Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2014-06-17

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

All times shown according to UTC.

Time Nick Message
00:31 rmoe joined #fuel
01:01 angdraug Kupo24z: it is, there are still cases where nova would use it (e.g. snapshotting a vm)
01:02 angdraug basically give it the rest of the boot drive, you're not supposed to use it for an OSD anyway
01:11 crandquist joined #fuel
01:28 xarses joined #fuel
01:29 crandquist joined #fuel
02:24 aleksandr_null joined #fuel
02:48 crandquist joined #fuel
04:00 crandquist joined #fuel
04:12 vkozhukalov joined #fuel
04:39 crandquist joined #fuel
04:39 crandquist joined #fuel
04:41 crandquist joined #fuel
04:44 xarses Kupo24z: as long as you are using ceph ephemeral, no
04:51 crandquist joined #fuel
05:22 crandquist joined #fuel
05:24 kobier joined #fuel
06:12 e0ne joined #fuel
06:14 al_ex3 joined #fuel
06:50 vkozhukalov joined #fuel
07:10 artem_panchenko joined #fuel
07:14 dnikishov joined #fuel
07:30 e0ne joined #fuel
07:47 sbartel joined #fuel
07:47 sbartel hello
07:47 sbartel do you know on which repository i can got neutron-bass package for fuel 4.1.1?
07:53 sbartel sorry neutron-lbaas not neutron-baas
08:04 saibarspeis joined #fuel
08:15 brain461 hello. on centos it does not come as separate package, it's the part of openstack-neutron rpm
08:16 mattymo wow what a huge netsplit
08:17 brain461 (I presume that you meant neutron-lbaas-agent package, correct?)
08:17 sbartel yes this is it
08:18 sbartel i look into http://fuel-repository.mirantis.com/fwm/4.1.1/ubuntu/pool/main/
08:18 sbartel but wasn't able to find it
08:18 skonno joined #fuel
08:18 sbartel I found that bug https://bugs.launchpad.net/fuel/+bug/1330610
08:18 sbartel but with no mention on how to retrieve it for ubuntu
08:20 artem_panchenko joined #fuel
08:20 xarses joined #fuel
08:20 rmoe joined #fuel
08:20 bogdando joined #fuel
08:20 obcecado joined #fuel
08:20 agordeev joined #fuel
08:20 zhiyan joined #fuel
08:20 strictlyb joined #fuel
08:20 gleam joined #fuel
08:21 aglarendil joined #fuel
08:21 book` joined #fuel
08:21 jbellone joined #fuel
08:21 izinovik joined #fuel
08:21 xenolog joined #fuel
08:21 vkramskikh joined #fuel
08:21 holser joined #fuel
08:21 monester joined #fuel
08:21 evg joined #fuel
08:21 mrasskazov joined #fuel
08:21 kat_pimenova joined #fuel
08:21 jkirnosova joined #fuel
08:21 apalkina joined #fuel
08:21 akislitsky joined #fuel
08:21 MiroslavAnashkin joined #fuel
08:21 MorAle joined #fuel
08:21 meow-nofer joined #fuel
08:21 mihgen joined #fuel
08:21 dburmistrov joined #fuel
08:21 nurla joined #fuel
08:21 justif joined #fuel
08:21 brain461 regarding ubuntu package, there's a bug https://bugs.launchpad.net/fuel/+bug/1330610 - you could subscribe to it to get status updates
08:22 sbartel ok thank you
08:22 brain461 welcome!
08:22 sbartel but in the bug it is said that "Customers have to obtain this package elsewhere in order to configire LBaaS."
08:22 brain461 yes, we will check it and make available from our own repos
08:22 sbartel do you know where elsewhere I have to look? ^^
08:23 sbartel i had the same issu for 4.1
08:23 sbartel and a developper told me it was packaged but not include as there is no puppet manifest for lbaas-agent
08:23 sbartel and send it to me
08:24 sbartel i guess it is packaged for  4.1.1 also
08:24 sbartel i am wrong?
08:24 obcecado_ joined #fuel
08:27 brain461 we have to recheck that
08:27 brain461 we'll provide updates in the bug
08:28 sbartel ok thank you
08:41 e0ne joined #fuel
08:44 artem_panchenko joined #fuel
08:44 xarses joined #fuel
08:44 rmoe joined #fuel
08:44 bogdando joined #fuel
08:44 agordeev joined #fuel
08:44 zhiyan joined #fuel
08:44 strictlyb joined #fuel
08:44 gleam joined #fuel
09:37 skonno_ joined #fuel
10:32 sbog joined #fuel
10:51 scroiset_ Hi Fuelers, I've a question about hardware sizing of a Fuel node ... to be able to manage around 200 nodes.
10:51 scroiset_ do you have any numbers somewhere ? is 12 CPU/ 32 Gb RAM could be sufficent ?
10:51 scroiset_ or should I deploy fuel on multi node (ie deploy components like astute or nailgun on 2 nodes)?
10:59 skonno joined #fuel
11:05 racingferret joined #fuel
11:07 evg scroiset_: hi
11:07 evg scroiset_: you can read about system requirements here:
11:08 evg scroiset_: http://docs.mirantis.com/openstack/fuel/fuel-5.0/pre-install-guide.html#system-requirements
11:10 evg scroiset_: i haven't  got such huge env, but i would pay more attantion for the controller nodes
11:21 racingferret hi guys, was wondering if you could shed some light on an error I'm getting:
11:21 racingferret http://paste.openstack.org/show/J3o0WFizNGc6YpRTbvwi/
11:22 racingferret I tried to add two more hosts to an existing environment, but when I deployed changes, the Fuel gave me an error and now I can't stop the deploy.
11:23 racingferret cobbler and rabbitmq are both running fine (yes, I tried rebooting)
11:25 skonno_ joined #fuel
11:29 scroiset_ evg: thanks for the links .. given that is it reasonnable to say that 8 cpu and 8Gb RAM are enought to handle 200 nodes ?
11:29 scroiset_ evg: and indeed the controller size is much more critical but it's another question ;)
11:30 skonno joined #fuel
11:30 e0ne joined #fuel
11:37 ddmitriev joined #fuel
11:59 evg scroiset_: nodes are deploying sequentially in 5.0, so i hope it's enough
12:05 racingferret actually, scratch that - I was a change that I made at the behest of out InfoSec guys.
12:22 scroiset_ evg: uhmm .. by sequentially do you mean the operating system bootstrap step (via PXE) or the whole provisioning process (puppet) ?
12:22 scroiset_ evg: I guess/wish the first step only is sequential
12:46 crandquist joined #fuel
13:00 dilyin joined #fuel
13:13 evg scroiset_: no, i mean provisioning,
13:15 evg scroiset_: http://docs.mirantis.com/openstack/fuel/fuel-5.0/release-notes.htm (sorry i can't give you the ancor in the doc, search "sequentially")
13:17 skonno joined #fuel
13:38 scroiset_ evo: ok thks, the sequential deployement concerns _only_ controllers, therefore deployments of computes nodes are in // ..
13:39 scroiset_ evo: ..  so the sizing of the fuel node depends of numbers of computes nodes, as I presumed initially.
13:51 ddmitriev joined #fuel
14:23 xdeller joined #fuel
14:30 evg scroiset_: yes, you're right
14:32 evg scroiset_: i was unclear, sorry
14:56 ddmitriev joined #fuel
14:58 crandquist joined #fuel
15:04 ddmitriev joined #fuel
15:07 scroiset_ evg: np, thank you for answering .. u've provided to me lot of pointers
15:13 casanch1 joined #fuel
15:24 blahRus joined #fuel
15:26 jobewan joined #fuel
15:31 DaveJ__ joined #fuel
15:49 xarses joined #fuel
16:04 ddmitriev joined #fuel
16:12 rmoe joined #fuel
16:35 ddmitriev joined #fuel
16:37 angdraug joined #fuel
17:00 casanch1_ joined #fuel
17:02 ddmitriev joined #fuel
17:57 e0ne joined #fuel
18:01 albionandrew joined #fuel
18:03 albionandrew xarses: quick question. Where do you get the common name cobbler in docker? For example if I do a docker ps -a I see a list of containers but no names. Yet if I do a dockerctl shell cobbler I can get in. Where do I find the name of the container>
18:03 albionandrew ?
18:12 xarses albionandrew: not sure, dockerctl is our invention. I asked matt for a list function. see https://review.openstack.org/#/c/96776/
18:15 e0ne joined #fuel
18:15 albionandrew xarses: that explains it. Thanks. Just developing our KB on it …
18:32 e0ne joined #fuel
18:50 e0ne joined #fuel
18:54 e0ne joined #fuel
18:55 vkozhukalov joined #fuel
19:06 albionandrew xarses: Any plans for Nagios ? I know it was in Fuel is it coming back?
19:07 xarses not for nagios, zabbix is on the roadmap
19:10 albionandrew xarses: Time frame?
19:15 xarses https://blueprints.launchpad.net/fuel/+spec/monitoring-system
19:16 xarses high priority for 5.1
19:17 CTWill joined #fuel
19:18 CTWill Hello, is there someone who can help me troubleshoot my fuel install?
19:19 albionandrew xarses: thanks
19:22 xarses CTWill: go ahead, We'll try out best
19:23 CTWill I just installed mirantis 5.0, deployed 8 nodes, 1 controller, 3 compute, 1 Storage Cinder storage and 3 Storage Ceph OSD
19:23 CTWill I can only deploy m1.tiny using the TestVM image only
19:24 CTWill error message I get is No valid host was found, Code 500
19:25 xarses CTWill you will need to inspect the nova-scheduler log to determine why it was unable to find a compute host
19:25 CTWill side note the health check has one failure at check stack auto scaling, not sure if that is important
19:26 CTWill that is on the compute node?
19:26 xarses you can also quickly glance at the nova service-list to see if any/all of the compute nodes show online (source /root/openrc ; nova service-list) from a controller
19:26 xarses no, it will be on one of the three controllers
19:27 xarses sorry, you have one
19:28 xarses it will be in your controller /var/log/nova-all.log if you have it, /var/log/nova/scheduler.log if you dont
19:28 xarses search for "no valid host"
19:30 CTWill >Jun 17 19:06:25 node-25 nova-nova.scheduler.utils WARNING: Failed to scheduler_run_instance: No valid host was found. Exceeded max scheduling attempts 3 for instance
19:30 CTWill node-25 nova-nova.scheduler.utils WARNING: Setting instance to ERROR state.
19:34 xarses output of nova service-list?
19:35 CTWill I do not see a log for /var/log/nova/scheduler.log
19:35 CTWill the pasted logs are from  /var/log/nova-all.log
19:36 xarses you will have one or the other depending on your logging settings
19:38 CTWill would this happen if I deployed a HA node cloud/
19:39 CTWill or Multi-node with HA?
19:40 xarses possibly, it really depends on why the filter for the nova-scheduler failed to select a compute host. typically, its constrained by nova-compute must be online, and must have enough cpu and memory for the flavor
19:40 CTWill 4 cpu
19:40 CTWill 8gb ram
19:41 CTWill 40gb hd
19:41 CTWill centos
19:41 CTWill compute nodes 8cpu 16gb ram 40gb hd
19:44 xarses CTWill: so can you paste the result of 'nova service-list'?
19:44 rmoe also nova hypervisor-show <hostname of a compute>
19:48 e0ne joined #fuel
19:51 CTWill I get an error message
19:51 CTWill [root@node-25 log]# nova service-list
19:51 CTWill ERROR: You must provide a username via either --os-username or env[OS_USERNAME]
19:51 xarses source ~/openrc
19:51 xarses then run the commands
19:52 CTWill +------------------+--------------------+----------+---------+-------+----------------------------+-----------------+
19:52 CTWill | Binary           | Host               | Zone     | Status  | State | Updated_at                 | Disabled Reason |
19:52 CTWill +------------------+--------------------+----------+---------+-------+----------------------------+-----------------+
19:52 CTWill | nova-cert        | node-25.domain.tld | internal | enabled | up    | 2014-06-17T19:52:25.000000 | -               |
19:52 CTWill | nova-consoleauth | node-25.domain.tld | internal | enabled | up    | 2014-06-17T19:52:27.000000 | -               |
19:52 CTWill | nova-scheduler   | node-25.domain.tld | internal | enabled | up    | 2014-06-17T19:52:28.000000 | -               |
19:53 CTWill | nova-conductor   | node-25.domain.tld | internal | enabled | up    | 2014-06-17T19:52:25.000000 | -               |
19:53 e0ne_ joined #fuel
19:53 CTWill | nova-network     | node-26.domain.tld | internal | enabled | up    | 2014-06-17T19:52:25.000000 | -               |
19:53 CTWill | nova-compute     | node-26.domain.tld | nova     | enabled | up    | 2014-06-17T19:52:25.000000 | -               |
19:53 CTWill | nova-network     | node-27.domain.tld | internal | enabled | up    | 2014-06-17T19:52:25.000000 | -               |
19:53 CTWill | nova-compute     | node-27.domain.tld | nova     | enabled | up    | 2014-06-17T19:52:26.000000 | -               |
19:53 CTWill | nova-network     | node-28.domain.tld | internal | enabled | up    | 2014-06-17T19:52:26.000000 | -               |
19:53 CTWill | nova-compute     | node-28.domain.tld | nova     | enabled | up    | 2014-06-17T19:52:26.000000 | -               |
19:53 CTWill +------------------+--------------------+----------+---------+-------+----------------------------+-----------------+
19:53 xarses CTWill: next time please use http://paste.openstack.org/ next time to prevent the channel flooding
19:54 CTWill ok sorry
19:54 CTWill its been a really long time sinec I used IRC
19:54 CTWill netcom days
19:54 xarses happens
19:55 xarses can you paste 'nova hypervisor-show node-26.domain.tld'
19:56 CTWill Paste #84341
19:58 rmoe I think it's a disk space issue
19:58 rmoe your compute nodes report 14GB free
19:58 rmoe but all flavros > small use a minimum of 20GB
19:59 xarses try booting from a cinder volume
19:59 CTWill So how much space should a compute node have I gave it 40gb
20:00 rmoe you gave it 40 just for compute storage?
20:00 rmoe or 40 including the OS?
20:00 CTWill 40 including os
20:00 xarses so, if you use ceph for ephemeral storage, then it would use the ceph pool, otherwise the ephemeral disk will come out of the vm volume on each compute node
20:01 rmoe or you can boot from a cinder volume
20:01 CTWill I have 3x Ceph OSD
20:01 MiroslavAnashkin It depends on flavor or image filesystem size, if you create instances from images. Sometimes terabytes of free stace is not enough
20:03 xarses CTWill: you would have to select "Use Ceph RBD for ephemeral instance storage" from the settings page prior to deployment to get the deployment engine to configure it.
20:03 CTWill hmm
20:04 CTWill sounds like I need to redoly
20:05 xarses there is a reset deployment button on the last tab, that will set you back to an un-deployed state (it will nuke the hdd and return the nodes to discovery, but keep your current settings and role assignments)
20:09 CTWill im going to nuke all the nodes and re-deploy
20:10 CTWill all the nodes are looking the same
20:20 casanch1_ joined #fuel
20:20 rmoe joined #fuel
20:20 jobewan joined #fuel
20:20 sbog joined #fuel
20:20 bogdando joined #fuel
20:20 agordeev joined #fuel
20:20 gleam joined #fuel
20:20 strictlyb joined #fuel
20:20 zhiyan_ joined #fuel
20:21 CTWill ok I am redeploying I have my fingers crossed
20:36 vkozhukalov joined #fuel
20:36 xdeller joined #fuel
20:39 e0ne joined #fuel
20:50 e0ne joined #fuel
20:52 casanch1 joined #fuel
21:03 e0ne joined #fuel
21:03 CTMotorsports joined #fuel
21:25 e0ne joined #fuel
21:26 e0ne joined #fuel
21:27 casanch1_ joined #fuel
22:42 fandi joined #fuel
22:43 CTWill joined #fuel
22:53 CTWill So after the rebuild, there is a lot more logs on the HA pair
22:54 CTWill the same problem is there, that I can not deploy anyting than m1.tiny or a different flavor, I am going into the logs now

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