| mandre | woohoo | 00:00 |
|---|---|---|
| *** openstackgerrit has quit IRC | 00:01 | |
| *** openstackgerrit has joined #kolla | 00:01 | |
| *** openstackstatus has quit IRC | 00:02 | |
| *** openstack has joined #kolla | 00:05 | |
| *** banix has quit IRC | 00:06 | |
| *** devlaps has quit IRC | 00:09 | |
| *** tpot has quit IRC | 00:17 | |
| *** sdake_ has joined #kolla | 00:38 | |
| *** sdake has quit IRC | 00:43 | |
| *** sdake_ has quit IRC | 00:53 | |
| *** jtriley has joined #kolla | 01:10 | |
| *** erkules_ has joined #kolla | 01:11 | |
| *** erkules has quit IRC | 01:14 | |
| *** jtriley has quit IRC | 01:17 | |
| *** unicell has quit IRC | 01:35 | |
| *** Kennnan2 is now known as Kennan | 01:37 | |
| *** marun has quit IRC | 01:52 | |
| *** tpot has joined #kolla | 01:55 | |
| *** sdake has joined #kolla | 01:58 | |
| *** tpot has quit IRC | 02:00 | |
| *** loth has quit IRC | 02:28 | |
| *** tpot has joined #kolla | 02:41 | |
| *** achanda has joined #kolla | 02:43 | |
| *** unicell has joined #kolla | 02:56 | |
| *** dims_ has quit IRC | 02:57 | |
| SamYaple | wooo | 03:13 |
| SamYaple | rc1 sdake | 03:13 |
| SamYaple | RC1! | 03:13 |
| SamYaple | sdake: i have pushed the ubuntu rc1 images to docker registry | 03:16 |
| sdake | nice | 03:33 |
| sdake | thanks for that sam | 03:33 |
| *** sdake has quit IRC | 03:38 | |
| *** gfidente has quit IRC | 03:39 | |
| SamYaple | only too 86 minutes btw | 03:40 |
| SamYaple | i pushed it in parralaler 10 | 03:40 |
| SamYaple | wow. parralel | 03:40 |
| *** sdake has joined #kolla | 03:41 | |
| *** sdake has quit IRC | 03:46 | |
| *** sdake has joined #kolla | 03:49 | |
| *** dims has joined #kolla | 03:57 | |
| *** dims has quit IRC | 04:03 | |
| *** diga has joined #kolla | 04:04 | |
| *** alisonh has quit IRC | 04:04 | |
| *** alisonh has joined #kolla | 04:14 | |
| *** sdake has quit IRC | 04:28 | |
| SamYaple | http://paste.openstack.org/show/474778/ | 05:00 |
| SamYaple | for those interested, those instructions above are tested ^^ and working with the images i pushed to the docker hub | 05:00 |
| *** CBR09 has joined #kolla | 05:12 | |
| CBR09 | Hello all | 05:13 |
| CBR09 | which version ubuntu images on docker hub? | 05:14 |
| CBR09 | kilo or liberty? | 05:14 |
| CBR09 | and rsyslog container on ubuntu doesn't work | 05:16 |
| CBR09 | https://bugs.launchpad.net/kolla/+bug/1501167 | 05:16 |
| openstack | Launchpad bug 1501167 in kolla "Rsyslog doesn't work on ubuntu" [Undecided,New] | 05:16 |
| SamYaple | CBR09: right now the only images pushed and working to docker hub would be the liberty-rc1 | 05:17 |
| SamYaple | CBR09: looks like an AUFS issue. not sure about that. it works for overlayfs and btrfs (ive personally tested) | 05:17 |
| SamYaple | CBR09: can you update teh bug with your kernel and docker versions? | 05:17 |
| CBR09 | my version docker 1.9 and kernel 3.19, both are ok? | 05:19 |
| SamYaple | I havent tested 1.9 but i think it would be fine yes | 05:19 |
| SamYaple | 3.19 should work for ubuntu and AUFS | 05:20 |
| SamYaple | CBR09: I am kicking a AUFS deploy now | 05:20 |
| SamYaple | should be about ~10 min | 05:21 |
| CBR09 | yes, but rsyslog doesn't work :( | 05:21 |
| CBR09 | yes, thanks you:) | 05:22 |
| SamYaple | CBR09: ive asked for exact command output in the bug, please update when you can :) | 05:24 |
| *** CBR09_ has joined #kolla | 05:29 | |
| CBR09_ | sorry Sam:D | 05:29 |
| CBR09_ | I'm using my phone | 05:29 |
| *** CBR09 has quit IRC | 05:29 | |
| CBR09_ | wait me for 15minutes | 05:30 |
| CBR09_ | :D | 05:30 |
| CBR09_ | Ah Sam, could you give me your skype nick?. My IP public workplace banned, not access IRC:( | 05:33 |
| SamYaple | i don't have a skype thing CBR09_ | 05:34 |
| SamYaple | CBR09_: I do have an irc proxy node... | 05:34 |
| *** inc0 has joined #kolla | 05:34 | |
| SamYaple | CBR09_: ok full new environment rekick with AUFS and kernel 3.19, no problems with rsyslog | 05:35 |
| inc0 | good moringing | 05:37 |
| inc0 | morning even | 05:37 |
| inc0 | yeah, I was testing everything on aufs + 3.19 | 05:37 |
| inc0 | so if you have different conf and you experience problems, do tell plz | 05:37 |
| *** CBR09_ has quit IRC | 05:42 | |
| inc0 | SamYaple, rsyslog was causing problems? | 05:48 |
| SamYaple | not on AUFS for me | 05:48 |
| SamYaple | looking into the error it looks like a wierd kernel bug that people "fix" by changing versions | 05:48 |
| SamYaple | havent seen a reason why yet | 05:48 |
| inc0 | SamYaple, I was meant to ask you, what's biggest deployment you've made using ansible? does ansible scale? | 05:52 |
| SamYaple | ansible _can_ scale yes, but you have to configure it as such | 05:52 |
| SamYaple | the biggest factor for larger deploys would probably be --forks | 05:53 |
| SamYaple | it defaults to 8, but if you have 800 nodes it will take 100 times longer | 05:53 |
| inc0 | based on that it's a single ansible node | 05:54 |
| inc0 | it would be cool to make this scallable, so you have ansible on node 1 which ssh to 10 different ansible hosts from which each run actual playbooks on N forks | 05:54 |
| SamYaple | inc0: you can configure ansible in pull mode | 06:01 |
| SamYaple | so each host runs pull | 06:01 |
| SamYaple | and you can adjust the forks | 06:02 |
| inc0 | so puppet-like stuff | 06:02 |
| SamYaple | 100 forks is fine | 06:02 |
| SamYaple | and the 100 times longer bit isnt completely accurate, it was just to give oy uan idea | 06:02 |
| inc0 | yeah I get that | 06:03 |
| SamYaple | also things like fact caching _can_ be helpful, though i tend to only use them if you have many many nodes. at the scale of 10-20 i would never consider it | 06:03 |
| *** itsuugo has joined #kolla | 06:04 | |
| inc0 | SamYaple, ansible question: what's best approach to situation in which I have 2 types of nodes in cluser and different interface names in each? | 06:16 |
| inc0 | if I use group_vars in inventory, that will work? | 06:16 |
| inc0 | in which I override globals.yml | 06:16 |
| SamYaple | inc0: i have been thinks about this alot | 06:17 |
| SamYaple | long term I think for that we would have to have overrides pernode, but you can set them in the inventory file | 06:17 |
| inc0 | well, per node would be hard, but since I have these grouped, inv and group_vars there should work | 06:17 |
| SamYaple | agreed that would be the more propery way to do it | 06:18 |
| SamYaple | esspecially at scale wher eyour inventory arguably _should_ be in a database | 06:18 |
| SamYaple | but thats ansible stuff outside the scope of kolla | 06:18 |
| inc0 | sure | 06:18 |
| inc0 | I'm deploying this right now so it's my personal inventory | 06:19 |
| SamYaple | but yea variables attached to groups would work fine | 06:19 |
| inc0 | also, ceph caching stuff we were talking about yesterday | 06:20 |
| inc0 | I spoke about it with our ceph team | 06:20 |
| SamYaple | oh boy | 06:20 |
| inc0 | and well...if you have ssd cache on top of hardware cache... | 06:20 |
| inc0 | you don't really need journal right? | 06:20 |
| SamYaple | im not sure i follow? my setup is as follows: | 06:20 |
| SamYaple | 4 spinning disk osds in 'slow-tier' | 06:21 |
| SamYaple | 2 ssh in 'fast-tier' | 06:21 |
| SamYaple | the journal is colocated on the disks | 06:21 |
| inc0 | how did you do tiering? | 06:21 |
| SamYaple | the fast-tier is used to cache the slow-tier | 06:21 |
| SamYaple | http://docs.ceph.com/docs/master/rados/operations/cache-tiering/ | 06:21 |
| inc0 | bache? | 06:21 |
| SamYaple | its pretty straight forward | 06:21 |
| SamYaple | no offical ceph | 06:21 |
| inc0 | ah, so in ceph | 06:21 |
| inc0 | ok | 06:22 |
| SamYaple | it also means i can do erasure coding for the underlying tier and still support rbd | 06:22 |
| SamYaple | currently earaure coding doesnt support rbd, but the cachetier makes the possible! | 06:22 |
| inc0 | noob question: whats erasure coding? | 06:22 |
| SamYaple | fancy cloudy raid | 06:23 |
| SamYaple | instead of 3 copies of the data you can get better reliability for only 1.4 copies | 06:23 |
| SamYaple | (numbers depend on erasure profile) | 06:23 |
| inc0 | wow | 06:24 |
| SamYaple | basically save a whole bunch of space _and_ make the data more resiliant | 06:24 |
| inc0 | I'll read about it, sounds like fun | 06:24 |
| inc0 | (damn I should become operator) | 06:24 |
| SamYaple | earasure coding was around before RAID but it wasn't used due to its overhead which is nothing nowadays | 06:24 |
| SamYaple | nah the only reason i get to play with tihs stuff is because i do it in my free time | 06:25 |
| inc0 | I play too much games :S | 06:25 |
| *** sdake has joined #kolla | 06:25 | |
| inc0 | morning sdake | 06:25 |
| sdake | yo | 06:26 |
| SamYaple | hey sdake | 06:26 |
| sdake | yo | 06:26 |
| SamYaple | tested a full deploy from docker hub with rc1 | 06:26 |
| SamYaple | its awesome | 06:26 |
| SamYaple | very quick | 06:26 |
| inc0 | brb, food | 06:26 |
| *** itsuugo has quit IRC | 06:28 | |
| *** sdake has quit IRC | 06:32 | |
| SamYaple | ha! inc0 https://01.org/intel%C2%AE-storage-acceleration-library-open-source-version/ | 06:38 |
| SamYaple | intel only erasure coding profile | 06:38 |
| *** erkules_ is now known as erkules | 06:55 | |
| *** erkules has quit IRC | 06:55 | |
| *** erkules has joined #kolla | 06:55 | |
| *** tpot has quit IRC | 07:05 | |
| inc0 | cool, I'll ask around about that | 07:07 |
| inc0 | sounds like fun | 07:07 |
| inc0 | SamYaple, for ceph I'm to prepare paritions or disks? | 07:08 |
| inc0 | it's partitions right? | 07:09 |
| *** Kennan has quit IRC | 07:23 | |
| *** Haomeng has quit IRC | 07:23 | |
| *** Haomeng has joined #kolla | 07:23 | |
| *** Kennan has joined #kolla | 07:25 | |
| *** sdake has joined #kolla | 07:25 | |
| *** shardy has joined #kolla | 07:29 | |
| SamYaple | inc0: it uses the entire disk | 07:31 |
| SamYaple | but you wipe the partition and apply a special label to the first partition to bootstrap it | 07:31 |
| *** itsuugo has joined #kolla | 07:34 | |
| inc0 | aaa ok | 07:48 |
| *** achanda has quit IRC | 07:48 | |
| *** mat425 has joined #kolla | 07:49 | |
| SamYaple | parted /dev/$DISK -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP 1 -1 | 07:50 |
| SamYaple | where $DISK is something like 'sdb' | 07:50 |
| SamYaple | inc0: ^ | 07:50 |
| *** gfidente has joined #kolla | 07:59 | |
| *** itsuugo has quit IRC | 08:08 | |
| *** athomas has joined #kolla | 08:16 | |
| SamYaple | inc0: raw text, orugh for ceph | 08:20 |
| SamYaple | http://paste.openstack.org/show/474788/ | 08:20 |
| SamYaple | oops, this one http://paste.openstack.org/show/474789/ | 08:21 |
| inc0 | yeah, I'm trying to write ansible to do it;) | 08:21 |
| inc0 | on 10 nodes with 22 disks each | 08:21 |
| SamYaple | and you can, but that wont ever be in Kolla due to its destructive nature | 08:21 |
| SamYaple | i expect people will automate this step | 08:22 |
| SamYaple | but i dont want to be in the habbit of wiping disks | 08:22 |
| SamYaple | inc0: http://paste.fedoraproject.org/273040/14436013/ | 08:23 |
| inc0 | no no no, it's side script | 08:23 |
| SamYaple | {{ disk }} should be {{ item }} | 08:23 |
| *** tfukushima has joined #kolla | 08:28 | |
| *** shardy_ has joined #kolla | 08:49 | |
| *** shardy has quit IRC | 08:51 | |
| *** shardy_ has quit IRC | 08:55 | |
| *** shardy has joined #kolla | 08:55 | |
| Kennan | hi SamYaple: I tried late kolla code, and find | 09:02 |
| Kennan | ironic-pxe:The command '/bin/sh -c echo 'source not yet available for ubuntu' && /bin/false' returnon-zero code: 1 | 09:02 |
| Kennan | did you try if ironic-pxe related image can be built from source ? | 09:02 |
| SamYaple | Kennan: yes it is not available on ubuntu *yet* | 09:03 |
| SamYaple | i have some code i was about ot push up | 09:03 |
| SamYaple | but the ansible bits dont work if i remember.. | 09:03 |
| Kennan | OK. ironic-pxe is newly added ? OK | 09:04 |
| SamYaple | yes past few days | 09:04 |
| *** mat425 has quit IRC | 09:17 | |
| inc0 | guys, when do we open mitaka branch? | 09:24 |
| SamYaple | inc0: i believe sdake said he was opening up master after tagging rc1 | 09:25 |
| SamYaple | but he may not have yet | 09:26 |
| SamYaple | yes he did | 09:26 |
| SamYaple | we have stable/liberty now | 09:26 |
| SamYaple | so master is effectively mitaka | 09:26 |
| inc0 | cool, I'm thinking about modifying your ceph so you can specify journal disk | 09:28 |
| inc0 | but we'll see if that's required | 09:28 |
| SamYaple | yea it does need to expand, but i want to be careful of how _much_ we expand it | 09:29 |
| SamYaple | this is Openstack, not ceph deploy yo uknow? | 09:29 |
| SamYaple | i mean deploying ceph like this is almost its own project | 09:29 |
| inc0 | well I guess | 09:31 |
| inc0 | but if we presume to deploy ceph, which we do | 09:31 |
| inc0 | we kinda introduced it into our scope | 09:31 |
| SamYaple | when this was initially purposed it was purposed as a 'basic ceph deploy' with the ability to use an external ceph cluster | 09:32 |
| SamYaple | in case it wasn't clear, seperate journal device is cool with me, but i want to avoid complication and it adds a good bit of complication to do seperate journals | 09:32 |
| *** dims has joined #kolla | 09:32 | |
| inc0 | but to be honest, deploying ceph like that would be quite cool stuff itself | 09:33 |
| inc0 | and I know it's a large topic | 09:33 |
| SamYaple | which we could break out that into its own project that kolla consumes | 09:33 |
| SamYaple | then yea we can add all that craziness in | 09:33 |
| inc0 | sure, that would work too | 09:33 |
| SamYaple | but im weary of maintaining that within kolla | 09:33 |
| inc0 | wanna be PTL Sam? | 09:34 |
| SamYaple | im not trying to be dicouraging, just keep that in mind please | 09:34 |
| SamYaple | well it wouldnt be an openstack project | 09:34 |
| inc0 | yeah, I know | 09:34 |
| inc0 | storage is whole new space of problems | 09:34 |
| SamYaple | since it doesn't have anything to do with openstack | 09:34 |
| inc0 | well, that depends | 09:34 |
| *** dims has quit IRC | 09:39 | |
| *** dims has joined #kolla | 09:42 | |
| SamYaple | depends on what? | 09:46 |
| SamYaple | i mean they are related, but if we break out ceph _completely_ then i dont know... | 09:47 |
| *** itsuugo has joined #kolla | 09:56 | |
| inc0 | politics | 10:04 |
| inc0 | quick question, delegate host raises authentication error for me | 10:04 |
| inc0 | doesn't it take info from inv? | 10:04 |
| SamYaple | yea it wont work for AIO and doesn't respect ssh_user according ot pbourke_ | 10:05 |
| SamYaple | idk | 10:05 |
| SamYaple | basically it initiates an ssh session to the host it delegates to | 10:05 |
| SamYaple | it doesnt respect ansible_connection=local or ansible_ssh_user=blah apparently | 10:05 |
| inc0 | it's multinode | 10:06 |
| SamYaple | ok. then its probably the ssh_user thing | 10:06 |
| inc0 | so I need to make paswordless? | 10:06 |
| SamYaple | probably | 10:06 |
| *** mat425 has joined #kolla | 10:08 | |
| *** tfukushima has quit IRC | 10:11 | |
| *** pbourke_ has quit IRC | 10:34 | |
| *** pbourke_ has joined #kolla | 10:35 | |
| *** shardy_ has joined #kolla | 10:39 | |
| *** shardy has quit IRC | 10:41 | |
| *** shardy_ has quit IRC | 10:45 | |
| *** shardy has joined #kolla | 10:45 | |
| mat425 | kolla installation on vagrant takes incredibly long time... :| | 10:49 |
| SamYaple | sorry mat425 i haven't tried the vagrant stuff yet | 10:50 |
| SamYaple | pbourke_ harmw_ mandre can you comment on speed of vagrant? | 10:50 |
| pbourke_ | a deploy of core services takes maybe about 3 mins | 10:51 |
| SamYaple | that can't be | 10:51 |
| SamYaple | thats way faster than even baremetal | 10:51 |
| SamYaple | i can do 6min in AIO on very fast baremetal | 10:51 |
| pbourke_ | maybe im being optimistic | 10:51 |
| SamYaple | with images prestaged | 10:51 |
| pbourke_ | i'll time a run now | 10:51 |
| mat425 | just the build-kolla command is running from 1.5h. but even the vagrant up cmd last about 1h | 10:52 |
| SamYaple | mat425: something wrong... gotta be | 10:52 |
| SamYaple | whats your environment? | 10:52 |
| mat425 | vcpu issue? the param is not set in vagrantfile | 10:53 |
| mat425 | i leave vmemory as is (4GB) | 10:53 |
| mat425 | (single node) | 10:53 |
| mat425 | xubuntu 14.04 64bit | 10:53 |
| SamYaple | are you out of memory? and swapping to the disk? | 10:54 |
| mat425 | sometimes it used swap | 10:55 |
| mat425 | not for the whole process | 10:56 |
| inc0 | damn these 2 ifaces are real pain | 11:01 |
| SamYaple | which ones? | 11:02 |
| pbourke_ | 7 mins :) | 11:03 |
| inc0 | fact that I don't have homogenous setup | 11:05 |
| SamYaple | hey pbourke_ | 11:09 |
| SamYaple | pbourke_: https://github.com/openstack/kolla/blob/master/docker/swift/swift-object-base/config-swift-object.sh | 11:09 |
| SamYaple | is that still used... | 11:09 |
| pbourke_ | no | 11:09 |
| SamYaple | ok removing | 11:09 |
| pbourke_ | there's some crud in swift that needs to go | 11:09 |
| SamYaple | well like that whole file i dont see where it gets called at all | 11:10 |
| pbourke_ | i dont think object-base is needed after we moved to ansible | 11:10 |
| mat425 | pbourke_ 7 min to deploy kolla on vagrant? | 11:10 |
| pbourke_ | mat425: yes, but no every service, just the ones needed for nova | 11:11 |
| pbourke_ | *not | 11:11 |
| mat425 | oh jeez... | 11:11 |
| pbourke_ | yours in longer? | 11:11 |
| pbourke_ | *is | 11:11 |
| mat425 | yes still running | 11:11 |
| pbourke_ | try boosting memory | 11:12 |
| pbourke_ | I would give minimum of 8GB in an AIO | 11:12 |
| mat425 | already reach the cap of my laptop :' | 11:12 |
| mat425 | i have 8 GB | 11:12 |
| mat425 | and 4GB for vagrant | 11:12 |
| pbourke_ | yeah its going to struggle I think | 11:13 |
| mat425 | and inside the vm still has 2GB free | 11:13 |
| mat425 | okok, i'm gonna try on a bigger one | 11:14 |
| *** itsuugo has quit IRC | 11:21 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 11:29 |
| openstackgerrit | Mick Thomspon proposed openstack/kolla: Add missing core library to murano https://review.openstack.org/229372 | 11:29 |
| SamYaple | wowzers a 700 line reduction in codebase size for removal of duplicate start.sh | 11:33 |
| inc0 | SamYaple, failed: [10.237.154.17] => (item=(11, {'device': '/dev/sdn', 'fs_uuid': u''})) => {"changed": true, "cmd": ["docker", "wait", "bootstrap_osd_11"], "delta": "0:00:00.034045", "end": "2015-09-30 07:42:36.637229", "failed": true, "failed_when_result": true, "item": [11, {"device": "/dev/sdn", "fs_uuid": ""}], "rc": 0, "start": "2015-09-30 07:42:36.603184", "stdout_lines": ["1"], "warnings": []} | 11:44 |
| inc0 | output of "Waiting for bootstrap containers to exit" | 11:44 |
| inc0 | can you make any sense out of it? | 11:44 |
| SamYaple | inc0: the bootstrap container should still exist | 11:46 |
| SamYaple | docker logs bootstrap_osd_11 | 11:46 |
| SamYaple | docker wait returns the exist code of the server | 11:46 |
| SamYaple | container* | 11:46 |
| SamYaple | so the container bombed out, but why it bombed out you will ahve to find out from the container | 11:47 |
| inc0 | Error connecting to cluster: PermissionError it's not too helpful tbh :S | 11:47 |
| SamYaple | yea thats just the output of the ceph client | 11:47 |
| SamYaple | thats not much i can help with other than guess | 11:47 |
| SamYaple | did _all_ containers go away? | 11:48 |
| inc0 | all bootstrap ones | 11:48 |
| SamYaple | you shoudl fix your networking :P | 11:48 |
| inc0 | librados: client.admin authentication error (1) Operation not permitted that was real problem, but still not helpful | 11:51 |
| SamYaple | ah so youve probably rekicked a few times | 11:51 |
| inc0 | seems like keys are broken, which might be the case cause I changed place of mons | 11:51 |
| inc0 | yeah I did | 11:51 |
| SamYaple | and you didnt removed /etc/kolla/ceph-mon | 11:51 |
| SamYaple | so you had old keys | 11:51 |
| inc0 | makes sense, let me fix that | 11:51 |
| *** jruano has joined #kolla | 11:52 | |
| inc0 | shouldn't ansible detect that it's supposed to change? | 11:52 |
| SamYaple | no you had old cluster keys | 11:52 |
| SamYaple | not like it didnt copy them | 11:52 |
| SamYaple | you generated them in an old cluster | 11:52 |
| *** dwalsh has joined #kolla | 12:17 | |
| *** dwalsh_ has joined #kolla | 12:17 | |
| *** sdake has quit IRC | 12:22 | |
| *** sdake has joined #kolla | 12:23 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 12:25 |
| *** mat425 has quit IRC | 12:27 | |
| SamYaple | inc0: also watch out, ansible will populate the /etc/fstab on your host box | 12:28 |
| inc0 | these are dedicated machines for this purpose so I don't care | 12:29 |
| inc0 | right now I'm still battling | 12:29 |
| inc0 | aaaand delegate-f***ing-host again | 12:29 |
| inc0 | this thing is broken | 12:30 |
| *** diga has quit IRC | 12:31 | |
| *** jruano has quit IRC | 12:31 | |
| *** rhallisey has joined #kolla | 12:32 | |
| *** jtriley has joined #kolla | 12:33 | |
| sdake | samyaple ping q re your patch | 12:34 |
| sdake | does the CMD in base always run if no CMD is specifiedin a child container? | 12:34 |
| *** banix has joined #kolla | 12:37 | |
| *** jtriley has quit IRC | 12:37 | |
| *** Kennan has quit IRC | 12:38 | |
| *** Kennan has joined #kolla | 12:39 | |
| SamYaple | sdake: correct, but im testing | 12:40 |
| SamYaple | oops +w was added | 12:40 |
| SamYaple | sdake: ill make sure it doesn't merge before you verify centos | 12:41 |
| SamYaple | inc0: yea so basically just make sure you have ssh keys for root pushed around | 12:42 |
| SamYaple | inc0: thats what our docs say. liberty requires root | 12:42 |
| SamYaple | things gonna break without it | 12:42 |
| inc0 | LM will | 12:42 |
| inc0 | I know that | 12:42 |
| inc0 | anyway, I'm already past that | 12:43 |
| SamYaple | still borked? | 12:43 |
| *** achanda has joined #kolla | 12:44 | |
| inc0 | yeah, now ceph_mon doesn't want to start, also we need to fix error handling there https://github.com/openstack/kolla/blob/master/ansible/roles/ceph/tasks/distribute_keyrings.yml#L5 | 12:44 |
| inc0 | no json object could be decoded is not very helpful | 12:44 |
| inc0 | but I'm past that as well, now debugging ceph_mon not starting | 12:45 |
| *** itsuugo has joined #kolla | 12:46 | |
| *** jtriley has joined #kolla | 12:47 | |
| openstackgerrit | Martin André proposed openstack/kolla: Fix wrong option name in kolla-build.conf https://review.openstack.org/229406 | 12:53 |
| inc0 | http://paste.openstack.org/show/474830/ SamYaple halp! | 12:55 |
| *** achanda has quit IRC | 13:00 | |
| inc0 | ceph_mon doesn't start | 13:07 |
| inc0 | ceph_mon_data is ok | 13:08 |
| *** jtriley has quit IRC | 13:12 | |
| inc0 | I've added this clobber opt and rebuilt container, might be worth checking out what have happened | 13:13 |
| *** dwalsh_ has quit IRC | 13:18 | |
| *** dwalsh has quit IRC | 13:18 | |
| SamYaple | inc0: stahp | 13:18 |
| SamYaple | inc0: what are yo udoing | 13:18 |
| SamYaple | cleanup your environment | 13:19 |
| SamYaple | you did not remove previous keys | 13:19 |
| *** dwalsh has joined #kolla | 13:19 | |
| *** banix has quit IRC | 13:23 | |
| SamYaple | mandre: well also need to cherry-pick that to the liberty branch | 13:23 |
| SamYaple | sdake: we need a proceedure around tracking backport things, how do other projects do this? | 13:24 |
| SamYaple | ive seen a few ways | 13:24 |
| openstackgerrit | Merged openstack/kolla: Fix wrong option name in kolla-build.conf https://review.openstack.org/229406 | 13:26 |
| *** itsuugo has quit IRC | 13:28 | |
| mandre | SamYaple, you're right, thanks for the reminder | 13:30 |
| *** banix has joined #kolla | 13:30 | |
| inc0 | SamYaple, where are these keys? | 13:31 |
| inc0 | /etc/kolla/ceph-mon on ansible host? | 13:31 |
| SamYaple | inc0: /etc/kolla/ceph-mon on all the monitor nodes | 13:34 |
| inc0 | ah | 13:34 |
| SamYaple | ansible ceph-mon -m file -a 'state=absent' -i /path/to/inventory | 13:34 |
| SamYaple | that should work.... | 13:34 |
| inc0 | I have just one mon node right now | 13:35 |
| *** dwalsh_ has joined #kolla | 13:35 | |
| SamYaple | ok. just fyi without 3 nodes bad things all around | 13:36 |
| SamYaple | i mean ceph can do aio but you have to hack the crush map to do it right and its a mess | 13:36 |
| inc0 | without 3 mons or 3 nodes? | 13:37 |
| inc0 | I mean, I have 1 mon and shitload of osd | 13:37 |
| inc0 | but ok, I'll add 2 more mons | 13:37 |
| SamYaple | inc0: yea 3 mons and many osds on several different hosts | 13:39 |
| SamYaple | (50 on one host will still violate the crush map placements stuffs) | 13:39 |
| *** jtriley has joined #kolla | 13:40 | |
| *** dims has quit IRC | 13:49 | |
| *** dims has joined #kolla | 13:49 | |
| *** banix has quit IRC | 13:56 | |
| *** itsuugo has joined #kolla | 14:08 | |
| *** shardy_ has joined #kolla | 14:13 | |
| *** jtriley has quit IRC | 14:14 | |
| *** shardy has quit IRC | 14:14 | |
| *** shardy_ has quit IRC | 14:18 | |
| *** shardy has joined #kolla | 14:19 | |
| *** athomas has quit IRC | 14:22 | |
| *** jruano has joined #kolla | 14:27 | |
| *** athomas has joined #kolla | 14:27 | |
| pbourke_ | SamYaple: do you have a neutron setup atm? | 14:33 |
| *** jtriley has joined #kolla | 14:35 | |
| sdake | pbourke_ is there a neutron problem? | 14:36 |
| pbourke_ | sdake: maybe | 14:36 |
| sdake | tell me more pls | 14:37 |
| pbourke_ | can anyone do an 'ip netns exec' on the network host | 14:37 |
| sdake | i donthave a deployment activec and have a presentation in 15 mins | 14:37 |
| sdake | but after I can look | 14:37 |
| pbourke_ | e.g. ip netns exec qdhcp-blah ip a | 14:37 |
| sdake | what is neutron doing or not doing propery? | 14:37 |
| pbourke_ | doing so gives me: seting the network namespace "qdhcp-46f7fce8-65a5-4123-859c-26682b811b3e" failed: Invalid argument | 14:38 |
| pbourke_ | now, this may not be a problem - it works from within the neutron containers themselves | 14:38 |
| pbourke_ | but if you restart the container it then gives the same error as above | 14:38 |
| pbourke_ | again, maybe benign, but it doesn't seem right | 14:39 |
| pbourke_ | Im having trouble getting an instance to get an ip via dhcp, which lead me down this path | 14:39 |
| sdake | ip address assignment has been wroking pretty solidly for me | 14:42 |
| sdake | i use centos binary tho | 14:42 |
| sdake | for the most part | 14:42 |
| sdake | oh if we only had functionaltesting :) | 14:42 |
| *** shardy_ has joined #kolla | 14:42 | |
| pbourke_ | yeah | 14:43 |
| sdake | pbourke_ do you notice a regression lately in neutron or has it always been this way? | 14:43 |
| pbourke_ | it probably does work for most people, otherwise there's be more shouting | 14:43 |
| pbourke_ | the network namespace error worries me slightly as I remember seeing it way back when we were beginning to containerise neutron | 14:43 |
| pbourke_ | and assumed it was sorted | 14:43 |
| *** inc0 has quit IRC | 14:44 | |
| *** shardy has quit IRC | 14:44 | |
| SamYaple | pbourke_: yes | 14:45 |
| SamYaple | pbourke_: no | 14:45 |
| SamYaple | pbourke_: you cannot do that from the host | 14:45 |
| pbourke_ | SamYaple: ok so that's expected | 14:46 |
| pbourke_ | SamYaple: what about restarting the container | 14:46 |
| SamYaple | yes, but we should add a docker volume to prevent those netns from showing up on the host like that | 14:46 |
| SamYaple | pbourke_: yea you can restart teh container | 14:46 |
| sdake | SamYaple I suggested that about a month ago but you said it wouldnt work so I never added it | 14:46 |
| SamYaple | pbourke_: fyi i justa added a comment about that https://review.openstack.org/#/c/229371/2/docker/neutron/neutron-base/ip_wrapper.py | 14:47 |
| SamYaple | sdake: i don't recall that... maybe i did try this before and it broke? | 14:47 |
| SamYaple | or maybe i miscommunicated? | 14:47 |
| sdake | not sure | 14:47 |
| sdake | i was pretty clear about addinga data container for neutron and you said itwouldl not function properly | 14:47 |
| sdake | for the netns files | 14:48 |
| *** shardy_ has quit IRC | 14:48 | |
| SamYaple | ohhhh i may have confused what yo uwere asking | 14:48 |
| SamYaple | i thought you were saying make it accessible from host | 14:48 |
| SamYaple | and no, that wouldnt make it accessible form host | 14:48 |
| SamYaple | just hide it from host | 14:48 |
| pbourke_ | can I get a summary of why they're not accesible from the host? | 14:48 |
| sdake | nah just to make one to hide it from the host | 14:48 |
| SamYaple | yes sorry. my fault | 14:48 |
| pbourke_ | or why they can't be | 14:48 |
| SamYaple | pbourke_: lol yea.... | 14:49 |
| SamYaple | its dockers fault | 14:49 |
| *** shardy has joined #kolla | 14:49 | |
| SamYaple | that ip_wrapper.py i linked you too works! | 14:49 |
| SamYaple | on docker 1.7.0-dev2..... | 14:49 |
| SamYaple | it was removed as a regressing in -dev3 | 14:49 |
| SamYaple | sdake was talking with someone to get it added back in.... sdake? | 14:49 |
| *** banix has joined #kolla | 14:49 | |
| sdake | need a bug link to communicate what needs to be done | 14:49 |
| SamYaple | its not a bug | 14:50 |
| sdake | need some type of reference | 14:50 |
| sdake | rather then dev3 added a regession | 14:50 |
| sdake | i thought i had it in my inbox but i can't find it | 14:50 |
| SamYaple | let me get the email i sent to you last time | 14:50 |
| SamYaple | this conversation seems familiar | 14:50 |
| sdake | right we have had it several times :) | 14:50 |
| sdake | and each time i think we end up at not being able to find the email with the references | 14:51 |
| sdake | you did sned it once i remember erading it | 14:51 |
| SamYaple | actually i think i dont look | 14:51 |
| sdake | but cant find it | 14:51 |
| sdake | outlook kind of sucks for searching | 14:51 |
| sdake | it had somethign to do with docker machine | 14:52 |
| pbourke_ | id like to understand it better | 14:52 |
| SamYaple | floker was bbroken | 14:52 |
| SamYaple | i think | 14:52 |
| SamYaple | pbourke_: i will explain in a minute | 14:52 |
| sdake | it could be that | 14:52 |
| SamYaple | going to get this information out | 14:52 |
| pbourke_ | ok thanks | 14:52 |
| sdake | SamYaple email to steven.dake@gmail.com | 14:52 |
| sdake | so i can search for it | 14:52 |
| sdake | put something like "docker regression" in the subject or something | 14:53 |
| pbourke_ | if anyone gets a few mins, could you please try this scenario: http://paste.openstack.org/show/474711/ (ignore the nova flavor its way too small) | 14:53 |
| sdake | it will take a lot of conversation to get it fixed up | 14:53 |
| sdake | if at all | 14:53 |
| SamYaple | sdake: https://github.com/openstack/kolla/commit/62e02fdbf4a4f714d008943c4e5c8512a93807ed | 14:57 |
| SamYaple | https://github.com/docker/docker/commit/ff770d33cdee5ab4dc0bdeb06dcfea49aa72ab8e | 14:57 |
| sdake | samyaple can you amil that to me | 14:57 |
| sdake | so i can find it later | 14:57 |
| SamYaple | yes which email | 14:57 |
| sdake | i have meetings all morning | 14:57 |
| sdake | steven.dake@gmail.com plz | 14:57 |
| SamYaple | hey guys, milestone | 14:59 |
| sdake | can someone look at this bug https://bugs.launchpad.net/kolla/+bug/1498694 | 14:59 |
| openstack | Launchpad bug 1498694 in kolla "neutro-server broken in kolla" [Critical,Triaged] | 14:59 |
| SamYaple | outside confirmation ceph works | 14:59 |
| SamYaple | sdake: i think thats delorean only | 15:00 |
| SamYaple | ok pbourke_ you around for a story? | 15:02 |
| pbourke_ | yup | 15:02 |
| SamYaple | story time then | 15:02 |
| SamYaple | when you create a network namespace $(ip netns add namepsace_name) it creates a mountpoint at /run/netns/namespace_name | 15:03 |
| SamYaple | it uses the for proc for that namespace | 15:03 |
| SamYaple | now when you create a container it creates a _new_ mountnamespace | 15:03 |
| SamYaple | it will inherit all host mounts if you bind it in | 15:03 |
| SamYaple | so if you bind in /run/netns with an already create mount namespace both host and container work | 15:04 |
| SamYaple | but if you create another one on the host, it wont propogate into the container | 15:04 |
| SamYaple | if you create one in the container it wont ever propogate up | 15:04 |
| SamYaple | (mounts are inherited) | 15:04 |
| SamYaple | well if you create your mount namespace when you create teh container in a certain way the host /run/netns _can_ propogate into the container on the fly | 15:05 |
| SamYaple | so new netns on the host can be accessed in the container | 15:05 |
| SamYaple | docker doesnt create it this special way, and they dont allow that to be configurable | 15:05 |
| pbourke_ | why dont they propagate | 15:05 |
| SamYaple | pbourke_: http://www.ibm.com/developerworks/library/l-mount-namespaces/index.html | 15:06 |
| SamYaple | great article on practicle usage of private mount namepsace | 15:06 |
| SamYaple | shared mount is what we want, private mount is what we have | 15:06 |
| pbourke_ | which is why we have fat container for agents right | 15:07 |
| SamYaple | yes | 15:07 |
| SamYaple | (ps i build custom binaries to make this work for me ;) | 15:07 |
| pbourke_ | why are we binding in /run then | 15:08 |
| *** sdake has quit IRC | 15:08 | |
| SamYaple | shared socket between containers i think | 15:08 |
| SamYaple | dont really recall | 15:09 |
| pbourke_ | right | 15:09 |
| SamYaple | either way we can put /run/netns in a volume for the agents container and make them invilible to the host | 15:09 |
| pbourke_ | ok think that makes sense thanks! | 15:09 |
| pbourke_ | will read the article also | 15:09 |
| SamYaple | yes it should make it more clear | 15:09 |
| pbourke_ | have you had any problems with instances not being able to get an ip via dhcp | 15:10 |
| *** sdake has joined #kolla | 15:10 | |
| SamYaple | nope | 15:10 |
| SamYaple | ive been kicking it all day with the rc1 images i pushed to docker hub | 15:10 |
| SamYaple | i did not push centos images | 15:10 |
| SamYaple | (i cant test them on ubuntu so i didnt push) | 15:10 |
| *** Hui has joined #kolla | 15:11 | |
| SamYaple | s/win 7 | 15:14 |
| *** dims_ has joined #kolla | 15:16 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 15:17 |
| *** itsuugo has quit IRC | 15:17 | |
| *** JohnN has joined #kolla | 15:19 | |
| *** dims has quit IRC | 15:20 | |
| sdake | SamYaple can you push the centos images anyway | 15:21 |
| sdake | and I'll test them | 15:21 |
| sdake | if they are bust, it wont be much harm | 15:21 |
| sdake | as we will push final images for release | 15:22 |
| SamYaple | sdake: sure but they arent build atm | 15:22 |
| SamYaple | i purges | 15:22 |
| Hui | Hi, Sdake, I have created a blueprint about consul. If you have time, could you help to review? | 15:22 |
| SamYaple | will build | 15:22 |
| sdake | hui I did review the patch | 15:23 |
| Hui | Thanks. | 15:23 |
| sdake | hui I didnt' provide a review inline but looked it over | 15:23 |
| SamYaple | sdake: i dont see a review | 15:23 |
| sdake | I have trouble understanding the value of consul for kolla compared to its relative complexity level which is high | 15:23 |
| sdake | like I said I didn't provide a review but looked itover | 15:23 |
| SamYaple | oh i see | 15:23 |
| SamYaple | sorry | 15:23 |
| sdake | no sweat | 15:24 |
| sdake | i think our messages wer emid-air :) | 15:24 |
| SamYaple | so happy right now everyone | 15:24 |
| SamYaple | CBR09 just reported a successful deployment with ceph | 15:24 |
| sdake | hui I read the spec about 5-10 times and I just don't see a compelling value to it | 15:24 |
| sdake | samyaple what is cbr09 | 15:25 |
| SamYaple | a person in hte channel (someitmes) | 15:25 |
| sdake | oh i see | 15:25 |
| Hui | Sdake, the basic idea is for dynamic service discovery | 15:25 |
| SamYaple | pbourke_: https://bugs.launchpad.net/kolla/+bug/1500858 i responded here. lets continue the discussion in the upgrade blueprint i think? | 15:26 |
| openstack | Launchpad bug 1500858 in kolla "Config changes do not cause a container restart when using COPY_ALWAYS" [Undecided,New] | 15:26 |
| sdake | samyaple https://bugs.launchpad.net/kolla/+bug/1501167 | 15:27 |
| openstack | Launchpad bug 1501167 in kolla "Rsyslog doesn't work on ubuntu" [Undecided,New] | 15:27 |
| SamYaple | closing now | 15:27 |
| SamYaple | it was apparmor | 15:27 |
| SamYaple | updating it literally right now | 15:27 |
| sdake | hui yes I just don't understand why kolla needs dynamic service discovery compared to the high degree of complexity | 15:27 |
| pbourke_ | SamYaple: sure, thanks for that | 15:27 |
| sdake | essentially it adds something that doesn't offer a lot of value with a high degree of complexity | 15:27 |
| sdake | the idea of kolla is to keep things simple so the development effort can scale well | 15:28 |
| sdake | hui for example, the HA system in kolla is highly complex in my opinion, but offers huge value | 15:28 |
| sdake | so I accept it :) | 15:28 |
| Hui | sdake, do you mean the HA blueprint under spec? | 15:29 |
| sdake | the gain is worth the pain | 15:29 |
| sdake | no i mean the ha implementation in general | 15:29 |
| sdake | hui we dont use specs in kolla typically | 15:29 |
| sdake | only for contentious items | 15:29 |
| sdake | that way we can have a common viewpoint | 15:29 |
| sdake | we typically rollcall vote specs | 15:29 |
| Hui | sdake, where can I find the HA doc in kolla? | 15:29 |
| sdake | the only one is in the specs directory | 15:30 |
| sdake | hui I realize you did alot of work with the consul dev, and maybe down the road it would be appropriate, but atm it seems like overkill for a simple problem | 15:30 |
| sdake | and I hope you aren't put off by lack of reception around the idea, but a whoel bunch of code was written without submitting a blueprint of communicating the idea was to be done | 15:31 |
| sdake | so the development could be guided | 15:31 |
| Hui | sdake, ok | 15:31 |
| sdake | hui clearly you are very sharp and understand kolla, so I like that :) | 15:32 |
| SamYaple | just to clarify something here. I didn't see what it added that we dont already do in one way or another. We already hot reload haproxy, while its not dynamic discovery it is still dynamicly updating the environment when running the playbook | 15:32 |
| sdake | its just that the complexity of consul I dont get | 15:32 |
| SamYaple | if you can make it add more value or explain how it already add more value i think youll find it more well recieved | 15:32 |
| sdake | agree with SamYaple | 15:32 |
| sdake | hui besides the complexity vs gain issue, the workflow is not ideal | 15:33 |
| Hui | sdake, could you give some guide on the workflow? | 15:34 |
| sdake | I want workflows that are essentially single commands that gte a huge bunch of output | 15:34 |
| sdake | what I mean is in the spec you wrote, to get an operational deployment several different commands must be run | 15:34 |
| sdake | atm, we hae one command that must be run to get a deployment | 15:34 |
| Hui | sdake, I use ansible tags to explain what you will expect at each step | 15:35 |
| sdake | yes, but i dont want the operator to have to understand how ansible tags work | 15:35 |
| Hui | but essetionally, you only need one ansible command to deploy | 15:35 |
| sdake | i want thme to run a command "deploy" and out the other end comes a deployment | 15:36 |
| sdake | i see, that was possbily a misunderstanding on my part | 15:36 |
| Hui | I add a variable enable_consul, so that you can choose whether or not use this feature | 15:36 |
| sdake | the fundametal issue though is the complexity vs the gain | 15:36 |
| Hui | ok | 15:36 |
| sdake | perhaps I don't understand the gain well enough | 15:36 |
| sdake | and as sam sai,d if that were better communicated itwould help | 15:37 |
| sdake | perhaps a mailinglist thread would help in this case | 15:37 |
| SamYaple | yea Hui i do see some advatanges here. please do pursue this so we can make an informed decision here | 15:38 |
| Hui | sdake, shall we use the launchpad as the mailinglist? | 15:38 |
| sdake | launchpad works too, we can discuss in the blueprint | 15:38 |
| sdake | typically what happens is I put blueprints into the discussion phase | 15:38 |
| sdake | and then discussion hapepns in launchpad | 15:39 |
| Hui | ok, sounds great | 15:39 |
| sdake | we have been super busy with liberty-3 and rc1, so wehavne't been on top of all the new blueprints or folowing that process | 15:39 |
| sdake | basically we have been locked down for about 6 weeks preparing for the release | 15:39 |
| Hui | : ) | 15:40 |
| sdake | hui it almost looks liek you had that code written based around the kolla single node deployment using compose | 15:41 |
| sdake | is that accurate? | 15:41 |
| Hui | sdake, no. The current implementation requres multi-node deployment | 15:41 |
| Hui | using ansible | 15:41 |
| sdake | ok thanks | 15:42 |
| sdake | i saw that, it just looked like it was baed upon older kolla work | 15:42 |
| Hui | the code is based on the master branch | 15:42 |
| sdake | yes I saw that | 15:42 |
| Hui | like Sam removed the config-external.sh | 15:42 |
| sdake | what I mean is it looks like it was started some time ago and then rebased | 15:42 |
| sdake | and ryan as well :) | 15:42 |
| sdake | hey folks if you triage a bug go ahead and set the priority and assign it to thenext rc | 15:43 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 15:43 |
| sdake | pbourke_ SamYaple you guys seem to help with the traiging so keep that in mind pls :) | 15:43 |
| sdake | and set to "triaged" | 15:43 |
| sdake | if your not sure what the priority shouldbe just make a best gues | 15:44 |
| sdake | i always error on the side of making the bug more critical even if its not confirmed | 15:44 |
| SamYaple | yea ive been cleaning up | 15:45 |
| sdake | once a bug enters the confirmed state the priority shoudl have a second pass | 15:45 |
| SamYaple | but i want to get any major changes done first | 15:45 |
| sdake | SamYaple sorry don't understand what you mean? | 15:45 |
| SamYaple | sorry wrong window | 15:46 |
| sdake | that would explain it :) | 15:46 |
| pbourke_ | jpeeler: but submitting reviews for master is the usual case right? | 15:51 |
| jpeeler | pbourke_: the policy is you must check master and submit first there if applicable i believe | 15:52 |
| jpeeler | honestly i forgot how the stable policy works as far as marking for backports, etc | 15:52 |
| pbourke_ | I think I dont understand the current branching policy | 15:52 |
| pbourke_ | is there a wiki page or something | 15:52 |
| jpeeler | should talk about it in the meeting i guess | 15:52 |
| pbourke_ | ah yeah that will do | 15:53 |
| sdake | what is the issue pbourke | 15:53 |
| pbourke_ | sdake: don't understand why reviews are being proposed for stable/liberty instead of master | 15:53 |
| sdake | they should be proposed for master first | 15:53 |
| jpeeler | that review doesn't apply to master | 15:54 |
| sdake | then once in master, git cherrypick to stable/liberty | 15:54 |
| pbourke_ | https://review.openstack.org/#/c/229429/ | 15:54 |
| sdake | i think thre ws a master version of that | 15:54 |
| jpeeler | yeah that one should have been cherry picked with -x | 15:55 |
| jpeeler | sdake: there was | 15:55 |
| sdake | right | 15:55 |
| jpeeler | https://wiki.openstack.org/wiki/StableBranch | 15:55 |
| pbourke_ | ok cool | 15:55 |
| pbourke_ | but I dont think 'git review' should target stable/liberty by default | 15:56 |
| jpeeler | that page even goes so far as to say that cherry pick with -x is needed to avoid a gerrit bug | 15:56 |
| jpeeler | pbourke_: it should in the liberty branch | 15:56 |
| sdake | as we are not a clinet/server project we will have slightly different policies towards backports | 15:56 |
| sdake | which we can define on our own | 15:56 |
| jpeeler | still, it's a good practice | 15:56 |
| sdake | that backport policy is targeted around 90% of the projects, which are client/server project | 15:56 |
| pbourke_ | jpeeler: aaaaaah | 15:58 |
| pbourke_ | jpeeler: sorry I get it now | 15:58 |
| sdake | pbourke ya stable/liberty should have the default gerrit set as it is in the depends-on patch | 15:58 |
| jpeeler | ok cool | 15:58 |
| pbourke_ | thbanks | 15:58 |
| *** inc0 has joined #kolla | 15:59 | |
| inc0 | wassup my dear kolla'gues? | 15:59 |
| sdake | lol | 15:59 |
| sdake | kolla-gues | 15:59 |
| *** bmace has quit IRC | 16:00 | |
| *** blahRus has joined #kolla | 16:00 | |
| inc0 | SamYaple, I | 16:00 |
| inc0 | I've given up on ceph | 16:00 |
| inc0 | I'll try again in few days maybe | 16:01 |
| inc0 | but that was because of mess I made | 16:02 |
| *** dims_ has quit IRC | 16:06 | |
| openstackgerrit | Hui Kang proposed openstack/kolla: Add consul containers for dynamic configuration and discovery. https://review.openstack.org/228719 | 16:06 |
| SamYaple | inc0: i understand | 16:10 |
| SamYaple | i have some cleanup aliases if you are interested | 16:10 |
| inc0 | not right now, I'll be off for 4 days - famili matters | 16:10 |
| inc0 | but hold on to them please | 16:10 |
| inc0 | this ceph is now my idee fixe, I will bring it to the end...his or mine | 16:11 |
| *** bmace has joined #kolla | 16:13 | |
| *** jtriley has quit IRC | 16:14 | |
| SamYaple | lol | 16:14 |
| inc0 | family** | 16:14 |
| sdake | team meeting in 3 mins | 16:27 |
| *** banix has quit IRC | 16:29 | |
| sdake | irc meeting in #openstack-meeting-4 | 16:31 |
| sdake | pbourke_ meeting time if your available | 16:32 |
| pbourke_ | sdake: thanks for the reminder! | 16:32 |
| *** dims_ has joined #kolla | 16:33 | |
| *** unicell has quit IRC | 16:39 | |
| *** banix has joined #kolla | 16:44 | |
| *** devlaps has joined #kolla | 16:50 | |
| *** achanda has joined #kolla | 16:58 | |
| *** achanda has quit IRC | 17:01 | |
| *** dwalsh_ has quit IRC | 17:01 | |
| *** jtriley has joined #kolla | 17:03 | |
| *** harmw_ is now known as harmw | 17:09 | |
| harmw | SamYaple: already did :) | 17:09 |
| *** jtriley has quit IRC | 17:09 | |
| inc0 | thanks harmw :) | 17:09 |
| harmw | welcome inc0, I guess :> | 17:09 |
| SamYaple | harmw: awesome! | 17:09 |
| SamYaple | exciting times | 17:09 |
| inc0 | so now I got +1 from all of cores;) cool! thank you guys! I really appreciate that | 17:09 |
| SamYaple | harmw: can you help me bring designate to ansible? | 17:09 |
| SamYaple | or can you do it :P | 17:10 |
| harmw | and excuse me for not being as active as I was some months ago | 17:10 |
| harmw | ansibalising designate is about the one biggest thing on my list :p | 17:10 |
| SamYaple | lol ok | 17:10 |
| SamYaple | ill trust you | 17:10 |
| SamYaple | so sdake to be clear on something | 17:10 |
| SamYaple | we are not backport something like designate? | 17:11 |
| SamYaple | that seems fairly nonintrusive to backport | 17:11 |
| harmw | designate is quote selfsupporting, indeed | 17:11 |
| harmw | no overly wicked dependencies or something | 17:11 |
| sdake | SamYaple I am not certain | 17:12 |
| SamYaple | i just mean if it doesnt change the container then the ansible code should be pretty backport flexible i think | 17:12 |
| sdake | my gut tells me if we don't we will end up lagging for liberty | 17:12 |
| *** unicell has joined #kolla | 17:12 | |
| sdake | myg ut also tells me it is dangerous to backport a bunch of code | 17:12 |
| SamYaple | but the way our roles work its super isolated | 17:12 |
| inc0 | well, this is pretty portable | 17:12 |
| sdake | right | 17:12 |
| SamYaple | so enable_designate: false would be the default obviously | 17:13 |
| sdake | like I said, we will end with different policies thn the typical client/server project | 17:13 |
| SamYaple | but if it doesnt affect other hosts.... | 17:13 |
| inc0 | so if someone have liberty stabe and wants designate, this should be pretty easy to do anyway | 17:13 |
| sdake | the one probelm is the /etc directory | 17:13 |
| SamYaple | just dont -2 it inc0 | 17:13 |
| SamYaple | sdake: agree. we need a good script for that | 17:13 |
| SamYaple | basically i want a script to run and generate/update /etc/kolla | 17:14 |
| SamYaple | so touch new files if needed, change globals.yml if they change | 17:14 |
| SamYaple | that kinda thing | 17:14 |
| SamYaple | preserve settings of course | 17:14 |
| sdake | ya its a shame we weren't able to write that already | 17:14 |
| inc0 | I won't just let's not rush | 17:14 |
| SamYaple | lol i was kidding inc0 dont let it get to your head | 17:14 |
| inc0 | and I'd rather keep enable_designate: no as default | 17:14 |
| inc0 | SamYaple, you did throw -2s freely:P | 17:15 |
| SamYaple | lies! look at my stats | 17:15 |
| SamYaple | 2.5% of my reviews were -2 | 17:16 |
| sdake | the /etc directory is almost like a database migration | 17:16 |
| inc0 | you have more -2s than few ptls I know | 17:16 |
| SamYaple | and 4 of those were for my patches! | 17:16 |
| sdake | because we use /etc almost as a database | 17:16 |
| inc0 | during all of their coreship (is this a word?) | 17:16 |
| SamYaple | sdake: yea. working on that | 17:16 |
| sdake | and its obvious database migration is not ideal during a backport | 17:16 |
| SamYaple | you need to look at stackalytics more. ireally dont | 17:16 |
| SamYaple | sdake: well only if you enable designate | 17:17 |
| SamYaple | if you dont no migration needed | 17:17 |
| SamYaple | if we are calling it that | 17:17 |
| sdake | lets do ourselves a favor and not call it migration | 17:17 |
| SamYaple | yes | 17:17 |
| SamYaple | please | 17:17 |
| sdake | lets make up a new word if need be :) | 17:17 |
| SamYaple | kollafication | 17:17 |
| inc0 | etcigration | 17:17 |
| sdake | etcgration | 17:17 |
| sdake | winning | 17:18 |
| inc0 | SamYaple, only thing I'm concerned about is that this is our first release and you don't get second "first impression" | 17:19 |
| inc0 | so please let's not rush things | 17:19 |
| SamYaple | non-default things don't impact first impressions | 17:19 |
| openstackgerrit | Jeff Peeler proposed openstack/kolla: WIP - tidy up syntax in all.yml https://review.openstack.org/229535 | 17:20 |
| SamYaple | but im not looking for change | 17:20 |
| SamYaple | we just have a way of being very consistent accross all branches | 17:20 |
| SamYaple | and we should use that advantage | 17:20 |
| inc0 | well, anyway, I'm seeing double, good sigh I should probably go off | 17:20 |
| sdake | this really only matters for liberty/mitaka, after that I think we will be right on top of the big tent in terms of delivery of ansible + containers | 17:21 |
| inc0 | have a good weekend guys, I'll be off for few days | 17:21 |
| sdake | so it may not be worth the risk for liberty/mitaka | 17:21 |
| sdake | take care inc0 | 17:21 |
| SamYaple | well lets not talk about things before they happen | 17:21 |
| SamYaple | night inc0 | 17:21 |
| *** inc0 has quit IRC | 17:21 | |
| SamYaple | but i am fairly confident there is no risk involved considering you don't even need to update your inventory file | 17:21 |
| SamYaple | ever comment out the swift or ceph hosts in an inventory file? try it its actually much cleaner | 17:22 |
| SamYaple | so a backport will be very clean. but well discuss later for sure! | 17:22 |
| jpeeler | SamYaple: thoughts? https://review.openstack.org/#/c/229535/1/ansible/group_vars/all.yml,cm | 17:23 |
| *** jmccarthy has quit IRC | 17:23 | |
| SamYaple | jpeeler: youre making dicts, its fine with me | 17:23 |
| jpeeler | if it's not clearer though, it's not worth doing | 17:24 |
| SamYaple | i see the advantage for organization | 17:24 |
| SamYaple | im not complaining :) | 17:24 |
| jpeeler | ok i'll go ahead and get it fixed up then | 17:25 |
| SamYaple | that is a breaking change though just fyi | 17:25 |
| SamYaple | that'll be harder to backport | 17:25 |
| SamYaple | the globals.yml will ahve to address things differently and then we expose having ot use dicts in globals.yml | 17:26 |
| jpeeler | oh, well maybe now's not the right time then | 17:26 |
| SamYaple | not saying no, but it is pretty late in the game to change this | 17:26 |
| SamYaple | other cores weigh in? | 17:26 |
| SamYaple | i guess the real question comes down to this: | 17:28 |
| SamYaple | so we want a 'flat' globals.yml without knowlege of yaml and dicts | 17:28 |
| SamYaple | or a structured one | 17:28 |
| jpeeler | i don't really want to go doing search/replace if we decide it's not worth the structure. but if you think it's worth doing, i'll go ahead and do it | 17:31 |
| SamYaple | jpeeler: im neutral. i like the change, but my question above stands. | 17:32 |
| SamYaple | check with other cores | 17:32 |
| jpeeler | sdake: do you have any opinion? | 17:32 |
| jpeeler | this was just something i came across while learning about ansible (which clearly is still not a lot) | 17:33 |
| sdake | jpeelr sorry was doing something else let me read scrollback | 17:34 |
| SamYaple | yea this is just dicts implemented in yaml | 17:34 |
| jpeeler | so it's not even an ansible thing, just yaml | 17:34 |
| SamYaple | jpeeler: did you know you can address dicts in ansible as {{ dict_name.key }} or {{ dict_name['key'] }} | 17:35 |
| SamYaple | jpeeler: correct | 17:35 |
| *** shakamunyi has quit IRC | 17:35 | |
| sdake | ok read scrollback | 17:35 |
| sdake | let me read patch | 17:35 |
| jpeeler | SamYaple: amazingly i did, but keep the tips of the day coming heh | 17:35 |
| SamYaple | :) | 17:35 |
| sdake | i like it | 17:35 |
| SamYaple | sdake: it will require a non-flat globals.yml | 17:36 |
| SamYaple | im neutral with that change. just fyi | 17:36 |
| sdake | SamYaple caan you expand on what that means | 17:36 |
| SamYaple | right now you do: | 17:36 |
| SamYaple | api_interface: 'eth0' | 17:36 |
| SamYaple | this would require - | 17:36 |
| *** banix has quit IRC | 17:36 | |
| SamYaple | netowrk: | 17:36 |
| SamYaple | api_interface: 'eth0' | 17:36 |
| SamYaple | in globals.yml | 17:37 |
| sdake | i understand | 17:37 |
| sdake | I like that, it groups things together | 17:37 |
| SamYaple | fair enough | 17:37 |
| SamYaple | i like the structure too | 17:37 |
| SamYaple | better now than later as well in my opinion | 17:37 |
| sdake | this is a PNR | 17:37 |
| SamYaple | so i can go either way, but if we do it, we need to do it now | 17:37 |
| *** achanda has joined #kolla | 17:37 | |
| SamYaple | PNR? | 17:37 |
| sdake | point of no return | 17:37 |
| SamYaple | yea | 17:38 |
| sdake | lets make sure everyone is on board jpeeler | 17:38 |
| sdake | please send an email to the mailing list | 17:38 |
| sdake | and explain the change | 17:38 |
| sdake | and see if anyone has any objections | 17:38 |
| sdake | beyond the obvious "why did you wait until the lsat possible minute" :) | 17:38 |
| jpeeler | i wish i had the change up during the meeting | 17:38 |
| SamYaple | yes | 17:39 |
| jpeeler | i waited because i thought this sort of thing could go in at the start of M | 17:39 |
| jpeeler | but forgot about backports and such | 17:39 |
| sdake | it basically makes the config more object oriented | 17:39 |
| SamYaple | yes | 17:39 |
| sdake | SamYaple from the context of ansible is there any downside? | 17:40 |
| SamYaple | nah. we already use a dict for the auth strings | 17:40 |
| SamYaple | works fine | 17:40 |
| sdake | in other words does it limit flexiblikty? | 17:40 |
| SamYaple | no but its a bit more confusing to override a variable | 17:40 |
| SamYaple | but anyone knowing they can do that should know how to do that | 17:40 |
| sdake | hwo do you do an override? | 17:41 |
| sdake | just specify x: | 17:41 |
| sdake | var: "xyz" ? | 17:41 |
| SamYaple | yes | 17:41 |
| sdake | jppeeler I'd like ithis in rc2 (2nd) if possible | 17:41 |
| sdake | unless someone in the cr team has a problem with it | 17:42 |
| jpeeler | so send out an email or go ahead and complete the review assuming people will be fine? | 17:42 |
| sdake | yes | 17:42 |
| SamYaple | jpeeler: both | 17:42 |
| sdake | unless you have something better to do :) | 17:42 |
| SamYaple | yea pbourke_ you around? | 17:42 |
| sdake | it is posible it could b ewasted effort | 17:42 |
| sdake | the only folks I think may find objection are oracle | 17:43 |
| SamYaple | hey guys. without maxing out my ceph cluster at all. I got 5,000 iops for 4k writes.... FROM INSIDE THE INSTANCE | 17:43 |
| SamYaple | i did that fio test on 8 instances at once | 17:43 |
| SamYaple | all doing 5,000 iops | 17:43 |
| sdake | that gear can do 50k iosp I think | 17:43 |
| SamYaple | i think the qemu driver is the bottlenexk | 17:44 |
| SamYaple | sdake: yes | 17:44 |
| sdake | yup memcpy | 17:44 |
| SamYaple | no its virtio-scsi driver | 17:44 |
| SamYaple | way better | 17:44 |
| SamYaple | but still not perfect | 17:44 |
| SamYaple | either way 900MB/s for seq large blocks | 17:44 |
| SamYaple | and 5,000 random 4k | 17:44 |
| SamYaple | iops | 17:44 |
| sdake | is the caching for ceph different to set up then using the disks as raw disks? | 17:46 |
| sdake | or does kolla figure it out | 17:46 |
| SamYaple | what do you mean? | 17:46 |
| sdake | SamYaple ^^ | 17:46 |
| sdake | i mean you have ceph which has 750s for caching | 17:46 |
| sdake | how doy ou set that up ? | 17:46 |
| sdake | does it require custom deploy steps? | 17:46 |
| SamYaple | ah. so you take your backing disks and ssds (what you want for cache) and bootstrap them the same way | 17:47 |
| SamYaple | parted /dev/nvme0n1 -s -- mklabel gpt mkpart KOLLA_CEPH_OSD_BOOTSTRAP 1 -1 | 17:47 |
| SamYaple | do that for all disks | 17:47 |
| SamYaple | then you have to do ceph things to setup the tiering | 17:47 |
| SamYaple | http://docs.ceph.com/docs/master/rados/operations/cache-tiering/ | 17:47 |
| sdake | is it possible to make automatic? | 17:47 |
| SamYaple | i wont no | 17:47 |
| sdake | like in mitaka | 17:47 |
| SamYaple | if you want to break out ceph into its own project, then yes | 17:48 |
| SamYaple | but it could easily balloon out paste all of the openstack code | 17:48 |
| SamYaple | ceph is complicated and storage is sensitive | 17:48 |
| sdake | is it something that can be documented instead of automatic? | 17:48 |
| SamYaple | of course | 17:48 |
| SamYaple | but in this case the ceph offical docs are far more useful | 17:49 |
| sdake | i do not want to break out ceph separately | 17:49 |
| SamYaple | right, so we need to be alot less flexible with the ceph 'automation' | 17:49 |
| SamYaple | im happy just getting the osds in the cluster | 17:49 |
| SamYaple | the maintianing performance is the operator all the way | 17:49 |
| SamYaple | sdake: I would be ok with a scripts/tools directory script/ansible that could do some of that automation | 17:51 |
| SamYaple | but again this could easily balloon out way past the openstack stuff | 17:51 |
| sdake | i was thinking something pretty inflexible, like specify the cache devices, specify the regular devices | 17:52 |
| sdake | and have kolla figure it out | 17:52 |
| sdake | like how you have the labels now | 17:52 |
| sdake | but with a second label for the cache devices | 17:52 |
| sdake | but I dont know the first thing about ceph or if that is feasible | 17:52 |
| SamYaple | yes i thought about that | 17:55 |
| SamYaple | its possible but ceph is complex | 17:55 |
| SamYaple | and im worried about a cluster with 200TB of data and breaking it.... | 17:55 |
| *** jtriley has joined #kolla | 17:57 | |
| sdake | well try to keep an open mind :) | 17:59 |
| sdake | possibly an optional feature for the future | 17:59 |
| SamYaple | yea no worries but its complex | 18:00 |
| sdake | I think it would be pertty sweet | 18:00 |
| SamYaple | because you have to redo the crush map | 18:00 |
| sdake | I dont even know what that means ;) | 18:00 |
| SamYaple | i know and thats my point | 18:00 |
| SamYaple | its do able. ill look into it | 18:00 |
| sdake | cool | 18:00 |
| sdake | would be a hot featur I think :) | 18:01 |
| *** shakamunyi has joined #kolla | 18:01 | |
| *** banix has joined #kolla | 18:01 | |
| sdake | something not flexible just "ya I want lazy-mans caching" | 18:01 |
| SamYaple | no substitue for a good ceph admin | 18:01 |
| SamYaple | more probe to breakage | 18:01 |
| sdake | if someone wants highly tuned storage they are oging to have ceph cats on staff and probably not run it in a container | 18:01 |
| *** loth has joined #kolla | 18:01 | |
| sdake | if they are a 100 person org, they are going to turn on poor mans storage setup | 18:01 |
| SamYaple | you say that but let me tell you running in a container actually _SOLVES_ some issues | 18:01 |
| sdake | rather not poor man lazy man | 18:02 |
| SamYaple | it gets rid of a nasty hack for looking up the osd id | 18:02 |
| SamYaple | double mounting | 18:02 |
| * SamYaple shakes a bit | 18:02 | |
| sdake | hey guys, inc0's vote was unanimous | 18:05 |
| sdake | inc0 welcome to the CR team! | 18:05 |
| sdake | oh I guess inc0 is not around | 18:05 |
| sdake | well anway maybe he will read the logs | 18:05 |
| Hui | could anyone point to me the HA design in Kolla? | 18:07 |
| sdake | I dont think there is any modern documentation on the HA design | 18:07 |
| sdake | our docs are really lagging hui | 18:07 |
| Hui | so to be HA, the current approach is to deploy multiple containers for a service. Is my understanding correct? | 18:08 |
| SamYaple | yes | 18:09 |
| Hui | Thanks Sam | 18:09 |
| sdake | samyaple I'm thinking for our demo at summit we should demo writing a file with the caching of ceph | 18:09 |
| sdake | to show how blindingly fast it is | 18:09 |
| SamYaple | sdake: as long as we can ssh to my gear :P | 18:09 |
| sdake | is it possible to access your cluster from the internets? | 18:09 |
| SamYaple | the whole world even! | 18:10 |
| Hui | If, say a container fails, there is no way to recover it now. Right? | 18:10 |
| SamYaple | I even have dns and icmp tunnels | 18:10 |
| SamYaple | Hui: the container will auto restart which *may* fix it | 18:10 |
| SamYaple | depends on failure | 18:10 |
| Hui | ok | 18:10 |
| sdake | ya i think ssh to your gear is the way to go, as long as we can setup a dns nat or something like that | 18:10 |
| sdake | you saw my slidedeck on tthe demo | 18:11 |
| sdake | that is easy to explain to folks | 18:11 |
| SamYaple | yes looks good | 18:11 |
| sdake | and it is understandable | 18:11 |
| SamYaple | so fyi we are doing cachemode=writeback for ceph | 18:11 |
| sdake | even though nobody si really going to run a netgear wrt router :) | 18:11 |
| sdake | right | 18:11 |
| SamYaple | this is safe for ceph (and only ceph) due to using librbd and not kernel pages | 18:12 |
| SamYaple | using that with anything else would be dangerous and live-migration could move stale-data | 18:12 |
| SamYaple | ceph fixed this in 2014 with librbd and qemu patches | 18:12 |
| SamYaple | in case anyone askes you :) | 18:12 |
| sdake | writeback is generall ysfe unless there is power loss | 18:13 |
| SamYaple | yea but its not actually safe with live-migration see above | 18:13 |
| sdake | in the 90s when I worked on raid stuff, we used writeback but we had battery backups on our raid cards | 18:13 |
| sdake | right | 18:13 |
| SamYaple | only with ceph can you actually issue a cache invalidate call | 18:13 |
| SamYaple | i still need to put in the dvr yes/no option in the configs for koll | 18:16 |
| SamYaple | a | 18:16 |
| *** openstackgerrit has quit IRC | 18:16 | |
| SamYaple | i do dvr here a bit when testing | 18:16 |
| *** openstackgerrit has joined #kolla | 18:16 | |
| SamYaple | it doesnt really matter for such a small test lab, but still | 18:16 |
| SamYaple | more people would try it if it were a yes/no option | 18:17 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 18:22 |
| SamYaple | ugh sdake ansible in kolla_ansible container is foobar | 18:42 |
| SamYaple | i guess some breakage upstream | 18:42 |
| SamYaple | locking down to a tag now | 18:42 |
| SamYaple | sdake: so there are more deps we can move to the openstack-common container | 18:43 |
| SamYaple | that should decrease size and improve build times | 18:44 |
| SamYaple | holy crap i need to sleep | 18:44 |
| *** dims_ has quit IRC | 18:46 | |
| *** openstackgerrit has quit IRC | 18:46 | |
| *** openstackgerrit has joined #kolla | 18:46 | |
| *** dims_ has joined #kolla | 18:46 | |
| sdake | samyaple groan | 18:46 |
| SamYaple | sdake: its cool. tested latest tag, all good | 18:47 |
| sdake | ok cool | 18:47 |
| sdake | what was the probllem then? | 18:47 |
| SamYaple | rc1 image upstream for ubuntu works | 18:47 |
| SamYaple | but the centos kolla_ansible container gonna be foobared | 18:47 |
| SamYaple | no idea | 18:47 |
| SamYaple | its stil la proble | 18:47 |
| sdake | i probalbybusted it when i was wirting code after being up for 30 hous | 18:48 |
| * sdake sighs | 18:48 | |
| sdake | i'll look into it today | 18:48 |
| SamYaple | sdake: no its upstream _ANSIBLE_ | 18:49 |
| sdake | oh i see | 18:49 |
| sdake | i was pretty sur eI teted that :) | 18:49 |
| sdake | which is why I was surprised when you said it | 18:49 |
| sdake | but now i get it | 18:49 |
| SamYaple | it literally just broke for me | 18:49 |
| SamYaple | im guessing latest commit or two | 18:49 |
| sdake | the build breaks ? | 18:50 |
| SamYaple | no build works, container breaks | 18:50 |
| sdake | ok | 18:50 |
| SamYaple | so no doing any commands in container | 18:50 |
| SamYaple | anyway ih ave patch testing now | 18:50 |
| *** achanda has quit IRC | 18:50 | |
| SamYaple | working thus far | 18:50 |
| *** achanda has joined #kolla | 18:53 | |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 18:58 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Lock down ansible version https://review.openstack.org/229579 | 18:58 |
| *** achanda has quit IRC | 19:02 | |
| *** jtriley_ has joined #kolla | 19:03 | |
| *** jtriley has quit IRC | 19:06 | |
| *** inc0 has joined #kolla | 19:08 | |
| inc0 | just come back to say thanks! I'm really excited;) | 19:09 |
| SamYaple | go for the gold! | 19:09 |
| sdake | inc0 grats bro! | 19:09 |
| inc0 | now the hard work starts:) | 19:10 |
| sdake | inc0 if you want a bit of coaching and have 5 minutes i can giveoyu the quick how to do the thing | 19:10 |
| sdake | do you have 5 minutes? | 19:10 |
| inc0 | sure | 19:10 |
| sdake | ok so basically one person comes alnog and +2s a review | 19:10 |
| sdake | if your the first person to a review, and you like the patch, go ahead and hit +2 | 19:11 |
| sdake | but do not change the workflow | 19:11 |
| sdake | if your the second person to a a review and it already has a +2, set +2 and workflow to +1 | 19:11 |
| sdake | do not ever approve a patch with only one review | 19:11 |
| sdake | no matter what the circiumstance | 19:11 |
| sdake | if you are part of the author chain you may not self +2 a patch | 19:11 |
| sdake | even if your only a coauthor | 19:12 |
| sdake | it requires an independent party to approve the patch | 19:12 |
| openstackgerrit | Sam Yaple proposed openstack/kolla: Common start.sh https://review.openstack.org/229371 | 19:12 |
| sdake | the one exception is if two core reviewers have already reviewed the patch | 19:12 |
| sdake | and its a rebase | 19:12 |
| sdake | and gerrit retains the other 2 +2s | 19:12 |
| SamYaple | sdake: https://review.openstack.org/#/c/229371/ | 19:12 |
| sdake | you may set +2/+a | 19:12 |
| SamYaple | plase test centos | 19:12 |
| SamYaple | ubutntu 100% | 19:12 |
| sdake | sam sure i'll pull it down now and reubild | 19:13 |
| sdake | hang tight | 19:13 |
| sdake | it takes about 45 mins | 19:13 |
| SamYaple | no prob im going to sleep | 19:13 |
| SamYaple | repatch if needed | 19:13 |
| sdake | enjoy | 19:13 |
| sdake | inc0 any questions? | 19:13 |
| inc0 | sdake, so far no, I've seen a lot of reviews so I think I get how it works;) | 19:14 |
| SamYaple | inc0: a good patch to flex your +2+1 :) https://review.openstack.org/#/c/229579/ | 19:14 |
| inc0 | I'll ask around if not | 19:14 |
| sdake | ya the only tricky one is the rebase | 19:14 |
| sdake | samyaple that is waiting on the gate | 19:14 |
| SamYaple | im asleeps now | 19:15 |
| sdake | word enjoy | 19:15 |
| inc0 | SamYaple, quick question...alpha? | 19:15 |
| inc0 | we pin to alpha? | 19:15 |
| SamYaple | inc0: its the only 2.0 tag avaialbe | 19:15 |
| sdake | that is all thee is | 19:15 |
| SamYaple | alpha1 and alpha2 | 19:15 |
| SamYaple | we need 2.0 | 19:16 |
| inc0 | for shade licence? I get it | 19:16 |
| SamYaple | i actually comment on this in the commit message | 19:16 |
| SamYaple | no nothing to do with shade | 19:16 |
| sdake | inc0 one other thing | 19:17 |
| sdake | i need to bring this up with al lthe crs | 19:17 |
| sdake | but i'd like people not to approve changes until the non-voting gates complete successfully | 19:17 |
| sdake | its annoying we can't just vote and move on | 19:18 |
| sdake | but that is the world we live in :) | 19:18 |
| inc0 | we need to think of good way of stable and reliable voting gates | 19:18 |
| sdake | i know how to do it | 19:19 |
| sdake | it involves getting infra to mirror the repos | 19:19 |
| sdake | going to tackle in mitaka | 19:19 |
| inc0 | I made first +2 ;) but not +w | 19:19 |
| sdake | nice | 19:19 |
| sdake | inc0 its complictaed by the fact that the ptl for infra changed | 19:20 |
| inc0 | new one will be less sympathetic? | 19:20 |
| inc0 | I can ask intel's CI guys to spare few machines for us | 19:20 |
| inc0 | if that's lack of hardware problem | 19:21 |
| *** jtriley_ has quit IRC | 19:32 | |
| sdake | its lack of people to implement it problem in the current infra | 19:32 |
| sdake | inc0 ^^ | 19:32 |
| sdake | new ptl not sure if sympathetic or not, but will have hands full with very busy project | 19:33 |
| sdake | so hard to get ear of ptl for this effort | 19:33 |
| *** jtriley has joined #kolla | 19:33 | |
| inc0 | I see | 19:33 |
| sdake | if you could find someone to magically implement it in the infra team, that would be fantastic :) | 19:34 |
| sdake | otherwise I'm going to hve to beg around | 19:34 |
| sdake | that is more important then having a functional gate imo | 19:34 |
| inc0 | or outside of infra, just accessible | 19:34 |
| inc0 | and dedicated for us | 19:34 |
| sdake | voting gate is critical | 19:34 |
| sdake | i'd like it to be in the infra | 19:34 |
| sdake | external ci can't vote | 19:34 |
| inc0 | no, I mean mirrors | 19:34 |
| sdake | oh right | 19:35 |
| sdake | ya doesn't matter where it is | 19:35 |
| sdake | but in infra woudl be better | 19:35 |
| sdake | it isn't that hard to mirror rpm repos | 19:35 |
| sdake | and deb repos | 19:35 |
| inc0 | yeah, and github is pretty stable | 19:35 |
| sdake | github doesn't do the mirroring | 19:35 |
| sdake | you have ot setup a http server | 19:35 |
| sdake | and run a cron job to mirror various rpm repos | 19:36 |
| sdake | the cron job runs a mirroring script | 19:36 |
| inc0 | well, that's something we'll talk about in Tokyo I guess | 19:36 |
| sdake | and we have to modify the build.py to add a specific set of mirrors | 19:36 |
| inc0 | I'll ask around if we can spare few machines from our community servers | 19:36 |
| sdake | the work is easy just need the mirrors running | 19:36 |
| sdake | doesn't even have ot be bare metal | 19:37 |
| inc0 | we have 3rd party CI in intel, maybe we could place mirrors there | 19:37 |
| sdake | could be virt | 19:37 |
| inc0 | anyway, I'm going off, cya and thanks again:) | 19:37 |
| sdake | enjoy :) | 19:37 |
| *** inc0 has quit IRC | 19:37 | |
| openstackgerrit | Jeff Peeler proposed openstack/kolla: Group variables by dictionary in all.yml https://review.openstack.org/229535 | 19:40 |
| sdake | jpeeler you sure cranked that out fast :) | 19:41 |
| jpeeler | you call that fast? | 19:41 |
| jpeeler | i still need to test it | 19:41 |
| sdake | looks groansihly long to review | 19:41 |
| sdake | set workflow so folks know you aren't done testing it | 19:41 |
| jpeeler | i did | 19:42 |
| sdake | cool | 19:42 |
| jpeeler | that was painful, so i really hope it goes in | 19:42 |
| sdake | it will be harder to review then it was to develop | 19:43 |
| sdake | you have the easy part of it :) | 19:43 |
| sdake | jpeeler please make a blueprint | 19:44 |
| jpeeler | i really hope nobody reviews it for two hours | 19:44 |
| sdake | that is how long it will take i suspect | 19:44 |
| sdake | i will review it for however long it takes | 19:44 |
| sdake | jpeeler hwere does kolla_internal_address come from? | 19:48 |
| sdake | why not kolla.internal_address? | 19:48 |
| jpeeler | that's from globals | 19:48 |
| sdake | i would expect globals to not have a flat structure? | 19:49 |
| jpeeler | i didn't touch that file | 19:49 |
| sdake | i think the intent is to touch that file | 19:49 |
| jpeeler | i'll need to test to see what that looks like | 19:51 |
| jpeeler | really didn't mean to spend all day on this | 19:52 |
| jpeeler | but i do think it's a good change... | 19:52 |
| jpeeler | actually i guess you're right | 19:53 |
| *** JohnN has quit IRC | 20:00 | |
| openstackgerrit | Jeff Peeler proposed openstack/kolla: Group variables by dictionary in all.yml https://review.openstack.org/229535 | 20:10 |
| *** shardy has quit IRC | 20:14 | |
| *** banix has quit IRC | 20:18 | |
| *** thumpba has joined #kolla | 20:25 | |
| *** tpot has joined #kolla | 20:52 | |
| *** itsuugo has joined #kolla | 20:53 | |
| *** thumpba has quit IRC | 21:01 | |
| *** thumpba has joined #kolla | 21:03 | |
| *** thumpba has quit IRC | 21:08 | |
| *** thumpba has joined #kolla | 21:09 | |
| *** jtriley has quit IRC | 21:20 | |
| *** dwalsh has quit IRC | 21:34 | |
| *** thumpba_ has joined #kolla | 21:40 | |
| *** thumpba has quit IRC | 21:43 | |
| *** thumpba_ has quit IRC | 21:46 | |
| *** thumpba has joined #kolla | 21:46 | |
| *** banix has joined #kolla | 21:55 | |
| *** sdake has quit IRC | 21:55 | |
| *** sdake has joined #kolla | 21:58 | |
| *** thumpba has quit IRC | 22:05 | |
| *** jruano has quit IRC | 22:10 | |
| *** diogogmt has joined #kolla | 22:16 | |
| *** jtriley has joined #kolla | 22:25 | |
| *** devlaps has quit IRC | 22:30 | |
| *** dims__ has joined #kolla | 22:30 | |
| *** dims_ has quit IRC | 22:33 | |
| *** jtriley has quit IRC | 22:55 | |
| *** alisonh has quit IRC | 23:03 | |
| *** alisonh has joined #kolla | 23:03 | |
| *** jtriley has joined #kolla | 23:04 | |
| *** jruano has joined #kolla | 23:08 | |
| *** jtriley has quit IRC | 23:25 | |
| *** diogogmt has quit IRC | 23:36 | |
| *** blahRus has quit IRC | 23:47 | |
| *** diogogmt has joined #kolla | 23:48 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!