Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-07-13

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

All times shown according to UTC.

Time Nick Message
00:38 code-R joined #fuel
00:39 elopez_ joined #fuel
00:40 code-R_ joined #fuel
00:56 Julien-zte joined #fuel
01:00 raunak joined #fuel
01:03 code-R joined #fuel
01:04 code-R_ joined #fuel
01:14 xarses joined #fuel
01:28 elopez joined #fuel
01:41 elopez_ joined #fuel
01:47 elopez__ joined #fuel
02:26 raunak joined #fuel
02:35 elopez joined #fuel
02:41 elopez_ joined #fuel
02:47 elopez_ joined #fuel
02:48 Sesso joined #fuel
02:54 elopez_ joined #fuel
02:57 elopez__ joined #fuel
03:19 elopez_ joined #fuel
03:21 elopez_ joined #fuel
03:29 elopez__ joined #fuel
03:34 elopez_ joined #fuel
03:40 elopez_ joined #fuel
04:20 raunak joined #fuel
04:31 javeriak joined #fuel
05:09 elopez joined #fuel
05:15 elo joined #fuel
05:15 elopez_ joined #fuel
05:19 maestropandy joined #fuel
06:03 Ravikiran_K joined #fuel
06:13 tatyana joined #fuel
06:13 javeriak joined #fuel
06:19 vkulanov joined #fuel
06:35 maestropandy joined #fuel
06:35 elo joined #fuel
06:36 elopez__ joined #fuel
06:40 elopez_ joined #fuel
06:42 cartik joined #fuel
06:46 elo joined #fuel
07:01 DavidRama joined #fuel
07:58 bgaifullin joined #fuel
08:01 pod2metra joined #fuel
08:04 noshankus joined #fuel
08:19 hkominos joined #fuel
08:29 yassine__ joined #fuel
08:34 apopovych Folks pls review nailgun patch https://review.openstack.org/#/c/288336/
08:38 pod2metra joined #fuel
08:48 code-R joined #fuel
08:49 code-R joined #fuel
08:51 code-R_ joined #fuel
09:12 javeriak joined #fuel
09:15 rtikunov joined #fuel
09:19 veena joined #fuel
09:21 DavidRama1 joined #fuel
09:24 code-R joined #fuel
09:32 ddmitriev joined #fuel
09:42 ekosareva joined #fuel
09:50 akscram joined #fuel
09:57 code-R joined #fuel
10:05 xarses_ joined #fuel
10:20 DavidRama joined #fuel
10:28 obcecado hi
10:28 obcecado does mos9 deploy the lbaas component?
10:34 kaliya joined #fuel
10:47 priya joined #fuel
10:57 zubchick Hi folks, please review small patch to railgun https://review.openstack.org/#/c/341387
10:57 zubchick railgun == nailgun
11:08 magicboiz joined #fuel
11:19 tatyana joined #fuel
11:57 kaliya joined #fuel
11:59 DavidRama joined #fuel
12:02 DavidRama joined #fuel
12:32 magicboiz joined #fuel
12:34 bgaifullin joined #fuel
12:35 v1k0d3n joined #fuel
12:38 DavidRama joined #fuel
12:44 flerfb0rt joined #fuel
12:49 vkulanov romcheg, can you, please, review two patches related to python-fuelclient (https://review.openstack.org/#/c/340219/, https://review.openstack.org/#/c/340030/)
12:49 romcheg vkulanov: I will
12:52 DavidRama joined #fuel
12:55 pod2metra joined #fuel
13:08 pod2metra joined #fuel
13:14 DavidRama joined #fuel
13:27 tatyana joined #fuel
13:36 ekosareva_ joined #fuel
13:40 ekosareva joined #fuel
13:42 AlexAvadanii I have a question about changes between fuel8 and fuel9, in terms of provisioning + deploy (i.e. mostly after this spec was imeplemented https://github.com/openstack/fuel-specs/blob/master/specs/9.0/graceful-stop-restart-deployment.rst)
13:43 AlexAvadanii we used UEFI with fuel8 (we added 2 small patches to fuel-agent that add the ESP partition flag + vfat partition on top of that); which worked very nicely with fuel, when we reset the env, the partitions were erased, and EFI boot entry was removed on each node
13:44 AlexAvadanii however, in fuel9, env reset and/or delete does not remove our EFI partition OR boot entry, leaving the enviromnet with old boot entires that point to non-existing partitions
13:44 ekosareva joined #fuel
13:45 AlexAvadanii any pointers on where to look to address this? or at least where this step belongs? env delete seems like a reasonable step to add the EFI boot entry cleanup to
13:46 mwhahaha AlexAvadanii: https://github.com/openstack/fuel-astute/blob/master/mcagents/erase_node.rb#L115-L123
13:46 mwhahaha well that's what it is today
13:46 mwhahaha but that's where we do the disk cleanup on node erasing
13:49 AlexAvadanii mwhahaha: thank you, that looks like a good point to start
13:49 AlexAvadanii btw, I really love what you did with the above spec, continuing a deploy after failure really speeds up testing and work in general
13:56 javeriak joined #fuel
13:59 Billias joined #fuel
14:19 ekosareva_ joined #fuel
14:44 xarses_ joined #fuel
14:44 ekosareva joined #fuel
14:51 code-R joined #fuel
14:55 code-R_ joined #fuel
14:57 cr0wrx joined #fuel
14:57 cr0wrx anyone know which logs I should investigate for why vip__public keeps stopping?
15:00 mwhahaha cr0wrx: pacemaker log, does the public gateway become unavailable?
15:00 mwhahaha that's generally why the vip__public stops
15:00 cr0wrx that's what I'm working on and am sure it's related...trying to add a second public gateway
15:01 cr0wrx I checked pacemaker.log and it just shows it starting and a bunch of info logs but nothing jumped out for why it went down
15:01 mwhahaha there's another log
15:01 mwhahaha let me go find the name
15:01 mwhahaha pengine.log
15:02 mwhahaha wait thats not it
15:02 mwhahaha yea it is that should have the errors in it
15:04 cr0wrx I don't have that file
15:05 mwhahaha check on the fuel master in the remote logs
15:06 raunak joined #fuel
15:10 cr0wrx hm not seeing it. Oh well, I see a handful of other issues.
15:11 mwhahaha it might be named something more cryptic :D depends on the version i think
15:13 Julien-zte joined #fuel
15:15 pod2metra warpc, evgenyl, dshulyak, kozhukalov review plz https://review.openstack.org/#/c/341593/
15:15 cr0wrx do you know when I should need to call neutron l3-agent-router-add ? I created a new router and it doesn't seem to be on l3 agent, so maybe I missed that step. But I haven't seen anyone calling that in guides online (they usually juts create network and router and set router gateway, maybe modify an ini file as well)
15:17 mwhahaha i think it's the last step
15:18 cr0wrx I've created routers before (internal) and didn't have to use that command, but they are assigned to l3 agent
15:22 mwhahaha i didn't find any reference in the fuel docs, but the redhat docs have it https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux_OpenStack_Platform/5/html/Installation_and_Configuration_Guide/Configuring_the_L3_Agent.html
15:26 cr0wrx well adding the router with that commands seems to have done some stuff I was messing with yesterday (router gateway now shows as up, ip namespaces exist (they didn't before), ovs flows were added for a port that was added (which I manually tried yesterday)...maybe neutron will take care of most of it
15:26 Anatolii joined #fuel
15:28 cr0wrx so one thing that seems fuel specific - do you know what the p_* network interfaces are for?
15:28 thumpba joined #fuel
15:29 cr0wrx It seems fuel spins one up and adds it to linux bridge (br-ex), and then uses it for ovs port as well (ie br-floating port to p_*)
15:29 cr0wrx why go through p_* instead of just adding br-ex directly to ovs port?
15:30 mwhahaha not sure
15:30 mskalski br-ex bridge has ip assigned if you add this directly to ovs then this will not work
15:31 mwhahaha aren't the p_* ports to ovs?
15:32 mskalski https://github.com/openvswitch/ovs/blob/master/FAQ.md "I created a bridge and added my Ethernet port to it, using commands"
15:32 cr0wrx er, I'm new to a lot of these concepts, but there is port in ovs for p_* and it's also been added to br-ex (brctl) but p_* is a network interface (/etc/network/interfaces.d/) it looks like
15:33 cr0wrx but the ifcfg file for it is different, containing ovs specific modifiers and whatnot, looks to be handled be /etc/network/openvswitch or something
15:34 cr0wrx anyways, just curious. I copied one of the existing p_* configs and slightly modified to make a second br-floating essentially and added to a new ovs bridge, it may have worked once I get through these other issues
15:34 cr0wrx trying to get a second floating ip range so we can separate prod/dev easier
15:37 Anatolii left #fuel
15:37 aneliubin joined #fuel
15:39 kaliya joined #fuel
16:02 raunak joined #fuel
16:13 Sesso joined #fuel
16:26 javeriak joined #fuel
16:45 Zer0Byte__ joined #fuel
16:46 DavidRama joined #fuel
16:47 DavidRama left #fuel
16:51 DavidRama joined #fuel
16:52 DavidRama left #fuel
16:53 Zer0Byte__ hi
17:04 thumpba joined #fuel
17:13 javeriak_ joined #fuel
17:13 cr0wrx Hi
17:20 krobzaur joined #fuel
17:21 krobzaur joined #fuel
17:22 krobzaur left #fuel
17:37 Sesso joined #fuel
17:52 Sesso_ joined #fuel
18:18 DavidRama joined #fuel
18:38 bgaifullin joined #fuel
18:38 elopez__ joined #fuel
18:46 raunak joined #fuel
18:53 pod2metra joined #fuel
18:59 elo joined #fuel
19:05 DavidRama joined #fuel
19:15 xarses joined #fuel
19:23 pod2metra joined #fuel
19:52 bgaifullin joined #fuel
20:06 raunak joined #fuel
20:24 neophy joined #fuel
20:49 tatyana joined #fuel
21:02 Sesso joined #fuel
21:19 Sesso joined #fuel
21:22 bgaifullin joined #fuel
21:36 AlexAvadanii fuel-agent guys: why debootstrap and not multistrap? it would speed up the process considerably, by obsoliting the necessity for apt-get dist-upgrade
21:37 AlexAvadanii I played with multistrap vs debootstrap for fuel a few months back, I was able to build bootstraps in 5min from a local mirror
21:38 mwhahaha do they take the same params? or is it alot to rework?
21:39 AlexAvadanii it leads to a few additional quirks you'd need to work around, so it would be a medium amount of rework
21:39 AlexAvadanii it's not as easy as replacing the debootstrap call and package with the new one
21:40 mwhahaha well that's probably why it wasn't looked at :D
21:40 AlexAvadanii but for continous integration, this would be a very deep cut in waiting time
21:40 mwhahaha this is true
21:40 mwhahaha that's part of fuel-agent i believe
21:40 AlexAvadanii yes, it is
21:41 AlexAvadanii when I was playing with it (in fuel 6.1), it was an experimental sh in fuel-main, then it got reworked into the python version in fuel agent
21:44 AlexAvadanii anyway, I would like to propose multistrap as a thing to look into, for the sake of total bandwidth and cycles we could spare
21:45 mwhahaha might be worth a mailing list message/blueprint or something
21:48 AlexAvadanii I have a few ideas for fuel-agent, if we are to change stuff there :) loop devices makes everything so much slower btw
21:48 AlexAvadanii and I understand the need to use them wehn building each filesystem image separately, but when there's only one partition, like for bootstrap, it just slows things down
21:50 Zer0Byte__ joined #fuel
21:57 elopez joined #fuel
22:04 Guest73487 joined #fuel
22:37 Egyptian joined #fuel
23:05 krogon joined #fuel
23:21 Sesso_ joined #fuel
23:49 Zer0Byte__ joined #fuel

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