Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel-dev, 2016-02-29

| Channels | #fuel-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:21 krobzaur_ joined #fuel-dev
02:48 ilbot3 joined #fuel-dev
02:48 Topic for #fuel-dev is now Fuel Development http://wiki.openstack.org/wiki/Fuel | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel-dev/ | gerrit traffic @ #fuel-tracker
05:06 javeriak joined #fuel-dev
05:12 javeriak_ joined #fuel-dev
06:15 javeriak joined #fuel-dev
06:28 _nadya_ joined #fuel-dev
06:38 fzhadaev joined #fuel-dev
06:49 igajsin joined #fuel-dev
06:50 igajsin left #fuel-dev
07:00 dpuchkin joined #fuel-dev
07:13 _nadya_ joined #fuel-dev
07:17 dshulyak joined #fuel-dev
07:48 kozhukal` joined #fuel-dev
07:55 amnk joined #fuel-dev
07:58 kozhukal` joined #fuel-dev
08:05 dkaigarodsev joined #fuel-dev
08:05 pasquier-s joined #fuel-dev
08:05 mattymo joined #fuel-dev
08:06 ddepaoli joined #fuel-dev
08:15 [HeOS] joined #fuel-dev
08:33 dguryanov|2 joined #fuel-dev
08:35 dguryanov|2 Hello, fuel-astute cores, could you, please, check https://review.openstack.org/#/c/279436/ ?
08:38 zubchick joined #fuel-dev
08:46 ekosareva joined #fuel-dev
08:58 lezbar joined #fuel-dev
08:59 javeriak joined #fuel-dev
09:07 javeriak joined #fuel-dev
09:13 salmon_ joined #fuel-dev
09:14 zubchick_ joined #fuel-dev
09:23 IvanKliuk__ joined #fuel-dev
09:24 MorAle_ joined #fuel-dev
09:26 zinovik joined #fuel-dev
09:29 amnk joined #fuel-dev
09:43 e0ne joined #fuel-dev
10:03 kaliya joined #fuel-dev
10:08 aglarendil dguryanov|2: this fix may be not needed if you use task-based deployment. we now wrap shell tasks into puppet execs which handle almost all that you need along with an ability to see the logs in puppet logs
10:12 dguryanov|2 aglarendil: Milestone in bug is Fuel for OpenStack 9.0, if it will be done before, then OK
10:13 aglarendil we already have task-based deployment enabled for 9.0 by default with an ability to fall back to granular deployment with old tasks version
10:14 dguryanov|2 I see, so can we close https://bugs.launchpad.net/fuel/+bug/1435603 ?
10:15 aglarendil I think, this commit could be useful for non-9.0 and for legacy deployments in 9.0 still using granular orchestration. So no, we should not close it, but the bug severity could be decreased, I think
10:15 amnk joined #fuel-dev
10:17 amnk joined #fuel-dev
10:27 rvyalov_ joined #fuel-dev
10:28 enikher joined #fuel-dev
10:32 bhaskarduvvuri joined #fuel-dev
10:34 IvanBerezovskiy joined #fuel-dev
10:43 bhaskarduvvuri If we need to add a new component debian packaging to http://mirror.fuel-infra.org/mos-repos/ubuntu, then what should be the process for the same
10:46 zerda joined #fuel-dev
10:51 enikher could someone point me to the code where restrictions of a fuel-plugin are read and interpreted?
11:12 xek joined #fuel-dev
11:15 ekosareva joined #fuel-dev
11:29 dims joined #fuel-dev
11:32 dims joined #fuel-dev
11:48 amnk joined #fuel-dev
12:00 skolekonov_ joined #fuel-dev
12:05 openstackstatus NOTICE: Infra currently has a long backlog. Please be patient and where possible avoid rechecks while it catches up.
12:12 dshulyak joined #fuel-dev
12:19 zubchick joined #fuel-dev
12:39 amnk joined #fuel-dev
12:40 amnk joined #fuel-dev
12:50 zubchick joined #fuel-dev
12:58 dklenov joined #fuel-dev
13:06 zubchick joined #fuel-dev
13:08 zubchick joined #fuel-dev
13:19 zubchick joined #fuel-dev
13:31 mgariepy joined #fuel-dev
13:33 bhaskarduvvuri joined #fuel-dev
14:06 bhaskarduvvuri joined #fuel-dev
14:20 krobzaur_ joined #fuel-dev
14:43 zubchick joined #fuel-dev
15:36 xarses joined #fuel-dev
15:55 javeriak joined #fuel-dev
15:57 angdraug joined #fuel-dev
16:28 amnk joined #fuel-dev
16:37 gariveradlt joined #fuel-dev
16:46 asilenkov_ joined #fuel-dev
16:46 dnikishov joined #fuel-dev
16:50 meow-nofer_ joined #fuel-dev
16:53 igajsin1 joined #fuel-dev
17:01 dshulyak joined #fuel-dev
17:23 mwhahaha hey folks, can I get some reviews on https://review.openstack.org/#/c/273737/ and https://review.openstack.org/#/c/273740/ they add functionaility to be able to report on idempotency
17:55 zigo mwhahaha: Hi there!
17:55 mwhahaha hey
17:56 zigo mwhahaha: I was wondering if you knew a bit about Cobbler's power management setup with Virtualbox.
17:57 zigo mwhahaha: My Debian port of the master node goes up to the reboot of slave nodes, and then it fails...
17:57 mwhahaha fails how so?
17:59 mwhahaha I'm not sure where the power management part comes in since we've switched to image based deploys several releases ago we don't use cobbler as part of the os provisioning process anymore. or are you just doing something else with cobbler?
18:02 zigo Sorry, the link to my IRC bouncer is super slow with packet loss... Looking forward next week being back with a decent internet connectivity, you quickly get tired of the Chinese LAN... :/
18:02 zigo Error
18:02 zigo Provision has failed. <class 'cobbler.cexceptions.CX'>:'power management type must be one of: ack_manual,alom,apc,apc_snmp,baytech,bladecenter,brocade,bullpap,cisco_mds,cisco_ucs,cpint,drac,drac5,eaton_snmp,egenera,eps,ibmblade,ifmib,ilo,ilo_mp,intelmodular,ipmilan,ldom,lpar,mcdata,na,nss_wrapper,rackswitch,rhevm,rsa,rsb,sanbox2,scsi,virsh,vixel,vmware,vmware_helper,vmware_soap,wti,xcat,xenapi,zvm'
18:02 zigo mwhahaha: That's what I get.
18:03 mwhahaha interesting
18:03 zigo mwhahaha: If I'm not wrong: power management comes into play once the IBP is done on the master node, and slave got to reboot.
18:03 mwhahaha i assume it's when we update cobbler to not provide the bootstrap image anymore and we're probably providing a bad value
18:04 zigo The message I get above is displayed directly in Fuel UI.
18:04 mwhahaha yea astute messages end up there
18:04 zigo I've checked /etc/cobbler/settings in a Fuel 9.0 CentOS VM, and didn't see anything else but the default value.
18:05 zigo So I wonder how cobbler is configured...
18:06 zigo BTW, it was fun to discover that I couldn't use another name but "ubuntu_bootstrap" for the bootstrap profile in cobbler! :)
18:06 zigo Though I could customize the name for the Jessie profile.
18:07 zigo mwhahaha: So it really is astute that's trying to reboot the discovery bootstrap machines?
18:07 zigo Probably I should investigate there?
18:07 mwhahaha it may not even be rebooting it
18:08 mwhahaha it might be occuring when we update cobbler
18:09 zigo mwhahaha: What do you mean "update cobbler" ?
18:09 mwhahaha updating the system in cobbler
18:10 mwhahaha so we us cobbler to distribute the bootstrap image, after we provision we update the system in cobbler to stop providing the bootstrap image so it boots from local
18:10 zigo mwhahaha: As in "cobbler system <something>" ?
18:10 mwhahaha we use the api but yea those things
18:10 mwhahaha i've not spent a ton of time in the astute code around this so i'd have to dig a bit to see what actually happens
18:11 mwhahaha i wonder if the cobbler version in debian is different enough from the version we use with centos such that we're not providing a required value or something
18:11 zigo mwhahaha: I happen to be the package maintainer for cobbler in Debian, so I can do what I want with it! :)
18:11 zigo It was a bitch package to write, btw.
18:12 zigo I'm running version 2.6.6.
18:12 mwhahaha we use 2.6.9
18:12 mwhahaha on centos
18:18 zigo I can upgrade, though I'm convinced it wont help.
18:18 zigo I noticed that it looks like we're using fence_ssh which is in fuel-library.
18:18 zigo Could it be that we're trying to use the ssh type of power management, but by default cobbler doesn't know it?
18:19 mwhahaha yea that's probably it
18:19 mwhahaha i just checked, we're setting to ssh
18:19 zigo mwhahaha: Sorry, I got disconnected from my bouncer AGAIN ... :(
18:19 zigo mwhahaha: Could you get the full command line / API call so I could reproduce it?
18:19 zigo (and fix cobbler...)
18:20 zigo Anyway, thanks a lot, that's already a lot of help.
18:20 mwhahaha like i said we use the api but let me see if i can find the equiavelent cobbler commant
18:20 zigo You gave me a very nice clue so I understood it was in astute...
18:20 zigo Cool ! :)
18:21 mwhahaha cobbler system edit --power-type=ssh --name=<node name>
18:21 zigo mwhahaha: What would I put in --name?
18:21 mwhahaha the node name
18:21 mwhahaha from cobbler system list
18:21 zigo mwhahaha: Yeah, but what? :)
18:21 zigo "1" ?
18:21 mwhahaha like node-1
18:21 zigo mwhahaha: As in, returned by "fuel node" ?
18:21 mwhahaha no use `cobbler system list`
18:22 mwhahaha http://paste.openstack.org/show/488615/
18:22 mwhahaha this is the list of cobbler power types from our version of cobbler, http://paste.openstack.org/show/488616/ (cobbler system edit --help)
18:22 zigo mwhahaha: It returns nothing for me... :(
18:23 mwhahaha probably because it's failing on the create
18:23 zigo (I mean, the cobbler system list)
18:23 mwhahaha because we probably populate the power type
18:23 mwhahaha so you could just create a dummy entry via cobbler system add
18:24 mwhahaha http://paste.openstack.org/show/488617/
18:24 mwhahaha there's a system report from a dev instance
18:25 _nadya_ joined #fuel-dev
18:47 zigo mwhahaha: I simply needed to add the /usr/sbin/fence_ssh binary, it seems. So I passed this stage. Now I get: Error
18:47 zigo Provision has failed. Too many nodes failed to provision"
18:48 mwhahaha oh how i hate our error messaging
18:49 zigo Yeah.
18:49 mwhahaha check the astute log to see if there any actual useful errors
18:51 e0ne joined #fuel-dev
18:54 zigo mwhahaha: What's happening: Unexpected error while running command.\nCommand: udevadm settle --quiet\nExit code: -6\nStdout: ''\nStderr: \"Code should not be reached 'Unknown argument' at ../src/udev/udevadm-settle.c:87, function adm_settle().
18:54 zigo I got the same error when I try to do that command on the slaves.
18:55 mwhahaha well that'd be a problem
18:56 zigo mwhahaha: It looks like --quiet isn't a valid argument, that's easy to patch ! :)
18:56 amnk joined #fuel-dev
18:57 mwhahaha interesting, it's valid on ubuntu
18:58 zigo mwhahaha: It looks like it's in fuel-agent... Patching it out! :)
18:59 * mwhahaha shrugs
19:01 obcecado joined #fuel-dev
19:11 dims_ joined #fuel-dev
19:13 angdraug joined #fuel-dev
19:21 _nadya_ joined #fuel-dev
19:28 javeriak_ joined #fuel-dev
19:41 zigo mwhahaha: Now, I get: "Unexpected error\nnone of the templates given were found: cloud_config_fuel_9.0_ubuntu.jinja2, cloud_config_debian_jessie_x86_64.jinja2, cloud_config_debian.jinja2, cloud_config_debian_jessie_x86_64.jinja2, cloud_config.jinja2\n",
19:41 zigo WTF is this? :/
19:42 zigo cloud-init-templates ...
19:42 zigo What are they for?
19:42 mwhahaha when the node boots we run cloud-init to put down some basic configuration
19:42 mwhahaha it's part of the provisioning the process
19:43 zigo mwhahaha: Ok, but when / how should I get it installed? As part of fuel-agent ?
19:43 mwhahaha sec
19:43 angdraug FYI https://bugs.launchpad.net/fuel/+bug/1551402
19:43 mwhahaha https://docs.mirantis.com/fuel/fuel-master/reference-architecture.html#image-based-provisioning
19:44 zigo Note: I'm also (one of) the maintainer(s) for cloud-init ! :)
19:44 angdraug zigo: good for you. now stop swearing please :p
19:44 zigo angdraug: Sorry! :)
19:44 mwhahaha let me go find those templates
19:44 zigo angdraug: As you see, I've done huge progress... :P
19:44 mwhahaha i thought they were part of nailgun but i can never remember where they are
19:45 mwhahaha should be part of fuel-agent
19:46 zigo angdraug: I'm well passed the IBP stage, now I'm solving issues one by one, it's just a mater of time and knowing fuel (which knowledge I seriously lack...).
19:46 zigo mwhahaha: Yes, they are.
19:46 zigo mwhahaha: Though I wonder where to get them installed.
19:46 mwhahaha https://github.com/openstack/fuel-agent/tree/master/cloud-init-templates
19:46 zigo mwhahaha: There's a cloud-init-templates in fuel-agent.
19:46 zigo Yeah, that's it! :)
19:46 zigo mwhahaha: But when do I put them in the fuel-agent package?
19:47 mwhahaha they need to be on the master
19:47 mwhahaha they are used as part of building the image to be applied which occurs on the master
19:47 zigo mwhahaha: I mean, in what path!
19:47 mwhahaha oh /usr/share/fuel-agent/cloud-init-templates/cloud_config_ubuntu.jinja2
19:47 zigo mwhahaha: In /etc/fuel-agent/cloud-init-templates ?
19:47 zigo Oh, ok.
19:47 mwhahaha yea
19:47 mwhahaha er no usr/share
19:47 zigo Got it.
19:48 zigo Doing that.
19:48 zigo :)
19:49 mwhahaha https://github.com/openstack/fuel-agent/blob/master/specs/fuel-agent.spec#L112
20:00 zigo Gosh, it looks like Fuel already partitioned the slave nodes! :)
20:00 * zigo tries a new deployment with the fuel-agent patch for cloud-init.
20:01 krobzaur joined #fuel-dev
20:12 dims joined #fuel-dev
20:30 _nadya_ joined #fuel-dev
20:59 angdraug joined #fuel-dev
21:23 dims joined #fuel-dev
21:44 dims joined #fuel-dev
21:52 dshulyak joined #fuel-dev
23:14 dims joined #fuel-dev
23:19 dshulyak joined #fuel-dev
23:58 salmon_ joined #fuel-dev

| Channels | #fuel-dev index | Today | | Search | Google Search | Plain-Text | summary