Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-03-29

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

All times shown according to UTC.

Time Nick Message
00:09 vsedelnik joined #fuel
00:23 eriswans How do I make the fuel master listen for ssh on its public (NON-admin/PXE) network interface? The sshd_config written by puppet has an explicit ListenAddress and I don't want to be fighting puppet.
00:25 fatdragon joined #fuel
00:42 xarses joined #fuel
00:57 fatdragon joined #fuel
01:06 Liuqing joined #fuel
01:37 fatdragon joined #fuel
01:40 fatdragon joined #fuel
01:47 fatdragon joined #fuel
01:51 Liuqing joined #fuel
02:11 vsedelnik joined #fuel
03:03 fedexo joined #fuel
03:12 vsedelnik joined #fuel
03:15 askb joined #fuel
04:13 vsedelnik joined #fuel
04:20 vsedelnik joined #fuel
05:41 dcs joined #fuel
05:43 zefciu joined #fuel
05:45 dcs hi all, pls hlp me. how i can use multiple external network this mos 7.0 HA scenario, pls link me document reference, if possible. very thx
05:49 kaliya joined #fuel
05:51 fatdragon joined #fuel
06:21 fBigBro joined #fuel
06:22 tuvenen joined #fuel
06:31 magicboiz joined #fuel
06:43 neilus joined #fuel
06:55 DevStok joined #fuel
07:00 l4ng1t joined #fuel
07:02 DevStok hi
07:02 DevStok I'm upload snapshots
07:03 DevStok I'm using KILO glance + swift
07:03 DevStok I'm facing problems with /srvnode/.... path
07:03 DevStok get full
07:09 criss2016 joined #fuel
07:10 samuelBartel joined #fuel
07:14 javeriak joined #fuel
07:16 neilus joined #fuel
07:21 fatdragon joined #fuel
07:32 venkat_ joined #fuel
07:32 javeriak joined #fuel
07:52 vsedelnik joined #fuel
07:52 vsedelnik joined #fuel
07:55 magicboiz joined #fuel
08:05 DevStok ddd
08:05 DevStok /dev/dm-4
08:07 akaszuba joined #fuel
08:09 kozhukalov dcs,  https://docs.mirantis.com/openstack/fuel/fuel-7.0/operations.html#configuring-multiple-cluster-networks
08:11 kozhukalov DevStok: could you please describe in more details the problem that you are facing?
08:15 DevStok I'm uploading a lot of snapshots with glance image-create
08:15 DevStok KILO glance + swift
08:15 DevStok HA
08:16 DevStok there is a path that get full during the process of uploading several snapshots
08:16 DevStok /dev/dm-4                   50G   17G   31G  35% /
08:17 DevStok so i have to delete /srv/node/*
08:20 javeriak joined #fuel
08:22 dcs kozhukalov: thx, but this doc about new environment, right? i have deployed env HA with GRE, and me want add new external network, it's possible?
08:22 fatdragon joined #fuel
08:23 pumphouse joined #fuel
08:23 elo joined #fuel
08:25 kozhukalov dcs, changing configuration of existing cluster is so called "day 2" operation (maintenance). fuel 7.0 does not support such operations, we are planning to support such operations in 9.0 release.
08:30 dcs kozhukalov: ok, i understood. but i can change this manyaly, in deployed env?
08:32 kozhukalov yes, you can do it manually
08:33 dcs kozhukalov: very thx
08:34 jmhale joined #fuel
08:34 aleksandr_null joined #fuel
08:39 kozhukalov dcs once cluster is deployed, fuel 7.0 does not change its configuration until you add some nodes (e.g. computes) to the cluster. when you add new nodes, you again have to click "deploy changes" button, but new node will be configured the same way as other nodes in the cluster (cluster configuration is stored in the database) and you will have to change
08:39 kozhukalov network configuration on the new node manually after it is deployed.
08:40 DevStok @72561EFA2E3A6628
08:40 DevStok @kozhukalov
08:42 kozhukalov DevStok: could you please ask your question in #glance channel, there are glance developers who could help. your question is more about glance + swift than about fuel
08:46 DevStok th
08:46 DevStok x
08:50 dcs kozhukalov: ok, i have 3 controller node and N compute and e.g., i will not add anymore controllers nodes so i can change configuration only exist controllers nodes, right?
08:52 dcs and yes, non-controllers nodes don't have external ip addresses
08:53 Wida joined #fuel
08:54 kozhukalov yes, if you change configuration on controllers and then add yet another compute node, and then click "deploy changes",  this manual configuration on computes should not be affected
08:54 kozhukalov sorry * this manual configuration on controllers should not be affected
08:58 vsedelnik joined #fuel
09:00 dcs kozhukalov: ok, and as much as i understood, i must add crm resource into pacemaker e.g. vip__vrouter_pub_X
09:02 kozhukalov let's ask deployment engineers. aglarendil, sbog, xenolog, could you please help with pacemaker related stuff? ^^
09:03 kozhukalov mwhahaha: mattymo ^^
09:09 dcs kozhukalov: thx, i left to think further....
09:12 obcecado joined #fuel
09:14 apalkina joined #fuel
09:14 vsedelnik joined #fuel
09:17 meow-nofer joined #fuel
09:18 e0ne joined #fuel
09:23 fatdragon joined #fuel
09:36 jmhale joined #fuel
09:52 bazilza joined #fuel
09:52 krogon joined #fuel
10:24 fatdragon joined #fuel
10:27 Liuqing joined #fuel
10:43 DevStok joined #fuel
10:43 DevStok hi
10:43 DevStok on the compute nova log i get
10:43 DevStok ImageNotFound: Image ea84cf4a-7cab-4526-aba0-39a917f0a369 could not be found.
10:43 DevStok so the vm failed to spawn
10:44 dcs joined #fuel
10:44 DevStok but the image has been uploaded successfully
10:47 DevStok the compute see the internal vip ip
10:47 DevStok ip:9292
10:53 mescanef joined #fuel
10:53 aglarendil_ joined #fuel
10:58 DevStok it's strange
10:58 DevStok with the last images uploaded I have no problem
10:58 DevStok but with the images uploaded at the first
10:58 DevStok I get error image not found
10:59 DevStok maybe wift doesnt appreciate my command : rm /srv/node/1/*
11:02 DevStok glance + swift set in this way is shitting bad
11:10 javeriak joined #fuel
11:14 Wida joined #fuel
11:18 askb joined #fuel
11:35 Liuqing joined #fuel
12:08 venkat_ joined #fuel
12:25 fatdragon joined #fuel
12:40 spyzalski joined #fuel
12:53 spyzalski joined #fuel
12:58 kozhukalov DevStok: when you run rm command you need to be aware what you do. although swift is distributed storage, it depends on replication factor if removal some files does not break data consistency. could you please contact swift team on #openstack-swift?
13:07 natarej joined #fuel
13:41 jmhale Hey all -- I have a question about deployment. I had a botched deployment and I'm trying to get the nodes back to a good state and bootstrap them again. When I PXE boot them, they all start with a "Boot sector signature not found (unbootable disk/partition?)". I've removed all nodes from the env in the UI. Any thoughts on where to look next?
13:44 kozhukalov jmhale: it should be enough to remove nodes from the cluster (via UI) and click "deploy changes". that is going to remove them from cobbler. then you can reboot them, they should boot with bootstrap image.
13:47 jmhale kozhukalov: if I remove all nodes from the env, there should there be no UUIDs listed in the cobbler container's /var/lib/tftpboot/pxelinux.cfg?
13:48 kozhukalov jmhale: if only you don't have other envs.
13:49 kozhukalov by default, when you don't have envs it is only /var/lib/tftpboot/pxelinux.cfg/default file should be there
13:50 jmhale Interesting. That's my problem then. I have a single env with no nodes, but all of the stale UUIDs are still there, in addition to default. When I removed the offending UUID, the node bootstraps normally.
13:50 jmhale I imagine that if I do the same to the rest of them, my remaining nodes will come back up
13:51 kozhukalov jmhale: i am not sure about the name 'default'
13:51 kozhukalov let me check
13:52 jmhale default is there for me (in 8.0, btw)
13:53 neilus joined #fuel
13:53 kozhukalov ahh, ok, then 'default' is the correct name. then please try to run 'dockerctl cobbler shell cobbler sync
13:55 kozhukalov or even restart cobbler container
13:55 jmhale running the sync now
14:00 DevStok left #fuel
14:04 vsedelnik joined #fuel
14:13 jmhale kozhukalov: ran the sync. Looks like it deleted all of the UUIDs, but then regenerated them
14:13 srwilkers_mobile joined #fuel
14:14 jmhale http://paste.openstack.org/show/492281/
14:17 kozhukalov jmhale: ok, could you please take a look at /var/log/docker-logs/astute/astute.log? there should be some info about interacting with cobbler
14:18 kozhukalov anyway, you can just use cobbler CLI to remove nodes manually
14:19 jmhale ok, that works. I'll remove them. Looking at astute.log as well, although nothing immediately jumps out at me
14:20 kozhukalov dockerctl cobbler shell cobbler system remove --name <name>
14:22 jmhale Yup, that did it. Although I had to modify the command: dockerctl shell cobbler cobbler system remove --name <name>
14:22 vsedelnik joined #fuel
14:23 dmellado joined #fuel
14:23 vsedelnik joined #fuel
14:27 fatdragon joined #fuel
14:31 jmhale Yeah, all of the nodes are bootstrapping normally now. Thanks for the assist!
14:32 kozhukalov jmhale: you are welcome
14:34 Reepicheep joined #fuel
14:36 claflico joined #fuel
14:37 xarses joined #fuel
14:40 fatdragon joined #fuel
14:48 vpleshakov joined #fuel
14:51 kozhukalov joined #fuel
14:53 jaypipes joined #fuel
15:03 ErikNeeley joined #fuel
15:24 askb_ joined #fuel
15:27 magicboiz joined #fuel
15:38 blahRus joined #fuel
15:46 angdraug joined #fuel
15:58 xarses joined #fuel
16:11 vsedelnik joined #fuel
16:25 Wida Hi! I redeploy Openstack 8.0 and try add IRONIC nodes. But I have tha same error when try 'ironic node-create'
16:26 Wida error: 'No valid host was found. Reason: No conductor service registered which supports driver fuel_ipmitool. (HTTP 400)'
16:27 Wida does anybody can help me?
16:34 elopez joined #fuel
16:39 mwhahaha Wida: I found, https://github.com/openstack/fuel-agent/blob/master/contrib/ironic/ironic-fa-deploy/README.rst
16:39 mwhahaha not sure if that helps
16:42 Wida mwhahaha: thx I try it.
16:42 Wida but why it not included in Mirantis Openstack image ?
16:43 mwhahaha there's supposed to be an ironic image thats uploaded i thought. but i've not played with ironic and fuel so i can't speak to it much
16:45 Wida and in Mirantis Openstack Docs, and Mirantis presentation video - nowhere nobody says about it
16:45 Wida mwhahaha: thx for help
16:46 mwhahaha sounds like we might have documentation gaps around this
16:48 jaypipes joined #fuel
16:48 dilyin joined #fuel
16:57 javeriak joined #fuel
17:02 fatdrago_ joined #fuel
17:04 elemoine- joined #fuel
17:05 Reepicheep_ joined #fuel
17:06 izinovik- joined #fuel
17:07 MiroslavAnashkin joined #fuel
17:07 mihgen joined #fuel
17:07 dhellmann joined #fuel
17:08 asvechnikov joined #fuel
17:08 vkramskikh joined #fuel
17:08 Kdecherf joined #fuel
17:08 akislitsky joined #fuel
17:09 jmhale joined #fuel
17:14 jaranovich joined #fuel
17:14 elemoine_ joined #fuel
17:16 holser_ joined #fuel
17:16 mmalchuk joined #fuel
17:17 smakar joined #fuel
17:17 igorbelikov joined #fuel
17:19 mkwiek joined #fuel
17:19 teran joined #fuel
17:19 zubchick joined #fuel
17:19 aglarendil joined #fuel
17:19 asaprykin joined #fuel
17:19 skath_ joined #fuel
17:21 warpc joined #fuel
17:22 dpyzhov joined #fuel
17:23 igorbelikov joined #fuel
17:28 ErikNeeley left #fuel
17:39 EZPZ joined #fuel
17:39 EZPZ left #fuel
17:39 EZPZ joined #fuel
17:47 e0ne joined #fuel
18:13 e0ne joined #fuel
18:23 javeriak joined #fuel
18:37 e0ne joined #fuel
18:43 fuel-slackbot joined #fuel
18:47 jaypipes joined #fuel
18:57 jaypipes_ joined #fuel
18:58 angdraug joined #fuel
19:03 e0ne joined #fuel
19:04 e0ne joined #fuel
19:15 akscram joined #fuel
19:43 Marcellin__ joined #fuel
19:43 Marcellin__ Hello, I created a patch to fix a bug about Fuel master deployment fails due to disk size. https://review.openstack.org/#/c/293054/ But I got a comment saying that it couldn't be merged to Master because we don't have docker. Please what should I do of that Patch?
19:49 mwhahaha good question let me take a look
19:52 mwhahaha i commented, you may want to poke Sergey for clarifications
19:59 Marcellin__ mwhahaha: Thank you!
21:00 kaliya joined #fuel
21:41 notmyname joined #fuel
21:43 notmyname I want to talk about the summit schedule. who would I talk to / what channel should I ask in?
21:50 mwhahaha anything in particular you want to talk about?
21:55 notmyname potential schedule swapping
22:01 mwhahaha that's a good question, angdraug kozhukalov mihgen might be able to assist
22:01 notmyname is he on IRC?
22:02 notmyname just wanted to discuss it directly before going to the ML thread
22:02 mwhahaha those were nicks of people here
22:02 notmyname heh, ok :-)
22:02 mwhahaha kozhukalov in europe so he's probably not arround. probably faster to do a ML
22:16 dslevin joined #fuel
23:05 xarses joined #fuel

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