Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2012-11-14

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:07 dachary joined #puppet-openstack
03:02 comptona joined #puppet-openstack
07:45 dachary joined #puppet-openstack
07:55 dachary joined #puppet-openstack
09:11 JoeJulian There's no way to use mysql::config::root_password inside the mysql providers unless I add it to the specific resource, right? For instance  https://gist.github.com/4071123
09:22 derekh joined #puppet-openstack
09:27 dachary joined #puppet-openstack
13:50 e1mer joined #puppet-openstack
13:50 e1mer joined #puppet-openstack
15:29 JoeJulian Eh, nevermind that question. I didn't realize that the forge module was so far behind github.
18:20 bodepd JoeJulian: yeah, we kind of suck at keeping those things up to date
18:33 dachary joined #puppet-openstack
19:00 JoeJulian Did keystone change this between essex and folsom? --service-id is now --service_id
19:02 dachary joined #puppet-openstack
20:17 comptona :'(
20:17 comptona I think the epel-testing packages have a bug
20:18 comptona specifically, it seems that the keystone package didn't have its "incubated copy of openstack-common" updated when it was built
20:20 comptona site-packages/keystone/config.py calls ConfigOpts with a 'project' parameter, but site-packages/keystone/openstack/common/cfg.py's ConfigOpts does not take such
20:27 JoeJulian Is that why it keeps trying to create the same users over and over again?
20:38 comptona maybe? the keystone service won't even start on my test box
20:39 JoeJulian openstack-keystone-2012.2-3.el6.noarch ?
20:40 comptona python-keystone-2012.2-4.el6.noarch
20:40 comptona /usr/lib/python2.6/site-packages/keystone/config.py and /usr/lib/python2.6/site-packages/keystone/openstack/common/cfg.py
20:41 comptona I'm going to go hunt down the source of that package and see how it's built to verify my theory
20:41 JoeJulian I'm still on -3 and it starts fine.
20:42 comptona let me see what the changelog is then
20:46 comptona huh, that's bizarre
20:46 comptona python-keystone is not a part of https://admin.fedoraproject.org/updates/FEDORA-EPEL-2012-13088/
20:55 comptona ok, I feel like I'm losing my mind
20:56 comptona epel-testing has a package called "python-keystone" for which there does not appear to be any SRPM, spec file, or documentation
20:59 JoeJulian No packages by that name in koji either...
20:59 comptona also, I can't roll back to python-keystone-2012.2-3 since it got deprecated when the current version was made
21:00 comptona man, I love openstack, but this sort of thing just makes me want to jump off a cliff and/or shovel money at vmware :-P
21:00 JoeJulian lol
21:00 JoeJulian lol at the vmware part, not the jumping off cliff part..
21:03 JoeJulian Ah, ok. The source rpm for python-keystone is openstack-keystone.
21:04 JoeJulian rpm -qi python-keystone
21:05 comptona but I grabbed the openstack-keystone SRPM and its spec file requires python-keystone
21:06 e1mer joined #puppet-openstack
21:06 comptona on the plus side, I'm not insane; the keystone-2012.1.3.tar.gz included with the openstack-keystone SRPM includes the bug I saw
21:08 comptona oh, I see now where the openstack-keystone spec file also defines the python-keystone package
21:36 comptona …and I can't file a ticket about the epel build without registering, and apparently the bugzilla isn't sending signup confirmation emails right now
21:36 comptona this was just a bad day to work on this :-P
21:39 JoeJulian I can file it. How can I hit that bug?
21:40 comptona running keystone-all: http://pastie.org/5379284
21:41 comptona RPM versions: http://pastie.org/5379288
21:42 comptona if I could get a hold of the old RPMs I'd downgrade and verify
21:42 pdurbin bodepd: i'm not quite sure what you mean by "this looks like an error on the server. Can you get a stack trace from the server?" at https://groups.google.com/a/puppetlabs.com/d/msg/puppet-openstack/nz8XHmEOMZs/aKXjHCNeTIUJ
21:42 comptona but I can't find them anywhere
21:46 pdurbin bodepd: i assume you mean strace -o file.out. is that right? this might be a bit tricky with multiple slaves to our puppet master and 1700 puppet clients checking in
22:02 samkottler joined #puppet-openstack
22:05 bodepd JoeJulian: yes. those things changed.
22:05 bodepd JoeJulian: in keystone, those changes have been made to the native types.
22:07 bodepd pdurbin: I built an openstack controller/compute with puppet 3 this week.
22:07 bodepd pdurbin: using puppet apply though.
22:07 bodepd pdurbin: can you try that out and see if it works?
22:07 JoeJulian I'm trying to do that right now.
22:08 bodepd JoeJulian: to run with puppet 3?
22:08 JoeJulian yep
22:08 bodepd let me know if you see pdurbin 's issue. I could not recreate.
22:08 bodepd I also used 3.0.x though.
22:08 bodepd what was the exact version you saw it on?
22:10 JoeJulian So far I'm getting stuck at all the keystone changes.
22:10 bodepd JoeJulian: really?
22:10 bodepd what version of the modules are you using?
22:10 bodepd keystone CLI changed, and do did hte modules
22:11 bodepd you may want to try out the openstack_dev_env
22:11 bodepd you can set
22:11 bodepd export OPERATINGSYTEM=redhat
22:11 JoeJulian There's a bunch of cli changes from essex to folsom and the provider's not made the changes.
22:11 bodepd and it will build based on whats in rhel-testing
22:11 bodepd what version of keystone are you using? that stuff was definitely fixed
22:12 bodepd what version of the keystone module
22:12 JoeJulian latest from github.
22:12 bodepd what branch?
22:12 bodepd master I assume?
22:13 bodepd what repo
22:13 JoeJulian Neither master nor folsom have the service-id to service_id change.
22:13 JoeJulian Once I saw that wasn't there I kind-of stopped looking.
22:14 JoeJulian yeah, the user hash has the fields in the wrong order too.
22:14 JoeJulian Did someone forget to push? ;)
22:15 bodepd have a look at this commit:
22:15 bodepd git show 7a39b106
22:15 bodepd the change I see was from service to service-id
22:15 bodepd did it change again?
22:16 bodepd also, I am definitley not seeing this in my environment.
22:17 JoeJulian comptona nailed it. Someone at fedora screwed up.
22:17 bodepd ah, they put grizzley in the package?
22:17 JoeJulian They put essex keystone into folsom.
22:17 bodepd fwiw,I do not see with epel-testing
22:17 comptona bodepd: what package version?
22:17 JoeJulian That's from epel-testing.
22:18 bodepd hmm. is this just today?
22:18 comptona I think they pushed a new build in the last ~3 days
22:18 bodepd let me run an install
22:18 bodepd I ran a build last night.Could be that I got a cached version of the packages. let me restart squid and try again
22:18 bodepd it'll take 10 to rebuild
22:20 JoeJulian openstack-keystone-2012.2-4.el6 apevec 2012-11-12 19:48:25
22:29 comptona bwahaha
22:29 comptona I spun up a new python-keystone RPM from the SRPM with all of the properly matched pieces
22:29 comptona and it works
22:29 comptona *lightning crackles*
22:37 JoeJulian hmm. From github.com/openstack/keystone the folsom branch has "./tools/sample_data.sh:    keystone endpoint-create --region RegionOne --service_id $NOVA_SERVICE" where the master branch has "./tools/sample_data.sh:    keystone endpoint-create --region RegionOne --service-id $NOVA_SERVICE" so does that mean that service-id is grizzly?
22:47 comptona JoeJulian: I'm not, myself, in the mood to trust anything from epel-testing at the moment :-P
22:48 comptona for what it's worth, the folsom keystoneclient on my box says "service-id" in its help text
22:58 e1mer joined #puppet-openstack
22:58 e1mer joined #puppet-openstack
23:01 JoeJulian Actually, you have grizzly comptona. The tar.gz file in the rpm has a last commit of 8154492da3d8b6e0e71eb7dcb5e2e3b06ddb09a7. That's the same commit that the openstack/keystone git tree is tagged 2012.2
23:02 comptona *headdesk*
23:29 e1mer joined #puppet-openstack

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary