Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel-dev, 2015-07-17

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

All times shown according to UTC.

Time Nick Message
00:00 mattymo joined #fuel-dev
00:02 mattymo joined #fuel-dev
00:04 mattymo joined #fuel-dev
00:06 mattymo joined #fuel-dev
00:09 mattymo joined #fuel-dev
00:11 mattymo joined #fuel-dev
00:13 mattymo joined #fuel-dev
00:15 mattymo joined #fuel-dev
00:17 mattymo joined #fuel-dev
00:19 mattymo joined #fuel-dev
00:21 mattymo joined #fuel-dev
00:23 mattymo joined #fuel-dev
00:25 mattymo joined #fuel-dev
00:27 mattymo joined #fuel-dev
00:29 mattymo joined #fuel-dev
00:31 mattymo joined #fuel-dev
00:33 mattymo joined #fuel-dev
00:35 mattymo joined #fuel-dev
00:37 mattymo joined #fuel-dev
00:39 mattymo joined #fuel-dev
00:39 tzn joined #fuel-dev
00:41 mattymo joined #fuel-dev
00:43 mattymo joined #fuel-dev
00:45 mattymo joined #fuel-dev
00:47 mattymo joined #fuel-dev
00:49 mattymo joined #fuel-dev
00:51 mattymo joined #fuel-dev
00:53 mattymo joined #fuel-dev
00:55 mattymo joined #fuel-dev
00:57 mattymo joined #fuel-dev
00:59 mattymo joined #fuel-dev
01:01 mattymo joined #fuel-dev
01:03 mattymo joined #fuel-dev
01:05 mattymo joined #fuel-dev
01:07 mattymo joined #fuel-dev
01:09 mattymo joined #fuel-dev
01:12 mattymo joined #fuel-dev
01:14 mattymo joined #fuel-dev
01:16 mattymo joined #fuel-dev
01:18 mattymo joined #fuel-dev
01:20 mattymo joined #fuel-dev
01:22 mattymo joined #fuel-dev
01:24 mattymo joined #fuel-dev
01:26 mattymo joined #fuel-dev
01:28 mattymo joined #fuel-dev
01:30 mattymo joined #fuel-dev
01:32 mattymo joined #fuel-dev
01:34 mattymo joined #fuel-dev
01:36 mattymo joined #fuel-dev
01:38 mattymo joined #fuel-dev
01:40 mattymo joined #fuel-dev
01:41 tzn joined #fuel-dev
01:42 mattymo joined #fuel-dev
01:45 mattymo joined #fuel-dev
01:48 ilbot3 joined #fuel-dev
01:48 Topic for #fuel-dev is now #fuel-dev Fuel dev docs http://docs.mirantis.com/fuel-dev/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel-dev/ | gerrit traffic @ #fuel-tracker
01:48 mattymo joined #fuel-dev
02:04 youellet_ joined #fuel-dev
03:32 dims_ joined #fuel-dev
03:42 tzn joined #fuel-dev
04:26 neophy joined #fuel-dev
05:05 trociny joined #fuel-dev
05:10 trociny ogelbukh, ping
05:10 ogelbukh trociny: hi
05:12 trociny ogelbukh, so do you deploy on old release, then upgrade the primary controller, restart the monitor and then deploy additional controller?
05:13 ogelbukh trociny: that's right
05:14 ogelbukh trociny: primary controller tries to update ceph config with a new mon and fails
05:15 ogelbukh trociny: because /etc/init.d/ceph can't handle process started by upstart
05:15 trociny I faced issues with the first restart after the upgrade. Something like you are observing -- the old process is not seen by new init scripts. So I had to kill it manually, and the next restarts work
05:16 trociny actually, using upstart is correct here, and using /etc/init.d/ceph is wrong
05:16 ogelbukh yes, if I, for example, stop it with 'stop ceph'
05:16 trociny I am going to fix fuel for this
05:16 ogelbukh and run with /etc/init.d.ceph start mon
05:16 ogelbukh then it can be handled
05:16 ogelbukh good to know
05:17 ogelbukh the only thing that confuses me is how come that it is started by upstart?
05:17 ogelbukh I mean, initially
05:17 trociny it is started by udev scripits
05:17 ogelbukh it is the same puppet module that supposedly starts it
05:18 ogelbukh hm, so it's not puppet?
05:29 ilbot3 joined #fuel-dev
05:29 Topic for #fuel-dev is now #fuel-dev Fuel dev docs http://docs.mirantis.com/fuel-dev/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel-dev/ | gerrit traffic @ #fuel-tracker
05:43 tzn joined #fuel-dev
05:48 alex_bh joined #fuel-dev
05:53 asheplyakov joined #fuel-dev
06:00 dims_ joined #fuel-dev
06:19 stamak joined #fuel-dev
06:37 mkwiek joined #fuel-dev
06:54 ub joined #fuel-dev
06:54 alex_bh joined #fuel-dev
07:10 dkusidlo joined #fuel-dev
07:25 aliemieshko_ joined #fuel-dev
07:25 dancn joined #fuel-dev
07:46 dklepikov joined #fuel-dev
07:47 tzn joined #fuel-dev
07:49 _tzn joined #fuel-dev
07:49 dims_ joined #fuel-dev
08:04 amnk joined #fuel-dev
08:27 salmon_ joined #fuel-dev
08:27 stamak joined #fuel-dev
08:30 evgeniyl___ joined #fuel-dev
08:40 romcheg joined #fuel-dev
08:40 monester joined #fuel-dev
08:41 diurchenko joined #fuel-dev
08:49 dkusidlo joined #fuel-dev
08:49 ashtokolov joined #fuel-dev
08:53 alex_didenko left #fuel-dev
08:59 HeOS joined #fuel-dev
09:01 e0ne joined #fuel-dev
09:02 dkusidlo joined #fuel-dev
09:07 romcheg joined #fuel-dev
09:11 DaveJ__ joined #fuel-dev
09:11 barthalion joined #fuel-dev
09:15 dims_ joined #fuel-dev
09:20 dkusidlo joined #fuel-dev
09:24 azemlyanov joined #fuel-dev
09:25 barthalion joined #fuel-dev
09:26 azemlyanov_ joined #fuel-dev
09:44 romcheg joined #fuel-dev
10:01 ashtokolov joined #fuel-dev
10:05 dims_ joined #fuel-dev
10:11 teran joined #fuel-dev
10:17 teran joined #fuel-dev
10:21 barthalion joined #fuel-dev
10:27 barthalion joined #fuel-dev
10:43 teran joined #fuel-dev
11:13 teran joined #fuel-dev
11:14 teran joined #fuel-dev
11:21 teran joined #fuel-dev
11:23 teran joined #fuel-dev
11:26 ashtokolov joined #fuel-dev
12:05 skolekonov joined #fuel-dev
12:24 dims_ joined #fuel-dev
12:31 prmtl joined #fuel-dev
12:35 dkusidlo joined #fuel-dev
12:36 teran_ joined #fuel-dev
12:36 ashtokolov joined #fuel-dev
12:38 tzn joined #fuel-dev
12:50 fuel-slackbot joined #fuel-dev
12:57 fuel-slackbot joined #fuel-dev
13:45 dims_ joined #fuel-dev
13:48 ashtokolov joined #fuel-dev
13:50 alex_didenko joined #fuel-dev
13:58 teran joined #fuel-dev
14:02 ashtokolov joined #fuel-dev
14:26 claflico joined #fuel-dev
14:35 ashtokolov joined #fuel-dev
14:36 skolekonov joined #fuel-dev
14:42 ashtokolov joined #fuel-dev
14:42 rmoe joined #fuel-dev
14:48 romcheg joined #fuel-dev
14:57 samuelBartel pasquier-s_, swann  hi
15:01 jaypipes joined #fuel-dev
15:03 ashtokolov joined #fuel-dev
16:10 ashtokolov joined #fuel-dev
16:20 angdraug joined #fuel-dev
16:21 neophy joined #fuel-dev
16:26 ashtokolov joined #fuel-dev
16:49 akasatkin left #fuel-dev
16:54 jobewan joined #fuel-dev
17:13 tzn joined #fuel-dev
17:52 romcheg joined #fuel-dev
17:52 ashtokolov joined #fuel-dev
17:55 stamak joined #fuel-dev
17:56 teran joined #fuel-dev
18:10 ashtokolov joined #fuel-dev
18:16 stamak joined #fuel-dev
18:21 xarses joined #fuel-dev
18:26 neophy joined #fuel-dev
18:29 e0ne joined #fuel-dev
18:44 stamak joined #fuel-dev
18:59 monester joined #fuel-dev
19:18 dims_ joined #fuel-dev
19:25 tzn joined #fuel-dev
19:30 dims__ joined #fuel-dev
19:41 blockedpipes xarses, xenolog: a moment to answer a question about https://review.openstack.org/#/c/178247/15/nailgun/nailgun/fixtures/openstack.yaml?
19:43 blockedpipes mwhahaha: you too, if you have a moment... ^^
19:44 mwhahaha i can guess with the best of them
19:45 blockedpipes mwhahaha: heh :) ok. so, this patch is adding support for VxLAN segmentation. And I'm wondering why line 391 is necessary there, since I assumed vlan_start would be specifically for the vlan segmentation type?
19:45 tzn joined #fuel-dev
19:47 mwhahaha I'd have to take some time to look at the code, but my guess would be it's related to validation/poor naming choice made previously
19:47 mwhahaha if you look at the GRE page it says vlan id
19:48 mwhahaha so it might be a similar thing
19:48 izinovik joined #fuel-dev
19:48 mwhahaha and i think in the UI we might be leveraging the same bit to generate those ids which was previously used for the vlan number generation
19:48 meow-nofer joined #fuel-dev
19:49 kat_pimenova joined #fuel-dev
19:49 vkramskikh joined #fuel-dev
19:50 xenolog joined #fuel-dev
19:50 nurla joined #fuel-dev
19:51 aglarendil joined #fuel-dev
19:51 * xarses looks
19:51 mrasskazov joined #fuel-dev
19:52 mihgen joined #fuel-dev
19:52 blockedpipes mwhahaha: ok, thx for the info
19:52 bdudko joined #fuel-dev
19:52 MorAle joined #fuel-dev
19:53 xarses blockedpipes: 391 is there because it doesn't differentiate between the two overlay providers at that point
19:53 sbog joined #fuel-dev
19:53 blockedpipes xarses: but what does it mean for VxLAN segmentation?
19:54 xarses nothing, it should be ignored
19:54 xarses the overlay type determines which is shown to the ui
19:54 xarses and which is used in the serializer
19:54 blockedpipes xarses: but that line is still necessary?
19:54 apalkina joined #fuel-dev
19:55 mwhahaha i think that's the validation part, we provide that for all of the sections but it's only used for some cases
19:55 xarses this yes, if you go up to 335, you can see that the entite section is for neutron
19:55 evg joined #fuel-dev
19:55 xarses the fixture doesn't care which overlay it is at this point
19:56 MiroslavAnashkin joined #fuel-dev
19:57 akislitsky joined #fuel-dev
19:57 xarses actually, we are outside the network section entirely
19:57 xarses erm
19:58 * xarses loads it in a text editor to read the indentation correctly
19:58 rvyalov joined #fuel-dev
19:59 xarses ok, yep, its part of neutron, just not networks
20:00 blockedpipes xarses: so is line 391 necessary then?
20:00 xarses yes
20:00 xarses it will be ignored in gre/tun
20:01 blockedpipes xarses: ok
20:01 * xarses looks for the place that ignores it
20:03 xarses blockedpipes: i realize now I'm looking at the wrong side
20:03 xarses you are asking about vlan_start
20:03 blockedpipes right.
20:03 xarses that's the vlan tag that will be set on the underlay interface
20:04 xarses (i thought you where asking about config: vlan_range)
20:05 xarses it will be used to configure the underlay interface
20:05 xarses all of them but public are on by default to help with POC/CI environments
20:05 blockedpipes xarses: I see... not exactly clear :)
20:05 xarses bad attribute name, I agree
20:06 xarses IIRC it's a hold over from when vlan_range was part of the network attribute
20:06 blockedpipes k
20:07 xarses it's the least of the annoying issues in tis section thought
20:07 xarses s/thought/though
20:08 blockedpipes reading this puppet-librarian patch from mwhahaha reminds me just how much I hate the Ruby language and especially the Ruby packaging world... :(
20:09 xarses explains the constipated nick today
20:09 blockedpipes xarses: heh, no that's just because my work queue is beyond full. :)
20:09 * mwhahaha gives ruby a dirty look
20:10 mwhahaha i'm also not a fan either
20:10 teran joined #fuel-dev
20:11 salmon_ joined #fuel-dev
20:12 xarses I'm not sure many are
20:24 HeOS joined #fuel-dev
20:28 mwhahaha VICTORY
20:28 mihgen mwhahaha: ?
20:29 angdraug joined #fuel-dev
20:29 mwhahaha finally figured out how to get librarian-puppet wedged into the fuel-library spec and have it work with our build jobs
20:29 mihgen :)
20:29 mihgen awesome!
20:29 mwhahaha only took like 12 tries
20:30 mihgen excellent, let’s see if we can sustain it!
20:30 angdraug for something that big, 12 tries is totally worth it :D
20:31 mwhahaha it'll total break when the fuel-library is built using something other than centos
20:31 mwhahaha but it might be an easier fix since we probably won't be on ruby 1.8.7
20:33 mihgen what is the hard dependency on centos there?
20:34 mwhahaha fuel-master is centos
20:34 mwhahaha so fuel-library has a spec file
20:34 mwhahaha and the build would now require ruby21 (centos package)
20:35 mwhahaha not the install, we're just building on centos so the spec assumes centos packages. it'll take some minor effort to port which build process change like that will have
20:50 blockedpipes mwhahaha, mihgen, xarses: https://review.openstack.org/#/c/178247/ could use a review if you get a chance. thx in advance.
20:51 xarses I added rmoe to it too
20:55 blockedpipes cheers
21:00 dims_ joined #fuel-dev
21:00 ashtokolov joined #fuel-dev
21:47 angdraug I'm still convinced that keeping ruby 1.8.7 in our centos repos is a cargo cult
21:49 mwhahaha well it's what ships default :/ so it's more of not having a clean way to upgrade it in centos.  I mean they do have software collections but it's just another python2.6
21:51 mwhahaha there is a ruby 1.9.3 in the SCL that would be a nice improvement at least over 1.8.7.
21:52 angdraug we already have ruby21 packages for centos, why do we need any other ruby versions?
21:53 mwhahaha only if we're choosing not to package it all
21:53 mwhahaha i mean to use something already provided
21:53 angdraug we don't have that option, we need ruby21 otherwise we wouldn't have had to package it
21:57 mwhahaha I don't think we explicitly use the 1.8.7 package for anything
21:57 angdraug exactly
21:58 angdraug what I'm going for is that we should just replace 1.8.7 with 2.1 and stop worrying about alternative ruby versions
21:58 angdraug btw have you seen Igor's email on [Fuel][Fuel-library] Using librarian-puppet to manage upstream fuel-library modules @openstack-dev?
21:59 mwhahaha i saw it, not sure what he was refering to
21:59 angdraug he was referring to using r10k to download upstream modules during ISO build
21:59 angdraug I agree that from build team's perspective that would be unacceptable
22:00 mwhahaha we wont' need to during iso build
22:00 angdraug so my question to you is, can I assume that the librarian solution you've put together doesn't introduce downloading anything from external sources into the build process?
22:00 angdraug good :D
22:00 mwhahaha the way i've wedged this in there, i've put it in as part of the rpm build
22:00 mwhahaha so you only need internets for the fuel-library rpm build
22:00 angdraug well, right now we're still building fuel-library rpm during iso build
22:01 angdraug and even if we get rid of that, i'd still prefer to download from a local clone of upstream and not from github
22:01 bookwar fetching something from internet at rpm build stage doesn't look good too
22:01 angdraug yup
22:02 mwhahaha yea i know, but it's how much do you want to do to package the world
22:03 mwhahaha i agree that we need a mirror/proxy to reduce the internet dep on the build process
22:03 mwhahaha but what i forsee is that puppet modules aren't rpm'd so we would have to take the time to build that out
22:04 mwhahaha and because we aren't installing to /etc/puppet/modules, how we use the modules (their location) will need to either change from an architectual standpoint or we need an alterantive solution
22:04 mwhahaha anyway i better go before my wife yells at me as its friday afternoon
22:04 * mwhahaha wanders off
22:07 angdraug we don't need them rpm'd we just need them cloned
23:53 xarses joined #fuel-dev

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