Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2012-11-08

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

All times shown according to UTC.

Time Nick Message
01:23 e1mer joined #puppet-openstack
01:23 e1mer joined #puppet-openstack
01:24 newptone joined #puppet-openstack
01:28 newptone_ joined #puppet-openstack
03:27 bodepd I saw that error when any of the service were not up and running correctly.
03:31 comptona joined #puppet-openstack
03:39 jtopjian joined #puppet-openstack
04:10 e1mer joined #puppet-openstack
04:10 e1mer joined #puppet-openstack
05:05 bodepd aimon: ping
05:16 aimon boped hi
05:20 e1mer joined #puppet-openstack
05:30 newptone joined #puppet-openstack
05:45 e1mer joined #puppet-openstack
05:45 e1mer joined #puppet-openstack
05:47 bodepd aimon: do you mind if I walk you through the process cleaning up that commit.
05:49 aimon sure that would be greatr
05:50 aimon boped ^
06:07 bodepd take that branch
06:08 bodepd run git log, figure out the commit before 7d475a4
06:09 bodepd > git reset <that_commit>
06:10 bodepd > git stash
06:10 bodepd assuming origin is puppetabs remote
06:10 bodepd git fetch origin
06:10 bodepd git reset --hard origin/master
06:10 bodepd ^^ that command is dangerous, make sure you dont have any unsaved work
06:10 bodepd git stash pop
06:10 bodepd then you can make a single commit
06:10 bodepd and push it
06:11 aimon ah interesting.. the brute force approach..
06:11 aimon ok will give ti a shot :)
06:11 bodepd that may not be the best workflow, but it'll work :)
06:11 bodepd you could probably also do an interactive reabase and drop those merge commits
06:12 aimon yeah did that.. it used to be like 24 commits (motley from failed squashes)
06:12 aimon but I like your method also
06:12 aimon its direct
06:14 bodepd also, you may want to change the test to use any value other than '/'
06:14 bodepd that dont actually test it is set b/c that is the default
06:45 aimon boped. Much cleaner!
06:45 aimon thx for the help
06:46 aimon thing I need to remember is to do rebase after fetch rather then merge
06:47 aimon that'll keep me from getting into this issue
07:03 bodepd yep, I make the same mistake sometimes.
07:45 aimon hoped what HA module are you guys using? puppetlabs-corosync?
07:45 aimon *boped
07:46 aimon ha turned off spell check to prevent me from calling you hoped :)
09:07 romero joined #puppet-openstack
09:41 dachary joined #puppet-openstack
10:04 \etc\bin joined #puppet-openstack
10:04 \etc\bin joined #puppet-openstack
10:59 derekh joined #puppet-openstack
11:10 e1mer joined #puppet-openstack
11:10 e1mer joined #puppet-openstack
11:32 newptone joined #puppet-openstack
11:47 dachary joined #puppet-openstack
15:36 derekh bodepd: just needed to make one addition to the cinder module for it to work on RHEL, https://github.com/derekhiggins/puppetlabs-cinder​/commit/cafd6595d965553d4b4b4df6ab4911e565dbb4c6   anybody know if its a Red Hat specific change?
15:36 derekh basically it adds the line include /etc/cinder/volumes/* to /etc/tgt/targets.conf
15:37 jtopjian it is, the ubuntu pkg automatically does that
15:37 jtopjian I used to have a similar block of code during the folsom milestones but removed it when the package started doing it
15:38 derekh ok, it doesn't look like this is going to be added to the epel packages, so I may need to wrap it in a conditional that checks OS family
15:38 derekh how do that sound to you?
15:39 jtopjian sounds good to me
15:39 jtopjian Ubuntu creates an /etc/tgt/conf.d directory - do the rpm packages do that by chance?
15:40 derekh nope, there was an attempt to put the include into /etc/tgt/conf.d/cinder.conf
15:40 derekh but
15:41 derekh the globing of /etc/cinder/volumes/* didn't work in the version of tgt being used
15:41 jtopjian ahh ok
15:41 derekh but the globing does work in the targets.conf file, go figure
18:21 bodepd jtopjian: welcome!
18:21 bodepd aimon: there are a few HA solutions built on top of the Puppet stuff.
18:21 bodepd look at the PR's on nova.
18:21 bodepd there is one for adding mysql galera support.
18:22 bodepd but it patches openstack to make it work active-active
18:24 bodepd the community still needs to consolidate HA and monitoring.
18:38 pdurbin any opinions on the "OpenStack High Availability Modules" at https://github.com/danehans/puppet-openstack ?
18:48 bodepd I know that Cisco put a lot of effort into those.
18:49 bodepd 2 comments
18:49 bodepd - they require patching openstack itself
18:49 bodepd I am pretty sure that those patches are excepted into grizzley
18:49 bodepd - it is a fork of the modules that I am working on
18:50 bodepd they diffs are in a PR on nova, but there is more work to do to get it accepted
18:50 bodepd - it only works for essex
18:50 bodepd I'm supposed to work with someone soon on getting everythign merged based on folsom
18:52 pdurbin ok. very interesting. yes, they did seem to be a fork of the code you're working on
18:52 * pdurbin needs to get bodepd's code running on physical hardware next :)
18:54 bodepd I know Daneyon, and have been following that work
18:55 bodepd pdurbin: it should be as easy as setting up a master
18:55 pdurbin bodepd: by master you mean openstack_controller? (from your vagrant file)
19:02 bodepd no, install a puppet master
19:03 bodepd it probably easiest if its a master/dhcp server/pxe server/
19:03 bodepd I started on some work to use razor for this. I can dig it up.
19:04 pdurbin oh, oh. we have puppet already set up. i just need to commit your latest code to my branch and push it up to my test environment
19:04 bodepd https://github.com/bodepd/vagrant-openstack-puppet
19:04 bodepd that is a really rough example of how to use razor to bootstrap OS
19:04 pdurbin interesting. thanks!
19:28 jtopjian I'm not knocking razor, but i have put a lot of work into cobbler for this: https://github.com/jtopjian/cobbler-admin
19:28 jtopjian There's a modified readme for my dair branch: https://github.com/jtopjia​n/cobbler-admin/tree/dair basically i spent way too many hours dealing with Ubuntu 12.04 changing ethX allocations
19:28 jtopjian one day i will write formal documents up on this stuff
19:33 pdurbin jtopjian: i use cobbler and like it :)
19:33 jtopjian likewise
19:33 jtopjian i still want to take a look at razor in the future, though
19:34 jtopjian but since cobbler does everything i need it to, it's on the backburner
19:34 pdurbin it just works. usually :)
19:35 jtopjian yup
21:22 aimon Boped the danehans HA modules look pretty clean
21:22 aimon Not the style that we need atm…. too much overhead
21:23 aimon we'll be getting to that next.. atm working on  simple 1:1 failover controller
21:23 aimon with iscsi back
21:24 aimon So far the corosync module is working nicely
21:36 bodepd this repo has an example drbd/corosync active/passive controller https://github.com/puppetlab​s/puppetlabs-openstack_admin
21:36 bodepd its not very robust unfortunately...
21:38 aimon yeah I am kinda anti drbd
21:38 aimon data loss waiting to happen
21:39 aimon we have a opensolaris san in the mix.. so use iscsi
21:40 aimon makes things really clean.. preconnect to target and just move mount and IP
21:56 bodepd yep, I'm also a little anti drbd.
21:56 bodepd just did that as a prototype of the easiest way possible.
22:00 aimon yeah it ids nice since you dint need a 3rd point
22:00 aimon but it scares me :)
22:00 aimon I hear its better then it used to be.. but..
22:17 e1mer joined #puppet-openstack
22:17 e1mer joined #puppet-openstack
22:38 bodepd has anyone looked at the new HA deployment guide?
22:38 bodepd written by Florian?
22:38 bodepd I'm working with them on a Puppet based openstack course.
22:39 bodepd so, there should be modules based on his recommendations early next year.
23:20 bodepd aimon: one more tip.
23:21 bodepd if you force push to a branch associated with a PR.
23:21 bodepd it auto-updates the PR
23:21 bodepd you don't have to close and reopen.
23:21 aimon yeah I noted that last night.. very handy :)
23:21 aimon thx
23:30 jtopjian joined #puppet-openstack

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