| *** dragondm has quit IRC | 00:11 | |
| *** vladimir3p has quit IRC | 00:13 | |
| *** jakedahn has quit IRC | 00:19 | |
| *** lorin1 has joined #openstack-dev | 00:21 | |
| *** lorin1 has left #openstack-dev | 00:22 | |
| *** mgius has quit IRC | 00:40 | |
| *** joonwon has quit IRC | 00:49 | |
| *** cole has quit IRC | 00:54 | |
| openstackjenkins | Project nova build #1,392: SUCCESS in 2 min 54 sec: https://jenkins.openstack.org/job/nova/1392/ | 00:54 |
|---|---|---|
| openstackjenkins | Tarmac: Reverted some changes to instance_get_all_by_filters() that was added in rev 1594. An additional argument for filtering on instance uuids is not needed, as you can add 'uuid: uuid_list' into the filters dictionary. Just needed to add 'uuid' as an exact_match_filter. This restores the filtering to do a single DB query. | 00:54 |
| openstackjenkins | Also updated ID/UUID mapping code to be a little more efficient, by returning a dictionary of 'ID: UUID'... vs a list. | 00:54 |
| openstackjenkins | Fixed a test that assumed list order. | 00:54 |
| openstackjenkins | A couple of typo fixes and a pep8 issue in trunk also fixed. | 00:54 |
| *** dragondm has joined #openstack-dev | 01:00 | |
| openstackjenkins | Project nova build #1,393: SUCCESS in 3 min 56 sec: https://jenkins.openstack.org/job/nova/1393/ | 01:05 |
| openstackjenkins | * Tarmac: Fixes migration for Mysql to drop the FK on the right table | 01:05 |
| openstackjenkins | * Tarmac: Reverted some changes to instance_get_all_by_filters() that was added in rev 1594. An additional argument for filtering on instance uuids is not needed, as you can add 'uuid: uuid_list' into the filters dictionary. Just needed to add 'uuid' as an exact_match_filter. This restores the filtering to do a single DB query. | 01:05 |
| openstackjenkins | Also updated ID/UUID mapping code to be a little more efficient, by returning a dictionary of 'ID: UUID'... vs a list. | 01:05 |
| openstackjenkins | Fixed a test that assumed list order. | 01:05 |
| openstackjenkins | A couple of typo fixes and a pep8 issue in trunk also fixed. | 01:05 |
| openstackjenkins | * Tarmac: Reworked the export command to be | 01:05 |
| openstackjenkins | nova-manage shell export --filename=somefile | 01:05 |
| openstackjenkins | This will output the somefile file in to the keystone-manage syntax that will then be used by a shellscript that is being proposed to keystone. | 01:05 |
| *** liquid_ has joined #openstack-dev | 01:14 | |
| *** jakedahn has joined #openstack-dev | 01:15 | |
| *** jdurgin has quit IRC | 01:20 | |
| *** mandela123 has joined #openstack-dev | 01:29 | |
| *** liquid_ has quit IRC | 01:30 | |
| *** joonwon has joined #openstack-dev | 01:30 | |
| *** jkyoung has joined #openstack-dev | 01:37 | |
| *** mgius has joined #openstack-dev | 01:42 | |
| *** HugoKuo has joined #openstack-dev | 01:57 | |
| *** bengrue has quit IRC | 02:22 | |
| *** HugoKuo has quit IRC | 02:25 | |
| *** HugoKuo has joined #openstack-dev | 02:25 | |
| *** dragondm_ has joined #openstack-dev | 02:39 | |
| *** dragondm has quit IRC | 02:39 | |
| *** dragondm_ is now known as dragondm | 02:39 | |
| *** hugokuo has joined #openstack-dev | 02:43 | |
| hugokuo | morning | 02:43 |
| *** zorzar has quit IRC | 02:51 | |
| *** rohitkarajgi has joined #openstack-dev | 02:54 | |
| *** jakedahn has quit IRC | 02:57 | |
| *** zorzar has joined #openstack-dev | 03:00 | |
| *** joonwon has quit IRC | 03:07 | |
| *** mandela123_ has joined #openstack-dev | 03:23 | |
| *** mandela123 has quit IRC | 03:23 | |
| *** mandela123_ is now known as mandela123 | 03:24 | |
| *** plu has quit IRC | 03:38 | |
| *** plu has joined #openstack-dev | 03:47 | |
| *** jkyoung has quit IRC | 03:50 | |
| *** joonwon has joined #openstack-dev | 03:51 | |
| *** lorin1 has joined #openstack-dev | 03:58 | |
| *** lorin1 has left #openstack-dev | 03:58 | |
| *** Rajaram has joined #openstack-dev | 04:05 | |
| *** novas0x2a|laptop has quit IRC | 04:12 | |
| *** rohitkarajgi has quit IRC | 04:16 | |
| *** dragondm has quit IRC | 04:18 | |
| *** mgius has quit IRC | 04:23 | |
| *** mandela123 has quit IRC | 04:36 | |
| *** mandela123 has joined #openstack-dev | 04:37 | |
| *** mandela123 has quit IRC | 04:51 | |
| *** mandela123 has joined #openstack-dev | 04:52 | |
| *** mandela123 has quit IRC | 04:59 | |
| *** mandela123 has joined #openstack-dev | 04:59 | |
| *** rohitkarajgi has joined #openstack-dev | 05:26 | |
| *** rnorwood has quit IRC | 05:32 | |
| rohitkarajgi | Hey nova:core - Please review https://code.launchpad.net/~rohitkarajgi/nova/libvirt_unittests | 05:38 |
| *** rohitkarajgi has quit IRC | 05:42 | |
| vishy | soren: are you around? | 06:26 |
| vishy | soren: I pinged mtaylor about the branches in milestone-proposed not getting merged by tarmac, but it looks like he didn't get to it yesterday. Any ideas? | 06:27 |
| vishy | soren: have 6 branches ready to go in https://code.launchpad.net/~hudson-openstack/nova/milestone-proposed/+activereviews | 06:27 |
| *** mandela123_ has joined #openstack-dev | 06:28 | |
| *** mandela123 has quit IRC | 06:30 | |
| *** mandela123_ is now known as mandela123 | 06:30 | |
| openstackjenkins | Project nova build #1,394: SUCCESS in 3 min 10 sec: https://jenkins.openstack.org/job/nova/1394/ | 06:34 |
| openstackjenkins | * Tarmac: Update migration 047 to dynamically lookup the name of the instance_id | 06:34 |
| openstackjenkins | fkey before dropping it. We can't hard code the name of the fkey since we | 06:34 |
| openstackjenkins | didn't name it explicitly on create. | 06:34 |
| openstackjenkins | * Tarmac: Fixes migration for Mysql to drop the FK on the right table | 06:34 |
| *** rcc has joined #openstack-dev | 06:37 | |
| *** jdag has quit IRC | 06:41 | |
| openstackjenkins | Project nova build #1,395: SUCCESS in 3 min 3 sec: https://jenkins.openstack.org/job/nova/1395/ | 06:44 |
| openstackjenkins | Tarmac: removes warning set forth in d3 for deprecated setting of bridge automagically | 06:44 |
| openstackjenkins | Project nova build #1,396: SUCCESS in 2 min 57 sec: https://jenkins.openstack.org/job/nova/1396/ | 06:56 |
| openstackjenkins | Tarmac: Fixed --uuid network command in nova-manage to desc to "uuid" instead of "net_uuid" | 06:56 |
| *** mnour has joined #openstack-dev | 07:08 | |
| *** reidrac has joined #openstack-dev | 07:15 | |
| soren | vishy: /me looks | 07:25 |
| soren | vishy: Erm.. I don't know where the nova-milestone-tarmac job has gone. | 07:25 |
| vishy | milestone page | 07:26 |
| vishy | there seems to be a list of jobs at the top | 07:26 |
| soren | Oh, it's still there. | 07:26 |
| *** Vek has quit IRC | 07:26 | |
| *** Vek has joined #openstack-dev | 07:26 | |
| soren | Weird. I used Jenkin's Search thing, but it didn't find it. | 07:26 |
| soren | vishy: Ok, so what's the problem exactly? Looking at https://code.launchpad.net/~hudson-openstack/nova/milestone-proposed/+activereviews none of them are approved, afaics. | 07:27 |
| vishy | they aren't now | 07:27 |
| vishy | because tarmac kicked them back with no error | 07:27 |
| soren | vishy: example? | 07:28 |
| vishy | soren: https://code.launchpad.net/~vishvananda/nova/secgroup-fixes-ms/+merge/76116 | 07:28 |
| *** stanchan has quit IRC | 07:28 | |
| soren | Oh, you backported that? Great! Thanks! | 07:28 |
| soren | vishy: Ok, I'll see if I can find out where that job ran and work out why it failed. | 07:28 |
| vishy | soren: np | 07:28 |
| soren | Gimme 10 minutes or so. | 07:28 |
| vishy | soren: cool thx | 07:28 |
| soren | Oh. | 07:32 |
| soren | That again. | 07:32 |
| soren | vishy: fixed. | 07:33 |
| soren | mtaylor, jeblair: nova-1 was screwed. It couldn't find nose, so I "sudo apt-get --reinstall install python-nose"ed. | 07:34 |
| vishy | soren: ah nice | 07:34 |
| vishy | k retrying props | 07:34 |
| soren | I re-approved the secgroups one. | 07:35 |
| *** rcc1 has joined #openstack-dev | 07:49 | |
| *** rcc has quit IRC | 07:51 | |
| openstackjenkins | Project nova-milestone build #57: SUCCESS in 25 sec: https://jenkins.openstack.org/job/nova-milestone/57/ | 07:52 |
| openstackjenkins | Tarmac: Fix a bug that would make spawning new instances fail if no port/protocol is given (for rules granting access for other security groups). | 07:52 |
| *** chemikadze has joined #openstack-dev | 07:52 | |
| chmouel | sandywalsh, jk0: anybody can merge https://github.com/rackspace/python-novaclient/pull/119 it's trivial and that would make people using OpenStack 1.0 (or RaX public cloud) very happy | 07:58 |
| smoser | ok. /me has to go to bed. but would really appreciate any nova-core reviewing https://code.launchpad.net/~smoser/nova/lp853330/+merge/76164 and https://code.launchpad.net/~smoser/nova/milestone-proposed.lp853330/+merge/76165 | 08:03 |
| *** negronjl has quit IRC | 08:07 | |
| *** negronjl has joined #openstack-dev | 08:12 | |
| *** jeffjapan has quit IRC | 08:29 | |
| *** AhmedSoliman has joined #openstack-dev | 08:33 | |
| *** negronjl has quit IRC | 08:46 | |
| *** joonwon has quit IRC | 08:46 | |
| *** negronjl has joined #openstack-dev | 08:49 | |
| openstackjenkins | Project nova build #1,397: SUCCESS in 3 min 48 sec: https://jenkins.openstack.org/job/nova/1397/ | 08:50 |
| openstackjenkins | * Tarmac: create disk.local the same way ephemerals are created (LP: #851145) | 08:50 |
| openstackjenkins | If the user did not specify '--block-device-mapping /dev/vdb=ephemeral0', then the first non-root device would end up being created differently then if they had. It would not have a filesystem on it. | 08:50 |
| openstackjenkins | This makes 'local_gb' in the libvirt connection created the same way that | 08:50 |
| openstackjenkins | it would be if it were named ephemeral0. | 08:50 |
| openstackjenkins | * Tarmac: Fixed --uuid network command in nova-manage to desc to "uuid" instead of "net_uuid" | 08:50 |
| openstackjenkins | * Tarmac: removes warning set forth in d3 for deprecated setting of bridge automagically | 08:50 |
| openstackjenkins | * Tarmac: Update migration 047 to dynamically lookup the name of the instance_id | 08:50 |
| openstackjenkins | fkey before dropping it. We can't hard code the name of the fkey since we | 08:50 |
| openstackjenkins | didn't name it explicitly on create. | 08:50 |
| *** negronjl has quit IRC | 08:55 | |
| *** negronjl has joined #openstack-dev | 09:00 | |
| openstackjenkins | Project nova-milestone build #58: SUCCESS in 18 sec: https://jenkins.openstack.org/job/nova-milestone/58/ | 09:02 |
| openstackjenkins | Tarmac: Fixed --uuid network command in nova-manage to desc to "uuid" instead of "net_uuid" | 09:02 |
| *** negronjl has quit IRC | 09:08 | |
| *** negronjl has joined #openstack-dev | 09:08 | |
| *** AhmedSoliman has quit IRC | 09:08 | |
| openstackjenkins | Project nova-milestone build #59: SUCCESS in 24 sec: https://jenkins.openstack.org/job/nova-milestone/59/ | 09:12 |
| openstackjenkins | Tarmac: Reworked the export command to be | 09:12 |
| openstackjenkins | nova-manage shell export --filename=somefile | 09:12 |
| openstackjenkins | This will output the somefile file in to the keystone-manage syntax that will then be used by a shellscript that is being proposed to keystone. | 09:12 |
| *** ahmed_ has joined #openstack-dev | 09:16 | |
| openstackjenkins | Project nova-milestone build #60: SUCCESS in 17 sec: https://jenkins.openstack.org/job/nova-milestone/60/ | 09:22 |
| openstackjenkins | Tarmac: backport of arping for floating ips. | 09:22 |
| *** wwkeyboard has joined #openstack-dev | 09:25 | |
| openstackjenkins | Project nova-milestone build #61: SUCCESS in 16 sec: https://jenkins.openstack.org/job/nova-milestone/61/ | 09:32 |
| openstackjenkins | Tarmac: removes warning set forth in d3 for deprecated setting of bridge automagically | 09:32 |
| *** rohitkarajgi has joined #openstack-dev | 09:33 | |
| openstackjenkins | Project nova-milestone build #62: SUCCESS in 14 sec: https://jenkins.openstack.org/job/nova-milestone/62/ | 09:41 |
| openstackjenkins | Tarmac: backport of lp:~yamahata/nova/lp851218 which was approved which fixes | 09:41 |
| openstackjenkins | https://bugs.launchpad.net/nova/+bug/851218 | 09:41 |
| uvirtbot | Launchpad bug 851218 in nova "no swap if block-device-mapping of /dev/vdb=swap" [Undecided,In progress] | 09:42 |
| openstackjenkins | Project nova build #1,398: SUCCESS in 4 min 0 sec: https://jenkins.openstack.org/job/nova/1398/ | 09:45 |
| openstackjenkins | Tarmac: Add iptables filter rules for dnsmasq (lp:844935) | 09:45 |
| openstackjenkins | On Fedora, the default policy for the INPUT chain in the filter table | 09:45 |
| openstackjenkins | is DROP. This means that DHCP and DNS request packets from the guest | 09:45 |
| openstackjenkins | get dropped. | 09:45 |
| openstackjenkins | Add these rules to allow the traffic through: | 09:45 |
| openstackjenkins | ��$> sudo iptables -t filter -A nova-network-INPUT -i br0 -p udp -m udp --dport 67 -j ACCEPT | 09:45 |
| smoser | soren, i addressed your comments and pushed to https://code.launchpad.net/~smoser/nova/lp853330/+merge/76164 | 09:45 |
| openstackjenkins | ��$> sudo iptables -t filter -A nova-network-INPUT -i br0 -p tcp -m tcp --dport 67 -j ACCEPT | 09:45 |
| openstackjenkins | ��$> sudo iptables -t filter -A nova-network-INPUT -i br0 -p udp -m udp --dport 53 -j ACCEPT | 09:45 |
| openstackjenkins | $> sudo iptables -t filter -A nova-network-INPUT -i br0 -p tcp -m tcp --dport 53 -j ACCEPT | 09:45 |
| vishy | core devs: rubber stamp review here: https://code.launchpad.net/~vishvananda/nova/fix-swap-manage/+merge/76181 | 10:02 |
| openstackjenkins | Project nova build #1,399: SUCCESS in 4 min 19 sec: https://jenkins.openstack.org/job/nova/1399/ | 10:21 |
| openstackjenkins | Tarmac: convert images that are not 'raw' to 'raw' during caching to node | 10:21 |
| openstackjenkins | Project nova-milestone build #63: SUCCESS in 19 sec: https://jenkins.openstack.org/job/nova-milestone/63/ | 10:22 |
| openstackjenkins | Tarmac: Add iptables filter rules for dnsmasq (lp:844935) | 10:22 |
| openstackjenkins | On Fedora, the default policy for the INPUT chain in the filter table | 10:22 |
| openstackjenkins | is DROP. This means that DHCP and DNS request packets from the guest | 10:22 |
| openstackjenkins | get dropped. | 10:22 |
| openstackjenkins | Add these rules to allow the traffic through: | 10:22 |
| openstackjenkins | $> sudo iptables -t filter -A nova-network-INPUT -i br0 -p udp -m udp --dport 67 -j ACCEPT | 10:22 |
| openstackjenkins | $> sudo iptables -t filter -A nova-network-INPUT -i br0 -p tcp -m tcp --dport 67 -j ACCEPT | 10:22 |
| openstackjenkins | $> sudo iptables -t filter -A nova-network-INPUT -i br0 -p udp -m udp --dport 53 -j ACCEPT | 10:22 |
| openstackjenkins | $> sudo iptables -t filter -A nova-network-INPUT -i br0 -p tcp -m tcp --dport 53 -j ACCEPT | 10:22 |
| openstackjenkins | Project nova-milestone build #64: SUCCESS in 16 sec: https://jenkins.openstack.org/job/nova-milestone/64/ | 10:32 |
| openstackjenkins | Tarmac: convert images that are not 'raw' to 'raw' during caching to node | 10:32 |
| smoser | anyone have thoughts on how i'd fix bug 854614 | 10:45 |
| uvirtbot | Launchpad bug 854614 in nova "metadata service local-hostname is not fqdn" [Undecided,New] https://launchpad.net/bugs/854614 | 10:45 |
| zykes- | jaypipes: ? | 10:49 |
| *** markvoelker has joined #openstack-dev | 11:03 | |
| *** bsza has joined #openstack-dev | 11:15 | |
| *** wwkeyboard has quit IRC | 11:28 | |
| *** lorin1 has joined #openstack-dev | 11:46 | |
| *** markmc has joined #openstack-dev | 11:49 | |
| *** wwkeyboard has joined #openstack-dev | 11:51 | |
| *** wwkeyboard has quit IRC | 11:57 | |
| soren | AAAARGH | 11:57 |
| soren | That is all. Move along. | 11:57 |
| vishy | soren: +1 | 12:00 |
| hugokuo | pound could only handle 600 request ? | 12:01 |
| hugokuo | but the proxy-server can handle more than 600 though | 12:02 |
| *** lorin1 has quit IRC | 12:04 | |
| *** lts has joined #openstack-dev | 12:06 | |
| *** dprince has joined #openstack-dev | 12:28 | |
| *** lorin1 has joined #openstack-dev | 12:29 | |
| *** mfer has joined #openstack-dev | 12:30 | |
| vishy | soren: are you here today? | 12:47 |
| *** lorin1 has left #openstack-dev | 12:49 | |
| *** hugokuo has quit IRC | 12:49 | |
| dprince | vishy: Heads up. I haven't had a working nova/glance/keystone setup since Friday. Some of it was packaging issues but things appear to be in free fall... | 12:50 |
| vishy | dprince: I hear the same thing from cloudbuilders sf | 12:51 |
| dprince | vishy: I had a subset of things working over the weekend. XenServer works. Libvirt w/ EC2 however hasn't been good. | 12:51 |
| vishy | aparently the last set of changes to keystone broke everything | 12:51 |
| dprince | vishy: just a bit frustrating since we have systems to catch these things. | 12:51 |
| vishy | dprince: you mean trunk? | 12:51 |
| vishy | dprince: I've basically abandoned trunk for the moment. Spending all my time with milestone | 12:52 |
| dprince | vishy: yes. I haven't tried diablo in awhile. I wouldn't be surprised if there were issues there too however. | 12:52 |
| vishy | dprince: we really need to focus on diablo. Release is in two days | 12:52 |
| dprince | vishy: awhile == 3/4 days. | 12:52 |
| vishy | dprince: I've been managing backported bugs, which is about all of the bandwidth I have | 12:53 |
| vishy | i'm trying to get SF to verify integration between the systems | 12:53 |
| vishy | they are running into the same issues as you with keystone | 12:53 |
| vishy | we have to get that fixed now | 12:53 |
| dprince | vishy: sure... but if we were really doing CI right trunk wouldn't break either. The frustrating thing is branches need to land in trunk first before they get into diablo. So I see us in bad shape. | 12:53 |
| vishy | dprince: It would be awesome if you guys could help test/fix diablo | 12:54 |
| dprince | vishy: getting ready to push a small keystone fix. | 12:54 |
| vishy | dprince: cool | 12:54 |
| *** bcwaldon has joined #openstack-dev | 12:55 | |
| openstackgerrit | Dan Prince proposed a change to openstack/keystone: Update validate_service_or_keystone_admin_token so that it doesn't cause exceptions if the admin or service admin haven't been configured. https://review.openstack.org/486 | 12:57 |
| *** rohitkarajgi has quit IRC | 12:58 | |
| *** Guest9842 has joined #openstack-dev | 12:59 | |
| Guest9842 | hi stackers, why i get 'n/a HTTP 400' after 'nova list' but euca-describe-instances lists correct instances | 13:00 |
| bcwaldon | Guest9842: your api probably didn't start up correctly | 13:00 |
| bcwaldon | Guest9842: look in nova-api.log | 13:00 |
| zykes- | any keystone folks here ? | 13:01 |
| Guest9842 | it seems working i get correct GET request-response visible after euca, let me try restart | 13:02 |
| Guest9842 | oh, i try restart nova-api | 13:04 |
| Guest9842 | (nova): TRACE: ImportError: cannot import name v1_1 | 13:05 |
| Guest9842 | but novaclient is installed and this was working yesterday | 13:05 |
| bcwaldon | Guest9842: can you show me the full traceback? | 13:08 |
| Guest9842 | 2011-09-20 06:04:27,631 CRITICAL nova [-] cannot import name v1_1 | 13:10 |
| bcwaldon | that's all you get? | 13:10 |
| Guest9842 | wait | 13:10 |
| Guest9842 | i paste in pastebin | 13:10 |
| bcwaldon | yessir | 13:10 |
| Guest9842 | http://paste.openstack.org/show/2481/ | 13:11 |
| bcwaldon | Ok, so it looks like you don't have an up-to-date novaclient on that machine | 13:12 |
| bcwaldon | can you tell me what version you have installed? | 13:12 |
| Guest9842 | 2.6.1~bzr89-0ubuntu0ppa1~maverick1 | 13:12 |
| bcwaldon | can you try to upgrade that package? | 13:13 |
| Guest9842 | yes | 13:14 |
| Guest9842 | itry add-apt-repository ppa:nova-core/trunk; apt-get update; apt-get install python-novaclient but it install same version, what am i wrong | 13:16 |
| bcwaldon | Guest9842: what's the output of apt-get install | 13:18 |
| *** rbasak has joined #openstack-dev | 13:19 | |
| Guest9842 | first i do apt-get remove python-novaclient and then install | 13:19 |
| Guest9842 | Setting up python-novaclient (2.6.1~bzr89-0ubuntu0ppa1~maverick1) ... | 13:20 |
| bcwaldon | hmm, maybe we havent pushed the latest ppas yet | 13:20 |
| bcwaldon | you can definitely install from pip and get the latest | 13:20 |
| *** kbringard has joined #openstack-dev | 13:20 | |
| Guest9842 | pip install ? thank you | 13:20 |
| bcwaldon | yep | 13:21 |
| Daviey | ttx: Wanted to talk to you about post release updates.. | 13:21 |
| zykes- | hmm, when you auth towards keystone shouldn't it get a list of the users roles ? | 13:21 |
| Daviey | .. Does it make sense for Ubuntu to just do it's own thing, and carry patches.. or us to fork the release and make our own fixes branch, or for openstack to to maintain a -fixes branch? | 13:22 |
| ttx | Daviey: are those the same quesions zul just asked me ? | 13:22 |
| *** amccabe has joined #openstack-dev | 13:22 | |
| zul | yes | 13:22 |
| ttx | Daviey: we won't maintain a -fixes branch, since it would not precisely match Ubuntu SRU guidelines | 13:22 |
| Daviey | ttx: Possibly.. PM's suck for this very reason. | 13:22 |
| ttx | so it would help nobody | 13:23 |
| vishy | Guest9842: did you apt-get update in between? | 13:23 |
| ttx | Daviey: between the two first options, whatever fits you best | 13:23 |
| *** glenc_ has joined #openstack-dev | 13:23 | |
| bcwaldon | Guest9842: just fyi, novaclient is at 2.6.4 in pypi | 13:23 |
| ttx | Daviey: given the LTS-1 nature of Diablo, I would say the former, but that's just me | 13:23 |
| Guest9842 | yes i do , but now i try pip install --upgrade python-novaclient and it is success | 13:23 |
| Guest9842 | thank you san | 13:23 |
| Daviey | ttx: So you have no problem with us essentially forking the release and creating a -fixes branch for our use? | 13:23 |
| ttx | Daviey: no -- it's not worse than carrying a patch tree | 13:24 |
| bcwaldon | Guest9842: you are welcome! | 13:24 |
| ttx | Daviey: at least if they are under DVCS they are reusable | 13:24 |
| Daviey | yeah.. flat patches can kinda suck. | 13:25 |
| Daviey | zul: thoughts? | 13:25 |
| zul | +1 from me | 13:25 |
| *** glenc has quit IRC | 13:26 | |
| zul | Daviey: we do bastard hybrids for SRUs anyways ;) | 13:28 |
| Daviey | :) | 13:29 |
| *** mnour has quit IRC | 13:42 | |
| *** Binbin has joined #openstack-dev | 13:42 | |
| *** mandela123 has quit IRC | 13:44 | |
| *** rmk` has joined #openstack-dev | 13:44 | |
| *** rmk has quit IRC | 13:46 | |
| *** rcc has joined #openstack-dev | 13:49 | |
| openstackgerrit | Anne Gentle proposed a change to openstack/openstack-manuals: Adds admin manuals for Identity, Image, Compute, and Object Storage https://review.openstack.org/487 | 13:50 |
| *** rcc1 has quit IRC | 13:51 | |
| soren | vishy: I am, yes, but was out for a bit to pick up a new phone. | 13:53 |
| *** errr has quit IRC | 13:53 | |
| *** smoser has quit IRC | 13:53 | |
| vishy | soren: cool | 13:53 |
| vishy | soren: I was just wondering what you think about libvirt 9.5 backports | 13:54 |
| *** deshantm has quit IRC | 13:54 | |
| vishy | soren: I'd really like to be able to use natty/oneric nova on cloudservers | 13:54 |
| *** glenc_ is now known as glenc | 13:55 | |
| *** errr has joined #openstack-dev | 13:55 | |
| *** smoser has joined #openstack-dev | 13:55 | |
| *** deshantm has joined #openstack-dev | 13:55 | |
| soren | vishy: I have a fix for that. | 13:56 |
| *** wwkeyboard has joined #openstack-dev | 13:56 | |
| *** Gordonz has joined #openstack-dev | 13:56 | |
| vishy | soren: there is a fix that doesn't require 9.5? | 13:56 |
| vishy | The bug states that the patch would be in that release | 13:57 |
| vishy | soren: but if you have a workaround I will be a happy panda | 13:57 |
| *** Rajaram has quit IRC | 13:58 | |
| *** Gordonz_ has joined #openstack-dev | 13:59 | |
| soren | vishy: 0.9.5 has my patch? | 14:00 |
| soren | Hm... I thought it wasn't going t obe there until the next one. | 14:00 |
| soren | Anyway, I don't mind a full 0.9.5 in the PPA. | 14:00 |
| *** Rajaram has joined #openstack-dev | 14:00 | |
| soren | libvirt is very well-behaved. | 14:00 |
| soren | ...and I'm perfectly comfortable debugging it, if there's a problem. | 14:00 |
| vishy | soren: 9.5 is the next one. It is in rc right now :) | 14:02 |
| soren | Oh, ok. | 14:02 |
| * soren lost track somewhere :) | 14:03 | |
| vishy | soren: is there a decent workaround besides building libvirt from source? | 14:03 |
| soren | vishy: No. | 14:03 |
| soren | Hm.. | 14:03 |
| soren | No. | 14:03 |
| *** Gordonz has quit IRC | 14:03 | |
| soren | It blows up because there's no PCI bus. | 14:03 |
| soren | For some reason that's considered fatal. | 14:04 |
| soren | Well, it used to. | 14:04 |
| mtaylor | soren: thanks for fixing nova-1 | 14:04 |
| soren | mtaylor: Sure thing. I hope I didn't destroy any debugging evidence you needed. | 14:04 |
| * soren is on a rather sketchy internet connectio nright now, by the way. | 14:05 | |
| *** Guest9842 has quit IRC | 14:06 | |
| mtaylor | soren: nope. that was a weirdness that happened a while back on the other machine when we transitioned from pip to apt packages for that. I should have thought to check it yesterday when ttx reenabled nova-1 | 14:06 |
| soren | mtaylor: Oh, perhaps the coverage job is screwed there, too, then. | 14:07 |
| ttx | mtaylor: how close are we to fixing bug 848978 (on milestone-proposed at least ?) | 14:07 |
| uvirtbot | Launchpad bug 848978 in openstack-ci "Nova fails to build in PPA for Lucid" [Medium,New] https://launchpad.net/bugs/848978 | 14:07 |
| mtaylor | soren: ok. all fixed | 14:10 |
| mtaylor | ttx: the fix should be in the nova packaging now | 14:10 |
| mtaylor | although lemme look at unittest2 real quick | 14:11 |
| ttx | mtaylor: still fails as of 3 hours ago | 14:11 |
| mtaylor | ttx: GRRRR | 14:11 |
| ttx | https://launchpad.net/~nova-core/+archive/milestone-proposed/+packages | 14:11 |
| dprince | soren/mtaylor: https://code.launchpad.net/~dan-prince/nova/ubuntu-nova-fix-dnsmasq-sudoers/+merge/76223 | 14:11 |
| mtaylor | ttx: gotcha | 14:11 |
| ttx | Missing build dependencies: python-unittest2 | 14:11 |
| dprince | https://bugs.launchpad.net/nova/+bug/854731 | 14:12 |
| uvirtbot | Launchpad bug 854731 in nova "PPA: nova_sudoers has bad dnsmasq line" [Medium,In progress] | 14:12 |
| mtaylor | dprince: should that also go to diablo? | 14:12 |
| dprince | Also. The libvirt issue vish mentioned above is what keeps me from using the lp:~openstack-ubuntu-packagers/nova/ubuntu. | 14:12 |
| dprince | I recently updated openstack VPC so that I can use trunk more closely so hopefully I'll find things like this a bit more quickly. | 14:13 |
| dprince | Essentially the PPA package hard coded the required libvirt to 0.8.8 which doesn't work on cloud servers. | 14:13 |
| dprince | So I do this now: | 14:13 |
| dprince | https://github.com/dprince/openstack_vpc/commit/8deec43ebed2bda3d995b4de4d101de7326ec7f4 | 14:13 |
| dprince | Anyway. I think that should go into diablo as well. Its a show stopper for anyone using the PPA packages. | 14:14 |
| mtaylor | great. | 14:16 |
| mtaylor | soren: you and vish were talking about libvirt earlier - do you have any opinions on the libvirt/nova-ppa/cloudservers issue? | 14:16 |
| vishy | that is axactly what we were discussing | 14:16 |
| vishy | we either need to use a backported 8.3 | 14:16 |
| vishy | or forward-ported that is | 14:17 |
| vishy | or a backported 9.5 (which isn't released) | 14:17 |
| vishy | it works fine on maverick in cloudservers which still has 8.3.1 available | 14:17 |
| mtaylor | hrm. | 14:17 |
| dprince | I always downgrade to 0.8.3 and that is how I get it working. I don't think that supports the latest LXC code in nova however. | 14:17 |
| vishy | but you can't force install the older package on natty | 14:18 |
| vishy | dprince: same | 14:18 |
| dprince | right. yep. holding me back as well. | 14:18 |
| mtaylor | I'd kind of personally prefer to move in the direction of 0.9.5 being in our ppa | 14:18 |
| mtaylor | but I don't know what that means for the oneiric release | 14:18 |
| vishy | mtaylor: well i don't think there is a huge use case outside of dev land for it | 14:19 |
| mtaylor | any idea when 0.9.5 is supposed to release? as in - is there any chance ubuntu would be able to update their oneiric version before oneiric drops? | 14:19 |
| mtaylor | vishy: for oneiric? | 14:19 |
| mtaylor | agree | 14:19 |
| vishy | so we could put it in trunk ppa and leave it out of milestone proposed | 14:19 |
| mtaylor | but I don't want to hamstring the ubuntu guys | 14:20 |
| vishy | mtaylor: 9.5 is in rc3 so i assume soon | 14:20 |
| mtaylor | vishy: AHH. yes. I see the words you're saying now | 14:20 |
| vishy | mtaylor: actually | 14:20 |
| vishy | libvirt-python-0.9.5-1.fc14.x86_64.rpm20-Sep-2011 08:29261K | 14:20 |
| vishy | apparently it is released | 14:21 |
| soren | I don't mind 0.9.5 in tha PPA's at all. | 14:21 |
| vishy | as of today | 14:21 |
| mtaylor | let's do that then | 14:21 |
| soren | ...but backporting the fix to older libvirts is simple. | 14:21 |
| soren | So we can do whatever. | 14:21 |
| soren | It's about the same amount of work. | 14:21 |
| ttx | vishy: should we backport https://code.launchpad.net/~doude/nova/lp838154/+merge/75692 to Diablo ? | 14:21 |
| ttx | bug 835154 | 14:22 |
| uvirtbot | Launchpad bug 835154 in ensemble "Ensemble should depend on ssh-client" [Medium,Fix released] https://launchpad.net/bugs/835154 | 14:22 |
| ttx | arh | 14:22 |
| ttx | bug 838154 | 14:22 |
| vishy | ttx: I think i did already | 14:22 |
| uvirtbot | Launchpad bug 838154 in nova "LXC instance without injected key doesn't get created" [High,Fix committed] https://launchpad.net/bugs/838154 | 14:22 |
| ttx | ah. let me doublecheck that for you | 14:22 |
| * soren disappears for bit to attend a friend's birthday party | 14:22 | |
| mtaylor | soren: let's put 0.9.5 in the ppa. do you know already if the patches you have in our current libvirt ppa package have been applied upstream? or should I walk through and check them all? | 14:22 |
| vishy | ttx: hmm i don't see it | 14:22 |
| vishy | nope not in | 14:23 |
| vishy | i'll backport | 14:23 |
| ttx | it used to be proposed -- but there was some issues with the proposal | 14:24 |
| ttx | you rejected it a few hours ago | 14:25 |
| vishy | ttx: right, it had merged trunk | 14:26 |
| vishy | i intended to backport it but apparently i missed it. | 14:27 |
| vishy | backported here: https://code.launchpad.net/~vishvananda/nova/lp838154-ms/+merge/76227 | 14:27 |
| *** jaypipes has quit IRC | 14:30 | |
| vishy | ttx: any other ones that you noticed? | 14:30 |
| *** joesavak has joined #openstack-dev | 14:31 | |
| *** ahmed_ has quit IRC | 14:32 | |
| *** vladimir3p has joined #openstack-dev | 14:32 | |
| vishy | mtaylor: are you doing the sudoers fix for packaging that dprince proposed? | 14:33 |
| vishy | it is causing issues for people in #openstack | 14:33 |
| *** dolphm has joined #openstack-dev | 14:33 | |
| ttx | vishy: Was wondering about https://bugs.launchpad.net/nova/+bug/837463 | 14:34 |
| uvirtbot | Launchpad bug 837463 in nova "Starting nova-compute fails on non-x86 architecture." [Low,Fix committed] | 14:34 |
| ttx | not so sure about impact | 14:34 |
| *** AhmedSoliman has joined #openstack-dev | 14:34 | |
| vishy | ttx: yeah that one seems totally reasonable | 14:35 |
| vishy | didn't realize that didn't make it in | 14:35 |
| dprince | ttx: Yes. That was zul's fix. Its fine. | 14:35 |
| vishy | ttx: i wonder when he merged trunk | 14:35 |
| vishy | i probably need to backport it | 14:35 |
| ttx | fwiw, I'm looking at the list of fixcommitted bugs that are not targeted to 2011.3 | 14:36 |
| *** zul has quit IRC | 14:36 | |
| ttx | vishy, dprince: maybe bug 834030 | 14:39 |
| uvirtbot | Launchpad bug 834030 in nova "'invalid literal for int' when listing a image snapshot" [Medium,Fix committed] https://launchpad.net/bugs/834030 | 14:39 |
| vishy | dprince: will that apply to milestone-proposed? | 14:40 |
| ttx | also the DB migration 047 issues only affect Essex, right | 14:41 |
| vishy | right | 14:41 |
| dprince | vishy: I'll look at it. You want a branch for that one? | 14:41 |
| vishy | dprince: if it doesn't depend on stuff that isn't in mp, then yes | 14:41 |
| ttx | that's all my sweep caught. We could really use a nice dashboard with the list of trunk commits and backport/ignore_with_comment actions :) That would ensure we don't miss any. | 14:42 |
| dprince | vishy: give me a bit. still bug searching and dealing w/ some keystone confusions. | 14:42 |
| * ttx will try to get that working for Essex release | 14:42 | |
| vishy | dprince: ok | 14:43 |
| *** zul has joined #openstack-dev | 14:43 | |
| vishy | ttx: yes, it has been a bit painful to manage | 14:43 |
| vishy | ttx: I'm beginning to come around to the idea of a 1 month freeze | 14:44 |
| vishy | :) | 14:44 |
| ttx | vishy: heh | 14:44 |
| ttx | freezes definitely make *our* job easier. | 14:44 |
| vishy | I want to force people to actually test it | 14:45 |
| vishy | not sure how to do it since all of the teams have internal requirements as well | 14:45 |
| ttx | vishy: or at least entice them. | 14:45 |
| vishy | beer? | 14:45 |
| vishy | torture? | 14:45 |
| vishy | both? | 14:45 |
| ttx | I'm more of a torture guy than a beer guy. | 14:45 |
| *** blamar_ has joined #openstack-dev | 14:46 | |
| nhm | vishy: beer torture, ie, make them drink coors light? | 14:46 |
| ttx | I'll continue to say things like that until we have a quotes page on the wiki. | 14:46 |
| *** jsavak has joined #openstack-dev | 14:46 | |
| openstackjenkins | Project nova-milestone build #65: SUCCESS in 16 sec: https://jenkins.openstack.org/job/nova-milestone/65/ | 14:47 |
| openstackjenkins | Tarmac: Authorize to start a LXC instance withour, key, network file to inject or metadata. | 14:47 |
| ttx | vishy: adding 834030 and 837463 to targets for tracking | 14:47 |
| vishy | k just approved | 14:47 |
| notmyname | ttx: I've been getting a lot of emails recently for nova merges into ~hudson-openstack/nova. the volume has really stepped up, and I'd prefer to not get these emails | 14:49 |
| *** joesavak has quit IRC | 14:50 | |
| ttx | notmyname: hmm, that's as part of release team. I can remove you from that team, at least temporarily (this stuff needs to be updated for Gerrit anyway) | 14:50 |
| ttx | notmyname: works for you ? | 14:50 |
| notmyname | ttx: ya, that would be fine :-) | 14:51 |
| notmyname | thanks | 14:51 |
| ttx | there must be a way to filter them out, but that will be faster | 14:51 |
| notmyname | ya, I can always set a rule, but it would end up being "mark as read, put in folder that I never open" | 14:51 |
| ttx | notmyname: done | 14:51 |
| * ttx has been a bit overwhelmed with bugmail recently too | 14:52 | |
| openstackgerrit | A change was merged to openstack/openstack-manuals: Adds admin manuals for Identity, Image, Compute, and Object Storage https://review.openstack.org/487 | 14:54 |
| *** code_franco has joined #openstack-dev | 14:58 | |
| dprince | Okay. Just confirmed this one. | 14:59 |
| dprince | https://bugs.launchpad.net/nova/+bug/854585 | 14:59 |
| uvirtbot | Launchpad bug 854585 in nova "nova list not showing networks info" [Undecided,In progress] | 14:59 |
| dprince | The OSAPI isn't showing networks in trunk. | 14:59 |
| mtaylor | dprince: if I hand you a new libvirt package, can you test it in nova_vpc before I shove it in the trunk ppa? | 15:01 |
| dprince | mtaylor: Sure. Nova isn't working great but I can kick the tires on it a bit I think with something. | 15:02 |
| mtaylor | dprince: k. the debian packaging just has a lot of patches - some of which have been applied upstream. I mainly wanted to ensure I hadn't missed anything major as far as that goes | 15:02 |
| mtaylor | ttx: bug 848978 should be sorted now | 15:04 |
| uvirtbot | Launchpad bug 848978 in openstack-ci "Nova fails to build in PPA for Lucid" [Medium,New] https://launchpad.net/bugs/848978 | 15:04 |
| mtaylor | vishy: any yeah, I'll pick up dprince's fix | 15:04 |
| *** dragondm has joined #openstack-dev | 15:06 | |
| openstackjenkins | Project nova-milestone build #66: SUCCESS in 15 sec: https://jenkins.openstack.org/job/nova-milestone/66/ | 15:06 |
| openstackjenkins | Tarmac: Backport of fix to for libvirt running on non-x86 architecture. | 15:06 |
| *** jdag has joined #openstack-dev | 15:10 | |
| dolphm | dprince: just had a meeting with ziad, yogi & joe... the consensus to was to use the backend-managed identifiers throughout all API requests (resources & body's)... which implies that auth { tenantId } will stay tenantId -- we just need to update both samples and sampledata.py to better illustrate tenant names (using strings that don't look like ints) | 15:10 |
| dprince | dolphm: Okay. Cool. So I filed this yesterday: | 15:12 |
| dprince | https://bugs.launchpad.net/keystone/+bug/854228 | 15:12 |
| uvirtbot | Launchpad bug 854228 in keystone "update docs and templates to use tenant IDs (instead of names)" [Undecided,In progress] | 15:13 |
| dprince | dolphm: I think we are on the same page then right? | 15:13 |
| dprince | dolphm: Essentially getting rid of the tenants named "1234" in the examples, etc. | 15:14 |
| dolphm | exactly | 15:14 |
| *** rnirmal has joined #openstack-dev | 15:14 | |
| dolphm | dprince: let me know if you run across any other inconsistencies like this -- fixing these kinks is my only goal today | 15:15 |
| dprince | dolphm: Do you want to take that ticket then? I won't get to it this morning. Probably tonight. | 15:16 |
| dolphm | dprince: done | 15:17 |
| dprince | dolphm: If this is on the top of your list then feel free to grab it is what I'm saying. Otherwise I'm happy to take it once I get a couple nova things done. | 15:17 |
| dprince | dolphm: great. Thank you sir. | 15:17 |
| dolphm | dprince: /salute | 15:17 |
| dprince | dolphm: Actually. If you guys could take a look at this it would help too: | 15:18 |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Changing/introducing actual extension json/xml snippets. https://review.openstack.org/484 | 15:18 |
| dprince | dolphm: https://review.openstack.org/#change,486 | 15:18 |
| *** cp16net has joined #openstack-dev | 15:19 | |
| dprince | dolphm: That is a show stopper for my keystone tests (I don't currently setup per service admin's) | 15:19 |
| dolphm | dprince: wouldn't it be more appropriate to only check for a role for service-admin's then? or you think they should both be optional | 15:20 |
| dprince | dolphm: I mean if they both get set to None your going to get a 400 back right? | 15:20 |
| dprince | dolphm: I can make service-admin optional and require the admin role. However we should still check for a NoneType role getting returned there as well. | 15:21 |
| dolphm | yes... something in the 400 range lol | 15:21 |
| openstackgerrit | A change was merged to openstack/keystone: Changing/introducing actual extension json/xml snippets. Adding updated documents. https://review.openstack.org/484 | 15:21 |
| *** zaitcev has joined #openstack-dev | 15:22 | |
| dolphm | dprince: i guess that's my question... if keystone is mis-configured (lacking an admin), I wouldn't want to suppress that | 15:22 |
| dolphm | dprince: but then the question should be: *is* keystone mis-configured if it doesn't have an admin? i suppose if your backend doesn't need management, you don't need an admin.. | 15:23 |
| dprince | dolphm: Sure. I can update the code to check for None and log an error if the admin role is missing then. I hate seeing stack traces and having to guess at the problem. | 15:23 |
| dolphm | dprince: that sounds like a better solution | 15:24 |
| dprince | dolphm: Okay. Give me a sec and I'll push a new one. Thanks again. | 15:24 |
| dolphm | dprince: np | 15:24 |
| *** reidrac has quit IRC | 15:28 | |
| zykes- | dolphm: why doesn't auth return roles when authenticating ? | 15:32 |
| zykes- | like in examples... | 15:33 |
| dolphm | zykes-: because our code is lagging behind the spec :) | 15:33 |
| zykes- | k | 15:33 |
| zykes- | how does services get roles now then? based on tokens? | 15:33 |
| dolphm | I believe so... GET /tokens/{token_id} | 15:35 |
| zykes- | and for a user then, since glance checks for Admin how does it get that ? | 15:35 |
| dolphm | zykes-: GET /users/{user_id}/roleRefs .... or GET /users/{user_id}/roles, depending on whether you're looking at docs implementation | 15:37 |
| dolphm | docs or* implementation | 15:37 |
| openstackgerrit | Dan Prince proposed a change to openstack/keystone: Update validate_service_or_keystone_admin_token so that it doesn't cause exceptions if the admin or service admin haven't been configured. https://review.openstack.org/488 | 15:38 |
| dprince | dolphm: try this one: https://review.openstack.org/#change,488 | 15:39 |
| dolphm | dprince: approved a couple minutes ago :) | 15:40 |
| openstackgerrit | A change was merged to openstack/keystone: Update validate_service_or_keystone_admin_token so that it doesn't cause exceptions if the admin or service admin haven't been configured. https://review.openstack.org/488 | 15:41 |
| dolphm | dprince: ^^^ | 15:41 |
| dprince | dolphm: thanks. That helps. | 15:42 |
| *** Rajaram_ has joined #openstack-dev | 15:43 | |
| *** Rajaram has quit IRC | 15:43 | |
| *** Rajaram_ is now known as Rajaram | 15:43 | |
| *** tryggvil has joined #openstack-dev | 15:46 | |
| *** jsavak has quit IRC | 15:56 | |
| vishy | dolphm: the keystone-import script looks like it is using the wrong patch | 15:57 |
| vishy | s/patch/path | 15:57 |
| vishy | dolphm: ./bin/keystone-manage doesn't seem like it will work because they are in the same directory... | 15:58 |
| vishy | dolphm: should it be ./keystone-manage ? | 15:58 |
| *** rnorwood has joined #openstack-dev | 16:00 | |
| dolphm | vishy: i think it's just intended to be called from outside the root directory, right? | 16:04 |
| dolphm | err, outside the bin/ directory | 16:04 |
| vishy | dolphm: but in that case it still won't work | 16:04 |
| mtaylor | ugh. there are a surprising number of patches in the libvirt packages which have NOT been applied upstream | 16:04 |
| vishy | unless you happen to be in the keystone directory | 16:04 |
| dolphm | that's what i mean | 16:04 |
| mtaylor | given that upstream went from 0.8.8 to 0.9.5 - I would say that somewhere there a process totally failed | 16:04 |
| vishy | but what if you install from packages | 16:04 |
| *** ameade has joined #openstack-dev | 16:04 | |
| vishy | dolphm: etc... | 16:05 |
| vishy | anyway just seems a little fragile | 16:05 |
| dolphm | vishy: agree | 16:05 |
| dolphm | vishy: do you need to call keystone-import while installing from a package? | 16:06 |
| dolphm | vishy: i see the intention, but i don't know what the business case is for keystone-import | 16:06 |
| vishy | dolphm: you need to use it when upgrading from an old nova install | 16:07 |
| vishy | nova-manage shell export auth.txt | 16:07 |
| vishy | keystone-import auth.txt | 16:07 |
| dolphm | vishy: i feel like i fixed this same bug in the old keystone-manage, prior to packaging efforts... lol | 16:08 |
| mtaylor | vishy: any chance you happen to know ... do we need 0.9.5 specifically? or is 0.9.4 good enough? | 16:08 |
| vishy | mtaylor: has to be 9.5 | 16:09 |
| mtaylor | vishy: k. thanks. good to know | 16:09 |
| dolphm | vishy: would something like this fix it: `dirname $0`/keystone-manage $line | 16:15 |
| mtaylor | dolphm: actually, as much as that gives me hives, I think it would solve both use cases - in tree and installed | 16:17 |
| vishy | yeah that might do it | 16:17 |
| dolphm | mtaylor: lol | 16:17 |
| dolphm | <-- not much of a sysadmin | 16:18 |
| mtaylor | dolphm: as a second patch, keystone-import could be written as a python script which imports keystone.manage, does the reading of input and calling of the right commands | 16:18 |
| mtaylor | :) | 16:18 |
| mtaylor | dolphm: something like this: http://pastebin.com/jQqs7uz7 | 16:22 |
| *** mdomsch has joined #openstack-dev | 16:22 | |
| mtaylor | dolphm: (sorry - needed a mental distraction from packaging) | 16:22 |
| dolphm | mtaylor: possible_topdir gives me hives, but otherwise +1 :) | 16:24 |
| mtaylor | well yeah - I just copied that bit from keystone-manage - and have been meaning to make you a patch to get rid of that anyway | 16:25 |
| openstackgerrit | Monty Taylor proposed a change to openstack/keystone: Fixed path issues with keystone-import. https://review.openstack.org/489 | 16:36 |
| mtaylor | dolphm: ^^ give that a stab | 16:37 |
| *** rmk` has quit IRC | 16:37 | |
| *** rmk has joined #openstack-dev | 16:39 | |
| dprince | jkoelker_: you there? | 16:39 |
| *** tudamp_ has joined #openstack-dev | 16:40 | |
| *** jdurgin has joined #openstack-dev | 16:40 | |
| *** jaypipes has joined #openstack-dev | 16:41 | |
| *** joesavak has joined #openstack-dev | 16:46 | |
| dprince | Question on commit 1594 where we removed the fkeys for instances and virtual interfaces. Why did we do that. | 16:49 |
| dprince | ? | 16:49 |
| *** ameade has quit IRC | 16:49 | |
| dprince | The OSAPI relied on the virtual_interfaces fkey to obtain the IP information. So now we get no IPs via the API. | 16:49 |
| dprince | The commit went in so fast I missed it in merge prop so I'm just trying to catch up here... | 16:50 |
| dprince | tr3buchet/jkoelker_: ^^^^^ | 16:51 |
| tr3buchet | dprince: it would only work with network data stored in the nova database | 16:52 |
| dprince | tr3buchet: So this is for network as a service then? melange? | 16:52 |
| tr3buchet | yes, exactly | 16:52 |
| dprince | tr3buchet: How should we obtain the IP info now then? | 16:52 |
| tr3buchet | now if you want IP info you must request it from a network host | 16:53 |
| annegentle | sandywalsh: around? Couple of zone and endpoint questions for you. | 16:53 |
| tr3buchet | using the network api | 16:53 |
| dprince | tr3buchet: Okay. Is anybody working on that? | 16:54 |
| tr3buchet | dprince we're in the process of working more features into the network api as we speak | 16:54 |
| tr3buchet | yes i am | 16:54 |
| dprince | tr3buchet: Okay. So do you want to take this then: https://bugs.launchpad.net/nova/+bug/854585 | 16:54 |
| uvirtbot | Launchpad bug 854585 in nova "nova list not showing networks info" [Critical,In progress] | 16:54 |
| dprince | tr3buchet: I can't do much of anything with nova trunk without IPs. | 16:55 |
| tr3buchet | sure, can do. this is essex right? | 16:55 |
| dprince | yes. But IMHO we really shouldn't have merged anything in this state into trunk (Essex or whatever) | 16:56 |
| mtaylor | ++ | 16:56 |
| dprince | It just pulls the rug out from too much. | 16:56 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Revised samples https://review.openstack.org/490 | 16:56 |
| tr3buchet | true. as far as I knew this worked with the osapi | 16:56 |
| *** ameade has joined #openstack-dev | 16:56 | |
| dprince | Not sure who is clicking the approve button on these things but we have systems to catch this stuff. | 16:56 |
| tr3buchet | or rather, it did not break the osapi | 16:56 |
| dprince | we should use them | 16:56 |
| * mtaylor really wants nova_vpc to become a merge blocker | 16:57 | |
| tr3buchet | there are several functional test suites that need to be leveraged better | 16:57 |
| mtaylor | ++ | 16:57 |
| tr3buchet | that could be leveraged better i mean | 16:57 |
| tr3buchet | this is a perfect example of something that would have been caught | 16:58 |
| mtaylor | yeah - part of the problem right now is that they are being run, but nothing is forcing anyone to look at the results. this shall hopefully soon be fixed | 16:58 |
| dprince | mtaylor: I'll try and work w/ you on that. Getting the VPC thing to help gate things. (That is if you want my help) | 16:59 |
| mtaylor | dprince: totally. it becomes WAY easier to engineer after thursday | 16:59 |
| *** heckj has joined #openstack-dev | 16:59 | |
| dprince | Is that gerrit time for nova? | 16:59 |
| *** stanchan has joined #openstack-dev | 16:59 | |
| mtaylor | yup | 16:59 |
| dprince | cool. | 16:59 |
| mtaylor | at which point adding in additional jenkins jobs to the gate becomes a piece of cake | 17:00 |
| *** stanchan has quit IRC | 17:00 | |
| dprince | tr3buchet: Can you take this then: https://bugs.launchpad.net/nova/+bug/854585 | 17:00 |
| uvirtbot | Launchpad bug 854585 in nova "nova list not showing networks info" [Critical,In progress] | 17:00 |
| *** stanchan has joined #openstack-dev | 17:00 | |
| dprince | tr3buchet: we need to fix it ASAP. | 17:00 |
| tr3buchet | dprince: yes | 17:00 |
| openstackgerrit | A change was merged to openstack/keystone: Revised samples https://review.openstack.org/490 | 17:00 |
| dprince | tr3buchet: thanks man | 17:00 |
| heckj | mtaylor: missed the earlier part of that conversation - when are we talking about shifting nova -> github? | 17:01 |
| mtaylor | heckj: this week. as soon as we cut diablo we'll be doing the gerrit migration | 17:02 |
| openstackgerrit | A change was merged to openstack/identity-api: Changes to support extensions document.Also bringing cramers wadltodocbook changes.Changes to extension names. https://review.openstack.org/466 | 17:02 |
| heckj | mtaylor: cool, thanks. | 17:02 |
| mtaylor | (we didn't want to migrate anything before that, because, well eek) | 17:02 |
| heckj | mtaylor: yeah, good idea! | 17:03 |
| zykes- | why update sample when it doesn't work in the keystone program ? | 17:03 |
| zykes- | dolphm: after your change it's not returning role names anymore for stuff like glance that uses Admin as a role? | 17:06 |
| dolphm | zykes-: we're updating the implementation around /tokens right now, and it will go in today to match our samples/contract | 17:07 |
| zykes- | kk | 17:08 |
| zykes- | meaning the roles in auth as well will be thrown in ? | 17:08 |
| dolphm | zykes-: definitely | 17:09 |
| zykes- | oh | 17:09 |
| *** zul has quit IRC | 17:14 | |
| *** cburgess has joined #openstack-dev | 17:14 | |
| *** code_franco has quit IRC | 17:15 | |
| *** code_franco has joined #openstack-dev | 17:16 | |
| *** zul has joined #openstack-dev | 17:18 | |
| openstackgerrit | A change was merged to openstack/keystone: Fixed path issues with keystone-import. https://review.openstack.org/489 | 17:19 |
| sandywalsh | annegentle, hey! | 17:25 |
| annegentle | hey! | 17:25 |
| *** mattfarina has joined #openstack-dev | 17:26 | |
| *** mfer has quit IRC | 17:26 | |
| *** mattfarina is now known as mfer | 17:26 | |
| annegentle | I think I got the answer I needed from anotherjesse, basically that when you have a single region you only have to label the endpoint using keystone-manage endpointTemplate | 17:26 |
| annegentle | its when you got to multiple regions and zones that you have to do much more config with keystone | 17:26 |
| dprince | dolphm: you there? | 17:36 |
| *** martine has joined #openstack-dev | 17:37 | |
| *** martine has quit IRC | 17:37 | |
| *** AhmedSoliman has quit IRC | 17:38 | |
| *** martine has joined #openstack-dev | 17:38 | |
| *** AhmedSoliman has joined #openstack-dev | 17:38 | |
| sandywalsh | annegentle, correct, you have to add 1 endpoint per zone (per tenant). The video gives a demo of this. | 17:39 |
| *** martine has quit IRC | 17:39 | |
| *** dragondm has quit IRC | 17:40 | |
| *** Rajaram has quit IRC | 17:42 | |
| *** xavicampa has joined #openstack-dev | 17:46 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Renaming GET /tokens/{token_id} response container to 'access' https://review.openstack.org/491 | 17:53 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Renaming GET /tokens/{token_id} response container to 'access' https://review.openstack.org/491 | 17:55 |
| zykes- | dolphm: but for the token stuff will it return role names or ? | 17:56 |
| dolphm | zykes-: working on it :) | 17:57 |
| dolphm | (role name & id) | 17:57 |
| dolphm | zykes-: working toward this: https://github.com/openstack/keystone/blob/master/keystone/content/common/samples/validatetoken.json | 17:58 |
| openstackgerrit | A change was merged to openstack/keystone: Renaming GET /tokens/{token_id} response container to 'access' https://review.openstack.org/491 | 17:58 |
| dprince | vishy: you there? Got a question for you on how to check for the admin keystone role in nova.... | 18:02 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Renaming 'roleRefs' container to 'roles' https://review.openstack.org/492 | 18:10 |
| *** novas0x2a|laptop has joined #openstack-dev | 18:13 | |
| *** mgius has joined #openstack-dev | 18:21 | |
| *** xtor_ has joined #openstack-dev | 18:26 | |
| vishy | dprince: hey | 18:29 |
| vishy | dprince: why are we getting role_ids now? | 18:29 |
| vishy | dprince: I guess we can have a flag for admin_roles | 18:30 |
| dprince | vishy: yep. | 18:30 |
| vishy | dprince: i really don't like that they are numeric ids though | 18:30 |
| dprince | vishy: you buy this: https://bugs.launchpad.net/nova/+bug/854917 | 18:31 |
| uvirtbot | Launchpad bug 854917 in nova "update keystone role check to use IDs" [High,In progress] | 18:31 |
| dprince | vishy: we'll have to do it conditionally however. Only if the auth type is keystone. | 18:31 |
| vishy | dprince: https://github.com/openstack/keystone/blob/master/keystone/content/common/samples/validatetoken.json | 18:31 |
| vishy | why not use role name? | 18:31 |
| vishy | that seems a lot more useful than the id | 18:32 |
| dprince | sure. I'm not sure its in there. But if it is we can use it. | 18:32 |
| dprince | not all the examples are actually implemented are they? | 18:32 |
| vishy | perhaps not | 18:34 |
| vishy | we need some version of keystone to set on | 18:34 |
| vishy | dolphm: are you guys planning on an actaual release | 18:34 |
| vishy | * actual | 18:34 |
| *** martine has joined #openstack-dev | 18:35 | |
| dprince | vishy: but if we would rather go that way we can focus on that instead. (getting role names into the validatetoken output) | 18:35 |
| vishy | dprince: I think that ultimately we should be using role names not ids | 18:35 |
| vishy | eventually nova is supposed to create roles in keystone right? | 18:36 |
| dprince | vishy: I'm inclined to make this work however. Its been a landside of failure this week as far as integration accross services. | 18:36 |
| vishy | dprince: sure get it working | 18:36 |
| dolphm | vishy: the plan is yes, 6 weeks from yesterday | 18:36 |
| dolphm | ^ tentative plan i guess | 18:36 |
| uvirtbot | dolphm: Error: "tentative" is not a valid command. | 18:36 |
| vishy | dolphm: sigh, so we don't have keystone for diablo | 18:37 |
| jaypipes | wwkeyboard: welcome aaron :) | 18:38 |
| wwkeyboard | Thank you! | 18:38 |
| zykes- | no keystone for diablo ? :| | 18:38 |
| jaypipes | wwkeyboard: remember: "embrace the chaos" :) | 18:38 |
| dolphm | vishy: i think a subset of keystone will be documented & stable, and happy by the end of today -- hopefully an important subset | 18:38 |
| wwkeyboard | jaypipes: Thats the life of a developer! | 18:39 |
| vishy | dolphm: are role names part of that subset? | 18:39 |
| dolphm | e.g. (GET/POST/HEAD/DELETE) /tokens | 18:39 |
| jaypipes | wwkeyboard: indeed | 18:39 |
| dolphm | vishy: i would like them to be! | 18:39 |
| rmk | Ahh, you guys are talking about an issue I raised yesterday. | 18:40 |
| rmk | No one was paying attention to me. :) | 18:40 |
| dolphm | psh | 18:40 |
| vishy | dolphm: can we get a tag called diablo-prerelase or diablo-stable or something? | 18:40 |
| dolphm | diablo-preview? lol | 18:41 |
| vishy | dolphm: Preferably packaged I need something that I can deploy with the diablo milestones | 18:41 |
| dprince | dolphm: if you get that it would be great. I'm not able to use roles with nova right now. | 18:41 |
| vishy | s/milestone/releases | 18:41 |
| dolphm | vague-hints-of-diablo | 18:41 |
| rmk | I was speaking to dolphm about this, keystone needs a branch for diablo | 18:41 |
| dprince | rmk: yesterday a different set of things hosed me. I just now got to this failure... | 18:41 |
| vishy | we have to be able to put in documentation about what to use and we need a stable package to install | 18:41 |
| rmk | dprince: No worries, I was just being opportunistic | 18:41 |
| rmk | keystone has not branched yet | 18:42 |
| zykes- | dolphm: will the final diablo release be "according to spec"? | 18:42 |
| vishy | rmk: the way that it is supposed to work is fairly well expressed in python-novaclient | 18:42 |
| rmk | vishy: sure, you're expecting a role name back, that's what I thought too | 18:43 |
| dolphm | zykes-: of course, but the date of "final" isn't concrete | 18:43 |
| rmk | https://github.com/openstack/keystone/commit/35e4b2579164e763654e507ec215e51c72e5ab22#commitcomment-598575 | 18:43 |
| mtaylor | dolphm, vishy: how about we put some "diablo" keystone packages somewhere | 18:43 |
| rmk | You need a branch first | 18:44 |
| vishy | mtaylor: we definitely need to do that | 18:44 |
| openstackgerrit | Alex Meade proposed a change to openstack/glance: Added min_disk and min_ram properties to images https://review.openstack.org/429 | 18:44 |
| rmk | There currently isn't anything other than master | 18:44 |
| vishy | i need something stable that I can deploy with diablo nova glance swift | 18:44 |
| rmk | I'm glad this came up, I was concerned about it yesterday | 18:44 |
| mtaylor | dolphm: let me know a good v and I'll get us some packages published somewhere | 18:44 |
| dolphm | mtaylor: will do | 18:45 |
| dolphm | mtaylor: i propose tagging it as reheated-spaghetti | 18:45 |
| vishy | mtaylor: we don't have the v yet :) | 18:45 |
| *** martine has quit IRC | 18:45 | |
| vishy | today mb :) | 18:45 |
| mtaylor | vishy: well yeah :) | 18:45 |
| mtaylor | dolphm: and no :) | 18:45 |
| mtaylor | dolphm: or yes. what do I care ;) | 18:45 |
| dolphm | mtaylor: i'm open to other suggestions lol | 18:45 |
| mtaylor | dolphm: olbaid ? | 18:45 |
| dolphm | mtaylor: is that german or something? | 18:46 |
| mtaylor | dolphm: it's diablo backwards | 18:46 |
| mtaylor | btw - in case anyone was curious - libvirt is ludicrously frafile | 18:47 |
| mtaylor | fragile | 18:47 |
| mtaylor | not frafile | 18:47 |
| mtaylor | ok - it's probably frafile too | 18:47 |
| rmk | mtaylor: Are you handling the ubuntu packages for keystone? | 18:47 |
| mtaylor | rmk: yup. | 18:47 |
| rmk | mtaylor: OK, so last time I checked (2 days ago), there were a few issues with the package | 18:47 |
| mtaylor | yes. I have not touched them at all ... have had a few larger fish to try | 18:47 |
| mtaylor | fry | 18:48 |
| rmk | mtaylor: Missing default config file in /etc/keystone and missing "content" directory | 18:48 |
| mtaylor | damn. I can't type today at all | 18:48 |
| rmk | It was looking in /usr/keystone for the content directories, which is probably also incorrect | 18:48 |
| mtaylor | yup. that's certainly incorrect | 18:50 |
| *** code_franco has quit IRC | 18:50 | |
| rmk | mtaylor: And the last thing was that the package should either include or populate the keystone database with enough data to be functional | 18:51 |
| dprince | rmk: you mean like an admin role? | 18:52 |
| rmk | It was starting with a completely empty database, which obviously doesn't work at all, since the installation instructions mention nothing about endpoints or endpoint templates | 18:52 |
| dprince | rmk: So what. Default the URLs to local host? | 18:53 |
| rmk | yes | 18:53 |
| rmk | That would be my recommendation | 18:53 |
| rmk | dprince: I'd suggest that the default data should include only what's needed for a single admin, single tenant with endpoints pointing at localhost | 18:53 |
| rmk | Right now there is some example data in there which isn't relevant to the installation itself | 18:54 |
| dprince | sure. sounds good | 18:54 |
| rmk | Are there any docs yet on how nova credentials should be supplied once api-paste for using keystone is applied? | 18:56 |
| rmk | I can't find anything on that | 18:56 |
| *** code_franco has joined #openstack-dev | 18:56 | |
| dprince | vishy: https://code.launchpad.net/~dan-prince/nova/fix_list_v11_snapshot_images-diablo/+merge/76274 | 18:57 |
| dprince | vishy: all set. had to port a couple minor conflicts in the tests. | 18:57 |
| *** med_out is now known as medberry | 18:59 | |
| rmk | Is there a milestone proposed ppa for keystone yet? | 19:00 |
| dolphm | mtaylor: is that german or something? | 19:01 |
| dolphm | mtaylor: oops | 19:01 |
| dolphm | wrong window for UP -> Enter | 19:01 |
| *** joesavak has quit IRC | 19:02 | |
| mtaylor | hahaha | 19:02 |
| *** mdomsch has quit IRC | 19:02 | |
| vishy | thanks dprince: going in | 19:04 |
| *** chomping has joined #openstack-dev | 19:10 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Renaming 'roleRef' container to 'role' https://review.openstack.org/493 | 19:13 |
| openstackjenkins | Project nova-milestone build #67: SUCCESS in 14 sec: https://jenkins.openstack.org/job/nova-milestone/67/ | 19:21 |
| openstackjenkins | Tarmac: Fixes an issue where 'invalid literal for int' would occur when listing images after making a v1.1 server snapshot (with a UUID). | 19:21 |
| openstackjenkins | v1.1 image id's are now treated as strings (not integer ID's). The v1.0 API still tries to treat image id's as integers but doesn't fail miserably if they are uuid's either. | 19:21 |
| *** dragondm has joined #openstack-dev | 19:25 | |
| ttx | dprince: If you have any spare cycle, I'd appreciate some basic smokestack-testing of the milestone-proposed branch. We backported so much that we could use the results :) | 19:37 |
| ttx | dprince: if you haven't already | 19:37 |
| dprince | ttx: we think alike. | 19:37 |
| dprince | ttx: just fired off a job now. | 19:37 |
| dprince | ttx: I'm using milestone-proposed for nova and glance. | 19:38 |
| dprince | ttx: for keystone I'm backing off to revision 0425fba. Had to add a quick fix to smokestack to not use my git co alias for that one. https://github.com/dprince/smokestack/commit/bf376e05284a2f9ec897b2ec9fa394f104b7fd2d | 19:38 |
| dprince | ttx: haven't done this combination before but I wrote the runner to support it so just give me a couple minutes to get it right. | 19:39 |
| dprince | BTW. smokestack results were so red this week due to trunk failures I just cleared it all out. Hopefully diablo is better off. | 19:40 |
| vishy | dprince: do you think we should tag and build packages from that revision? | 19:42 |
| vishy | call them diablo-compatibility packages? | 19:42 |
| dprince | vishy: haven't fully gotten it working yet. But that is right before the role ID stuff so it should work I think. | 19:42 |
| *** AhmedSoliman has quit IRC | 19:43 | |
| vishy | dprince: the tenant_id/tenant_name user_id/user_name is kind of a breaking change as well | 19:46 |
| openstackgerrit | Anne Gentle proposed a change to openstack/openstack-manuals: Adds xml ids to all sections and chapters, adds identity info to Compute guide https://review.openstack.org/494 | 19:47 |
| openstackgerrit | Verification of a change to openstack/glance failed: Added min_disk and min_ram properties to images Fixes LP Bug#849368 https://review.openstack.org/429 | 19:48 |
| dprince | vishy: I suppose it could be yes. I see that as a migration issue. If you are doing a green field deployment it can work. | 19:50 |
| *** rcc1 has joined #openstack-dev | 19:50 | |
| *** rcc has quit IRC | 19:51 | |
| openstackgerrit | A change was merged to openstack/glance: Added min_disk and min_ram properties to images Fixes LP Bug#849368 https://review.openstack.org/429 | 19:51 |
| openstackgerrit | Brad Hall proposed a change to openstack/quantum: Update openvswitch plugin README https://review.openstack.org/474 | 19:53 |
| vishy | dprince: I'd like to have the migration scripts we ship with diablo match whatever version of keystone we lock on as diablo-compatible | 19:55 |
| dprince | vishy: I'd be fine backing off that one as well. My test setups worked after that (with my keystone-manage change) but it might be more upgrade freindly not to have that. | 19:56 |
| mtaylor | ttx, soren: based on that last meeting, could I convince you guys that we should set the version in setup.py to the next pypi milestone version - and not just the next full release version? (basically the same system as now, just do it as part of each milestone release?) | 19:56 |
| dprince | vishy/ttx: I gotta run. I'll let you know how the diablo tests go later this evening. | 19:57 |
| mtaylor | that way we wouldn't have to do any magic when we do a pypi upload, and during the period between final milestone and release we could still have the version set to the right thing so we still achieve "no step right before release" | 19:58 |
| *** dprince has quit IRC | 19:58 | |
| openstackjenkins | Project nova build #1,400: SUCCESS in 3 min 42 sec: https://jenkins.openstack.org/job/nova/1400/ | 20:00 |
| openstackjenkins | * Tarmac: Fix the display of swap units in nova manage. | 20:00 |
| openstackjenkins | * Tarmac: convert images that are not 'raw' to 'raw' during caching to node | 20:00 |
| openstackjenkins | * Tarmac: Add iptables filter rules for dnsmasq (lp:844935) | 20:00 |
| openstackjenkins | On Fedora, the default policy for the INPUT chain in the filter table | 20:00 |
| openstackjenkins | is DROP. This means that DHCP and DNS request packets from the guest | 20:00 |
| openstackjenkins | get dropped. | 20:00 |
| openstackjenkins | Add these rules to allow the traffic through: | 20:00 |
| openstackjenkins | ��$> sudo iptables -t filter -A nova-network-INPUT -i br0 -p udp -m udp --dport 67 -j ACCEPT | 20:00 |
| openstackjenkins | ��$> sudo iptables -t filter -A nova-network-INPUT -i br0 -p tcp -m tcp --dport 67 -j ACCEPT | 20:00 |
| openstackjenkins | ��$> sudo iptables -t filter -A nova-network-INPUT -i br0 -p udp -m udp --dport 53 -j ACCEPT | 20:00 |
| openstackjenkins | $> sudo iptables -t filter -A nova-network-INPUT -i br0 -p tcp -m tcp --dport 53 -j ACCEPT | 20:00 |
| openstackjenkins | * Tarmac: create disk.local the same way ephemerals are created (LP: #851145) | 20:00 |
| openstackjenkins | If the user did not specify '--block-device-mapping /dev/vdb=ephemeral0', then the first non-root device would end up being created differently then if they had. It would not have a filesystem on it. | 20:00 |
| openstackjenkins | This makes 'local_gb' in the libvirt connection created the same way that | 20:00 |
| openstackjenkins | it would be if it were named ephemeral0. | 20:00 |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Changing authenticate request content xml as well as json. https://review.openstack.org/495 | 20:00 |
| *** joesavak has joined #openstack-dev | 20:01 | |
| ttx | mtaylor: do you need to decide that now or it can wait until after diablo release ? | 20:02 |
| mtaylor | ttx: oh gosh no - no need for any changes this second | 20:04 |
| mtaylor | ttx: it would be good to chat about it before we actually cut the first essex milestone | 20:05 |
| ttx | sure. Plenty of time before that :) | 20:05 |
| ttx | at least until the design summit | 20:05 |
| *** anotherjesse has joined #openstack-dev | 20:05 | |
| mtaylor | yes | 20:06 |
| ttx | Meeting in 45 min in #openstack-meeting | 20:16 |
| ttx | mtaylor: release-related PPAs look ok now, thanks ! | 20:20 |
| *** mfer has left #openstack-dev | 20:21 | |
| annegentle | I'm getting 500 errors from Jenkins when I try to create or update jobs. anyone else seeing this? | 20:21 |
| vishy | Daviey: ping | 20:21 |
| *** mdomsch has joined #openstack-dev | 20:22 | |
| dolphm | annegentle: i think jenkins fell asleep | 20:23 |
| rmk | Can someone please answer my earlier question? What credentials does nova (via command line) expect once you're using keystone? | 20:23 |
| openstackgerrit | David Cramer proposed a change to openstack/identity-api: Fix broken build https://review.openstack.org/496 | 20:23 |
| annegentle | poke 'im! | 20:24 |
| rmk | The entire zip/novarc process seems to have no effect in that case. | 20:24 |
| markvoelker | Any dashboard folks around who might know the current status of https://bugs.launchpad.net/openstack-dashboard/+bug/848396 ? | 20:24 |
| uvirtbot | Launchpad bug 848396 in openstack-dashboard "Port Service Catalog to novaclient from openstackx" [High,Confirmed] | 20:24 |
| vishy | rmk: the creds are wrong for keystone | 20:24 |
| rmk | sorry what does that mean | 20:25 |
| rmk | doesn't work? | 20:25 |
| vishy | rmk: you'll need to use the keystone url for auth | 20:25 |
| vishy | it means that nova doesn't know that keystone exists so it will not generate reasonable values | 20:25 |
| vishy | but the variables are the same | 20:25 |
| rmk | thats fine I dont mind querying keystone manually but what is nova expecting I pass it | 20:26 |
| zul | vishy: hey question for you why do you guys need libvirt 0.9.5? | 20:26 |
| vishy | NOVA_USERNAME, NOVA_PROJECT_ID | 20:26 |
| vishy | NOVA_AUTH_URL | 20:26 |
| vishy | NOVA_APIKEY | 20:27 |
| vishy | IIRC | 20:27 |
| vishy | almost | 20:27 |
| vishy | it is NOVA_URL and NOVA_API_KEY | 20:27 |
| vishy | rmk: essentially the same vars from the bottom of the old rc | 20:28 |
| ttx | vishy: I think you can push https://code.launchpad.net/~vishvananda/nova/fix-swap-manage-ms/+merge/76182 now | 20:28 |
| vishy | the values will just be wrong or unknown to nova | 20:28 |
| rmk | Yeah I'm looking at it now. API key is what exactly with keystone? A valid admin token? | 20:29 |
| rmk | That's the issue, I'm not sure what's supposed to come from where | 20:29 |
| vishy | no | 20:29 |
| vishy | password | 20:29 |
| rmk | ok | 20:29 |
| vishy | AUTH_URL should be http://keystone_ip/keystone_port/auth/2.0/ | 20:30 |
| vishy | er keystone_ip:keystone_port | 20:30 |
| rmk | testing now | 20:30 |
| annegentle | API key for keystone is the password I believe. Odd, I know. | 20:31 |
| annegentle | ah you got it | 20:31 |
| rmk | http://paste.openstack.org/show/NjST6ZaczaLYOE9eJqBu/ | 20:35 |
| rmk | that's not working | 20:36 |
| *** Gordonz_ has quit IRC | 20:36 | |
| vishy | sorry rmk | 20:36 |
| vishy | the value i gave you above is for NOVA_URL | 20:36 |
| vishy | there is no AUTH_URL | 20:36 |
| vishy | also, do you have an openstack tenant? | 20:36 |
| rmk | yes | 20:37 |
| vishy | ok | 20:37 |
| rmk | keystone throws a 404 on /auth/2.0/ | 20:37 |
| dolphm | rmk: where did that path come from? | 20:38 |
| *** glenc has quit IRC | 20:38 | |
| rmk | dolphm: vishy just gave that to me a few lines up | 20:38 |
| dolphm | rmk: ah... out of the box, keystone uses http://127.0.0.1:5000/v2.0/ | 20:38 |
| *** glenc has joined #openstack-dev | 20:38 | |
| vishy | rmk: maybe i should actually look stuff up instead of trying to type from memory | 20:39 |
| vishy | :) | 20:39 |
| vishy | listen to dolphm :) | 20:39 |
| rmk | no worries | 20:39 |
| dolphm | (just passing through) | 20:39 |
| rmk | sorry to be asking in the dev channel | 20:39 |
| rmk | I don't think any users really know how any of this works yet | 20:39 |
| dolphm | can you use a query string in a DELETE call? | 20:40 |
| dolphm | e.g. DELETE /users/{user_id}/roles/{role_id}?tenant={tenant_id} | 20:40 |
| *** markvoelker has quit IRC | 20:44 | |
| Daviey | Does anyone know why socat was picked over netcat? Doesn't netcat do the things we use socat for? | 20:44 |
| *** misheska has quit IRC | 20:46 | |
| *** joesavak has quit IRC | 20:47 | |
| *** joesavak has joined #openstack-dev | 20:51 | |
| ttx | vishy: I think you can push https://code.launchpad.net/~vishvananda/nova/fix-swap-manage-ms/+merge/76182 now ? | 20:52 |
| vishy | ttx: i approved it 23 minutes ago | 20:52 |
| vishy | don't know why it hasn't merged yet | 20:52 |
| ttx | mtaylor: Status Code: 500 | 20:53 |
| ttx | Exception: | 20:53 |
| ttx | Stacktrace: | 20:53 |
| ttx | hudson.util.IOException2: : No space left on device | 20:53 |
| mtaylor | ttx: yay! | 20:53 |
| ttx | ^ on Jenkins | 20:53 |
| uvirtbot | ttx: Error: "on" is not a valid command. | 20:53 |
| vishy | woo! | 20:53 |
| vishy | that is a big boom | 20:53 |
| vishy | we win! | 20:53 |
| mtaylor | yup. that's out of disk | 20:53 |
| rmk | Still can't get it to auth =\ | 20:54 |
| ttx | mtaylor: looks like somesthing is not cleaned up regularly :) | 20:54 |
| mtaylor | ttx: I agree with you | 20:54 |
| *** misheska has joined #openstack-dev | 20:55 | |
| *** rbasak has quit IRC | 20:55 | |
| ttx | mtaylor: I prefer it happening now than in 2 days | 20:55 |
| *** misheska has quit IRC | 20:55 | |
| mtaylor | ttx: ++ | 20:56 |
| *** markvoelker has joined #openstack-dev | 20:58 | |
| Daviey | vishy: do you know why we went for socat over netcat? | 20:58 |
| vishy | Daviey: no | 20:58 |
| vishy | Daviey: it is only used in ajax proxy right? | 20:58 |
| Daviey | believe so | 20:58 |
| *** troytoman-away is now known as troytoman | 20:58 | |
| vishy | Daviey: probably no real reason then, let me ask the author real quick | 20:59 |
| Daviey | awesome, thanks | 20:59 |
| ttx | Meeting starts in 1 min in #openstack-meeting, please join | 20:59 |
| vishy | Daviey: do we have dnsmasq-utils? | 21:00 |
| *** misheska has joined #openstack-dev | 21:00 | |
| *** joesavak has quit IRC | 21:00 | |
| Daviey | vishy: we do. | 21:00 |
| vishy | whoot! | 21:00 |
| vishy | so we can update the packages now? | 21:00 |
| Daviey | already done! | 21:01 |
| Daviey | :) | 21:01 |
| Daviey | and it's now in Ubuntu Oneiric | 21:01 |
| vishy | Daviey: both the diablo and the trunk packaging has been updated? | 21:02 |
| rmk | I'm at a loss about this keystone/nova integration | 21:02 |
| * vishy does a little jig. That bug has annoyed me for a long time | 21:02 | |
| dolphm | rmk: new error? | 21:02 |
| rmk | dolphm: Invalid credentials is what's being thrown back | 21:03 |
| Daviey | zul: you pushed to both, or just diablo? | 21:03 |
| * annegentle joins rmk in the "lost" column | 21:03 | |
| rmk | How do projects and tenants relate? | 21:03 |
| zul | diablo | 21:03 |
| vishy | project_id == tenant_id | 21:03 |
| zul | ill update turnk tonight | 21:03 |
| rmk | so they have to match? | 21:03 |
| dolphm | rmk: can you send me your keystone request? | 21:03 |
| dolphm | headers & body | 21:03 |
| rmk | sure let me see how I can get at that | 21:03 |
| rmk | nova is making the request so it's not straightforward | 21:04 |
| vishy | rmk: depending on which middleware stack you are using | 21:04 |
| vishy | it should either a) create a project if it doesn't exist when it gets the tenant_id | 21:04 |
| vishy | or b) just set the tenant_id as the project_id in the context | 21:04 |
| vishy | from an end user perspective the experience is the same | 21:05 |
| vishy | project_id == tenant_id | 21:05 |
| rmk | Here's another point of confusion. It's really tenant_name == project_id | 21:08 |
| rmk | Because tenant_id is an int | 21:08 |
| *** sleepsonthefloor has quit IRC | 21:08 | |
| *** sleepsonthefloor has joined #openstack-dev | 21:10 | |
| *** jshepher has joined #openstack-dev | 21:10 | |
| dolphm | rmk: tenant_id is a unicode string* | 21:11 |
| rmk | sorry | 21:11 |
| dolphm | rmk: :) | 21:11 |
| rmk | im looking at the database tables | 21:11 |
| dolphm | rmk: the SQL backend uses int's as PK's, but the contract exposes the PK as a unicode string to support backends using whatever they want as PK's (LDAP uses distinguished names or whatever) | 21:12 |
| *** lts has quit IRC | 21:16 | |
| *** xavicampa has quit IRC | 21:16 | |
| vishy | rmk: well tenant_id used to be a string | 21:18 |
| vishy | dolphm: actually that is going to cause some problematic migrations | 21:19 |
| rmk | Either way I think your point is to use whatever was used for keystone-manage tenant add <whatever> | 21:19 |
| rmk | Which still doesn't work for me ;) | 21:19 |
| dolphm | vishy: explain? | 21:20 |
| vishy | dolphm: existing nova tenants are using a "name" | 21:20 |
| vishy | when we migrate we are giving a name but we are getting back an id | 21:20 |
| vishy | so what about existing instances for example | 21:21 |
| vishy | the project_id field will not match the id being sent back by keystone | 21:21 |
| vishy | so it will be very broken | 21:21 |
| dolphm | vishy: solution... GET /tenants?name={tenant_name} ? | 21:21 |
| openstackgerrit | A change was merged to openstack/quantum: Update openvswitch plugin README https://review.openstack.org/474 | 21:21 |
| openstackjenkins | Project nova-milestone build #68: SUCCESS in 34 sec: https://jenkins.openstack.org/job/nova-milestone/68/ | 21:22 |
| openstackjenkins | Tarmac: Fix the display of swap units in nova manage. | 21:22 |
| vishy | dolphm: do we get tenant_id and tenant_name back? | 21:22 |
| vishy | dolphm: from validate? | 21:23 |
| vishy | dolphm: if so we can continue to use tenant_name for ownership | 21:23 |
| dolphm | vishy: yes, i would like ID's *and* names to be returned wherever applicable for any object that has both | 21:23 |
| rmk | wouldn't both always exist? | 21:23 |
| dolphm | dolphm: so, all requests (except for ?name={*_name} queries) would be performed by ID, while all responses would contain both the ID and Name | 21:24 |
| dolphm | vishy: ^ | 21:24 |
| dolphm | (i have a habit of directing messages on IRC to myself..) | 21:25 |
| Vek | better than having a server direct messages on IRC to yourself. | 21:25 |
| vishy | dolphm: we're discussing keystone in meeting if you want to participate | 21:26 |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: GET /tokens/{token_id}: Exposing both role ID's and Name's https://review.openstack.org/497 | 21:34 |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Changes to support HEAD - checkToken call. https://review.openstack.org/498 | 21:42 |
| *** cp16net has quit IRC | 21:43 | |
| *** bcwaldon has quit IRC | 21:44 | |
| *** jshepher has quit IRC | 21:45 | |
| *** jshepher has joined #openstack-dev | 21:45 | |
| dolphm | mtaylor: jaypipes: is jenkins still KO'd? | 21:46 |
| mtaylor | dolphm: it should be better now | 21:46 |
| openstackgerrit | A change was merged to openstack/keystone: Renaming 'roleRefs' container to 'roles' https://review.openstack.org/492 | 21:52 |
| openstackgerrit | Mark Gius proposed a change to openstack/swift: Changes swift-ring-builder exit codes https://review.openstack.org/499 | 21:53 |
| *** ameade has quit IRC | 21:53 | |
| dolphm | mtaylor: thanks | 21:54 |
| openstackgerrit | A change was merged to openstack/keystone: Renaming 'roleRef' container to 'role' https://review.openstack.org/493 | 21:54 |
| *** anotherjesse has quit IRC | 21:57 | |
| *** anotherjesse has joined #openstack-dev | 21:57 | |
| openstackgerrit | A change was merged to openstack/keystone: GET /tokens/{token_id}: Exposing both role ID's and Name's https://review.openstack.org/497 | 21:57 |
| *** bengrue has joined #openstack-dev | 22:01 | |
| *** galstrom has joined #openstack-dev | 22:05 | |
| *** jakedahn has joined #openstack-dev | 22:07 | |
| openstackgerrit | Yogeshwar Srikrishnan proposed a change to openstack/keystone: Changes to support HEAD - checkToken call. https://review.openstack.org/498 | 22:07 |
| *** jshepher has quit IRC | 22:08 | |
| *** galstrom has quit IRC | 22:09 | |
| *** novas0x2a|laptop has quit IRC | 22:11 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Changes to support HEAD - checkToken call. https://review.openstack.org/498 | 22:15 |
| *** medberry is now known as med_out | 22:16 | |
| *** mdomsch has quit IRC | 22:16 | |
| dolphm | I'm headed home to work on keystone til I fall asleep... i'll be on this channel all night if anyone needs me | 22:17 |
| *** dolphm has quit IRC | 22:17 | |
| *** rnorwood has quit IRC | 22:19 | |
| *** code_franco has quit IRC | 22:20 | |
| rmk | Well I guess that answers all my keystone questions, I won't worry about it for now ;) | 22:20 |
| vishy | ttx, nova-core: branch is here: https://code.launchpad.net/~vishvananda/nova/remove-keystone-middleware/+merge/76297 | 22:21 |
| *** amccabe has quit IRC | 22:23 | |
| rmk | I won't lie, I was scared an unstable keystone being a requirement for deployment | 22:24 |
| *** bsza has quit IRC | 22:26 | |
| vishy | rmk: well at least we had a deployment option | 22:26 |
| vishy | now it is ??? | 22:26 |
| *** lorin1 has joined #openstack-dev | 22:27 | |
| *** lorin1 has left #openstack-dev | 22:27 | |
| *** rnirmal has quit IRC | 22:31 | |
| *** lorin1 has joined #openstack-dev | 22:32 | |
| Daviey | vishy: merge conflict. | 22:33 |
| vishy | Daviey: sigh, sometimes I don't understand bzr merges | 22:35 |
| vishy | fixing | 22:35 |
| *** lorin1 has left #openstack-dev | 22:36 | |
| Daviey | vishy: I'd take bzr merges over svn anyway day :) | 22:36 |
| vishy | yeah | 22:36 |
| vishy | i don't get it though there were no changes in that code | 22:37 |
| vishy | why did it conflict the whole thing? | 22:37 |
| *** anotherjesse has quit IRC | 22:37 | |
| Daviey | vishy: your branch was based on r1541, and trunk is -r1605.. that is quite a split. | 22:39 |
| vishy | yes | 22:39 |
| vishy | but that code hasn't been touched | 22:39 |
| vishy | it also failed into milestone-proposed | 22:39 |
| vishy | which is not nearly as many revs | 22:39 |
| Daviey | vishy: blame -r1556 | 22:41 |
| Daviey | http://pb.daviey.com/k9OS/ | 22:42 |
| *** nati has joined #openstack-dev | 22:42 | |
| vishy | interesting | 22:42 |
| vishy | ok that explains it i guess | 22:42 |
| vishy | thx | 22:42 |
| * Daviey will miss bzr :( | 22:43 | |
| vishy | Daviey: you prefer it to git? | 22:44 |
| Daviey | vishy: soren defined git well as a patch tracking system, bzr providing history of 'how you got there' is quite a nice feature IMO. | 22:45 |
| Daviey | git is a Maildir on cocaine. | 22:45 |
| Daviey | </troll> | 22:46 |
| Daviey | Although, git bisect is much more mature than bzr's IME. | 22:46 |
| Daviey | vishy: Did you get an answer why socat was picked over netcat? | 22:47 |
| vishy | Daviey: no reason | 22:48 |
| vishy | Daviey: if netcat can do the same thing | 22:48 |
| vishy | then it should be fine | 22:48 |
| vishy | Daviey: are you guys testing/using ajaxconsole? | 22:49 |
| vishy | we sort of abandonded in favor of vnc console which is much more generally useful, so I have no idea if it even works | 22:49 |
| zykes- | ajaxconsole as in serial ? | 22:51 |
| Daviey | vishy: we /did/ a long time ago, not tried it recently. | 22:51 |
| Daviey | on todo. | 22:51 |
| Daviey | hmm, others might have and not commnicated it tho | 22:52 |
| zykes- | why isn't thjat supported :/ | 22:52 |
| *** kbringard has quit IRC | 22:52 | |
| Daviey | vishy: ajaxconsole support copy/paste right? | 22:52 |
| Daviey | vnc console is bitmaps, no? | 22:52 |
| vishy | Daviey: yes good point, you probably can't copy/paste | 22:53 |
| vishy | but you do have to set up a serial console in the guest for it to work | 22:53 |
| vishy | it being ajaxconsole | 22:54 |
| vishy | and it can't do anything with windows for example | 22:54 |
| *** gabrielhurley has joined #openstack-dev | 22:57 | |
| *** troytoman is now known as troytoman-away | 23:01 | |
| Daviey | vishy: I've not even sniffed windows to see how it works :/ | 23:01 |
| *** jeremyb_ has joined #openstack-dev | 23:03 | |
| *** jeremyb_ has quit IRC | 23:05 | |
| *** jeremyb_ has joined #openstack-dev | 23:05 | |
| openstackgerrit | Rick Harris proposed a change to openstack/glance: Documenting nova_to_os_env.sh tool https://review.openstack.org/500 | 23:07 |
| *** RobertLaptop has quit IRC | 23:07 | |
| *** jeremyb has quit IRC | 23:07 | |
| s1rp | jaypipes: just pushed documentation for nova_to_os_env.sh up to gerrit | 23:07 |
| *** jeffjapan has joined #openstack-dev | 23:08 | |
| *** jeremyb_ is now known as jeremyb | 23:09 | |
| *** RobertLaptop has joined #openstack-dev | 23:14 | |
| *** blamar_ has quit IRC | 23:17 | |
| *** wwkeyboard has quit IRC | 23:24 | |
| *** dolphm has joined #openstack-dev | 23:29 | |
| *** stanchan has quit IRC | 23:31 | |
| dolphm | if anyone wants to help keystone in getting to a releasable state, i would think the easiest route would be to propose a review with some new (failing) functional tests asserting any missing behavior that keystone needs immediately | 23:35 |
| *** Binbin has quit IRC | 23:40 | |
| openstackgerrit | Dolph Mathews proposed a change to openstack/keystone: Changing authenticate request content xml as well as json. https://review.openstack.org/495 | 23:44 |
| *** heckj has quit IRC | 23:45 | |
| *** vladimir3p has quit IRC | 23:47 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!