Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2015-03-23

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

All times shown according to UTC.

Time Nick Message
01:08 Longgeek joined #fuel
01:16 LiJiansheng joined #fuel
02:04 xdeller joined #fuel
02:29 gongysh joined #fuel
03:06 lee__ joined #fuel
03:24 adanin joined #fuel
03:47 lee__ joined #fuel
04:30 Longgeek joined #fuel
06:03 zerda joined #fuel
07:03 artem_panchenko left #fuel
07:04 artem_panchenko joined #fuel
07:08 adanin joined #fuel
07:42 sovsianikov joined #fuel
08:17 pbeskow joined #fuel
08:46 julien_ZTE joined #fuel
08:51 tzn joined #fuel
09:02 e0ne joined #fuel
09:04 corepb joined #fuel
09:04 alecv joined #fuel
09:21 HeOS joined #fuel
09:32 gongysh joined #fuel
09:34 evgeniyl___ joined #fuel
09:49 admin0 joined #fuel
09:49 saibarspeis joined #fuel
09:54 e0ne joined #fuel
10:00 ddmitriev joined #fuel
10:03 hyperbaba joined #fuel
10:04 stamak joined #fuel
10:06 subscope joined #fuel
10:14 e0ne joined #fuel
10:34 samuelBartel joined #fuel
10:35 samuelBartel_ joined #fuel
10:35 samuelBartel joined #fuel
10:41 tzn joined #fuel
10:42 ddmitriev joined #fuel
11:18 vtzan joined #fuel
11:19 vtzan Hello team, I am trying to create an HA enviroment on virtualbox with 5.1.1 release and i always get "deployment timeout". with non-HA setup everything goes smoothly. thank you
11:24 aliemieshko <vtzan>hello, where  and when you get "deployment timeout". Please describe more detail
11:44 saibarspeis joined #fuel
12:12 vtzan aliemieshko, hi, after about an hour the deployment times out with this message on top of fuel web interface
12:13 vtzan aliemieshko, it creates the nodes, it installs the ubuntu os on the nodes and in step of installing openstack it times out
12:14 vtzan aliemieshko, :just FYI : network verification passes without problem.
12:16 kaliya vtzan: virtual or hardware?
12:18 vtzan kaliya, : virtualbox
12:18 vtzan running on 8 core cpu, 32GB Ram and ssd drive
12:19 kaliya vtzan: I experienced something like this in KVM when I was trying to deploy on qcow2, I switched to raw storage
12:20 vtzan the problem is on the enviroment deployment. not image creation...
12:21 kaliya vtzan: in my case, deployment got stuck during the OpenStack provisioning phase
12:21 kaliya but I had no time to dig and debug alot
12:40 aliemieshko please describe your HA configuration(number controllers, computes, their roles, configuration)
12:58 vtzan aliemieshko, : 3 controllers, Ubuntu OS, ceph for all, 2 compute nodes, GRE network configuration.
13:04 aliemieshko how much ram on each controllers,computes ?
13:12 vtzan aliemieshko, 2 GB Ram, 1 CPU, 192GB HD
13:13 vtzan aliemieshko, do you think i might need extra ram for the controllers?
13:14 aliemieshko yes, i think
13:14 vtzan ok let me try
13:16 vtzan aliemieshko, 4GB should be ok?
13:16 aliemieshko at least i think
13:19 vtzan ok lets rock. i gave them 2 cpus and 8GB Ram for each controller.
13:33 saibarspeis joined #fuel
13:45 mattgriffin joined #fuel
13:51 e0ne joined #fuel
13:53 claflico joined #fuel
14:24 samuelBartel joined #fuel
14:36 jobewan joined #fuel
14:38 stamak joined #fuel
15:15 blahRus joined #fuel
15:23 emagana joined #fuel
15:36 vtzan aliemieshko, Deployment has failed. Timeout of deployment is exceeded.
15:37 kaliya vtzan: try raw for disk images
15:38 aliemieshko can you attach log with error ?
15:39 aliemieshko from node where is error you got
15:40 tzn joined #fuel
15:42 jaypipes joined #fuel
15:43 mattgriffin joined #fuel
15:50 rmoe joined #fuel
15:50 vtzan check here: http://paste.openstack.org/show/195430/
15:51 vtzan the fuel dashboard error is : Deployment has failed. Timeout of deployment is exceeded.
15:55 emagana joined #fuel
15:57 bogdando joined #fuel
15:57 stamak joined #fuel
15:57 emagana joined #fuel
16:00 emagana joined #fuel
16:13 tzn_ joined #fuel
16:23 MiroslavAnashkin vtzan: How many Ceph OSD nodes/devices have you configured?
16:38 vtzan 2 compute nodes with 1 osd each one. replication is set to 2
16:38 vtzan the procedure stacks before provisioning the compute/ceph osd nodes
16:39 vtzan MiroslavAnashkin, only the first node becomes ready and the other two end up in error state
16:41 MiroslavAnashkin These ruby errors happen on the very start. Please try to generate a diagnostic snapshot or share the full log where did you get the errors
16:41 vtzan MiroslavAnashkin, : https://www.dropbox.com/s/i10g0wanknfn02a/fuel-snapshot-2015-03-23_15-36-58.tgz?dl=0
16:41 vtzan diagnostic snapshot
16:57 emagana joined #fuel
17:01 emagana joined #fuel
17:11 mattgriffin joined #fuel
17:20 xarses joined #fuel
17:26 DaveJ__ joined #fuel
17:29 MiroslavAnashkin vtzan: Oh, you have combined controllers with Ceph (monitors) with MongoDB...
17:30 MiroslavAnashkin It might work on very powerful environment.
17:54 MiroslavAnashkin vtzan: Well. looks like it is failed by timeout. We have such issue with MuSQL/Galera/Xtrabackup/Corosync in Ubuntu under Virtualbox.
17:54 MiroslavAnashkin Did you tried 6.0?
17:54 MiroslavAnashkin 6.0 has less issues under VBox
17:56 MiroslavAnashkin For 5.1.1 VBox NICs are not able to process a lot of packets and initial Galera sync under Ubuntu works very slow. And you added Ceilometer and set MongoDB to run on controllers
17:57 MiroslavAnashkin MongoDB may slowdown the most powerful hardware. Ceilometer creates a lot of data samples and it came fast when single request to Mongo processes 5-6 millions of records and serializes all to BSON.
17:58 MiroslavAnashkin So, it is highly recommended to use separate Mongo nodes
18:13 emagana joined #fuel
18:20 xarses joined #fuel
18:28 emagana joined #fuel
18:54 tzn joined #fuel
19:04 wiza joined #fuel
19:04 avlasov joined #fuel
19:04 martineg_ joined #fuel
19:04 sovsianikov joined #fuel
19:04 zimboboyd joined #fuel
19:12 HeOS joined #fuel
19:46 dpyzhov joined #fuel
20:21 e0ne joined #fuel
20:52 jaypipes joined #fuel
21:31 mattgriffin joined #fuel

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