Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel, 2014-08-19

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

All times shown according to UTC.

Time Nick Message
00:00 lordd_ joined #fuel
00:13 rmoe joined #fuel
01:12 xarses joined #fuel
01:38 mattgriffin joined #fuel
01:48 ilbot3 joined #fuel
01:48 Topic for #fuel is now Fuel 5.0.1 for Openstack: https://wiki.openstack.org/wiki/Fuel | Paste here http://paste.openstack.org/ | IRC logs http://irclog.perlgeek.de/fuel/
01:50 tatyana joined #fuel
01:58 fandi joined #fuel
02:04 wap joined #fuel
02:08 wap hi all, im trying fuel 5.0 now, with bonding
02:08 wap i found this error
02:09 wap http://ur1.ca/i03m7
02:09 wap AttributeError: 'NodeBondInterface' object has no attribute 'ip_addr'
02:10 wap maybe bonding still not working?
02:16 wap i also found this error when adding node with role
02:16 wap http://ur1.ca/i03nj
02:26 Krast joined #fuel
03:00 geekinutah joined #fuel
03:06 aleksandr_null joined #fuel
04:16 ArminderS joined #fuel
04:53 dpyzhov joined #fuel
05:21 e0ne joined #fuel
05:27 e0ne joined #fuel
05:32 wayneeseguin joined #fuel
05:34 e0ne joined #fuel
05:40 e0ne joined #fuel
05:50 e0ne joined #fuel
05:53 adanin joined #fuel
05:56 tatyana joined #fuel
05:58 e0ne joined #fuel
06:14 Krast joined #fuel
06:32 lordd_ joined #fuel
06:43 pasquier-s joined #fuel
06:48 kupo24z joined #fuel
06:50 e0ne joined #fuel
07:09 e0ne joined #fuel
07:09 ArminderS joined #fuel
07:10 HeOS joined #fuel
07:21 hyperbaba_ joined #fuel
07:26 Longgeek joined #fuel
07:36 ArminderS https://launchpad.net/fuel/+milestones is still showin 5.0.1 as "not yet released"
07:36 ArminderS probably someone need to put in milestone dates for next releases too
07:41 mihgen ArminderS: yep, thanks. Fuel devops will do it
08:05 Longgeek joined #fuel
08:06 Longgeek joined #fuel
08:08 homegrown joined #fuel
08:22 Krast joined #fuel
08:43 Longgeek joined #fuel
08:49 Longgeek joined #fuel
09:17 e0ne joined #fuel
09:39 nimion joined #fuel
09:58 adanin joined #fuel
10:00 bookwar joined #fuel
10:53 Longgeek joined #fuel
11:02 dshulyak joined #fuel
11:03 dshulyak wap: hi, about bug with bonds
11:04 dshulyak wap: you reproduced it on 5.0 or 5.0.1 version ?
11:15 e0ne joined #fuel
11:26 Longgeek joined #fuel
11:29 lordd_ joined #fuel
11:57 nimissa joined #fuel
11:57 wap hi dshulyak, are u still there?
11:57 wap im reproduce on 5.0
12:03 adanin joined #fuel
12:05 wap hi all, seems it known bug: https://bugs.launchpad.net/fuel/+bug/1328163
12:05 wap it says fixed in newer version of fuel
12:06 wap where i can download the fuel latest stable version of fuel?
12:06 wap fuel 5.1.1 maybe?
12:13 Arminder latest is 5.0.1 and you can get from here -> https://software.mirantis.com/terms-and-conditions/
12:33 hyperbaba__ joined #fuel
13:24 pasquier-s joined #fuel
13:25 racingferret joined #fuel
13:29 jaypipes joined #fuel
13:46 racingferret Hi guys, I've just tried to upgrade from 5.0.0 to 5.0.1 using the upgrade package, but it failed...
13:46 racingferret The output of the upgrade can be found here: http://paste.openstack.org/show/97311/
13:50 MiroslavAnashkin Sorry, what was the error message? Looks like the output is truncated.
13:51 racingferret yeah, I just noticed.  I'm trying pastebin... 1 sec.
13:52 racingferret this has the full log: http://pastebin.com/pcjxLL7B
13:53 MiroslavAnashkin Great, thank you!
13:53 evgeniyl racingferret: looks like docker is dead
13:54 evgeniyl racingferret: please, try to run `docker ps`
13:54 racingferret This is the first error:
13:54 racingferret 2014-08-19 13:51:50 ERROR 28134 (upgrade) DockerUpgrader: failed to upgrade: "UnixHTTPConnectionPool(host='localhost', port=None): Max retries exceeded with url: /run/docker.sock/v1.10/containers/24e3e91d97ddb99f492a39b03946dc31ea1da006a08224e1bcf63e3b757be34b/stop?t=10 (Caused by <class 'httplib.BadStatusLine'>: )"
13:54 racingferret all the previous docker containers seemed to upgrade without an issue.
13:55 racingferret hmm, docker isn't running
13:55 pasquier-s_ joined #fuel
13:55 evgeniyl racingferret: try to run it /etc/init.d/docker start
13:56 racingferret it died during the upgrade...
13:56 racingferret would it cause any problems running the upgrade again?
13:57 evgeniyl racingferret: yes, try to run docker manually as I described above
13:57 evgeniyl racingferret: and you will be able to run upgrade again
13:57 racingferret OK, I'll give it another go now...
13:57 evgeniyl racingferret: bug make sure that it's really running, try to run `docker ps` or some other command which uses docker's api
13:57 evgeniyl s/bug/but/
14:02 MiroslavAnashkin racingferret: Please check the docker first. Please run `docker ps` on dead master node :-)
14:04 racingferret MiroslavAnashkin: I have restarted docker and I'm re-running the upgrade.  I'll let you know how it goes.
14:09 geekinutah joined #fuel
14:31 mattgriffin joined #fuel
14:57 racingferret OK, still getting an error during the upgrade:
14:57 racingferret 2014-08-19 15:36:58 ERROR 14444 (upgrade) DockerUpgrader: failed to upgrade: "500 Server Error: Internal Server Error ("Cannot destroy container 12c2b560174ded55adf7badae3288a0aadd5fa62649c5275aec426d63ce3a921: Driver devicemapper failed to remove root filesystem 12c2b560174ded55adf7badae3288a0aadd5fa62649c5275aec426d63ce3a921: Error running removeDevice")"
14:58 xarses joined #fuel
14:58 racingferret seems it can't remove /var/www/nailgun from cobbler
15:00 racingferret I tried doing a "docker rm -f" on cobbler, but that fails also, since the first part of the upgrade copies from fuel-core-5.0-cobbler:/var/lib/cobbler/config to /var/lib/fuel_upgrade/5.0.1/cobbler_configs
15:00 racingferret potential ordering problem?
15:00 racingferret FYI, this a vanilla 5.0 install
15:01 kaliya racingferret: `dockerctl restart cobbler`
15:02 angdraug joined #fuel
15:06 mattymo racingferret, some volumes are mounted read only
15:06 mattymo docker rm -f removes a container
15:07 mattymo rm -f inside a container tries to remove files
15:07 mattymo dockerctl destroy cobbler would work. it runs a special umount command
15:13 ArminderS joined #fuel
15:17 pasquier-s joined #fuel
15:18 lordd_ joined #fuel
15:31 agordeev joined #fuel
15:33 e0ne joined #fuel
16:03 kat_pimenova joined #fuel
16:03 jkirnosova joined #fuel
16:05 homegrown left #fuel
16:09 rmoe joined #fuel
16:16 xarses joined #fuel
16:34 jobewan joined #fuel
16:40 dhblaz joined #fuel
16:46 e0ne joined #fuel
16:47 dhblaz When did fuel get the 10 node non-production use clause in the end user agreement?
16:47 dhblaz And is it acceptable to compile our own ISO and use it for produciton?
16:48 dhblaz Appears to be hanging
16:48 dhblaz oops wrong window
16:51 pasquier-s_ joined #fuel
16:53 e0ne joined #fuel
17:07 e0ne joined #fuel
17:18 xarses dhblaz: According to one of the biz guys I spoke with on our side is that the agreement has not changed, just that the new page makes it more clear.
17:18 dhblaz It was not part of the 3.x agreement
17:18 dhblaz As best I can remember
17:18 dhblaz xarses: thanks for asking
17:18 ArminderS- joined #fuel
17:20 angdraug joined #fuel
17:20 xarses dhblaz: according to who i spoke with its apparently been there since ~ October, but I'll ask around about it again
17:28 dhblaz xarses: I signed up for access 25 September 2013.  This was the first time since then I had to accept a new agreement.  I’m pretty sure I downloaded 5.0 without accepting the new agreement.  Thanks again for your help with this matter.
17:34 kupo24z xarses are there any major issues with 5.0.2 currently?
17:34 kupo24z I built an ISO for our internal testing but if its going to be changing majorly i'll wait
17:35 xarses kupo24z: not that I'm aware of, but I'm slightly behind on reviewing bugs
17:35 kupo24z xarses: any issues in 5.0.2 exist in the 5.0.1 tag correct?
17:38 xarses kupo24z: they would be in the 5.0.x series with 5.0.2 milestone
17:38 xarses correct
17:46 kupo24z are these rabbitmq errors normal in iso creation? http://pastebin.mozilla.org/6056176
17:47 dhblaz joined #fuel
17:52 xarses kupo24z: no they are not
17:54 kupo24z Anything I should check?
17:54 kupo24z It looks like it went on to make the ISO after that
17:56 nimion1 joined #fuel
18:28 youellet joined #fuel
18:45 e0ne joined #fuel
18:49 HeOS joined #fuel
19:05 kupo24z xarses: did the HA fixes get implemented into 5.1 yet? or do you have the launchpad link for it?
19:06 xarses kupo24z: for the most part, I thin the rabbit issue is still outstanding
19:09 kupo24z is that under mos or fuel?
19:09 kupo24z ah found it https://bugs.launchpad.net/mos/+bug/1341656
19:09 xarses at this point, everything is committed to the public repo
19:09 kupo24z Looks like it was built yesterday
19:10 kupo24z cool
20:14 geekinutah joined #fuel
21:10 e0ne joined #fuel
22:02 xarses joined #fuel
22:55 adanin joined #fuel
23:32 Bello2 joined #fuel
23:33 Bello2 hi guys, trying out fuel 5.0 in ha deployment. (3 controller, 3 storage, 2 compute). notice that some instances stale/unreachable when one compute is taken down.
23:34 Bello2 is this by design or was setup incorrectly. if it is, is there a way getting around this or recovering from this?
23:35 kupo24z Bello2: HA is only for the controller services, for HA compute/instances you need to use alternate methods
23:35 xarses Bello2: openstack doesn't have the DRS like feature of VMWare. It's currently your responsability to have application level HA between the compute nodes
23:35 Bello2 ah i see
23:36 xarses depening on your deployment configuration, you may be able to re-schedule, or even migrate a node after the compute fails
23:36 Bello2 oh? how do i do that
23:36 kupo24z xarses: is this related to the messaging issues or something different? this is an error from issueing a deployment http://pastebin.mozilla.org/6060195
23:36 Bello2 i did try nova evacuate/migrate but that does not seems to work
23:37 xarses Bello2: no, that likely wont work once the compute is down
23:37 kupo24z evacuate should work; http://docs.openstack.org/user-guide-admin/content/nova_cli_evacuate.html
23:37 xarses if you are using ceph+ephemeral disks, or boot from volume, there is a nova command to redeploy
23:38 Bello2 i am using ceph
23:38 xarses otherwise the root disk is ephemeral and migration wont work, but you can still re-instantiate
23:39 Bello2 xarses: i am using ceph, maybe what i need is to boot from volume?
23:39 Bello2 and what would be that command to redeploy?
23:40 xarses kupo24z: it appears to be a rabbitmq issue yes, It looks to be related but might be worth creating a bug for if you can upload the support bundle
23:41 kupo24z xarses: network verify failed as well, same Socket Error
23:42 xarses Bello2: I don't have a running deployment at the time to compare, but it sounds like redeploy or reschedule reprovision or something
23:42 Bello2 ok will find it, thanks xarses
23:42 Bello2 also, do you mind explaining a little bit more about using volume?
23:44 kupo24z Bello2: Evacuation using Ceph should be fixed in 6.0-6.1 with Juno, https://bugs.launchpad.net/nova/+bug/1340411 & https://bugs.launchpad.net/nova/+bug/1250751 FYI
23:45 Bello2 kupo24z: is that included in fuel 5.0?
23:45 kupo24z No
23:46 kupo24z We had to do a workaround see comment on first report
23:46 Bello2 gotit, i guess my first step is to make sure i use this rbd for my instances
23:48 kupo24z xarses: should that bug be created under fuel or mos?
23:48 xarses kupo24z: fuel is fine, we will move it if needed
23:50 Bello2 so just to be clear. in order for me to be able to evacuate, i have to create a volume and make sure the instance is boot from it that way it will exist in the storage instead of the compute node?
23:53 kupo24z xarses: Can't create diag snapshot, says socket closed :(
23:53 xarses o.0
23:54 xarses da heck
23:54 xarses was that 500 from fuel?
23:54 kupo24z It was in the 'show logs' link in the deployment error
23:54 xarses ah, yes it is
23:54 xarses glossed over
23:55 xarses thats a seperate error, you rabbit instance in the fuel master is broken
23:55 kupo24z Bello2: Both ceph ephemeral and ceph volume should be able to start up after a host is down
23:56 kupo24z ah yes it is
23:56 kupo24z http://pastebin.mozilla.org/6060453
23:56 xarses use supervisorctl docker-rabbit restart
23:57 xarses if that doesn't get it going, then ya we very much should file a bug for this
23:58 kupo24z is it supervisorctl restart docker-rabbit?
23:58 xarses likely
23:58 kupo24z docker-rabbit: ERROR (no such process)
23:58 xarses hehe
23:58 xarses having errors in the puppet log for fuel is bad, sorry i didn't notice this earlier
23:59 kupo24z ah its docker-rabbitmq

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