Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-09-28

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

All times shown according to UTC.

Time Nick Message
00:29 xarses joined #fuel
00:42 natarej__ joined #fuel
00:44 Reepicheep joined #fuel
00:46 ipsecguy joined #fuel
01:19 ipsecguy joined #fuel
01:48 ilbot3 joined #fuel
01:48 Topic for #fuel is now Fuel 9.0 (Mitaka) https://www.fuel-infra.org/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel/
01:50 ipsecguy joined #fuel
01:58 fatdragon joined #fuel
02:29 natarej__ joined #fuel
02:36 ipsecguy joined #fuel
03:35 Zer0Byte__ joined #fuel
03:45 ipsecguy joined #fuel
04:06 failart_ joined #fuel
04:22 raunak joined #fuel
06:51 vkulanov joined #fuel
07:02 [HeOS] joined #fuel
07:17 cartik joined #fuel
08:11 fatdragon joined #fuel
08:25 vkulanov ikalnytskyi, akasatkin can you please review (fuelclient) backport https://review.openstack.org/#/c/377971/
08:26 ekosareva joined #fuel
08:29 code-R joined #fuel
08:31 code-R_ joined #fuel
08:58 Julien-z_ joined #fuel
09:12 fatdragon joined #fuel
09:38 code-R joined #fuel
09:44 manik joined #fuel
09:44 manik Hai frnds
09:44 manik I have one doubt
09:44 manik could you please clarify
09:45 manik if I need to generate password in plugins
09:45 manik hw can I do that
09:49 code-R_ joined #fuel
10:01 mskalski manik: https://github.com/openstack/fuel-plugin-tacker/blob/master/environment_config.yaml#L4-L7
10:03 manik mskalski: tq :)
10:29 vkulanov joined #fuel
11:02 vkramskikh folks, who can i ask about keystone on master node? i'm trying to make Fuel UI to use API v3, but cannot authenticate - I get 401 (using v2 API I authenticate fine). probably user and project domains are not "default"?
11:03 vkramskikh the same request for controller issues a token without any problems
11:03 vkramskikh examples from http://docs.openstack.org/developer/keystone/api_curl_examples.html also don't work with master node keystone
11:12 ddmitriev joined #fuel
11:13 fatdragon joined #fuel
12:02 ekosareva joined #fuel
12:08 mskalski vkramskikh: on fuel master I see that default_domain_id is set to non-standard value: grep default_domain_id /etc/keystone/keystone.conf
12:08 vkramskikh mskalski: thanks! is it unique for every master node installation?
12:30 vkramskikh i see it is
12:37 Julien-zte joined #fuel
12:37 mskalski vkramskikh: do you found where it is generated?
12:37 vkramskikh mskalski: no I didn't, i just check the value on another master node and it was different
12:39 mskalski vkramskikh: ok, but as far as I understood it is only fallback for clients who are not aware of domains, domains are simply not configured on fuel master
12:41 vkramskikh still didn't find a way to get a token just with user/pass/project - without providing domains to v3 I get http 400 with "Expecting to find domain in project - the server could not comply with the request since it is either malformed or otherwise incorrect."
12:45 aarefiev joined #fuel
12:49 failart_ joined #fuel
12:55 mskalski vkramskikh: when I look at the /var/log/puppet/bootstrap_admin_node.log I see that 'fuel' domain should be created
12:57 vkramskikh mskalski: doesn't work, I don't see it in keystone db also
13:04 byrdog55 joined #fuel
13:08 Julien-zte joined #fuel
13:14 fatdragon joined #fuel
13:21 Damjanek left #fuel
13:24 Julien-zte joined #fuel
13:44 dmeltsaykin joined #fuel
13:45 dmeltsaykin fuelclient fellows, please review https://review.openstack.org/#/c/378610/ - fixes tests
13:52 GMAzrael joined #fuel
13:53 GMAzrael pasquier-s: any lucky since yesterday?
14:03 GMAzrael LMA Folks: http://paste.openstack.org/show/583305/
14:07 ddmitriev joined #fuel
14:24 Julien-zte joined #fuel
14:25 xarses joined #fuel
14:27 mskalski vkramskikh: https://gist.github.com/michalskalski/072e9bac392d4bc8ccb49e7be9df2447 nailgun password is in hiera: hiera keystone
14:35 vkramskikh mskalski: that's cool, but what i'm trying to do is to make fuel-ui use keystone v3 api instead of v2. fuel-ui should use "admin" user, and it doesn't seem to work with "fuel" domain
14:36 mskalski vkramskikh: you just need to replace nailgun with admin, and services with admin
14:37 mskalski vkramskikh: {"auth": {"scope": {"project": {"domain": {"name": "fuel"}, "name": "admin"}}, "identity": {"password": {"user": {"domain": {"name": "fuel"}, "password": "admin", "name": "admin"}}, "methods": ["password"]}}}
14:37 mskalski vkramskikh: but user can change admin password
14:38 AlexeyAbashkin joined #fuel
14:38 vkramskikh oh! it works
14:38 vkramskikh I wonder what I was doing wrong
14:38 vkramskikh thanks, I'll try to figure it out
14:39 fatdragon joined #fuel
14:40 vkramskikh oh I see
14:40 vkramskikh I was trying to specify domain id instead of domain name
14:41 mskalski yeach and default_domain_id is actually id of fuel domain
15:06 pasquier-s GMAzrael: regarding you last paste, pls see https://bugs.launchpad.net/lma-toolchain/+bug/1616765
15:08 GMAzrael pasquier-s: seems like my single node LMA is not binding to the VIP's. Can't ping or curl those addresses, as detailed in the tgz I sent yesterday
15:13 GMAzrael pasquier-s: maybe similar to this? https://bugs.launchpad.net/lma-toolchain/+bug/1594722?
15:14 pasquier-s GMAzrael: can you paste the output of "netstat -apn" from the LMA node?
15:18 GMAzrael http://paste.openstack.org/show/583331/
15:19 pasquier-s GMAzrael: you've got haproxy running in the main namespace, that's not correct
15:19 GMAzrael I blame puppet :D
15:19 GMAzrael I haven't touched any of the deployment
15:19 GMAzrael How would I change it?
15:20 pasquier-s GMAzrael: what does "/etc/init.d/haproxy status" say?
15:21 GMAzrael running
15:22 pasquier-s pls run "/etc/init.d/haproxy stop"
15:22 GMAzrael done
15:23 xek__ joined #fuel
15:23 pasquier-s can u check if it works now?
15:25 GMAzrael Error Connection Reset
15:25 GMAzrael same as before
15:27 pasquier-s pls run "crm resource restart p_haproxy" just to be sure
15:29 GMAzrael http://paste.openstack.org/show/583333/
15:32 pasquier-s oh, something's wrong with Pacemaker... Not good.
15:32 GMAzrael its a default deploy....so I am pretty lost
15:34 pasquier-s can you check the pacemaker status ( /etc/init.d/pacemaker status)?
15:35 GMAzrael root@node-19:~# service pacemaker status
15:35 GMAzrael pacemakerd is stopped
15:35 GMAzrael root@node-19:~# service pacemaker start
15:36 GMAzrael Starting Pacemaker Cluster Manager[  OK  ]
15:36 GMAzrael root@node-19:~# crm resource restart p_haproxy
15:36 GMAzrael INFO: ordering p_haproxy to stop
15:36 GMAzrael No messages received in 30 seconds.. aborting
15:36 GMAzrael WARNING: can't find DC
15:36 pasquier-s lets give a bit of time to Pacemaker for getting ready
15:37 GMAzrael root@node-19:~# crm resource restart p_haproxy
15:37 GMAzrael INFO: ordering p_haproxy to stop
15:37 GMAzrael INFO: ordering p_haproxy to start
15:37 GMAzrael still unable to reach the nagios vip
15:46 Vijayendra joined #fuel
16:01 cartik joined #fuel
16:22 kur1j joined #fuel
16:23 kur1j does Mirantis support Openstack, end to end?
16:23 kur1j so if we have GPUs that we need to support will they support that to get it to work within OS?
16:36 GMAzrael From Mirantis sales "we are only servicing clients with 100+ nodes"
16:37 GMAzrael unless you buy their full service rack
16:38 GMAzrael At least, that is what we were told about getting support for our 11 node cluster
16:39 kur1j so basically if we have 100+ nodes we can't even get service?
16:39 GMAzrael support
16:40 kur1j whats the full service rack?
16:41 GMAzrael "unlocked appliance" on their website
16:41 GMAzrael w
16:42 kur1j not sure what that even means
16:44 GMAzrael https://www.mirantis.com/products/unlocked-appliances/
16:45 GMAzrael basically you buy the rack, servers, switches preconfigured by Mirantis to drop into your space
16:46 kur1j ah, so basically couldn't get GPUs anyway because they probably don't offer them
16:48 GMAzrael If you are spending the money on their appliance, you might be able to get them to support the cards.
16:50 kur1j well unforunately we have a GPU appliance from Cirrascale so I highly doubt they will be wanting to support that.
16:50 kur1j I'm running into what I think are bugs in libvirt with not releasing the GPUs correctly
16:56 code-R joined #fuel
16:58 raunak joined #fuel
17:02 Zer0Byte__ joined #fuel
17:35 xarses joined #fuel
18:07 Vijayendra joined #fuel
18:13 GMAzrael pasquier-s: any more thoughts?
19:02 ekosareva joined #fuel
19:03 failart_ joined #fuel
20:10 kbaegis joined #fuel
20:11 kbaegis Hi all.  I have a fuel 9.0 master who is failing on deploy. Passes the networking test, but I get a "provision has failed to execute hook" and a string of repos
20:11 kbaegis Anyone seen this before?
20:11 kbaegis Is the repo generation now a requirement?
20:16 mwhahaha check the image building log in the master
20:17 raunak joined #fuel
20:24 kbaegis mwhahaha: Thanks
20:24 kbaegis mwhahaha: Where is that log?
20:25 kbaegis found it
20:25 kbaegis nvm
20:26 kbaegis There's a gpg error . . .
20:26 kbaegis mirror.fuel-infra.org holdback is signed with an invalid key
20:27 kbaegis Scratch that, the key is unpublished
20:27 kbaegis There were also issues with auth for init-system-helpers and dh-python
20:31 kbaegis Nothing seems critical
20:49 mwhahaha so if the image built the next question would be if the nodes were provisioned correctly, are they still on the bootstrap via their console?
21:34 failart_ joined #fuel
22:14 kbaegis Yes, they were stuck on the discovery/bootstrap image
22:14 kbaegis Creating a local repo fixed the issue. Possibly a slow connection
22:24 kbaegis Having an issue with Ceph
22:25 kbaegis Failed Task[enable_rados/1]
22:28 kbaegis Anyone seen this before? :/
22:28 kbaegis Not on slack atm
22:35 kbaegis joined #fuel
22:40 kbaegis joined #fuel
22:52 kbaegis joined #fuel
23:01 raunak joined #fuel
23:20 raunak joined #fuel
23:56 rmoe joined #fuel

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