Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2016-04-08

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

All times shown according to UTC.

Time Nick Message
00:11 xarses joined #fuel
01:11 cartik joined #fuel
01:18 fatdragon joined #fuel
01:19 cartik joined #fuel
01:26 richoid joined #fuel
01:48 ilbot3 joined #fuel
01:48 Topic for #fuel is now Fuel 8.0 (Liberty) https://www.fuel-infra.org/ | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel/
02:22 richoid1 joined #fuel
02:24 gongysh joined #fuel
02:52 richoid joined #fuel
04:20 javeriak joined #fuel
04:51 askb joined #fuel
04:57 fedexo joined #fuel
05:23 javeriak joined #fuel
05:25 reddy joined #fuel
05:36 krypto joined #fuel
05:38 krypto hello all,i have provisioned one of the fuel slave but slaves dont have puppet files synced from master.How can i forcefully tell slaves to sync puppet files from fuel master
05:55 javeriak_ joined #fuel
06:09 surinder_ joined #fuel
06:54 akaszuba joined #fuel
07:01 e0ne joined #fuel
07:04 byrdog55 joined #fuel
07:04 hyperbaba joined #fuel
07:06 hyperbaba Hi there , I just want to report a problem with freshly deployed liberty using fuel 8. I used ssl connection to endpoints. When I choose ceph for all storage related services, the public object-store endpoint is created with http and not https thus rendering object-store service unavailable.
07:07 hyperbaba i did change the endpoint url in the database to https and object store started working
07:08 criss2016 joined #fuel
07:17 hyperbaba Another point, why is there a disk files on compute node in /var/lib/nova/instances/_base when i am using rbd for ephemeral instances?
07:24 yottatsa joined #fuel
07:27 zefciu joined #fuel
07:39 hyperbaba joined #fuel
08:10 javeriak joined #fuel
08:34 javeriak_ joined #fuel
08:34 yassine_ joined #fuel
08:40 e0ne joined #fuel
08:42 yottatsa joined #fuel
08:45 javeriak joined #fuel
08:49 kozhukalov krypto, one of predeployment tasks is to sync puppet files from the master node, it is not a part of provisioning. you could use rsync command to push puppet files to slaves
08:52 kozhukalov hyperbaba: thanks a lot for this info about ssl, sbog could you please comment on this ssl related issue?
08:54 hyperbaba kozhakulov: Here is the thing. I've deployed liberty using ceph for all. Also checked out to use https for endpoint access. When system was deployed, the public endpoind for object-store stated http instead of https. I've looked into haproxy config for radosgw and haproxy expected ssl connection instead of plain http. By changing endpoint to https object-store started working (from horizon interface atleast)
08:55 kozhukalov hyperbaba: could you please check if ceph volume mounted to this directory /var/lib/nova?
08:55 hyperbaba kozhukalov: xml file states that drive is RBD:compute ...
08:56 hyperbaba kozhukalov: maybe it's a leftover from rbd-import because the image was qcow2?
08:57 hyperbaba kozhukalov: Nevertheless, it's a issue becuase /var/lib/nova is a small partition on compute.... the entire idea of ceph is not to have anything local on the compute disk
08:58 yottatsa joined #fuel
09:00 kozhukalov hyperbaba: i will check it out, thanks for the information
09:01 hyperbaba kozhukalov: while we are here can i ask you some other questions?
09:02 javeriak joined #fuel
09:03 kozhukalov hyperbaba: sure
09:05 hyperbaba kozhukalov: while configuring the environment, i could not find any VPNaaS and LBaaS services to "checkout". They existed in my previous 5.1 deployment. I had users using them. Now they are missing. I found that there is now concept of plugins for some services, but none for liberty?
09:06 hyperbaba kozhukalov: I need to put this cloud back in production ASAP and i am missing crutial services for my customers. Is there some workaround for those two sevices?
09:20 ArminderSingh joined #fuel
09:39 kozhukalov hyperbaba: we don't have so much resources to support all these plugins, as far as I know vpnaas for 8.0 will be available soon, but be are not planning to support lbaas plugin for 8.0
09:39 hyperbaba not even octavia?
09:39 hyperbaba kozhukalov: not even octavia?
09:40 hyperbaba kozhukalov: So i can add this plugin later on when available... Does this need a redeployment of the envronment or can be just added to the existing environment?
09:40 irinapov joined #fuel
09:41 kozhukalov irinapov could you please comment on this ^^
09:45 irinapov redeployment will be needed for the plugins for 8.0 as those are networking, not applications
09:45 irinapov I  mean, for VPNaaS or LBaaS
09:48 yottatsa joined #fuel
10:02 javeriak joined #fuel
10:20 e0ne joined #fuel
10:29 hyperbaba irinapov: Why is so. It's a simple installation , regardless of the service type. If it can be done mannualy on the existing platforms why entire redeployment is needed? In this way none of fuel deployed system can't be used in production. I allready had to kill my production 5.1 platform because upgrade was not possible to the newer. The entire idea of the cloud is continuity.....
10:32 javeriak joined #fuel
11:36 byrdog55_ joined #fuel
11:56 javeriak_ joined #fuel
12:04 yottatsa joined #fuel
12:05 javeriak joined #fuel
12:10 hyperbaba I had to change disks in one node from unalocated pool. but after reboot fuel does not see the change but shows the earlier state. How can I remedy that?
12:36 spyzalski joined #fuel
12:41 javeriak joined #fuel
12:42 yassine joined #fuel
12:58 e0ne joined #fuel
13:55 hyperbaba I've managed to resolve the disk issue and assigned the node as third controller.. Run the "deploy changes" and now the other controller is in state "installing openstack" stuck at 19% for the last hour
13:56 dslevin1 left #fuel
13:57 kozhukalov hyperbaba: did you take a look at /var/log/docker-logs/remote/node-X/puppet.log ?
13:57 kozhukalov are there any errors ?
13:58 kozhukalov also /var/log/docker-logs/astute/astute.log could give some clues
13:59 hyperbaba i don't see any errors
13:59 hyperbaba and i see that pupet.log on the named node is constanly populated with new lines
14:00 hyperbaba it's unstuck while we are talking
14:00 hyperbaba now there is "installing openstack" on all three controllers
14:02 kozhukalov about "day 2" operations (reconfiguring cluster without re-deployment), you are absolutely right, in 9.0 cluster settings tab will be unlocked and re-configuration will be possible
14:02 yassine joined #fuel
14:02 kozhukalov we re-implemented deployment engine
14:02 kozhukalov now it is fully data driven and task based
14:03 kozhukalov when we need to reconfigure something we just calculate diff between 2 graphs and run only those tasks which needed
14:04 kozhukalov so, in 9.0 the situation with life cycle management will be better
14:06 dslevin1 joined #fuel
14:16 Mike_G__ joined #fuel
14:23 jaypipes joined #fuel
14:25 javeriak joined #fuel
14:27 javeriak_ joined #fuel
14:29 shams joined #fuel
14:29 shams left #fuel
14:30 hyperbaba joined #fuel
14:46 koma joined #fuel
14:48 xarses joined #fuel
15:13 asvechnikov joined #fuel
15:15 claflico joined #fuel
15:38 javeriak joined #fuel
15:42 fatdragon joined #fuel
15:51 fedexo joined #fuel
15:54 javeriak_ joined #fuel
16:32 aleksandr_null joined #fuel
16:56 e0ne joined #fuel
17:12 xarses_ joined #fuel
17:15 e0ne joined #fuel
18:12 javeriak joined #fuel
18:22 e0ne joined #fuel
18:26 e0ne joined #fuel
18:36 javeriak joined #fuel
18:58 hyperbaba joined #fuel
19:03 LanceHaig joined #fuel
19:38 dslevin1 left #fuel
19:54 dslevin1 joined #fuel
20:07 bookwar joined #fuel
20:11 seeg_ joined #fuel
21:04 ppradhan joined #fuel
21:54 ppradhan left #fuel
23:56 krobzaur_ joined #fuel

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