Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-03-18

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

All times shown according to UTC.

Time Nick Message
00:21 fedexo joined #fuel
00:26 vsedelnik joined #fuel
01:01 Jeffrey4l_ joined #fuel
01:25 krypto joined #fuel
01:27 vsedelnik joined #fuel
01:45 xarses joined #fuel
02:00 krypto i am trying to add new controller nodes to existing controller nodes but i get "Could not evaluate: Primitive 'neutron-metadata-agent' was not found in CIB!"
02:10 askb joined #fuel
02:28 vsedelnik joined #fuel
03:24 skamithi13 joined #fuel
03:28 vsedelnik joined #fuel
04:14 skamithi14 joined #fuel
04:22 skamithi13 joined #fuel
04:28 treaki__ joined #fuel
04:29 vsedelnik joined #fuel
04:45 skamithi14 joined #fuel
04:55 bildz joined #fuel
05:15 vsedelnik joined #fuel
05:34 krypto joined #fuel
05:34 vsedelnik joined #fuel
05:40 krypto joined #fuel
05:43 [HeOS] joined #fuel
05:44 vsedelnik joined #fuel
06:01 skamithi13 joined #fuel
06:41 fzhadaev_ joined #fuel
06:47 tzn joined #fuel
07:35 e0ne joined #fuel
07:41 akaszuba_ joined #fuel
07:41 akaszuba joined #fuel
07:58 pbrooko joined #fuel
08:11 neilus joined #fuel
08:15 pbrooko_ joined #fuel
08:17 pbelamge joined #fuel
08:30 neilus joined #fuel
08:36 vsedelnik joined #fuel
08:38 vsedelnik joined #fuel
08:56 neilus joined #fuel
09:17 yantarou joined #fuel
09:18 charz hi!
09:19 charz Just a simple question in fuel 8.0, is ther no swift cluster setup on controller nodes? I saw a option to enable ceph as object storage (swift api)
09:22 pbrooko joined #fuel
09:26 pbelamge joined #fuel
09:36 sc-rm joined #fuel
09:38 sc-rm pasquier-s: Now after running over the night after I tried out https://bugs.launchpad.net/mos/+bug/1497961/comments/15 the WARN went away, but now it’s back becasuse the queue lma_notifications.info keeps growing
09:40 sovsianikov joined #fuel
09:40 pasquier-s sc-rm, iiuc you need to restart the lma_collector service on the controllers, run this from one controller: crm resource restart lma_collector
09:41 sc-rm pasquier-s: Got the same idea, and for now it seems to work
09:42 sovsianikov sc-rm, check pcs status and then rabbitmqctl cluster_status on all controller and show output booth command
09:42 pasquier-s sc-rm, this is a side effect of this bug: https://bugs.launchpad.net/fuel-plugins/+bug/1503251
09:43 sc-rm pasquier-s: I see, makes sense, because I applied the other fix, then lma stopped. Good to know :-)
09:46 palexster joined #fuel
10:00 magicboiz joined #fuel
10:03 HeOS joined #fuel
10:08 neilus1 joined #fuel
10:09 pasquier-s sc-rm, in fact lma didn't stopped but it disconnected from rabbitmq and couldn't reconnect without a restart, this is fixed in the 0.9 upcoming release of lma
10:18 magicboiz joined #fuel
10:43 neilus joined #fuel
11:08 e0ne joined #fuel
11:16 pbrooko joined #fuel
11:32 slo_ making some changes within cinder.conf on controller node. Do I just use scp to copy the same configuration to all other openstack nodes or is there another, more proper way to do it?
11:32 pbelamge joined #fuel
11:52 skamithi14 joined #fuel
11:58 vsedelnik joined #fuel
12:05 pbrooko_ joined #fuel
12:26 rmoe_ joined #fuel
12:27 neilus1 joined #fuel
12:49 vsedelnik joined #fuel
13:01 neilus joined #fuel
13:14 pbrooko joined #fuel
13:24 fzhadaev joined #fuel
13:35 pbrooko joined #fuel
13:53 pbrooko joined #fuel
13:55 e0ne joined #fuel
14:30 pbrooko joined #fuel
14:35 claflico joined #fuel
14:46 BillC joined #fuel
14:52 neouf joined #fuel
14:56 HeOS joined #fuel
15:21 askb_ joined #fuel
15:22 blahRus joined #fuel
15:30 e0ne joined #fuel
15:50 CheKoLyN joined #fuel
15:54 xarses joined #fuel
16:49 neilus joined #fuel
16:51 askb joined #fuel
17:25 angdraug joined #fuel
17:31 kur1j_ joined #fuel
17:37 e0ne joined #fuel
17:42 kur1j_ I'm struggling with the networking of a single controller deployment. The system has 4 nics 1 for fuel network, 1 for public (192.168.1.0/24), 1 for storage (192.168.10.0/24), 1 for management (192.168.20.0/24) and last nic for private (192.168.30.0/24) and vlan on same nic (192.168.40.0/24). The system deploys successfully. However I can't get to the horizon interface at 192.168.1.3 from a machine that is on that network. I can ping t
17:59 vsedelnik joined #fuel
18:05 dslevin joined #fuel
18:15 HeOS joined #fuel
18:51 magicboiz joined #fuel
19:03 dslevin joined #fuel
19:13 Tatyanka_Leontov left #fuel
19:21 dslevin joined #fuel
19:22 e0ne joined #fuel
19:41 krobzaur joined #fuel
19:47 kur1j_ .
20:15 e0ne joined #fuel
20:56 daniel2 left #fuel
21:01 krypto joined #fuel
21:11 ohyeahhuh what is 192.168.1.0 on your network segment?
21:11 ohyeahhuh kur1j_:
21:24 mbiciunas_ joined #fuel
21:24 kur1j_ what do you mean?
21:24 kur1j_ that is my "lan"
21:25 charz_ joined #fuel
21:25 ukinau- joined #fuel
21:25 Damjanek_ joined #fuel
21:25 seeg- joined #fuel
21:25 zimboboy- joined #fuel
21:25 Verilium joined #fuel
21:25 kur1j_ Internet -->pfsense router--->lan (192.168.1.0/24)
21:26 pbrzozowski_ joined #fuel
21:26 dhellmann joined #fuel
21:26 JiYu joined #fuel
21:27 gbraad joined #fuel
21:30 krypto hi can some one help  me find all the tasks that can be excuted against a node-id
21:31 kur1j_ ohyeahhuh:
21:31 ggolin kur1j_: the only thing i can think of is that this nic is not on the right subnet
21:34 kur1j_ ggolin: on the controller node I can ping to the gateway
21:35 kur1j_ and I can ping the controller node from gateway
21:35 kur1j_ I just can't ping the HAProxy virtualhost
21:36 ggolin can you nmap it?
21:37 kur1j_ inside of the HAProxy virtual ip netns HAProxy and if I do if config it shows 192.168.1.3
21:37 vsedelnik joined #fuel
21:37 kur1j_ however from within that namespace I can't ping the 192.168.1.1 gateway
21:37 kur1j_ nmap from which machine?
21:40 ggolin from wherever you're trying to hit horizon
21:42 kur1j_ well I know I can't ping in but I'll try that
21:43 ggolin and you did the network verification test and all that, right?
21:43 kur1j_ I only have one machine
21:43 kur1j_ so it doesn't work
21:44 ggolin you only have one machine?
21:44 ggolin one node in the env?
21:45 kur1j_ correct
21:45 kur1j_ just trying to get the controller setup
21:45 ggolin it seems to me that the shortest critical path here is to add another node so that you could test your network connectivity settings.
21:45 ggolin short of that its shots in the dark, im afraid.
21:46 kur1j_ ok fair enough
21:46 kur1j_ what is the minimum network configuration needed? Would I need the management and private networks?
21:48 kur1j_ bleh I'm on Windows no nmap
21:48 ggolin yes and yes
21:48 kur1j_ storage?
21:49 kur1j_ I don't have vlans as I can't access switches to change settings
21:49 kur1j_ so I'm having to use multiple separate network switches
21:50 ggolin so then dont use vlans
21:50 ggolin it would make sense to skim the planning guide, imho
21:50 kur1j_ yeah but I don't have enough network ports
21:50 kur1j_ I did
21:50 ggolin just to get the taste of what is required and what you can do
21:51 ggolin oh good
21:51 ggolin are you just testing things out or?
21:51 ggolin (if you're testing id recommend using the vm scenarios, those get you up and running rather quickly)
21:52 kur1j_ I am just setting it up as a proof of concept
21:52 kur1j_ for now
21:52 kur1j_ and once I know more be deploying it on 32+ nodes
21:53 kur1j_ vm scenarios?
21:53 ggolin oh wait this is #fuel, not #mirantis
21:54 kur1j_ yes lol
21:54 ggolin https://docs.mirantis.com/openstack/fuel/fuel-6.1/virtualbox.html
21:54 ggolin i was referring to this
21:56 kur1j_ hmmm didn't see this...I was using kvm to do something similar
21:57 kur1j_ fuel on a vm
21:57 kur1j_ create another vm and let fuel bootstrap it
22:09 vsedelnik joined #fuel
22:15 kur1j_ ggolin: what should be acting as the management network gateway?
22:15 kur1j_ or does it not matter?
22:18 ggolin pardon me for the non-answer, but presumably the gateway on that network? :)
22:19 kur1j_ well what if there isn't one? it's just a switch
22:23 rmoe joined #fuel
22:59 ggolin is it a required setting? i dont remember.
23:18 Damjanek_ joined #fuel
23:26 kur1j ggolin: on the public side interface yes, on the other networks no
23:26 kur1j im drawing this out to make better sense of it
23:26 ggolin yes that is a good idea
23:43 neouf joined #fuel
23:44 krypto joined #fuel
23:44 e0ne joined #fuel

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