| *** dragondm has joined #openstack-dev | 00:04 | |
| openstackgerrit | Liem Nguyen proposed a change to openstack/keystone: x.509 client authentication https://review.openstack.org/1038 | 00:13 |
|---|---|---|
| *** rods has quit IRC | 00:19 | |
| *** adjohn has joined #openstack-dev | 00:24 | |
| *** Turicas has quit IRC | 00:25 | |
| *** dragondm has quit IRC | 00:35 | |
| *** hugokuo has joined #openstack-dev | 00:40 | |
| *** adjohn has quit IRC | 01:10 | |
| *** ChanServ sets mode: +v _cerberus_ | 01:21 | |
| *** tsuzuki_ has joined #openstack-dev | 01:37 | |
| *** lorin1 has left #openstack-dev | 02:34 | |
| *** lorin1 has quit IRC | 02:34 | |
| *** nati2 has joined #openstack-dev | 02:38 | |
| *** nati2 has quit IRC | 02:46 | |
| *** adjohn has joined #openstack-dev | 02:54 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Replace socat with netcat. https://review.openstack.org/729 | 02:54 |
| *** mattray has joined #openstack-dev | 03:02 | |
| *** mattray has quit IRC | 03:02 | |
| *** pyhole has quit IRC | 03:18 | |
| *** pyhole has joined #openstack-dev | 03:19 | |
| *** dolphm has joined #openstack-dev | 03:28 | |
| *** dolphm has quit IRC | 03:32 | |
| *** dolphm has joined #openstack-dev | 03:37 | |
| *** nati2 has joined #openstack-dev | 03:46 | |
| *** sdake has joined #openstack-dev | 03:51 | |
| *** chmouel has quit IRC | 04:11 | |
| *** chmouel has joined #openstack-dev | 04:11 | |
| *** adjohn has quit IRC | 04:11 | |
| openstackgerrit | Naveed Massjouni proposed a change to openstack/nova: Adding support for retrying glance image downloads. https://review.openstack.org/1017 | 04:25 |
| *** sdake has quit IRC | 04:40 | |
| ironcamel | Why does my reply to a gerrit comment remain as (DRAFT)? | 04:42 |
| ironcamel | so annoying | 04:42 |
| *** dolphm has quit IRC | 04:57 | |
| *** ben_ has joined #openstack-dev | 05:22 | |
| *** ben_ has left #openstack-dev | 05:23 | |
| *** nati2 has quit IRC | 05:37 | |
| *** sdake has joined #openstack-dev | 05:38 | |
| *** sdake has quit IRC | 05:44 | |
| *** timr has quit IRC | 05:46 | |
| *** Arminder has joined #openstack-dev | 05:47 | |
| *** tsuzuki_ has quit IRC | 05:58 | |
| *** timr has joined #openstack-dev | 06:03 | |
| *** nati2 has joined #openstack-dev | 06:34 | |
| *** tsuzuki_ has joined #openstack-dev | 06:37 | |
| *** sannes1 has joined #openstack-dev | 06:45 | |
| openstackgerrit | Razique Mahroua proposed a change to openstack/compute-api: Fix bug 845012 https://review.openstack.org/1039 | 07:25 |
| uvirtbot | Launchpad bug 845012 in openstack-manuals "OSAPI Doc for Update Server needs to be updated" [Undecided,New] https://launchpad.net/bugs/845012 | 07:25 |
| *** martines has quit IRC | 07:27 | |
| *** martines has joined #openstack-dev | 07:28 | |
| *** nati2 has quit IRC | 07:28 | |
| openstackgerrit | Razique Mahroua proposed a change to openstack/openstack-manuals: Update Swift doc : # Fix bug 791584 # Update formatting https://review.openstack.org/1040 | 07:56 |
| uvirtbot | Launchpad bug 791584 in openstack-manuals "Documentation missing detail" [Low,In progress] https://launchpad.net/bugs/791584 | 07:56 |
| *** winston-d has left #openstack-dev | 08:00 | |
| *** winston-d has joined #openstack-dev | 08:00 | |
| *** rbasak has joined #openstack-dev | 08:01 | |
| *** tsuzuki_ has quit IRC | 08:04 | |
| *** tsuzuki_ has joined #openstack-dev | 08:06 | |
| openstackgerrit | Razique Mahroua proposed a change to openstack/openstack-manuals: Fix bug 869091 https://review.openstack.org/1041 | 08:14 |
| uvirtbot | Launchpad bug 869091 in openstack-manuals "Broken link on openstack-compute/admin/content/configuring-the-vnc-proxy.html" [Undecided,Fix committed] https://launchpad.net/bugs/869091 | 08:14 |
| *** Tribaal|cooking is now known as Tribaal | 08:17 | |
| *** dillon-w has joined #openstack-dev | 08:22 | |
| *** corrigac_ has quit IRC | 08:26 | |
| *** darraghb has joined #openstack-dev | 08:52 | |
| *** jeffjapan has quit IRC | 08:52 | |
| *** Oneiroi has joined #openstack-dev | 08:55 | |
| *** dillon-w has quit IRC | 09:02 | |
| *** tsuzuki_ has quit IRC | 09:05 | |
| *** viraptor has joined #openstack-dev | 09:07 | |
| openstackgerrit | Razique Mahroua proposed a change to openstack/image-api: Update Object Storage documentation # Fixes bug 854159 # Update formatting https://review.openstack.org/1042 | 09:29 |
| uvirtbot | Launchpad bug 854159 in openstack-manuals "suggestions re OpenStack Object Storage Developer Manual" [Undecided,New] https://launchpad.net/bugs/854159 | 09:29 |
| *** hernad has joined #openstack-dev | 09:52 | |
| *** rods has joined #openstack-dev | 10:03 | |
| *** tsuzuki_ has joined #openstack-dev | 10:29 | |
| *** tomize has left #openstack-dev | 10:36 | |
| *** tomize has joined #openstack-dev | 10:36 | |
| Oneiroi | I've put up a script here: https://github.com/Oneiroi/openstack-debug to gather debug data and push to paste.openstack.org with anyluck it will help people to suppor tothers by getting info quickly, this is a 5 minute script and only grabs nov adata at the moment will add more later | 10:41 |
| zykes- | just for nova Oneiroi | 10:41 |
| Oneiroi | zykes-: yes at the moment | 10:42 |
| zykes- | i can help and patch it if you need to grab keystone stuff etc as well | 10:42 |
| Oneiroi | I will add more later :) | 10:42 |
| Oneiroi | awesome by all means get a pull request on github and I'll pull it in | 10:42 |
| Oneiroi | anyway kids to feed now I will bbl | 10:42 |
| *** Oneiroi has quit IRC | 10:42 | |
| *** jiva has quit IRC | 10:46 | |
| *** jiva has joined #openstack-dev | 10:47 | |
| *** mmetheny_ has joined #openstack-dev | 10:50 | |
| *** mmetheny has quit IRC | 10:50 | |
| *** mmetheny_ is now known as mmetheny | 10:50 | |
| openstackgerrit | Soren Hansen proposed a change to openstack/nova: Extend test_virt_driver to also test libvirt driver. https://review.openstack.org/1043 | 11:06 |
| zykes- | hey soren, got some time ? | 11:06 |
| soren | zykes-: On my way to lunch, actually. | 11:07 |
| zykes- | ah, i got an error where compute fails on run_instance | 11:07 |
| zykes- | http://paste.openstack.org/show/2829/ | 11:07 |
| zykes- | when you get time | 11:07 |
| openstackgerrit | aababilov proposed a change to openstack/keystone: Fix tenant_id for Role https://review.openstack.org/1044 | 11:13 |
| *** tsuzuki_ has quit IRC | 11:54 | |
| soren | zykes-: It's hard to tell what's going on there. It would seem like there's a bug in the exception logging code that ends up raising another exception, so we never get to see the real exception. | 12:00 |
| zykes- | suggestion to what to do ? | 12:02 |
| *** markvoelker has joined #openstack-dev | 12:03 | |
| soren | zykes-: Working on it, hang on. | 12:04 |
| soren | zykes-: Which version is this? | 12:04 |
| zykes- | 2011.3 | 12:05 |
| zykes- | in other words diablo from Griddynamics packages | 12:05 |
| soren | Oh. | 12:05 |
| zykes- | with keystone trunk | 12:06 |
| zykes- | since i couldn't get other stuff to work with stabl/diablo release of it | 12:06 |
| Kiall | zykes-: If its of any use .. I have stable/diablo .deb's for ubuntu oneiric of everything bar swift @ https://launchpad.net/~managedit/+archive/openstack .. | 12:09 |
| zykes- | Kiall: i'm running @ EL6 | 12:09 |
| zykes- | so no debs :( | 12:10 |
| Kiall | Ah - So useless for you ;) | 12:10 |
| zykes- | ;) | 12:10 |
| zykes- | atm yeah, not wanting to start @ scratch again | 12:10 |
| zykes- | i'm like 1 cm from completing a demo | 12:10 |
| Kiall | lol .. | 12:10 |
| Kiall | Well .. It took me 1.5 hours to deploy 3 nodes with those packages.. and most of that was driving to the datacenter ;) | 12:11 |
| soren | zykes-: Can I see your nova.conf, please? | 12:12 |
| zykes- | sure | 12:13 |
| zykes- | http://paste.pocoo.org/show/497306/ | 12:14 |
| soren | Ah. | 12:14 |
| soren | It's your own fault. :) | 12:14 |
| soren | --logging_context_format_string=%(asctime)s %(name)s: %(levelname)s [%(request_id)s %(user)s %(project)s] %(message)s | 12:15 |
| soren | There's no such thing as ½(user)s | 12:15 |
| *** zul has quit IRC | 12:15 | |
| zykes- | kinda, but i'm using standard config, haven't changed that :) | 12:16 |
| soren | zykes-: Then complain to your distributor. It's not standard OpenStack config. | 12:19 |
| soren | A standard nova.conf is an empty one. | 12:20 |
| zykes- | hehe :) | 12:21 |
| zykes- | it might be kinda my own fault, since i use newer keystone version then the one they package | 12:21 |
| *** nhubbard_away is now known as nhubbard | 12:29 | |
| sannes1 | If I have a qcow2 image working with qemu-kvm, can I use this directly with openstack or does it require additional magic? | 12:40 |
| sannes1 | ops, wrong chan | 12:40 |
| *** lts has joined #openstack-dev | 12:40 | |
| *** lorin1 has joined #openstack-dev | 12:42 | |
| *** crobinso has joined #openstack-dev | 12:52 | |
| Kiall | review.openstack.org down for anyone else? | 12:52 |
| *** joesavak has joined #openstack-dev | 12:53 | |
| Kiall | humm .. its back now | 12:53 |
| *** cmagina has joined #openstack-dev | 12:53 | |
| *** dillon-w has joined #openstack-dev | 12:54 | |
| *** markmc has joined #openstack-dev | 12:58 | |
| *** popux has joined #openstack-dev | 13:03 | |
| soren | zykes-: No comment :) | 13:05 |
| zykes- | any dashboard dev around here ? | 13:08 |
| *** dolphm has joined #openstack-dev | 13:09 | |
| *** kbringard has joined #openstack-dev | 13:17 | |
| *** wwkeyboard has joined #openstack-dev | 13:20 | |
| *** jsavak has joined #openstack-dev | 13:21 | |
| *** kbringard has quit IRC | 13:24 | |
| *** kbringard has joined #openstack-dev | 13:25 | |
| *** joesavak has quit IRC | 13:25 | |
| *** dolphm has quit IRC | 13:26 | |
| *** dprince has joined #openstack-dev | 13:33 | |
| *** dillon-w has quit IRC | 13:34 | |
| dprince | jeblair: Are the jenkins slaves offline this morning? | 13:34 |
| jeblair | dprince: lemme check | 13:45 |
| *** kbringard has left #openstack-dev | 13:45 | |
| *** kbringard has joined #openstack-dev | 13:45 | |
| *** dolphm has joined #openstack-dev | 13:47 | |
| jeblair | dprince: it looks like mtaylor moved the nova job to the new pip-based builder | 13:47 |
| jeblair | dprince: i think that may not have worked as well as he hoped | 13:47 |
| dprince | jeblair: Okay. Thanks for taking a look. | 13:50 |
| *** kbringard has quit IRC | 13:50 | |
| jeblair | dprince: we should be able to get it fixed in a few mins | 13:54 |
| *** joesavak has joined #openstack-dev | 13:55 | |
| *** jsavak has quit IRC | 13:55 | |
| *** kbringard has joined #openstack-dev | 13:56 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Raise InsufficientFreeMemory https://review.openstack.org/953 | 13:56 |
| openstackgerrit | Verification of a change to openstack/nova failed: Snapshots/backups can no longer happen simultaneously. Tests included. https://review.openstack.org/955 | 13:57 |
| openstackgerrit | Verification of a change to openstack/nova failed: bug 861310 https://review.openstack.org/957 | 13:58 |
| uvirtbot | Launchpad bug 861310 in nova "Scheduler hangs if filtered host list is empty." [Undecided,Fix committed] https://launchpad.net/bugs/861310 | 13:58 |
| *** Oneiroi has joined #openstack-dev | 13:58 | |
| *** Oneiroi has quit IRC | 14:03 | |
| dprince | jeblair: another question for you. How might I go about getting an SSH key on the gerrit box that lets me run queries only (read access). | 14:03 |
| *** Oneiroi has joined #openstack-dev | 14:03 | |
| jeblair | dprince: email me the public key and the username you want and i'll set it up | 14:03 |
| zykes- | soren: is there any way that i can terminate a "pending" instance? euca-terminate-instances <id> doesn't budge | 14:04 |
| *** dolphm has quit IRC | 14:05 | |
| dprince | jeblair: Excellent. Thank you sir. | 14:05 |
| *** sannes1 has quit IRC | 14:06 | |
| *** dragondm has joined #openstack-dev | 14:14 | |
| *** cp16net has joined #openstack-dev | 14:16 | |
| *** vladimir3p has joined #openstack-dev | 14:17 | |
| *** mdomsch has joined #openstack-dev | 14:18 | |
| soren | zykes-: That should be the right way to do it. If it doesn't work, that's a bug. | 14:19 |
| openstackgerrit | A change was merged to openstack/nova: Raise InsufficientFreeMemory https://review.openstack.org/953 | 14:20 |
| soren | jeblair: I had a change that I "had to" abandon. After that, whenever I tried to push to the branch (wanting to revive it), it was rejected. | 14:20 |
| jeblair | soren: use the web interface to "restore" the change | 14:21 |
| soren | jeblair: A couple of questions: I abandoned it, because I knew it was broken, and I didn't want anyone wasting time reviewing it. I couldn't find any useful way to mark the change as work in progress, so I abandoned it. | 14:21 |
| soren | jeblair: Where? | 14:21 |
| *** sdake has joined #openstack-dev | 14:21 | |
| jeblair | soren: same place as abandon | 14:21 |
| jeblair | soren: we'd like to add a WIP state to gerrit. for now, i think what you did is the closest thing we have | 14:22 |
| soren | I forget where that was :) | 14:22 |
| soren | Oh. | 14:22 |
| soren | I see it now. | 14:22 |
| jeblair | soren: go to the change in gerrit, and it's a button where the 'review' button normally is | 14:23 |
| jeblair | ok | 14:23 |
| soren | Hm... Yeah, I guess I can do that. What I did instead was to remove the Change-Id from the commit so that it would create a new one, so now it's registered as a different change. | 14:23 |
| zykes- | soren: http://pastebin.com/5E3JFNxh < shall i make a bug of it ? | 14:23 |
| jeblair | that works too | 14:23 |
| soren | I just lose some history. | 14:23 |
| soren | There wasn't really any to speak of anyway, so meh. | 14:23 |
| soren | But good to know for next time. | 14:24 |
| soren | jeblair: Thanks. | 14:24 |
| jeblair | soren: you're welcome | 14:24 |
| Kiall | zykes-: it looks like you didnt tail enough lines .. the error is not show in that pastebin | 14:24 |
| Kiall | It'll most likely be in nova-compute.log .. | 14:24 |
| zykes- | Kiall: there is no error. that's the funny part | 14:24 |
| Kiall | weird .. is nova-compute actually running? | 14:25 |
| soren | Yeah, it receives the request. | 14:25 |
| zykes- | it was "instances" that was stuck in pending because of a log format error in nova.conf earlier, i fixed that and now i can't cleanup the pending instancse | 14:25 |
| Kiall | soren: as yes .. it does.. | 14:26 |
| *** tudamp_ has quit IRC | 14:26 | |
| Oneiroi | markmc: el6 branch no longer has .init scripts :S | 14:27 |
| Oneiroi | for glance that is checking others | 14:27 |
| markmc | hmm | 14:27 |
| zykes- | Oneiroi: .init ? | 14:27 |
| * markmc looks | 14:27 | |
| zykes- | want me to file a bug soren ? | 14:28 |
| Oneiroi | zykes-: /etc/init.d/openstack-glance etc ... | 14:28 |
| zykes- | Oneiroi: what else does it have? | 14:29 |
| markmc | Oneiroi, my bad, sorry - fixed now | 14:30 |
| Oneiroi | zykes-: fedora uses a different system, it has a service specification instead | 14:30 |
| Oneiroi | markmc: ^_^ cool keystone and nova look ok will put a build through shortly, is there an extra step I need to take to start getting these into epel-testing ? | 14:31 |
| markmc | Oneiroi, yes | 14:31 |
| openstackgerrit | Verification of a change to openstack/nova failed: Raise InsufficientFreeMemory https://review.openstack.org/953 | 14:32 |
| openstackgerrit | A change to openstack/nova has been rejected: Deallocate ip if build fails. https://review.openstack.org/999 | 14:34 |
| markmc | Oneiroi, was waiting for the follow up question for comic effect ... :) | 14:34 |
| markmc | Oneiroi, no idea what that step is | 14:34 |
| openstackgerrit | A change to openstack/nova has been rejected: Ensure non-default FLAGS.logfile_mode is properly converted to an octet. https://review.openstack.org/997 | 14:34 |
| Oneiroi | lol sorry was pushing a build | 14:34 |
| markmc | Oneiroi, maybe "fedpkg update" just works? | 14:34 |
| Oneiroi | *shrug* | 14:34 |
| Oneiroi | will try it in a sec | 14:35 |
| openstackgerrit | A change to openstack/nova has been rejected: Enforce snapshot cleanup. https://review.openstack.org/1000 | 14:35 |
| soren | zykes-: Yes, please. | 14:35 |
| zykes- | it's done | 14:37 |
| soren | markmc: Looking at https://review.openstack.org/#change,960 I wonder how we should handle this in the general case. | 14:41 |
| Oneiroi | markmc: spec file for glance still using .service install lines | 14:41 |
| openstackgerrit | Verification of a change to openstack/nova failed: Snapshots/backups can no longer happen simultaneously. Tests included. https://review.openstack.org/955 | 14:41 |
| markmc | soren, how do you mean? | 14:41 |
| *** nhubbard is now known as nhubbard_away | 14:41 | |
| soren | Having unversioned dependencies in pip-requires is fine during development. We want to keep up with the latest stuff, but when we cut a release, it seems like a good idea to have a known working set of dependencies listed in there. | 14:42 |
| markmc | oh, right | 14:42 |
| soren | Otherwise, the day after release, our release might not work anymore. | 14:42 |
| *** rnirmal has joined #openstack-dev | 14:43 | |
| soren | It's at everyone else's whim. That sounds bad. | 14:43 |
| soren | :) | 14:43 |
| markmc | what I don't like about versioned deps is that they get stale quickly | 14:43 |
| markmc | but what you say makes sense | 14:43 |
| markmc | maybe version the dependencies after RBP | 14:43 |
| soren | Possibly. | 14:43 |
| markmc | but during development, all should be unversioned unless there's a good reason | 14:44 |
| soren | HEre's a thought: | 14:44 |
| soren | When we construct the tarball, we could detect what the current version is of all those things, and embed the specific versions in the pip-requires found in the tarball. | 14:45 |
| dprince | jeblair: anything I need to do to resend the FlavorID branch? (review 859) | 14:45 |
| *** med_out is now known as medberry | 14:45 | |
| soren | All automated QA at least used to use the tarball rather than an SCM checkout. | 14:45 |
| soren | I don't know if that's still the case, but it should be. | 14:45 |
| *** rnirmal has quit IRC | 14:45 | |
| Oneiroi | markmc: yeh looks like fedpgk update does the push into bohdi | 14:45 |
| markmc | soren, you wouldn't want to regenerate when doing a release from the stable branch later | 14:45 |
| *** rnirmal has joined #openstack-dev | 14:46 | |
| soren | markmc: Good point. No, I wouldn't. | 14:46 |
| Kiall | BTW, Should this make it into diablo/stable? https://review.openstack.org/#change,628 | 14:46 |
| soren | markmc: Yeah, very good point. A simple script someone could run at will to version all these things would probably be a better approach. | 14:47 |
| Kiall | (It fixes https://bugs.launchpad.net/nova/+bug/834633 as near as I can tell) | 14:47 |
| uvirtbot | Launchpad bug 834633 in nova "--auto_assign_floating_ip=true crashes on logout" [Medium,In progress] | 14:47 |
| jeblair | dprince: you can +2 the review again | 14:47 |
| soren | Kiall: I'd be very, very hesitant to do that. | 14:47 |
| dprince | jeblair: will do. Thanks | 14:47 |
| jeblair | dprince: or you can trigger in jenkins: https://jenkins.openstack.org/gerrit_manual_trigger/ | 14:47 |
| Kiall | soren, would a simpler fix for the bug make it in? If so, I can sort that out .. | 14:48 |
| soren | Kiall: IT's one thing that upgrading from one release to another is error prone. I don't want to move code between logical components for a point release. Being able to do rolling upgrades for those should really be a requirement, IMO. | 14:48 |
| markmc | Kiall, best bet would be to work with trey to figure out what a minimal fix might be | 14:49 |
| openstackgerrit | Verification of a change to openstack/nova failed: bug 861310 https://review.openstack.org/957 | 14:49 |
| uvirtbot | Launchpad bug 861310 in nova "Scheduler hangs if filtered host list is empty." [Undecided,Fix committed] https://launchpad.net/bugs/861310 | 14:49 |
| jeblair | dprince: gerrit account created | 14:49 |
| Kiall | markmc: is trey typically on IRC? | 14:50 |
| markmc | Oneiroi, hmm, sorry about the screwup - I'll fix | 14:50 |
| dprince | jeblair: thanks. | 14:52 |
| * markmc points Kiall at tr3buchet | 14:52 | |
| markmc | Oneiroi, ok, I think it might be finally fixed | 14:53 |
| Oneiroi | cool :) | 14:54 |
| *** rkukura has quit IRC | 14:54 | |
| Oneiroi | also: https://admin.fedoraproject.org/updates/openstack-nova-2011.3-6.el6 :) | 14:54 |
| *** dak has joined #openstack-dev | 14:55 | |
| *** dubsquared has joined #openstack-dev | 14:55 | |
| markmc | Oneiroi, great stuff, kudos | 14:57 |
| Oneiroi | markmc: pulling the rpm onto a 6.0 vm now, give it a test run | 14:57 |
| markmc | Oneiroi, cool. is that libvirt 0.8.7? | 14:58 |
| Oneiroi | markmc: 0.8.1-27 it would seem | 14:59 |
| markmc | Oneiroi, ok; curious to hear how you get on | 14:59 |
| markmc | Oneiroi, the diablo release PPAs have 0.8.8 backported, so that might be the minimum required | 15:00 |
| *** dubsquared has quit IRC | 15:02 | |
| openstackgerrit | James E. Blair proposed a change to openstack/glance: Add .gitreview config file for gerrit. https://review.openstack.org/1045 | 15:02 |
| openstackgerrit | James E. Blair proposed a change to openstack/identity-api: Add .gitreview config file for gerrit. https://review.openstack.org/1046 | 15:02 |
| openstackgerrit | James E. Blair proposed a change to openstack/image-api: Add .gitreview config file for gerrit. https://review.openstack.org/1047 | 15:02 |
| annegentle | jeblair: or mtaylor: how can I set up a Diablo branch for the openstack-manuals project? | 15:02 |
| jeblair | annegentle: just a min. | 15:03 |
| annegentle | k, thanks | 15:03 |
| *** code_franco has joined #openstack-dev | 15:04 | |
| *** jdag has quit IRC | 15:05 | |
| tr3buchet | Kiall: how can i help | 15:05 |
| *** dubsquared has joined #openstack-dev | 15:06 | |
| openstackgerrit | A change was merged to openstack/keystone: updating keystone developer documentation updating docstrings to remove errors in automodule generation updating setup.py to generate source documentation blueprint keystone-documentation bug 843056 https://review.openstack.org/1013 | 15:08 |
| uvirtbot | Launchpad bug 843056 in keystone "Design Documentation" [High,Confirmed] https://launchpad.net/bugs/843056 | 15:08 |
| *** kbringard has quit IRC | 15:08 | |
| *** uncleofthestick has joined #openstack-dev | 15:08 | |
| jeblair | annegentle: currently os-doc-core can +2 reviews in openstack-manuals, do you anticipate anyone else needing +2 ability for the diablo branch? for instance, the stable release maintainers can additionally +2 patches to the stable/diablo branches of the projects themselves. | 15:09 |
| annegentle | ah, good thinking. Right now I think the groups should be identical but if I identify a stable release maintainer it's good to know we can add another. | 15:09 |
| Kiall | tr3buchet: https://review.openstack.org/#change,628 | 15:11 |
| Kiall | that looks like it fixes https://bugs.launchpad.net/nova/+bug/834633 | 15:11 |
| uvirtbot | Launchpad bug 834633 in nova "--auto_assign_floating_ip=true crashes on logout" [Medium,In progress] | 15:11 |
| Kiall | tr3buchet: I was wanting to try and get that part of the review merged to diablo/stable .. | 15:12 |
| *** kbringard has joined #openstack-dev | 15:12 | |
| jeblair | annegentle: is it okay if the stable maint team also having +2 on your diablo branch now? or do you want exclusively doc-core? | 15:19 |
| *** nhubbard_away has quit IRC | 15:19 | |
| annegentle | jeblair: the more stability pursuers, the merrier. Add 'em. | 15:19 |
| *** nhubbard_away has joined #openstack-dev | 15:19 | |
| *** nhubbard_away is now known as nhubbard | 15:19 | |
| jeblair | annegentle: where do you want the branch to start from? | 15:20 |
| jeblair | annegentle: the current head or some other commit? | 15:20 |
| annegentle | jeblair: we can start from HEAD, it's all clean | 15:20 |
| jeblair | annegentle: okay, there's a stable/diablo branch now | 15:21 |
| annegentle | jeblair: ok, great. Do you have instructions for specifically sending branches for review to it? | 15:21 |
| jeblair | annegentle: you almost certainly want to use the new git-review for that (I think you said you are) | 15:21 |
| annegentle | (assuming that the "normal workflow" is for HEAD) | 15:21 |
| annegentle | jeblair: yep, new git review is in place | 15:22 |
| annegentle | so all my collaborators should also use the new git-review as well? | 15:22 |
| jeblair | annegentle: 'git review stable/diablo' | 15:22 |
| annegentle | ok thanks, updating my instructions on the wiki then | 15:23 |
| jeblair | other than that, it should be the same (except instead of checkout master, 'checkout stable/diablo') | 15:23 |
| annegentle | jeblair: and then do we just tag bugs "diablo" or some such? | 15:23 |
| Oneiroi | markmc: some odd requirements nova requires glance, any idea as to why they are supposed to be independant | 15:23 |
| dprince | jeblair: looks like tests passed but the flavors branch still doesn't land: https://jenkins.openstack.org/job/nova/121707/console | 15:23 |
| jeblair | annegentle: http://wiki.openstack.org/StableBranch | 15:24 |
| jeblair | annegentle: that's the work in progress for maintaining the stable code branches | 15:24 |
| jeblair | annegentle: may have useful info for you as well abotu bugs | 15:24 |
| markmc | Oneiroi, hmm, yeah - that came through from the original grid dynamics packaging | 15:24 |
| markmc | Oneiroi, I guess the idea was "yum install openstack-nova" would get you everything | 15:24 |
| markmc | Oneiroi, agree we could kill it; could you file a bug? | 15:25 |
| Oneiroi | indeed, but can't do that atm | 15:25 |
| Oneiroi | yeh will do | 15:25 |
| jeblair | dprince: hrm | 15:25 |
| annegentle | jeblair: thanks, just what I was looking for. Not sure if I need to track all doc changes in a bug? | 15:26 |
| annegentle | jeblair: off the top of your head, is there a wiki page explaining git-review (how to install it, etc)? | 15:27 |
| annegentle | just one more note, the StableBranch page is quite Nova-specific right now. What about Keystone? | 15:27 |
| jeblair | annegentle: probably don't need to track all changes in a bug, their needs are probably a bit different to yours | 15:28 |
| jeblair | annegentle: the docs here are it for now i think: https://github.com/openstack-ci/git-review | 15:29 |
| jeblair | annegentle: they're starting with nova, they'll add others as they get going | 15:29 |
| annegentle | jeblair: ok on all three, thank you very much :) As I read more, glance is in there too. Just didn't know if Keystone is hoping to do the same? dolph or yogi or zns may want to take a look. | 15:30 |
| jeblair | dprince: i'm asking mtaylor to look into that | 15:33 |
| jeblair | mtaylor: ^ | 15:34 |
| *** adjohn has joined #openstack-dev | 15:34 | |
| *** dolphm has joined #openstack-dev | 15:35 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Convert instancetype.flavorid to string https://review.openstack.org/859 | 15:37 |
| mtaylor | dprince: ignore that above | 15:38 |
| annegentle | jeblair: oh one more question. How do I reconfigure this Job: https://jenkins.openstack.org/view/Openstack-manuals/job/openstack-admin-manuals-diablo/ on Jenkins to go to stable/diablo? I guess there's some incantation under "Branches to build?" | 15:41 |
| openstackgerrit | A change was merged to openstack/swift: Fixing bug with x-trans-id. Will now be set on all incoming requests to proxy and trans-ids will not be reused. https://review.openstack.org/866 | 15:41 |
| jeblair | annegentle: yes, you'll want to set that to stable/diable, and then change the job for trunk to only build branch "master" | 15:44 |
| annegentle | ok thanks | 15:44 |
| openstackgerrit | A change was merged to openstack/nova: Convert instancetype.flavorid to string https://review.openstack.org/859 | 15:45 |
| Oneiroi | hmmmm markmc: openstack-nova required libvirt >= 0.82 and for the moment I can only get 0.81-72 :-/ going to have a hunt around | 15:48 |
| Oneiroi | not sure I want to remove the requirement in the spec _just incase_ | 15:48 |
| mtaylor | dprince: merged | 15:49 |
| markmc | Oneiroi, I don't know where the >= 0.82 requirement came from; remove it temporarily, do a scratch build and see what breaks with older libvirt | 15:49 |
| Oneiroi | markmc: ah yeh that'd make sense :) | 15:49 |
| markmc | Oneiroi, although, it came from grid dynamics who did RHEL packaging initially ... | 15:49 |
| markmc | Oneiroi, so it might just mean "RHEL 6.0 doesn't work" | 15:49 |
| Oneiroi | :-( | 15:50 |
| openstackjenkins | Project nova-docs build #1,519: SUCCESS in 4 min 37 sec: https://jenkins.openstack.org/job/nova-docs/1519/ | 15:51 |
| openstackjenkins | brian.waldon: Convert instancetype.flavorid to string | 15:51 |
| openstackgerrit | aababilov proposed a change to openstack/keystone: Fix tenant_id for Role https://review.openstack.org/1048 | 15:54 |
| zykes- | soren: looked at my bug ? :p | 15:57 |
| openstackgerrit | A change to openstack/keystone has been rejected: Fix tenant_id for Role https://review.openstack.org/1048 | 15:57 |
| openstackgerrit | aababilov proposed a change to openstack/keystone: Fix tenant_id for Role https://review.openstack.org/1044 | 15:57 |
| *** popux has quit IRC | 16:08 | |
| *** adjohn has quit IRC | 16:08 | |
| *** mnour has joined #openstack-dev | 16:14 | |
| openstackgerrit | A change was merged to openstack/nova: Upgrade pip in the venv when we build it. https://review.openstack.org/972 | 16:16 |
| openstackgerrit | Brian Waldon proposed a change to openstack/glance: Removing duplicate mox install in pip-requires https://review.openstack.org/1049 | 16:17 |
| *** rnorwood has joined #openstack-dev | 16:20 | |
| Oneiroi | markmc: http://paste.openstack.org/show/2839/ < so I should raise that as a bug and fix it yeh? or just fix it ? | 16:20 |
| openstackjenkins | Project nova-docs build #1,520: SUCCESS in 4 min 22 sec: https://jenkins.openstack.org/job/nova-docs/1520/ | 16:20 |
| openstackjenkins | Monty Taylor: Upgrade pip in the venv when we build it. | 16:20 |
| openstackgerrit | A change was merged to openstack/nova: Refactoring nova.tests.api.openstack.test_flavors https://review.openstack.org/908 | 16:24 |
| *** nati2 has joined #openstack-dev | 16:25 | |
| markmc | Oneiroi, which one? the class not found or the ENOENT? | 16:27 |
| markmc | Oneiroi, not sure what's going on there ? | 16:27 |
| Oneiroi | markmc: post install of the rpm into 6.0 get that back | 16:27 |
| markmc | Oneiroi, oh, interesting | 16:27 |
| markmc | Oneiroi, fix it if you can figure it out, file a bug otherwise | 16:28 |
| Oneiroi | indeed, seems to think that glance.client.Client could not be laoded | 16:28 |
| markmc | Oneiroi, if the fix makes sense in F16, push there first and merge into el6 | 16:28 |
| Oneiroi | loaded* | 16:28 |
| markmc | right | 16:28 |
| markmc | what's trying to load it in %post though? | 16:28 |
| Oneiroi | I _think_ it's the nova-manage db sync | 16:29 |
| markmc | ah | 16:29 |
| Oneiroi | but as it's tied into a glance class could explain why the dependency is there | 16:29 |
| markmc | well, maybe that's why the Requires: openstack-glance was there | 16:29 |
| markmc | should be Requires(post): python-glance then, I guess | 16:30 |
| zykes- | markmc: is it decided on which version of keystone one should grab to fedora ? | 16:30 |
| markmc | zykes-, the 2011.3 tag in git | 16:30 |
| zykes- | ah | 16:30 |
| openstackjenkins | Project nova-docs build #1,521: SUCCESS in 4 min 29 sec: https://jenkins.openstack.org/job/nova-docs/1521/ | 16:30 |
| openstackjenkins | brian.waldon: Refactoring nova.tests.api.openstack.test_flavors | 16:31 |
| zykes- | markmc: didn't that have some issues working with all components though ? | 16:31 |
| *** zaitcev has joined #openstack-dev | 16:31 | |
| markmc | zykes-, did it? | 16:32 |
| zykes- | just think i had some problems with it and swift markmc | 16:36 |
| zykes- | though i can test afterwards again and see | 16:36 |
| markmc | ok, thanks | 16:36 |
| * markmc hasn't tested it yet | 16:36 | |
| zykes- | dolphm: ? | 16:37 |
| Oneiroi | markmc: updated the bug, not 100% sure but looks like an upstream bug | 16:40 |
| *** timr has quit IRC | 16:40 | |
| *** jdurgin has joined #openstack-dev | 16:41 | |
| openstackgerrit | A change was merged to openstack/nova: Use UUIDs instead of IDs for OSAPI servers. https://review.openstack.org/964 | 16:44 |
| *** blamar_ has joined #openstack-dev | 16:46 | |
| dprince | 5~5~5~5~ | 16:46 |
| *** blamar has quit IRC | 16:47 | |
| *** blamar_ is now known as blamar | 16:47 | |
| zykes- | Daviey: are you here ? | 16:47 |
| Oneiroi | markmc: verfied the bug in the code, raising a launchpad bug | 16:47 |
| zykes- | Oneiroi: which bug ? | 16:48 |
| Daviey | zykes-: seems so :) | 16:48 |
| Oneiroi | zykes-: https://bugzilla.redhat.com/show_bug.cgi?id=748490 | 16:48 |
| uvirtbot | bugzilla.redhat.com bug 748490 in openstack-nova "openstack-nova dependant on openstack-glance" [Medium,Assigned] | 16:48 |
| zykes- | Daviey: Unexpected error: list indices must be integers, not str seen that before? | 16:48 |
| *** mnour has quit IRC | 16:49 | |
| Daviey | zykes-: in what component | 16:49 |
| Daviey | ? | 16:49 |
| zykes- | @ the front page when authing | 16:49 |
| larissa | zykes-: Error: "the" is not a valid command. | 16:49 |
| Oneiroi | zykes-: just posting the launchpad one hang on a sec | 16:49 |
| Daviey | zykes-: dashboard? | 16:49 |
| Oneiroi | https://bugs.launchpad.net/nova/+bug/880964 | 16:49 |
| uvirtbot | Launchpad bug 880964 in nova "nova-manage db sync requires glance" [Undecided,New] | 16:49 |
| Oneiroi | that's the one | 16:49 |
| zykes- | Daviey: correct, thought you where working on that ? ;) | 16:50 |
| Daviey | I don't think so.. | 16:50 |
| Daviey | certainly of interest tho | 16:50 |
| openstackjenkins | Project nova-docs build #1,522: SUCCESS in 4 min 41 sec: https://jenkins.openstack.org/job/nova-docs/1522/ | 16:51 |
| openstackjenkins | mark.washenberger: Use UUIDs instead of IDs for OSAPI servers. | 16:51 |
| zykes- | damned Daviey :) | 16:51 |
| zykes- | what team are you on ? | 16:51 |
| openstackgerrit | Chmouel Boudjnah proposed a change to openstack/swift: Add support for OS Auth API version 2 https://review.openstack.org/704 | 16:51 |
| Daviey | zykes-: Distro :), Ubuntu Server Team :) | 16:52 |
| zykes- | ah | 16:52 |
| openstackgerrit | A change was merged to openstack/glance: Removing duplicate mox install in pip-requires https://review.openstack.org/1049 | 16:53 |
| Daviey | zykes-: certainly interested in that bug. | 16:54 |
| zykes- | Daviey: which bug ? | 16:54 |
| *** Oneiroi has quit IRC | 16:55 | |
| Daviey | zykes-: the one you just quoted. | 16:59 |
| zykes- | ah | 16:59 |
| openstackgerrit | A change was merged to openstack/glance: Add .gitreview config file for gerrit. https://review.openstack.org/1045 | 17:00 |
| zykes- | markmc: with the package from griddynamics i get "FAILED" when doing a start on keystone, known issue? | 17:01 |
| dolphm | zykes-, can you be more specific? | 17:02 |
| *** uncleofthestick has quit IRC | 17:02 | |
| zykes- | dolphm: the init script fails @ redhat | 17:02 |
| zykes- | "Starting OpenStack Keystone: [FAILED]" | 17:03 |
| dolphm | zykes-, anything in the logs? | 17:04 |
| zykes- | nada | 17:04 |
| openstackgerrit | Verification of a change to openstack/nova failed: Make snapshots private by default https://review.openstack.org/954 | 17:05 |
| zykes- | nothing in either syslogs or keystones log | 17:05 |
| *** btorch has joined #openstack-dev | 17:08 | |
| *** bengrue has joined #openstack-dev | 17:08 | |
| btorch | dolphm: hi, got a question about the swift endpoints on keystone | 17:08 |
| zykes- | dolphm: the syntax for keystone in diablo's endpointtemplate is ? "$cmd endpointTemplates add RegionOne swift https://swift.stack.lan:8080/v1/AUTH_%tenant_id% https://swift.adm.lan:8080/ https://127.0.0.1:8080/v1/AUTH_%tenant_id% 1 0" doesn't wanna work | 17:08 |
| dolphm | btorch, i can try to answer | 17:09 |
| zykes- | markmc: funky that it just "fails" | 17:11 |
| zykes- | oh, nvm, was a permissions error | 17:11 |
| btorch | dolphm: so the swift endpoint is something like "http://swift.publicinternets.com/v1/AUTH_%tenant_id%" and the "tenant_id" is just a number that is incremented serially ? | 17:12 |
| dolphm | btorch, that depends on the backend you're using, but in the default sqlalchemy backend, yes, those are just auto incremented pk's | 17:12 |
| dolphm | the LDAP backend uses names, and there's been discussion of migrating the default sql-backend to using uuid's | 17:13 |
| btorch | btorch: what you mean by the backend ? | 17:13 |
| dolphm | btorch, keystone is intended to be more of an identity interface than an identity store - so you can point keystone to (for example) an LDAP server to act as the identity-backend | 17:14 |
| btorch | dolphm: ah ok .. that's what I was wondering cause right now using incremental numbers you are loosing the ability to migrate accounts to other clusters that may or may not belong to the same organization | 17:14 |
| *** reed has joined #openstack-dev | 17:15 | |
| dolphm | btorch, i'd like to see us using uuid's in essex, but the concern is about producing "non-user friendly" url's | 17:15 |
| *** lorin1 has quit IRC | 17:16 | |
| *** darraghb has quit IRC | 17:18 | |
| chmouel | I think account name would make more sense for swift as this would make easy to migrate between different auth server | 17:19 |
| chmouel | but it seems that keystone doesnt have a %tenant_name% template | 17:19 |
| *** adjohn has joined #openstack-dev | 17:21 | |
| zykes- | dolphm: keystone-manage endpoint <tenantname> <numerical endpoint template id> should be ok ? | 17:21 |
| btorch | chmouel: perhaps different auth servers but what about account migration ? obviously one could implement a wrapper around the whole tenant thing | 17:22 |
| zykes- | the tenant_id's in db ends up being 0 | 17:22 |
| btorch | zykes-: that would be a good option to have | 17:22 |
| *** joesavak has quit IRC | 17:23 | |
| *** novas0x2a|laptop has joined #openstack-dev | 17:23 | |
| chmouel | btorch: yeah right i meant account migration as well | 17:24 |
| zykes- | btorch: ? | 17:25 |
| openstackgerrit | Verification of a change to openstack/nova failed: Snapshots/backups can no longer happen simultaneously. Tests included. https://review.openstack.org/955 | 17:26 |
| zykes- | dolphm: keystone diablo | 17:29 |
| zykes- | when you do keystone-manage endpoint add <tenant> <endpointtemplate> is both of those numerical id's ? | 17:30 |
| zykes- | yeah, they are | 17:30 |
| zykes- | :p | 17:30 |
| dolphm | chmouel, including tenant_name is a good idea -- definitely a missing feature | 17:30 |
| * zykes- deskbashes his head | 17:31 | |
| btorch | what is this admin_url supposed to be within the endpoint ? | 17:31 |
| dolphm | zykes-, <tenant> should be the name, according to the code i'm looking at | 17:33 |
| dolphm | (trunk) | 17:33 |
| zykes- | dolphm: i trunk yeah, not diablo | 17:33 |
| zykes- | there's it's the id | 17:33 |
| dolphm | zykes-, that's awesome. | 17:34 |
| openstackgerrit | Verification of a change to openstack/nova failed: Fixed bug lp850602. Adding backing file copy operation on kvm block migration. https://review.openstack.org/956 | 17:34 |
| zykes- | now then the funny part, now after i downgraded keystone on my swift proxy and keystone node i get that my account doesn't exist anymore | 17:36 |
| dolphm | zykes-, downgraded from what to what? | 17:37 |
| zykes- | trunk > diablo in keystone | 17:37 |
| *** rkukura has joined #openstack-dev | 17:37 | |
| *** crobinso has quit IRC | 17:38 | |
| *** rkukura has quit IRC | 17:39 | |
| dolphm | I'd almost recommend sticking with trunk for now -- i don't think we've done anything backwards incompatible with diablo, and i'd like to see us release a "2011.3.1" or something | 17:40 |
| zykes- | when i use swift>> Kiall.... | 17:40 |
| zykes- | sigh, this is a mess | 17:40 |
| Kiall | humm? | 17:40 |
| zykes- | dolphm: i tried that but for some reason it breaks dashboard... | 17:40 |
| zykes- | Kiall: keystone stuff is a mess | 17:41 |
| Kiall | works perfectly for me, except for the EC2 api -_- | 17:41 |
| *** devcamcar has joined #openstack-dev | 17:42 | |
| Kiall | actually .. EC2 API works now :) | 17:42 |
| zykes- | Kiall: you using swift / glance ? | 17:42 |
| Kiall | (Cant believe i forgot to restart nova-api after editing the keystone ec2_token middleware! dooh!) | 17:43 |
| Kiall | glance, but not swift | 17:43 |
| openstackgerrit | Verification of a change to openstack/nova failed: bug 861310 https://review.openstack.org/957 | 17:43 |
| uvirtbot | Launchpad bug 861310 in nova "Scheduler hangs if filtered host list is empty." [Undecided,Fix committed] https://launchpad.net/bugs/861310 | 17:43 |
| *** rbasak has quit IRC | 17:43 | |
| zykes- | Kiall: then you can't say that "everything" works fine... | 17:43 |
| Kiall | Well, i did qualify it with "for me" ;) | 17:44 |
| btorch | how come "user list" doesn't give back the username ? (using "stable" branch) | 17:45 |
| openstackgerrit | tylesmit proposed a change to openstack/quantum: blueprint quantum-packaging https://review.openstack.org/1050 | 17:46 |
| zykes- | btorch: in keystone ? | 17:46 |
| btorch | yeah | 17:46 |
| zykes- | works fine here | 17:46 |
| zykes- | though it doesn't show name | 17:46 |
| zykes- | ah, correct yeah :p | 17:46 |
| zykes- | hence one more reason i was using essex branch | 17:47 |
| *** lorin1 has joined #openstack-dev | 17:47 | |
| zykes- | dolphm: any idea on when a fully tested version will be out ? | 17:48 |
| *** zaitcev has quit IRC | 17:49 | |
| *** martines has quit IRC | 17:49 | |
| openstackgerrit | Chmouel Boudjnah proposed a change to openstack/swift: Add support for OS Auth API version 2 https://review.openstack.org/704 | 17:50 |
| *** btorch has left #openstack-dev | 17:50 | |
| openstackgerrit | Verification of a change to openstack/nova failed: Enforce snapshot cleanup. https://review.openstack.org/958 | 17:51 |
| dolphm | zykes-, essex? lol | 17:52 |
| dolphm | or "now," depending on your standards | 17:53 |
| zykes- | ;p | 17:53 |
| Kiall | lol .. yea .. essex is a good while away by my standards ;) | 17:53 |
| zykes- | funky, global endpoints not showing in keystone diablo for my admin user | 17:55 |
| zykes- | this is entertraining | 17:55 |
| *** jtran has joined #openstack-dev | 17:55 | |
| *** viraptor has quit IRC | 17:56 | |
| zykes- | Kiall: wonder how one is supposed to get a fully working setup with keystone | 17:57 |
| *** martines has joined #openstack-dev | 17:57 | |
| jtran | hey all. in regards to roles. if i want PER PROJ roles for 'user1' and 'netadmin' role , do i have to grant them global AND per project? that doesn't seem like a lot of sense to me | 17:57 |
| Kiall | zykes-: by installing ubuntu and using my PPA ;) https://launchpad.net/~managedit/+archive/openstack | 17:58 |
| Kiall | (other than swift .. but thats next on my list to package) | 17:58 |
| dolphm | zykes-, find bug -> fix bug -> repeat until keystone "works" | 17:58 |
| zykes- | dolphm: i don't have time to fix bugs, i can report though.. | 17:59 |
| dolphm | zykes-, that's half the battle! | 17:59 |
| zykes- | all was working with latest trunk until i tried dashboard :/ | 17:59 |
| dolphm | keystone should work with dashboard | 18:00 |
| Kiall | I've only had to fix 1 bug in keystone .. https://github.com/managedit/keystone/blob/master/debian/patches/fix-keystone-ec2.patch | 18:00 |
| Kiall | No other hacks were needed for it, but, I havent gone near swift | 18:00 |
| *** bcwaldon has joined #openstack-dev | 18:00 | |
| zykes- | Kiall: so in other words you haven't check that everything works ? | 18:00 |
| dolphm | zykes-, no one has :) | 18:01 |
| Kiall | yes, everything I've tested appears to be working. | 18:01 |
| Kiall | any pointers to stuff you think is broken? | 18:01 |
| zykes- | Kiall: i'm not testing on ubuntu so dunno with your packages | 18:03 |
| Kiall | I know, I meant broken in keystone rather than packages | 18:03 |
| zykes- | Kiall: for one i used keystone trunk earlier, then with swift when i did stat -v it worked fine, now i just get account not found | 18:05 |
| Kiall | that doesnt sound like a bug, more a "whoops i messed up the db" (I could be wrong .. I've not touched the master branches since they are inherently unstable) | 18:06 |
| rmk | keystone itself is really not the problem so far as I have seen | 18:06 |
| rmk | Does anyone have nova/keystone working with the EC2 api? | 18:07 |
| rmk | That's the main piece I am missing at the moment. | 18:07 |
| rmk | And of course glance doesn't use the OS_ environment variables for auth. | 18:08 |
| sandywalsh | dolphm, ping | 18:11 |
| *** joesavak has joined #openstack-dev | 18:11 | |
| sandywalsh | joesavak, dolphm, GET tokens/xxxx/endpoints is giving me an unauthorized now ... I see there were some changes to the code, have the semantics changed? | 18:11 |
| Kiall | rmk: yes, i do | 18:12 |
| rmk | Kiall: It could just be a function of adding credentials to keystone in my case | 18:12 |
| Kiall | rmk: devstack has a keystone_data.sh file that shows how to add the credentials to keystone .. | 18:13 |
| rmk | cheers I'll take a look | 18:13 |
| rmk | Kiall: Did you patch glance in your PPA to fix the keystone auth issues? | 18:14 |
| Kiall | and if your running the diablo branch, you need this patch .. https://github.com/managedit/keystone/blob/master/debian/patches/fix-keystone-ec2.patch | 18:14 |
| Kiall | rmk: havent ran into any auth issues? | 18:14 |
| zykes- | Kiall: isn't that false for keystone diablo ? diablo version when adding endpoints takes int id instead of tenant name ? | 18:14 |
| Kiall | zykes-: devstack uses the diablo branches... | 18:15 |
| Kiall | or something close to them anyway | 18:15 |
| Kiall | They use these repos+bracnhes https://github.com/cloudbuilders/devstack/blob/master/stackrc | 18:15 |
| Kiall | branches* | 18:15 |
| rmk | I think what's most frustrating is the lack of official stable diablo branches of every project. | 18:15 |
| Kiall | rmk: well .. there is for all the major repos | 18:16 |
| rmk | keystone, glance, nova, dash? | 18:16 |
| *** zaitcev has joined #openstack-dev | 18:16 | |
| Kiall | yup | 18:16 |
| mtaylor | rmk: keystone was not an openstack project for diablo, and neither was dash | 18:16 |
| dolphm | sandywalsh, reading | 18:16 |
| Kiall | my PPA packages are based on all the diablo/stable branches | 18:16 |
| zykes- | Kiall: Kiall one by one yeah, but is there that's been tested together ? | 18:16 |
| mtaylor | but nova, glance and swift all have a stable/diablo branch | 18:16 |
| mtaylor | zykes-: that's the idea | 18:16 |
| mtaylor | zykes-: there are, however, bugs - as well as holes in testing currently | 18:17 |
| Kiall | mtaylor: yes, but keystone and dash both have diablo branches that, barring some minor fixes, do work... | 18:17 |
| mtaylor | Kiall: ++ | 18:17 |
| sandywalsh | dolphm, nm, my test program is working, I suspect there's something funky going on with novaclient ... stay tuned | 18:17 |
| dolphm | mtaylor, while keystone wasn't officially "core" for diablo, i don't consider that an excuse to produce a shoddy project | 18:17 |
| mtaylor | dolphm: I agree | 18:17 |
| rmk | dolphm said what I wanted to say. | 18:18 |
| zykes- | i'm giving up soon .. not sure on what to use for keystone with diablo and dashboard to get it to work | 18:18 |
| sandywalsh | dolphm, separate question though ... it would be nice to get the admin interface from the service catalog ... otherwise we have to put extra flags in nova and novaclient | 18:19 |
| sandywalsh | the port 35357 stuff | 18:19 |
| *** rkukura has joined #openstack-dev | 18:20 | |
| rmk | sandywalsh: Question about multiple zones/keystone. Is the proper way still to add references to said zones via nova? Or is mapping multiple nova endpoints via keystone the appropriate method? | 18:20 |
| sandywalsh | rmk, the zone name needs to have an endpoint in keystone ... and then you still need to add it to nova via 'nova zone-add' | 18:21 |
| sandywalsh | rmk, but when you add it, you add the URL for keystone, not the child zone. It'll use the keystone Service Catalog to find the child zone URL | 18:21 |
| dolphm | sandywalsh, it's been discussed that the admin url should be returned if your user has the Admin role, i have no clue if that's implemented or not though | 18:21 |
| rmk | OK that makes sense.. | 18:21 |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: Creating uuid -> id mapping for S3 Image Service https://review.openstack.org/988 | 18:22 |
| rmk | I'm trying to find the specific documentation reference for keystone/scheduler I was curious about. Basically referencing functionality which would be present once keystone was available, which I'm not sure current is. | 18:23 |
| *** zaitcev has quit IRC | 18:24 | |
| rmk | "NOTE: currently the GET /servers command is not Zone-aware since all operations done in child Zones are done via a single administrative account. Therefore, asking a child Zone to GET /servers would return all the active instances ... and that would not be what the user intended. Later, when the Keystone Auth system is integrated with Nova, this functionality will be enabled." | 18:24 |
| rmk | That. :) | 18:24 |
| zykes- | so what should i do to get a fully working stack from swift to dashboard with keystone ? | 18:24 |
| dolphm | sandywalsh, on the GET tokens/xxxx/endpoints issue -- which versions/commits/whatever of keystone are you comparing? | 18:24 |
| annegentle | zykes-: the cloudbuilders branches did not work? | 18:24 |
| zykes- | annegentle: you mean for keystone ? | 18:24 |
| annegentle | zykes-: or are you on RHEL or sumpin? | 18:25 |
| zykes- | annegentle: correct | 18:25 |
| sandywalsh | rmk, the child zone calls should be done as the user making the call now. I'm currently working on a bug where keystone isn't letting that happen though. | 18:25 |
| annegentle | zykes-: the whole stack of cloudbuilders branches work together | 18:25 |
| rmk | sandywalsh: Thus, if a user were to do a nova list against the parent zone, would they see all their instances across zones? | 18:25 |
| sandywalsh | dolphm, trunk | 18:25 |
| rmk | ok so not yet then | 18:25 |
| annegentle | zykes-: I'm still trying to figure out the best way to document a working set for Diablo... | 18:25 |
| sandywalsh | rmk, correct, if you're working with trunk | 18:25 |
| zykes- | damned | 18:26 |
| rmk | Working with diablo-stable | 18:26 |
| sandywalsh | rmk, in the midst of a large zones refactoring to unify much of this | 18:26 |
| *** zaitcev has joined #openstack-dev | 18:27 | |
| zykes- | annegentle: cause i got everything working until i got to dashboard | 18:27 |
| zykes- | with trunk | 18:27 |
| sandywalsh | dolphm, I see what's going on ... just to ensure it's wrong. | 18:28 |
| dolphm | sandywalsh, ? | 18:28 |
| annegentle | zykes-: ah ok | 18:28 |
| sandywalsh | dolphm, with the .../endpoints call | 18:28 |
| dolphm | my fault or yours? ;) | 18:28 |
| sandywalsh | about to find out :) | 18:28 |
| *** adjohn has quit IRC | 18:28 | |
| zykes- | which is dolphm's suggestion as well | 18:29 |
| rmk | sandywalsh: I joined the orchestration team to hopefully participate in this as much as I can. :) | 18:29 |
| sandywalsh | rmk, excellent! | 18:30 |
| zykes- | but the cloudbuilders one does that use swif ? | 18:30 |
| zykes- | nope :( | 18:30 |
| *** bhall_ has quit IRC | 18:31 | |
| jtran | hey all. in regards to roles. if i want PER PROJ roles for 'user1' and 'netadmin' role , do i have to grant them global AND per project? that doesn't seem like a lot of sense to me | 18:31 |
| *** bhall has joined #openstack-dev | 18:32 | |
| zykes- | annegentle: what's your suggestion ? | 18:35 |
| *** adjohn has joined #openstack-dev | 18:39 | |
| *** ohnoimdead has joined #openstack-dev | 18:39 | |
| annegentle | zykes-: sorry I'm still in diablo-land | 18:39 |
| zykes- | annegentle: has anyone verified that keystone works with diablo core projects? | 18:40 |
| zykes- | or can i pack my bags and give u | 18:40 |
| notmyname | zykes-: what are you looking for in swift? packages? | 18:40 |
| *** lorin1 has quit IRC | 18:40 | |
| *** lorin1 has joined #openstack-dev | 18:41 | |
| *** lorin1 has quit IRC | 18:41 | |
| notmyname | zykes-: I've had keystone working with swift before, but it seems that support comes and goes as the keystone trunk gets updated | 18:41 |
| *** lorin1 has joined #openstack-dev | 18:41 | |
| joesavak | we tagged 2011.3 and branched Diablo - trunk will become essex... | 18:42 |
| zykes- | notmyname: i'm looking at getting a full diablo stack working with keystone, simple as that | 18:42 |
| zykes- | but to now all my attempts has failed, when 1 version works with 1 project another thing breaks | 18:42 |
| dolphm | zykes-, have you tried devstack or cloudbuilders? | 18:43 |
| dolphm | notmyname, keystone trunk should be fairly reliable now | 18:43 |
| zykes- | i'm not wanting to reinstalling OS'es | 18:43 |
| zykes- | but i'll try trunk again | 18:44 |
| zykes- | and work around the dashboard issue | 18:44 |
| notmyname | zykes-: the swift trunk is stable and the auth integration as never changed. as long as the keystone-provided middleware works, it should work with any version of swift | 18:44 |
| notmyname | dolphm: good to know. it's been a while since I last looked at keystone :-) | 18:45 |
| sandywalsh | dolphm, ok, here's what happening. I'm authing as an admin, but when I ask for /tokens/xxxx/endpoints it says no permission. However, if I use the 999777666555 token it works. | 18:45 |
| zykes- | notmyname: i tried with stable/diablo and that didn't work at all | 18:45 |
| notmyname | zykes-: the swiftauth middleware? | 18:45 |
| zykes- | tokenauth | 18:45 |
| notmyname | zykes-: tokenauth has never worked with swift (and the keystone docs have been incorrect for as long as I can remember) | 18:46 |
| zykes- | as it says on keystone's page | 18:46 |
| zykes- | what should one use then notmyname with swift? | 18:46 |
| notmyname | zykes-: xtoddx wrote swiftauth to use with swift. I assume it's still in keystone | 18:47 |
| zykes- | notmyname: i can't see that when i do "locat swiftauth" no.. | 18:47 |
| openstackgerrit | James E. Blair proposed a change to openstack/keystone: Add .gitreview config file for gerrit. https://review.openstack.org/1051 | 18:48 |
| zykes- | ah, swift_auth | 18:48 |
| xtoddx | zykes-: there is keystone middleware for swift, chmouel also wrote a version at https://github.com/cloudbuilders/swift-keystone2 | 18:48 |
| xtoddx | not sure its status though | 18:48 |
| openstackgerrit | James E. Blair proposed a change to openstack/netconn-api: Add .gitreview config file for gerrit. https://review.openstack.org/1052 | 18:48 |
| zykes- | now i learn this.. | 18:48 |
| *** jwhonce has joined #openstack-dev | 18:49 | |
| openstackgerrit | James E. Blair proposed a change to openstack/nova: Add .gitreview config file for gerrit. https://review.openstack.org/1053 | 18:49 |
| *** zul has joined #openstack-dev | 18:49 | |
| zykes- | i'm going to cry now if it all works after this... | 18:50 |
| openstackgerrit | James E. Blair proposed a change to openstack/object-api: Add .gitreview config file for gerrit. https://review.openstack.org/1054 | 18:50 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-chef: Add .gitreview config file for gerrit. https://review.openstack.org/1055 | 18:51 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-integration-tests: Add .gitreview config file for gerrit. https://review.openstack.org/1056 | 18:51 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-manuals: Add .gitreview config file for gerrit. https://review.openstack.org/1057 | 18:52 |
| openstackgerrit | James E. Blair proposed a change to openstack/openstack-puppet: Add .gitreview config file for gerrit. https://review.openstack.org/1058 | 18:52 |
| openstackgerrit | James E. Blair proposed a change to openstack/quantum: Add .gitreview config file for gerrit. https://review.openstack.org/1059 | 18:52 |
| annegentle | zykes-: oh don't cry :) | 18:53 |
| *** f-yang has joined #openstack-dev | 18:53 | |
| *** fred_ has joined #openstack-dev | 18:53 | |
| openstackgerrit | James E. Blair proposed a change to openstack/swift: Add .gitreview config file for gerrit. https://review.openstack.org/1060 | 18:53 |
| sandywalsh | dolphm, I am getting a warning "no service admin role defined" ... is that something new? | 18:53 |
| *** fred_ is now known as Guest3051 | 18:53 | |
| *** f-yang has quit IRC | 18:54 | |
| *** Guest3051 has quit IRC | 18:54 | |
| *** f-yang has joined #openstack-dev | 18:54 | |
| dolphm | sandywalsh, sort of -- i defer to yogi on service admin stuff though | 18:55 |
| sandywalsh | dolphm, ok, np ... I suspect this is the problem ... I'll run it down. | 18:55 |
| dolphm | sandywalsh, from keystone.conf: keystone-service-admin-role = KeystoneServiceAdmin #Role that allows to perform service admin operations. | 18:56 |
| dolphm | sandywalsh, from keystone/test/sampledata.py 18: ('role', 'add', 'KeystoneServiceAdmin'), | 18:57 |
| dolphm | sandywalsh, from keystone/test/sampledata.py 20: ('role', 'grant', 'KeystoneServiceAdmin', 'serviceadmin'), | 18:57 |
| sandywalsh | dolphm, yup, just saw that ... will make the changes. thx | 18:58 |
| *** rnorwood has quit IRC | 18:58 | |
| sandywalsh | dolphm, don't know how that's going to work with novaclient ... will I need to tell it about 2 accounts? the admin and the service admin? :/ | 18:58 |
| dolphm | sandywalsh, i hope not.. Admin is a superset of ServiceAdmin | 19:00 |
| openstackgerrit | A change was merged to openstack/keystone: Add .gitreview config file for gerrit. https://review.openstack.org/1051 | 19:00 |
| notmyname | jeblair: that proposed patch is very light on details on why it's needed :-) | 19:00 |
| sandywalsh | dolphm, that's what I would have assumed. Doesn't seem to be the case | 19:00 |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: Creating uuid -> id mapping for S3 Image Service https://review.openstack.org/988 | 19:00 |
| *** negronjl has joined #openstack-dev | 19:01 | |
| jeblair | notmyname: there was a recent mailing list thread. would you like me to amend the message with more details? | 19:02 |
| zykes- | dolphm: do you think that essex will have a more integrated keystone and working out of the box ? | 19:02 |
| dolphm | zykes-, absolutely | 19:02 |
| dolphm | if only because more is relative | 19:02 |
| zykes- | cause this is just painful :( | 19:03 |
| notmyname | jeblair: it wouldn't be bad (so we can have the reasoning all in one place with the code) | 19:03 |
| jeblair | notmyname: no problem, coming right up | 19:03 |
| dolphm | zykes-, any problem / solutions you find -- open bugs / write blueprints / whatever -- we need to get all that into docs and make fixes! | 19:03 |
| zykes- | dolphm: i'm using diablo atm | 19:04 |
| zykes- | of keystone.. | 19:04 |
| zykes- | though i'm not sure what to use anymore | 19:04 |
| zykes- | notmyname: using swift_auth should one still use -A <keystone url here> when using the swift command ? | 19:05 |
| dolphm | i'm just pointing out that trunk should still be backwards compatible with diablo, and should remain so for another week or two | 19:05 |
| notmyname | zykes-: yes | 19:05 |
| openstackgerrit | James E. Blair proposed a change to openstack/swift: Add .gitreview config file for gerrit. https://review.openstack.org/1060 | 19:06 |
| zykes- | dolphm: with dashboard as well ? | 19:06 |
| notmyname | zykes-: chmouel just submitted a patch to add auth 2.0 compat (ie keystone) to the swift binary. | 19:06 |
| openstackgerrit | James E. Blair proposed a change to openstack/quantum: Add .gitreview config file for gerrit. https://review.openstack.org/1059 | 19:07 |
| notmyname | zykes-: so, depending on keystone, the current swift binary may or may not work | 19:07 |
| dolphm | zykes-, when was the last time you ran devstack? | 19:07 |
| openstackgerrit | James E. Blair proposed a change to openstack/nova: Add .gitreview config file for gerrit. https://review.openstack.org/1053 | 19:07 |
| zykes- | i'm not on devstack dolphm i'm trying to make a demo for work | 19:07 |
| zykes- | which atm i'm stuck and confused | 19:08 |
| dolphm | isn't that what cloudbuilders & devstack are good at? | 19:08 |
| openstackgerrit | A change was merged to openstack/openstack-integration-tests: Add .gitreview config file for gerrit. https://review.openstack.org/1056 | 19:08 |
| zykes- | dolphm: swift, nova, glance, dashboard | 19:08 |
| zykes- | basically | 19:08 |
| *** rnorwood has joined #openstack-dev | 19:11 | |
| *** chomping has quit IRC | 19:11 | |
| *** mmetheny has quit IRC | 19:12 | |
| zykes- | dolphm: is anyone running swift with keystone in prod atm ? | 19:14 |
| notmyname | zykes-: I don't know of any swift+keystone prod clusters | 19:15 |
| dolphm | zykes-, not that i'm aware of | 19:16 |
| dolphm | i just noticed devstack doesn't install swift... i thought it did. | 19:16 |
| jeblair | dolphm: i think there's a branch working on that | 19:16 |
| dolphm | jeblair, good to know | 19:16 |
| openstackgerrit | Mark Washenberger proposed a change to openstack/nova: Server metadata must support server uuids. https://review.openstack.org/1061 | 19:19 |
| dolphm | zykes-, keystone + dashboard in 287 seconds: http://204.232.203.84/ (username: admin, password: asdf) | 19:20 |
| zykes- | dolphm: diablo release ? | 19:21 |
| dolphm | zykes-, trunk | 19:21 |
| zykes- | maybe i should do checkout on trunk instead ;p | 19:22 |
| zykes- | looks to work easier | 19:22 |
| zykes- | honestly until now i've got a belief for a better life in essex | 19:22 |
| dolphm | zykes-, i can't speak for other projects (nova-compute is failing in that devstack deploy!) | 19:22 |
| *** mmetheny has joined #openstack-dev | 19:23 | |
| dolphm | (in devstack) what's creating the "invisible_to_admin" tenant, which is clearly not invisible to admin? | 19:25 |
| zykes- | notmyname: i tried swift_auth | 19:25 |
| *** negronjl has quit IRC | 19:25 | |
| zykes- | AttributeError: 'NoneType' object has no attribute 'find' when using | 19:25 |
| openstackgerrit | Liem Nguyen proposed a change to openstack/keystone: Added SSL tests and other stuff from Dolph's comments https://review.openstack.org/1062 | 19:28 |
| openstackgerrit | Brian Lamar proposed a change to openstack/keystone: Adding the concept of creating a Keystone HTTP client in Python which can be used in Keystone and imported from Keystone to allow for easier Keystone integration. https://review.openstack.org/1063 | 19:28 |
| openstackgerrit | Jay Pipes proposed a change to openstack/glance: Adds Driver Layer to Image Cache https://review.openstack.org/1064 | 19:28 |
| dolphm | thank you brian lamar, for writing a keystone client! | 19:33 |
| *** dprince has quit IRC | 19:38 | |
| *** nati2 has quit IRC | 19:39 | |
| *** nati2 has joined #openstack-dev | 19:39 | |
| *** f-yang has quit IRC | 19:40 | |
| openstackgerrit | A change was merged to openstack/nova: Server metadata must support server uuids. https://review.openstack.org/1061 | 19:40 |
| openstackjenkins | Project nova-docs build #1,523: SUCCESS in 4 min 35 sec: https://jenkins.openstack.org/job/nova-docs/1523/ | 19:46 |
| openstackjenkins | mark.washenberger: Server metadata must support server uuids. | 19:46 |
| *** markwash has joined #openstack-dev | 19:52 | |
| *** zul has quit IRC | 19:53 | |
| openstackgerrit | Jay Pipes proposed a change to openstack/glance: Adds Driver Layer to Image Cache https://review.openstack.org/1064 | 19:53 |
| *** zaitcev has quit IRC | 19:57 | |
| zykes- | notmyname: i tried with swift_auth but then i just get unauthorized | 20:00 |
| zykes- | the localhost:8080 is that supposted to point to the service api or admin api of keystone ? | 20:00 |
| dolphm | that's a VERY old port for the keystone service API | 20:01 |
| dolphm | which is now on 5000 by default | 20:01 |
| dolphm | where did you find that? | 20:01 |
| notmyname | zykes-: I have not run keystone in a long time. however, here is what is in my proxy config for the keystone filter: http://paste.openstack.org/show/2845/ | 20:01 |
| zykes- | ah, there it worked :) | 20:03 |
| zykes- | thanks so much notmyname | 20:03 |
| zykes- | cause i tried with swift-keystone2 but that didn't work | 20:04 |
| notmyname | zykes-: I think that's still a work in progress (right chmouel?) | 20:05 |
| *** dolphm_ has joined #openstack-dev | 20:11 | |
| openstackgerrit | Brian Waldon proposed a change to openstack/nova: Creating uuid -> id mapping for S3 Image Service https://review.openstack.org/988 | 20:12 |
| zykes- | looking forward to esseix then ;) | 20:14 |
| zykes- | but then the network will be reamped with quantum so who knows ;) | 20:14 |
| *** zaitcev has joined #openstack-dev | 20:15 | |
| *** novas0x2a|laptop has quit IRC | 20:18 | |
| *** novas0x2a|laptop has joined #openstack-dev | 20:19 | |
| *** sannes1 has joined #openstack-dev | 20:19 | |
| sannes1 | I got a review on one of my suggested patches which asks me to add myself to the Authors file. Should I do this as part of the patch? My name contains UTF8 characters, should I use that or make due with ascii? Since I have several patches pending, should I add myself in all of them and hope it merges or just choose one and hope for the best? | 20:22 |
| zykes- | dolphm: funky with with dashboard is that: "o = urlparse(token.serviceCatalog['identity'][0]['adminURL'])" | 20:25 |
| *** negronjl has joined #openstack-dev | 20:28 | |
| *** joesavak has quit IRC | 20:29 | |
| openstackgerrit | Liem Nguyen proposed a change to openstack/keystone: Changed to relative paths per Dolph's comments https://review.openstack.org/1065 | 20:29 |
| rmk | Hmm, I don't see a diablo branch/tag for keystone. | 20:31 |
| dolphm_ | rmk: on launchpad or github? | 20:32 |
| zykes- | there is rmk stable/diablo | 20:32 |
| rmk | github | 20:32 |
| dolphm_ | rmk: because it's conspicuously missing on launchpad, but there's a tag in github | 20:32 |
| dolphm_ | rmk: https://github.com/openstack/keystone/tags | 20:32 |
| rmk | doh my fault, thanks | 20:33 |
| zykes- | ok dolphm so latest trunk works fine with nova + swift + latest glance trunk | 20:33 |
| dolphm_ | rmk: np | 20:33 |
| zykes- | nova and swift from diablo | 20:33 |
| dolphm_ | zykes-: YAY thanks for the feedback | 20:33 |
| zykes- | now i get an error with dashboard ;p | 20:33 |
| dolphm_ | zykes-: i probably can't help, but what's the error? | 20:33 |
| zykes- | http://paste.openstack.org/show/2846/ | 20:35 |
| vladimir3p | Hi All, have anybody experienced issues with simultaneous instance create & delete calls (or at least executed with relatively short interval)? The result of it is that instance is in "Deleted" state, but actually running | 20:41 |
| clayg | vladimir3p: ping | 20:43 |
| *** liemmn has joined #openstack-dev | 20:43 | |
| *** dolphm_ has quit IRC | 20:44 | |
| liemmn | @dolphm... Thanks for the feedback on the code review... I am unclear on the fact that you like absolute path (or not).... Absolute or relative? | 20:44 |
| larissa | liemmn: Error: "dolphm..." is not a valid command. | 20:44 |
| liemmn | dolphm... Thanks for the feedback on the code review... I am unclear on the fact that you like absolute path (or not).... Absolute or relative? | 20:45 |
| *** yoda has joined #openstack-dev | 20:47 | |
| vladimir3p | clayg: pong | 20:47 |
| *** sannes1 has quit IRC | 20:47 | |
| clayg | vladimir3p: why is extra specs not sufficient | 20:47 |
| clayg | re: adding auxiliary specs to volume create | 20:47 |
| vladimir3p | clayg: let's imagine that you would like to create a volume that will be sitted at the same place where volume #X | 20:48 |
| vladimir3p | clayg: you can't set it up in volume types | 20:48 |
| vladimir3p | clayg: actually, it is possible, but we will have volume types per volume :-) | 20:48 |
| *** lorin1 has quit IRC | 20:48 | |
| clayg | sry then naming... VolumeMetadata | 20:49 |
| clayg | you can already add volume meta on volume create - just tell them to stick it in there, and subclass scheduler to look at it | 20:49 |
| vladimir3p | yeah, meta-data might be one place to store it. But we, for instance, using meta-data to store volume-specific params, and don't want to use it for scheduler | 20:50 |
| vladimir3p | I was just thinking that passing additional parameters to scheduler (per volume basis) might be beneficial | 20:50 |
| zykes- | dolphm: is the serviceCatalog different from Diablo > trunk ? | 20:50 |
| clayg | idk if it's worth mucking with the api | 20:51 |
| chmouel | notmyname: yeah still a work in progress | 20:51 |
| clayg | meta is already there... | 20:51 |
| *** yoda has quit IRC | 20:51 | |
| vladimir3p | clayg: yeah, I understand your point. And agree with it. Some vendors may decide to implement it in such a way | 20:52 |
| chmouel | zykes-: but would love to know what error you are getting | 20:52 |
| vladimir3p | clayg: btw, do you have some time to talk about scheuler ... | 20:52 |
| clayg | but zadara doesn't need affinity either? so you're implimenting it but you have no need for it - let citrix or hp do it since they have the use case | 20:52 |
| clayg | sure pm | 20:52 |
| vladimir3p | clayg: yes, we don't need it :-) trying to simplify other folks life | 20:53 |
| zykes- | why doesn't zadara need it ? | 20:53 |
| clayg | you're such a sweetheart | 20:53 |
| *** jakedahn has joined #openstack-dev | 20:54 | |
| *** zaitcev has quit IRC | 20:58 | |
| zykes- | chmouel: http://paste.openstack.org/show/2847/ | 20:59 |
| chmouel | zykes-: that's dashboard stuff right? I am talking about the swift-keystone2 middleware | 21:00 |
| zykes- | ah chmouel | 21:01 |
| zykes- | yeah but i got swift_auth to work, what's the diff ? | 21:01 |
| zykes- | it's swift-keystone2 with swift from diablo | 21:01 |
| zykes- | i can test if you want | 21:01 |
| chmouel | handle ACL container sync and code more in line with swauth/tempauth default auth server | 21:02 |
| chmouel | zykes-: yes please that would be great | 21:02 |
| chmouel | zykes-: (from github.com/cloudbuilders/swift-keystone2 | 21:02 |
| *** nhubbard is now known as nhubbard_away | 21:04 | |
| zykes- | is there anyone here that's a keystone dev? | 21:05 |
| dolphm | zykes-, besides me? | 21:05 |
| zykes- | ehm, dashboard dev sorry dolphm | 21:05 |
| *** nati2 has quit IRC | 21:05 | |
| dolphm | /salute | 21:05 |
| *** nati2 has joined #openstack-dev | 21:05 | |
| zykes- | or aren't any of them online ;p | 21:06 |
| rmk | zykes-: I found the latest dashboard to not be working. | 21:06 |
| rmk | Use an older rev., | 21:06 |
| zykes- | rmk: i know, it doesn't work with older nova | 21:07 |
| rmk | use commit 601fc1b716829f1ab8257116e3257c06050e3bb2 | 21:07 |
| rmk | what? | 21:07 |
| rmk | and do you have openstackx added to the api? | 21:07 |
| zykes- | yeah, dashboard looked for some stuff under / but it was added to /admin | 21:07 |
| zykes- | chmouel: is keystone_url = http://keystone:35357/v2.0 correct ? | 21:07 |
| dolphm | liemmn: I was just alluding to the fact that the relative paths in previous default configuration files have failed when people copy them into package-based installs -- absolute paths should be fine in the example keystone.conf, while i believe we need relative paths in the template.conf's for functional tests | 21:08 |
| chmouel | zykes-: yep | 21:09 |
| zykes- | then i do swift -v -A http://keystone.stack.lan:5000/v2.0 -U zykes -K testing stat -v | 21:09 |
| chmouel | yep | 21:09 |
| chmouel | no | 21:09 |
| chmouel | sorry :) you hitting the 2.0 with swift which doesnt support it yet (i sent a patch for that) | 21:10 |
| zykes- | http://paste.openstack.org/show/2848/ < is the error given then | 21:10 |
| chmouel | change 2.0 by 1.0 | 21:10 |
| zykes- | uh, what should i use thne ? | 21:10 |
| zykes- | in the .conf or both? | 21:10 |
| chmouel | just the client | 21:10 |
| *** rkukura has quit IRC | 21:10 | |
| zykes- | then i get forbidden | 21:10 |
| *** rkukura has joined #openstack-dev | 21:11 | |
| chmouel | and you sure you get this username:password working? | 21:11 |
| *** novas0x2a|laptop has quit IRC | 21:11 | |
| *** zaitcev has joined #openstack-dev | 21:11 | |
| chmouel | try to set this in your log : http://pastie.org/private/r1omwctphlde2festvpa in your [DEFAULT] section | 21:12 |
| chmouel | sorry i mean proxy-server.conf | 21:12 |
| zykes- | yeah, if i switch back to swift_auth it's all dandy | 21:12 |
| chmouel | weird if you set the DEBUG can you see if you have any error there? | 21:13 |
| *** novas0x2a|laptop has joined #openstack-dev | 21:13 | |
| zykes- | nothing in messages no | 21:14 |
| zykes- | then i change back to swift_auth and restart proxy and all is ok | 21:14 |
| chmouel | it should be /var/log/syslog by default on ubuntu | 21:14 |
| zykes- | messages on redhat | 21:15 |
| chmouel | do you have memcached installed on proxy-server? | 21:15 |
| zykes- | yep | 21:15 |
| chmouel | if you use that script http://pastie.org/private/elgnpdpwcb6slno8qhy7xw with your username password url can you paste me the output ? | 21:17 |
| chmouel | the url is the keystone url ie: keystne.stack.lan | 21:17 |
| *** lts has quit IRC | 21:18 | |
| zykes- | i knwo | 21:19 |
| zykes- | http://paste.pocoo.org/show/497600/ | 21:19 |
| zykes- | dolphm: keystone servicecatalog from diablo > trunk any difference ? | 21:21 |
| chmouel | and what do you get back from curl -H 'X-Auth-Token: 451eb8d2-fc0f-4d0a-9a10-8761c5dd2c11' https://swift.stack.lan:8080/v1/AUTH_1 | 21:21 |
| dolphm | zykes-, not off the top of my head | 21:22 |
| zykes- | chmouel: you mean when using keystone2 ? | 21:23 |
| chmouel | zykes-: yes on proxy-server and launching that curl command against it (actually add -v to verbose it the headers) | 21:24 |
| zykes- | blimy | 21:24 |
| zykes- | i restarted memcache | 21:24 |
| zykes- | and then the proxy | 21:25 |
| zykes- | or not, had swiftauth in my config | 21:25 |
| zykes- | rmk: you think trunk db should work ? | 21:26 |
| zykes- | http://pastie.org/2752913 | 21:27 |
| openstackgerrit | Brian Lamar proposed a change to openstack/glance: Removed 'mox==0.5.0' and replaced with just 'mox' in tools/pip-requires. https://review.openstack.org/1066 | 21:27 |
| *** nati2 has quit IRC | 21:27 | |
| openstackgerrit | Liem Nguyen proposed a change to openstack/keystone: x.509 client authentication https://review.openstack.org/1038 | 21:28 |
| openstackgerrit | Brian Lamar proposed a change to openstack/glance: Glance redirect support for clients. https://review.openstack.org/1067 | 21:29 |
| zykes- | chmouel: ? | 21:30 |
| chmouel | zykes-: humm reading it | 21:32 |
| liemmn | dolphm: Thanks for the help! Patchset#2 uploaded. | 21:32 |
| dolphm | liemmn, awesome | 21:32 |
| chmouel | zykes-: it's strange that you don't have anything coming up in the logs any chance i can see your proxy-server.conf? | 21:33 |
| zykes- | sure.. | 21:33 |
| zykes- | nothing secret there anyways | 21:33 |
| dolphm | liemmn, minor note, i guess you lost your original commit message (about x 509) | 21:33 |
| liemmn | Dang it! lol | 21:34 |
| zykes- | chmouel: http://paste.pocoo.org/show/497610/ | 21:34 |
| chmouel | ny chance you can remove 12 to 14 lines ? | 21:35 |
| chmouel | and swift-init proxy reload | 21:35 |
| zykes- | sure | 21:36 |
| zykes- | and then do ? | 21:36 |
| chmouel | can you see if you have debugs in logs now? | 21:37 |
| zykes- | http://paste.openstack.org/show/2850/ is the stuff that's in the logs | 21:37 |
| chmouel | i'm not sure why it doesnt show the DEBUG on your server | 21:38 |
| chmouel | i get something like that in my logs http://pastie.org/private/mtqskkmsmks6de4xfuloig | 21:39 |
| zykes- | anything in config maybe? | 21:40 |
| dolphm | liemmn, just a heads up, i have one more minor comment, and jenkins is going to fail your patch because of pep8 violations | 21:40 |
| dolphm | liemmn, but i'm going to +2 it anyway, just so you can see the process | 21:41 |
| zykes- | what's the cert stuff for ? | 21:41 |
| chmouel | zykes-: what do you mean for having ssl on proxy? | 21:41 |
| liemmn | It's for x.509 client authentication from the middeware to keystone service: https://blueprints.launchpad.net/keystone/+spec/2-way-ssl | 21:42 |
| zykes- | it was for liemmn chmouel | 21:42 |
| chmouel | zykes-: k any chance you can reneable account_management in there? | 21:42 |
| zykes- | chmouel: is there any debug stuff i can put in the conf ? | 21:42 |
| dolphm | zykes-, out of the box, keystone likes to spew your passwords in plaintext across the network :) | 21:42 |
| zykes- | that does ? | 21:43 |
| chmouel | zykes-: well log_level = DEBUG would do | 21:43 |
| chmouel | zykes-: but that depend of your configuration | 21:43 |
| chmouel | zykes-: i'm not sure how redhat does stuff these days | 21:43 |
| chmouel | dolphm: hash-passwords = True by default or so it seems? | 21:44 |
| openstackgerrit | Verification of a change to openstack/keystone failed: Updated code based on Dolph's comments. https://review.openstack.org/1038 | 21:44 |
| dolphm | chmouel, that is password hashing in the database, in case your database in compromised | 21:44 |
| chmouel | dolphm: ah yeah sorry i got confused | 21:45 |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Fixes to user calls.Also making code changes to get rid of the default tenant concept.Getting rid of cranky comments. https://review.openstack.org/1068 | 21:45 |
| dolphm | ChanServ, no worries | 21:45 |
| dolphm | liemmn, so now that it failed ^^^ see above ^^^, you can submit a patch with your old commit message, pep8 fixes, and if you don't mind, add some additional comments to keystone.conf | 21:46 |
| dolphm | liemmn, (`pip install pep8` and `pep8 keystone` if you've never used it) | 21:47 |
| liemmn | dolphm: Will do... Thanks again. | 21:47 |
| dolphm | liemmn, you can also reference the build page: https://jenkins.openstack.org/job/keystone-pep8/333/violations/? | 21:47 |
| dolphm | liemmn, np -- i'm out for the day -- good luck! | 21:48 |
| zykes- | chmouel: should we give up or ? | 21:48 |
| chmouel | zykes-: this is i'm sure something obvious it just the logging that bugs me why swift is not logging debugs message in there | 21:49 |
| openstackgerrit | A change to openstack/keystone has been rejected: Fixes to user calls.Also making code changes to get rid of the default tenant concept.Getting rid of cranky comments. https://review.openstack.org/1068 | 21:50 |
| zykes- | :/ | 21:50 |
| zykes- | wonder what | 21:51 |
| openstackgerrit | A change to openstack/keystone has been rejected: Fixes to user calls.Also making code changes to get rid of the default tenant concept.Getting rid of cranky comments. https://review.openstack.org/1068 | 21:52 |
| zykes- | i wonder what | 21:52 |
| openstackgerrit | A change was merged to openstack/glance: Removed 'mox==0.5.0' and replaced with just 'mox' in tools/pip-requires. https://review.openstack.org/1066 | 21:52 |
| *** yogirackspace has joined #openstack-dev | 21:52 | |
| *** liemmn has quit IRC | 21:52 | |
| chmouel | let me check if some of swift ops have any ideas | 21:53 |
| *** liemmn has joined #openstack-dev | 21:54 | |
| yogirackspace | is jesse in? | 21:54 |
| openstackgerrit | A change to openstack/keystone has been rejected: Fixes to user calls.Also making code changes to get rid of the default tenant concept.Getting rid of cranky comments. https://review.openstack.org/1068 | 21:54 |
| *** anotherjesse has joined #openstack-dev | 21:55 | |
| zykes- | dolphm: is there any use yet with endpointing ? | 21:56 |
| yogirackspace | anotherjesse:Iam trying to move aweay from concept of default tenant that is tied to user | 21:57 |
| anotherjesse | yogirackspace: blueprint? | 21:57 |
| anotherjesse | bug? | 21:57 |
| yogirackspace | yogirackspace: anotherjesse:Refer new contract.We dont have tenantId as part of user any more | 21:57 |
| openstackgerrit | Johannes Erdfelt proposed a change to openstack/nova: Fix concurrency of XenAPI sessions https://review.openstack.org/1069 | 21:58 |
| *** liemmn has quit IRC | 21:58 | |
| yogirackspace | anotherjesse: Should we have another discussion? | 21:58 |
| anotherjesse | yogirackspace: are you referring to my comment in the gerrit that: | 21:59 |
| anotherjesse | 1) you aren't using proper git commit message formats | 21:59 |
| anotherjesse | 2) you shouldn't make schema changes without using migrations | 21:59 |
| yogirackspace | jesse Iam pinging you directly | 21:59 |
| anotherjesse | since I've not reviewed the code as far as functionality | 22:00 |
| *** kbringard has quit IRC | 22:02 | |
| vladimir3p | vishy: ping | 22:03 |
| openstackgerrit | Brian Lamar proposed a change to openstack/glance: Glance redirect support for clients. https://review.openstack.org/1067 | 22:08 |
| *** mfer has joined #openstack-dev | 22:09 | |
| *** mfer has quit IRC | 22:10 | |
| vishy | vladimir3p: one sec, on phone | 22:12 |
| *** cp16net has quit IRC | 22:14 | |
| *** dubsquared has quit IRC | 22:14 | |
| *** novas0x2a|laptop has quit IRC | 22:15 | |
| *** novas0x2a|laptop has joined #openstack-dev | 22:16 | |
| *** wwkeyboard has left #openstack-dev | 22:17 | |
| *** negronjl has quit IRC | 22:18 | |
| vishy | vladimir3p: sup? | 22:20 |
| vladimir3p | vishy: We are seeing quite a strange behavior related to relatively quick instance creation/deletion - instance is spawned, but is marked as "deleted" in the DB. Especially frequently it happens in our automatic tests. | 22:20 |
| vladimir3p | vishy: In some "normal" cases run_instance is interrupted by one of multiple _instance_update calls (there is an instance get and it stops with an exception) | 22:20 |
| vishy | vladimir3p: is this current code? | 22:21 |
| vladimir3p | but in some cases (when run_instance already entered "spawning" phase) it is running | 22:21 |
| vladimir3p | it is diablo code | 22:21 |
| vladimir3p | was it changed after that? | 22:21 |
| vishy | vladimir3p: there are definitely some issues with the "state machine" | 22:21 |
| vladimir3p | actually stable/diablo | 22:21 |
| vladimir3p | yeah, I'm trying to understand if we should create a bug & fix it with some mutex-type of lock | 22:22 |
| vishy | some stuff was changed after, I think we got rid of instance_set_state recently | 22:22 |
| vishy | vladimir3p: if you see an easy place to fix it | 22:22 |
| vladimir3p | I suppose we could modify "serializer" and make it "per instance" | 22:22 |
| vladimir3p | in this case deletion will wait for completion of create | 22:22 |
| vishy | perhaps we need a utils.synchronized | 22:23 |
| vishy | vladimir3p: there was a blueprint about fixing that | 22:23 |
| vladimir3p | the global one will be bad | 22:23 |
| vladimir3p | ah, really? | 22:23 |
| vladimir3p | have not seen it | 22:23 |
| vladimir3p | in general we can go with utils.sync, but because it is global it will affect oither instances | 22:23 |
| vladimir3p | would like to make it "per instance" | 22:24 |
| *** yogirackspace has left #openstack-dev | 22:24 | |
| vladimir3p | vishy: if you will find a ref to bug or BP, can oyu pls forward it to me? | 22:24 |
| vladimir3p | thanks | 22:24 |
| vishy | yeah there was a discussion about it by ray hookway | 22:25 |
| vishy | vladimir3p: http://wiki.openstack.org/nova-vm-state-management | 22:28 |
| vishy | vladimir3p: this is going to be addressed by the orchestration team? | 22:28 |
| vladimir3p | vishy: not sure that orchestration will focus on concurrency issues, but it might be definitely one of areas | 22:31 |
| vladimir3p | vishy: yeah, now I remember this topiv, but it was more about better states handling | 22:31 |
| vishy | vladimir3p: yeah that is one of the results though | 22:32 |
| vladimir3p | vishy: for example, he propose to allow termination in Building phase | 22:32 |
| vladimir3p | ok, I will probably file the bug and supply the fix for it, what do you think? | 22:32 |
| vishy | sure | 22:32 |
| vishy | vladimir3p: yes, but the updating code will be modified to check to make sure the state hasn't changed | 22:33 |
| vishy | vladimir3p: and to cleanup if it has | 22:33 |
| openstackgerrit | Johannes Erdfelt proposed a change to openstack/nova: Fix concurrency of XenAPI sessions https://review.openstack.org/1069 | 22:43 |
| *** code_franco has quit IRC | 22:44 | |
| *** rkukura has joined #openstack-dev | 22:47 | |
| *** markmc has quit IRC | 22:49 | |
| *** adjohn has quit IRC | 22:56 | |
| *** mwhooker_ has joined #openstack-dev | 22:57 | |
| *** mwhooker has quit IRC | 23:01 | |
| *** mwhooker_ is now known as mwhooker | 23:01 | |
| openstackgerrit | Sumit Naiksatam proposed a change to openstack/quantum: New tests are being adding to the Diablo code (Cisco L2-Network plugin), and some fixes in the case where the tests were failing. https://review.openstack.org/1070 | 23:02 |
| openstackgerrit | Liem Nguyen proposed a change to openstack/keystone: Updated code based on Dolph's comments. https://review.openstack.org/1038 | 23:04 |
| *** anotherjesse has quit IRC | 23:06 | |
| *** rnorwood has quit IRC | 23:06 | |
| *** rnirmal has quit IRC | 23:07 | |
| *** bcwaldon has quit IRC | 23:17 | |
| *** blamar has quit IRC | 23:19 | |
| *** bcwaldon has joined #openstack-dev | 23:22 | |
| *** negronjl has joined #openstack-dev | 23:22 | |
| *** misheska has joined #openstack-dev | 23:22 | |
| *** bencherian has joined #openstack-dev | 23:24 | |
| *** misheska has joined #openstack-dev | 23:24 | |
| *** jeffjapan has joined #openstack-dev | 23:32 | |
| *** bcwaldon has quit IRC | 23:37 | |
| *** anotherjesse has joined #openstack-dev | 23:39 | |
| *** joonwon has joined #openstack-dev | 23:39 | |
| *** jwhonce has quit IRC | 23:46 | |
| *** jakedahn has quit IRC | 23:47 | |
| *** jakedahn_ has joined #openstack-dev | 23:47 | |
| *** bencherian has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!