Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2015-02-16

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

All times shown according to UTC.

Time Nick Message
03:10 jaypipes joined #fuel
06:50 monester_laptop joined #fuel
07:02 dklepikov joined #fuel
07:06 sambork joined #fuel
07:12 aliemieshko_ joined #fuel
07:15 xarses joined #fuel
07:20 evgeniyl___ joined #fuel
07:31 okosse joined #fuel
07:38 samuelbartel_ joined #fuel
07:40 saibarspeis joined #fuel
07:56 julien_ZTE joined #fuel
08:02 sambork joined #fuel
08:18 HeOS joined #fuel
08:22 teran joined #fuel
08:31 teran joined #fuel
08:32 sambork joined #fuel
09:00 Longgeek joined #fuel
09:03 andriikolesnikov joined #fuel
09:09 kaliya joined #fuel
09:13 avgoor joined #fuel
09:21 stamak joined #fuel
09:26 ddmitriev joined #fuel
09:33 monester_laptop joined #fuel
09:48 alecv joined #fuel
10:08 sc-rm left #fuel
10:13 tzn joined #fuel
10:20 e0ne joined #fuel
10:49 neophy joined #fuel
11:55 ddmitriev1 joined #fuel
11:57 e0ne joined #fuel
12:14 alexd__ joined #fuel
12:17 duculete hello, I am trying do add new puppet module in Fuel. I want to be able to sync new module on a controller, then apply the new module to that controller (only that module)
12:17 duculete can someone help with sync and apply commands example?
12:20 saibarspeis I see your question :P
12:21 saibarspeis You have to wait sometime, not everyone is paying attention
12:21 kaliya duculete: to apply, `puppet apply /path/to/your/puppet.pp`. Instead, after a cluster deployment, the master is not acting as a puppet master anymore. So you will need to sync manually (e.g. rsync)
12:22 duculete ah ok, so manually/rsync module to controller
12:22 duculete then puppet apply
12:22 duculete thanks
12:23 kaliya Yep, or `puppet apply -e 'Puppet::method'
12:23 kaliya rysinc itself is used as a pre-deployment hook, but then it's disabled, once the cluster is deployed and operational
12:32 sambork joined #fuel
12:43 sc-rm joined #fuel
12:51 azemlyanov joined #fuel
12:58 azemlyanov joined #fuel
13:10 azemlyanov joined #fuel
13:16 e0ne joined #fuel
13:24 andriikolesnikov joined #fuel
13:34 duculete i am trying to add new deb to fuel repo and rebuild the Packages
13:35 duculete for some reason i got on the controller the following error:  Failed to fetch http://192.168.1.1:8080/ubuntu/fuelweb/x86_64/dists/precise/main/binary-amd64/Packages  Hash Sum mismatch
13:35 duculete any ideas?
13:36 kaliya duculete: please try to ask in #fuel-dev
13:39 Bomfunk joined #fuel
13:50 e0ne joined #fuel
13:51 codybum joined #fuel
13:55 andriikolesnikov joined #fuel
13:55 codybum Is @xarses listening :)
13:55 kaliya hi codybum
13:55 codybum Hi kaliya
13:56 codybum I have had an ongoing issue that we have been fighting through for a few weeks.
13:56 codybum Interested in joining the struggle?
13:56 rbowen joined #fuel
13:56 kaliya Yes, but later. Now I'm having a session with a customer
13:57 codybum Sure. Hit me up when you finish
13:57 kaliya is it about vm.swappiness?
14:05 okosse joined #fuel
14:10 codybum hi @kaliya: I have changed many things.  vm.swappiness seemed to help, but the issue remains
14:10 miroslav_ joined #fuel
14:11 codybum I have also create a separate physical management network, changed timeout values, disabled firewalls and offloading.
14:18 codybum It seems regardless of what I do, I will many many errors like this: "oslo.messaging._drivers.impl_rabbit [-] Failed to publish message to topic 'reply_d96fc31901e6408da8429241744ea725': Socket closed", especially under neutron-server.
14:18 codybum Oh. I have also restricted using pcs all rabbitmq and neutron services to a single node.
14:33 Longgeek joined #fuel
14:40 Longgeek joined #fuel
14:43 sc-rm dklepikov: I was testing some more today, If I do look closer at one of the hosts. # hdparm -t /dev/sdb gives me 103Mb/sec where doing # ceph tell osd.8 bench gives me only 55Mb/sec
14:44 sc-rm dklepikov: But the funny part is that all 3 disks in this host is of the same type and all have journal placed on an SSD, but the perform differently
14:45 claflico joined #fuel
14:47 dklepikov sc-rm : try to run bench few times and compare speed
14:55 sc-rm dklepikov: I’ll do some more benchmarking then :-)
14:57 sc-rm dklepikov: Do you know of any plans of having calamari as a deployment option like zabbix, so one have some more information about the cluster deployed
14:58 jetole joined #fuel
15:00 dklepikov sc-rm : Unfortunately I do not have such information
15:04 sc-rm dklepikov: np :-) Just tried to set it up, but calamari complained about incomplete ceph configuration. But I doubt it, since ceph is running as it should right now…
15:05 teran_ joined #fuel
15:08 dklepikov sc-rm :  incomplete in what? does calamari show why it decide so?
15:10 sc-rm dklepikov: That is my problem, it does not say why it decides so. But I’ll try to dig some more around to figure out what it’s problem is. http://calamari.readthedocs.org/en/latest/operations/minion_connect.html is not giving a lot of information about how or where to look for information on the ser why it decides not to detect the ceph configuration.
15:11 sc-rm dklepikov: But it’s first going to be tomorrow, have to go for now
15:13 dklepikov sc-rm : ok
15:31 blahRus joined #fuel
16:04 jkirnosova joined #fuel
16:27 codybum I've got the RabbitMQ blues
16:37 e0ne joined #fuel
16:51 teran joined #fuel
17:04 clauded joined #fuel
17:09 clauded left #fuel
17:09 championofcyrodi joined #fuel
17:11 miroslav_ joined #fuel
17:26 stamak joined #fuel
17:31 Bomfunk joined #fuel
17:31 dburmistrov joined #fuel
17:31 wayneeseguin joined #fuel
17:31 64MABZ0GB joined #fuel
17:31 SamYaple joined #fuel
17:31 meow-nofer_ joined #fuel
17:31 seeg joined #fuel
17:31 byrdnuge joined #fuel
17:53 codybum joined #fuel
18:05 jobewan joined #fuel
18:39 championofcyrodi Any reason I cannot add the Controller+Telemetry - MongoDB roles as new controllers?
18:39 championofcyrodi Currently running 1 controller with HA
18:41 championofcyrodi basically when i add a controller through the fuel ui, the mongodb option is disabled
18:53 HeOS joined #fuel
18:56 e0ne joined #fuel
18:56 championofcyrodi let me re-phrase, we only added 1 controller to our new cluster, running in HA... we also chose 'Telemetry' role at that time.
18:56 championofcyrodi now it looks like we cannot rung telemetry (mongod) in HA since it's already installed???
19:00 kupo24z left #fuel
19:26 andriikolesnikov joined #fuel
19:28 xdeller joined #fuel
19:28 dhblaz joined #fuel
19:53 teran joined #fuel
19:59 MiroslavAnashkin It is Fuel limitation. It expects separate MongoDB instance per each controller due to a lot of info, produced by telemetry.
20:22 monester_laptop joined #fuel
21:37 claflico do most people not install controllers in HA mode? it's seems that doing so causes more complexity and requires you to sacrifice some other stuff
21:38 claflico i.e. I'm attempting to move from nova to neutron so that I can utilize the load-balancer plugin
21:38 claflico deploy an environment with just one controller and it is available as a setting but deploy in HA and the setting is gone
21:56 jobewan joined #fuel
22:08 Longgeek joined #fuel
22:11 jobewan joined #fuel
22:22 tzn joined #fuel
22:35 mattgriffin joined #fuel
22:50 daniel3 joined #fuel
23:16 tobiash joined #fuel

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