:30 |
| jgriffith | dtroyer: Yeah, just occured to me that might be the case | 20:30 |
| jgriffith | dtroyer: I remembered the *new* model after that submission :) | 20:30 |
| dtroyer | jgriffith: can you see any case where the CINDER_ROOTWRAP var might be needed outside this function? if not it looks good | 20:32 |
| *** bencherian has joined #openstack-dev | 20:32 | |
| jgriffith | dtroyer: I don't see any reason, this would have to be run any time cinder service is enabled based on what I saw first glance | 20:33 |
| * jgriffith thinking | 20:33 | |
| jgriffith | dtroyer: If I'm correct in my thinking that configure_cinder would have to be run any time cinder is enabled then we should be good | 20:34 |
| jgriffith | dtroyer: Without the service running there's no other need/use for cinder | 20:34 |
| jgriffith | rootwra | 20:34 |
| jgriffith | dtroyer: I'll resubmit if you don't see any other issues? | 20:35 |
| dtroyer | jgriffith: the nova version sets root_helper=sudo $NOVA_ROOTWRAP in nova.conf, does cinder have an equivalent? | 20:36 |
| jgriffith | dtroyer: I believe that's the "new" model that we don't have implemented yet | 20:36 |
| jgriffith | dtroyer: lemme double check | 20:36 |
| jgriffith | dtroyer: Ahhh... very nice catch! | 20:38 |
| *** dmarkey has quit IRC | 20:44 | |
| jgriffith | dtroyer: http://paste.openstack.org/show/19517/ Not positive on the iniset syntax | 20:45 |
| *** rkukura has quit IRC | 20:46 | |
| *** rkukura has joined #openstack-dev | 20:46 | |
| jgriffith | dtroyer: found it... ok, so needs quotes I'm thinking | 20:48 |
| *** littleidea has quit IRC | 20:49 | |
| *** littleidea has joined #openstack-dev | 20:50 | |
| *** ayoung has quit IRC | 20:50 | |
| *** anniec has quit IRC | 20:52 | |
| *** EmilienM has quit IRC | 20:53 | |
| dtroyer | jgriffith: back…it'll need quotes, I'm not sure I tested inset with args with spaces, let me check | 20:55 |
| *** anderstj has joined #openstack-dev | 20:55 | |
| *** anderstj has quit IRC | 20:55 | |
| dtroyer | jgriffith: yes, it's even in the test script…should work fine with one set of quotes | 20:56 |
| *** timello has quit IRC | 20:59 | |
| *** timello has joined #openstack-dev | 20:59 | |
| jgriffith | dtroyer: good, that's what I submitted :) | 20:59 |
| jgriffith | dtroyer: https://review.openstack.org/#/c/10073/5 | 20:59 |
| dtroyer | jgriffith: you're close, real close... | 21:00 |
| jgriffith | dtroyer: killin me :) | 21:00 |
| dtroyer | jgriffith: the creation of CINDER_CONF_DIR needs to come first | 21:00 |
| jgriffith | dtroyer: BTW something is funky with diff view not moving it up | 21:00 |
| jgriffith | dtroyer: hehe | 21:01 |
| dtroyer | also, collapse the blank line at the end to one | 21:01 |
| jgriffith | dtroyer: Yeah, if you do the full file view you'll see it's actually moved up before the ini set | 21:01 |
| jgriffith | dtroyer: Ok, got ya | 21:01 |
| *** ecarlin has quit IRC | 21:01 | |
| *** JStoker has quit IRC | 21:02 | |
| jgriffith | dtroyer: last one | 21:03 |
| dtroyer | jgriffith: sooooo close... | 21:04 |
| dtroyer | jgriffith: it looks like you copied lines 94-99 rather than cut them | 21:04 |
| jgriffith | dtroyer: Yep | 21:04 |
| dtroyer | jgriffith: so in the current patch set, remove 91-100 | 21:04 |
| *** thovden has quit IRC | 21:05 | |
| dtroyer | jgriffith: nothing like late Friday coding to start the weekend | 21:05 |
| *** dmarkey has joined #openstack-dev | 21:05 | |
| jgriffith | dtroyer: This is what happens when I get in a hurry and think something is a "simple" change | 21:06 |
| *** rnirmal has quit IRC | 21:07 | |
| dtroyer | jgriffith: and that's the reason devops-types like Read-Only-Friday | 21:07 |
| jgriffith | dtroyer: Ok... now that should really ebe it this time | 21:07 |
| jgriffith | dtroyer: Hmm... that's not a bad idea | 21:07 |
| *** cp16net|away is now known as cp16net | 21:07 | |
| dtroyer | jgriffith: I say it's a go before anything else happens | 21:08 |
| jgriffith | dtroyer: Agreeed!! | 21:08 |
| jgriffith | dtroyer: Thanks! | 21:09 |
| jgriffith | jeblair: thanks to you as well, I'll run recheck after this hits | 21:09 |
| *** eglynn_ has quit IRC | 21:11 | |
| *** ecarlin has joined #openstack-dev | 21:12 | |
| jeblair | jgriffith: awesome! | 21:16 |
| jgriffith | jeblair: We'll see | 21:18 |
| jgriffith | jeblair: BTW, how did you determine a root/sudo problem? | 21:19 |
| * jgriffith wants to learn to fish | 21:19 | |
| jeblair | jgriffith: combing the log files that are archived in jenkins | 21:19 |
| *** dmarkey has quit IRC | 21:19 | |
| *** dmarkey has joined #openstack-dev | 21:20 | |
| jgriffith | jeblair: So I just missed the *right* screen log | 21:20 |
| jgriffith | figures | 21:20 |
| jeblair | yeah, and i think it ended up interleaved with the rest of the output (like it went to stderr instead of stdout) so it didn't really jump out at you. | 21:20 |
| jgriffith | jeblair: Well, good to know for the future. That data's there, just have to find it | 21:21 |
| *** dmarkey has quit IRC | 21:21 | |
| jeblair | jgriffith: yeah, and if there's any other data we need to pull back off the host into jenkins, we can add it to what devstack-gate collects. | 21:21 |
| *** atdxs has quit IRC | 21:28 | |
| *** JStoker has joined #openstack-dev | 21:28 | |
| *** troytoman is now known as troytoman-away | 21:31 | |
| *** dubsquared has joined #openstack-dev | 21:31 | |
| creiht | jgriffith: just fyi: https://bugs.launchpad.net/nova/+bug/1027263 | 21:34 |
| uvirtbot | Launchpad bug 1027263 in nova "Nova volume api volume list and volume detail list are the same" [Undecided,New] | 21:34 |
| jgriffith | creiht: thanks | 21:35 |
| creiht | Not sure how you guys are tagging bugs that are in both nova volume and cinder | 21:35 |
| jgriffith | creiht: I'm manually doing an "also affects project" | 21:35 |
| creiht | ahh ok | 21:35 |
| creiht | I'll do that for now on | 21:35 |
| jgriffith | creiht: Are you just doing a curl call for this? | 21:36 |
| creiht | jgriffith: yeah I was just going through validating the api calls | 21:36 |
| jgriffith | creiht: K | 21:36 |
| *** mdomsch has quit IRC | 21:44 | |
| *** littleidea has quit IRC | 21:45 | |
| *** dmarkey has joined #openstack-dev | 21:47 | |
| *** littleidea has joined #openstack-dev | 21:52 | |
| mtaylor | bcwaldon: around? | 21:54 |
| Vek | can I get some reviews on https://review.openstack.org/#/c/9997/ please? | 21:56 |
| jaypipes | dansmith: screw it... I played around with damn decorators all friggin day and was a total waste of time... there are so many decorators inside decorators inside decorators with nova it is tough to tell what level of inception you are in at times, and adding parameterized decorators (which require a decorator object with an __init__ method) just gets ugly with all the fucntools.wrapped() decorators... | 21:56 |
| jaypipes | dansmith: so I give up... and you can count me as +1 in your patch. | 21:57 |
| *** jog0 has joined #openstack-dev | 21:57 | |
| jaypipes | meanwhile, I'm going to have a beer. very frustrated. | 21:57 |
| *** jaypipes has quit IRC | 21:57 | |
| Vek | yeah, decorators are screwy. | 21:57 |
| Vek | when I have the chance, I try to write decorators that only meddle with attributes on the functions; that's a lot easier to do and understand... | 21:58 |
| Vek | but of course there's always a need to have decorators that wrap the function call... | 21:58 |
| *** mnewby has quit IRC | 21:59 | |
| *** mnewby has joined #openstack-dev | 21:59 | |
| *** salgado has quit IRC | 22:04 | |
| *** dubsquared has quit IRC | 22:06 | |
| bcwaldon | mtaylor: nope, email plz | 22:06 |
| mtaylor | bcwaldon: bah | 22:07 |
| mtaylor | bcwaldon: email too slow | 22:07 |
| mtaylor | bcwaldon: I've already stopped caring | 22:07 |
| bcwaldon | mtaylor: like a boss | 22:08 |
| *** ecarlin_ has joined #openstack-dev | 22:08 | |
| mtaylor | bcwaldon: hell yeah | 22:08 |
| *** ecarlin has quit IRC | 22:08 | |
| *** ecarlin_ is now known as ecarlin | 22:08 | |
| *** steveb_ has joined #openstack-dev | 22:12 | |
| *** anderstj has joined #openstack-dev | 22:12 | |
| *** kbringard has left #openstack-dev | 22:14 | |
| *** zaneb has quit IRC | 22:16 | |
| *** littleidea has quit IRC | 22:20 | |
| *** mnewby has quit IRC | 22:21 | |
| *** rpedde is now known as rpedde_away | 22:21 | |
| *** sacharya has quit IRC | 22:22 | |
| *** dachary1 has joined #openstack-dev | 22:30 | |
| *** issackelly has quit IRC | 22:31 | |
| *** dachary has quit IRC | 22:31 | |
| *** mnewby has joined #openstack-dev | 22:31 | |
| *** dachary1 is now known as dachary | 22:32 | |
| jgriffith | creiht: You still around? | 22:32 |
| *** littleidea has joined #openstack-dev | 22:32 | |
| *** steveb_ has quit IRC | 22:32 | |
| *** Slower has quit IRC | 22:33 | |
| jgriffith | jeblair: ping | 22:33 |
| *** jog0 has quit IRC | 22:37 | |
| ttx | jgriffith: fire me an email so that I don't forget to look into that | 22:44 |
| ttx | (cinder-rootwrap) | 22:44 |
| ttx | travel today is likely to erase all memory from it | 22:44 |
| jgriffith | ttx: sounds good, hopefully what we've done will work temporarily at least | 22:44 |
| *** thingee has quit IRC | 22:44 | |
| creiht | jgriffith: howdy | 22:46 |
| creiht | what's up? | 22:46 |
| *** dachary1 has joined #openstack-dev | 22:46 | |
| jgriffith | creiht: Just posted a question about your detail bug | 22:46 |
| creiht | k | 22:47 |
| *** dachary has quit IRC | 22:47 | |
| *** dachary1 is now known as dachary | 22:50 | |
| creiht | jgriffith: updated | 22:51 |
| *** PotHix has quit IRC | 22:51 | |
| creiht | jgriffith: tldr: it should work just like /servers and /servers/detail do | 22:52 |
| jgriffith | creiht: git it, thanks. Just wanted it explicitly pointed out | 22:52 |
| creiht | yeah | 22:52 |
| creiht | and I'm out... have a good weekend :) | 22:52 |
| *** PotHix has joined #openstack-dev | 22:53 | |
| jgriffith | creiht: Thanks.. you too | 22:53 |
| *** issackelly has joined #openstack-dev | 22:53 | |
| *** datsun180b has quit IRC | 22:54 | |
| jgriffith | Any tempest folks still around? | 22:55 |
| *** bencherian has quit IRC | 23:00 | |
| *** adjohn has joined #openstack-dev | 23:00 | |
| *** bencherian has joined #openstack-dev | 23:01 | |
| *** anderstj has quit IRC | 23:02 | |
| *** bencherian has quit IRC | 23:02 | |
| *** adjohn has quit IRC | 23:02 | |
| *** dtroyer is now known as dtroyer_zzz | 23:04 | |
| *** blamar has quit IRC | 23:08 | |
| *** edygarcia has quit IRC | 23:15 | |
| *** salv-orlando has joined #openstack-dev | 23:19 | |
| *** jgriffith is now known as jgriffith_away | 23:21 | |
| *** dtroyer_zzz is now known as dtroyer | 23:21 | |
| *** dmarkey has quit IRC | 23:26 | |
| *** dmarkey has joined #openstack-dev | 23:27 | |
| *** danwent has joined #openstack-dev | 23:29 | |
| *** mnewby has quit IRC | 23:30 | |
| *** anniec has joined #openstack-dev | 23:30 | |
| *** mnewby has joined #openstack-dev | 23:30 | |
| *** armaan has joined #openstack-dev | 23:31 | |
| *** armaan has left #openstack-dev | 23:31 | |
| *** arosen has quit IRC | 23:33 | |
| *** adjohn has joined #openstack-dev | 23:33 | |
| *** danwent has quit IRC | 23:37 | |
| *** mnewby has quit IRC | 23:40 | |
| *** steveb_ has joined #openstack-dev | 23:41 | |
| *** issackelly has quit IRC | 23:47 | |
| *** ecarlin_ has joined #openstack-dev | 23:48 | |
| jeblair | jgriffith_away: lots of progress; it looks like cinder passed the devstack exercises, but failed one of the tempest tests. | 23:48 |
| *** arosen has joined #openstack-dev | 23:49 | |
| *** bencherian has joined #openstack-dev | 23:51 | |
| *** ecarlin has quit IRC | 23:52 | |
| *** markmcclain has joined #openstack-dev | 23:52 | |
| *** markmcclain has quit IRC | 23:52 | |
| *** ecarlin_ has quit IRC | 23:52 | |
| *** sacharya has joined #openstack-dev | 23:54 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!