Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel-dev, 2014-07-30

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

All times shown according to UTC.

Time Nick Message
00:36 rmoe joined #fuel-dev
04:07 adanin joined #fuel-dev
04:16 Longgeek joined #fuel-dev
04:18 Longgeek joined #fuel-dev
04:19 Longgeek joined #fuel-dev
04:49 xarses joined #fuel-dev
05:15 dshulyak_ joined #fuel-dev
05:34 IvanBerezovskiy joined #fuel-dev
05:49 e0ne joined #fuel-dev
06:19 sambork joined #fuel-dev
06:29 e0ne joined #fuel-dev
06:46 taj joined #fuel-dev
06:46 taj joined #fuel-dev
06:48 salmon_ joined #fuel-dev
07:04 pasquier-s joined #fuel-dev
07:20 mihgen aglarendil: around?
07:21 mihgen aglarendil: is it " /Stage[main]/Neutron::Agents::Ml2::Ovs/Service[neutron-ovs-agent-service] (err): Could not evaluate: resource p_neutron-openvswitch-agent not found" - known issue?
07:21 aglarendil where did you find it ?
07:21 mihgen centos latest bvt
07:21 mihgen can file bug about it..
07:22 aglarendil well, it could happen because other dependencies failed
07:22 mihgen could it be caused by service neutron-server restart' returned 143 ?
07:22 aglarendil centos bvt is failing due to race condition in keystone provider - this can be a dependency for neutron agents
07:23 mihgen didn't notice anything about keystone failures
07:23 aglarendil fix was merged, I guess
07:23 aglarendil let me look
07:23 mihgen aglarendil: it was
07:24 mihgen aglarendil: but after bvt run :)
07:24 aglarendil mihgen: yep
07:24 mihgen aglarendil: well I think you better or someone from your team to take a look into logs still..
07:24 mihgen might be some other issue
07:24 aglarendil centos will also continue failing due to neutron bug
07:25 mihgen aglarendil: I actually partly confirmed another issue, https://bugs.launchpad.net/fuel/+bug/1348839, latest ubuntu bvt is likely failed because of this
07:26 mihgen there were mcagent retries, it was slow env
07:26 aglarendil mihgen: are you sure that load average was at least around 20 there
07:26 aglarendil mihgen: it may be due to the load of the env
07:26 aglarendil mihgen: did you see cib or crmd process consuming all the CPUs?
07:27 aglarendil mihgen: btw centos bvt is failing due to neutron bug
07:27 mihgen aglarendil: could be load of env, I've provided my comment here about it: https://bugs.launchpad.net/fuel/+bug/1350016/comments/2
07:27 mihgen aglarendil: I didn't have live env to take a look about CPU, it was from logs
07:27 mihgen so I think we just need to take a closer look
07:27 mihgen I hope it's just env
07:27 aglarendil mihgen: I am sure that it is an env
07:28 aglarendil mihgen: centos bvt failed because of https://bugs.launchpad.net/fuel/+bug/1350045
07:29 mihgen aglarendil: if you know it's mos, we need to file tickets against mos, not fuel..
07:30 mihgen you can easily change it to: affect mos, deaffect fuel
07:30 aglarendil I am not sure it's mos. I just assigned it to mos-neutron to take care of it
07:30 aglarendil if they confirm it is their's bug, I will deaffect Fuel
07:32 mihgen ok. what about patch https://review.openstack.org/#/c/110140/1 - does it affect anything at all?
07:34 aglarendil it's useless. I abandoned it
07:51 adanin joined #fuel-dev
07:54 e0ne joined #fuel-dev
07:57 al_ex11 joined #fuel-dev
07:57 sambork joined #fuel-dev
08:02 skolekonov joined #fuel-dev
08:08 bookwar joined #fuel-dev
08:09 sambork joined #fuel-dev
08:14 ylobankov joined #fuel-dev
08:52 e0ne joined #fuel-dev
09:08 tsufiev hi, guys! how is called log file of web server which serves Fuel Web UI on master node?
09:10 evgeniyl tsufiev: static files are served by nginx, if you need api logs, then you need nailgun service
09:13 tsufiev evgeniyl, I suppose I need nailgun logs. Is it at /var/log/docker-nailgun.log?
09:14 evgeniyl tsufiev: not really, it's a log file which use supervisor to log message when it runs container, what you need is /var/log/docker-logs/nailgun/api.log or /var/log/docker-logs/nailgun/app.log
09:15 tsufiev evgeniyl, thank you
09:15 tnurlygayanov joined #fuel-dev
09:15 tnurlygayanov hi there!
09:15 evgeniyl tsufiev: your are welcome
09:16 evgeniyl s/your/you
09:16 artem_panchenko joined #fuel-dev
09:16 tnurlygayanov probably someone could know the reason of this error during the deployment: http://paste.openstack.org/show/89114/
09:25 shauno joined #fuel-dev
09:27 ogelbukh hi
09:30 ogelbukh is it known bug that iptables configuration on compute nodes prevents block migration (at least under centos), even though otherwise it is correctly enabled everywhere?
09:31 prmtl_ joined #fuel-dev
09:31 Longgeek_ joined #fuel-dev
09:37 adanin joined #fuel-dev
09:44 rshon_ joined #fuel-dev
09:44 aedocw_ joined #fuel-dev
09:45 brain461 joined #fuel-dev
09:45 brain461 left #fuel-dev
09:45 jkirnosova joined #fuel-dev
09:46 kat_pimenova joined #fuel-dev
09:55 mihgen ogelbukh: hi Oleg
09:55 mihgen I've heard about some bug regarding it, I think different one
09:55 mihgen can you please create bug and we will get someone to take a look?
09:56 mihgen tnurlygayanov: holser might be able to help with http://paste.openstack.org/show/89114/
10:02 ogelbukh mihgen: hi )
10:02 e0ne joined #fuel-dev
10:02 ogelbukh mihgen: sure, I will
10:03 ogelbukh just trying to understand if there are other reasons for t to fail
10:14 tnurlygayanov ping holser :)
10:16 ogelbukh mihgen: you may be talking about this one: https://bugs.launchpad.net/nova/+bug/1334164
10:17 mihgen ogelbukh: nope. I don't really know if this is what you hit. What version are you on?
10:18 ogelbukh 5.0
10:18 ogelbukh I already hit both )
10:28 igoryozhikov left #fuel-dev
10:37 brain461 joined #fuel-dev
10:52 sambork joined #fuel-dev
11:07 bogdando show should I run fuel-astute rspecs?
11:08 bogdando I cannot install 'raemon', '0.3.0' with bundler, looks like this gem is not in https://rubygems.org/ :(
11:08 bogdando Could not find gem 'raemon (= 0.3.0) ruby', which is required by gem 'astute (>= 0) ruby', in any of the sources.
11:09 bogdando found it ... https://github.com/stackforge/fuel-web/blob/master/docs/develop/env.rst
11:10 dshulyak_ joined #fuel-dev
11:15 ikalnitsky joined #fuel-dev
11:21 al_ex11 joined #fuel-dev
11:50 al_ex11 joined #fuel-dev
11:58 evgeniyl warpc_: ^
11:59 warpc_ bogdando: evgeniyl; yes, in doc we have instruction about it
12:00 evgeniyl warpc_: "Could not find gem 'raemon (= 0.3.0) ruby',"
12:01 warpc_ evgeniyl: yes, exactly this case
12:01 evgeniyl warpc_: oh, sorry, now I see.
12:21 aglarendil neutron duplicate agent bug has been triaged :-)
12:23 dpyzhov aglarendil: please raise priority for https://bugs.launchpad.net/fuel/+bug/1342172 and https://bugs.launchpad.net/fuel/+bug/1333828 or move them to 6.0
12:26 warpc_ evgeniyl: no problem. I asked author to build it and upload to rubygems, but he did not answer
12:35 sambork joined #fuel-dev
12:40 odyssey4me joined #fuel-dev
12:42 prmtl joined #fuel-dev
12:42 akasatkin joined #fuel-dev
13:10 IvanBerezovskiy left #fuel-dev
13:18 jobewan joined #fuel-dev
13:39 sambork joined #fuel-dev
13:40 prmtl joined #fuel-dev
13:46 al_ex11 joined #fuel-dev
13:50 dshulyak_ joined #fuel-dev
13:55 al_ex11 joined #fuel-dev
14:02 evgeniyl warpc_: are you around? https://bugs.launchpad.net/fuel/+bug/1348331 what else we need to do to solve this ticket?
14:04 al_ex11 joined #fuel-dev
14:05 warpc_ evgeniyl: I will in net at 20.00 and write you
14:05 evgeniyl warpc_: ok, thanks
14:14 prmtl_ joined #fuel-dev
14:17 ikalnitsky evgeniyl: warpc_: as for this bug - https://bugs.launchpad.net/fuel/+bug/1348331 - maybe we should change description? because our tarball contains this images, but in unusual path. still, we need to fix our kickstarts to use them, not the defaults one.
14:18 al_ex12 joined #fuel-dev
14:21 evgeniyl ikalnitsky: I thought warpc_ already fixed it, and it seems to work https://github.com/stackforge/fuel-library/commit/368ac101f36bf954921bc17c66bf763deb3faa5c
14:21 taj joined #fuel-dev
14:21 evgeniyl ikalnitsky: here is patch for astute https://github.com/stackforge/fuel-astute/commit/63c88288231b9ec04d942f3a2004dfe51d216e61
14:41 ikalnitsky evgeniyl: these patches are intended to use new repos during provision. for instance, we have 2014.1.1-5.1 and we want to deploy a NEW environment. Before this patches, the mcagents/netchecks/etc was installed from the default repo (/var/www/nailgun/centos/...), not the /var/www/nailgun/2014.1.1-5.1/centos/...
14:43 ikalnitsky evgeniyl: so we still need a patch to use new images, not the old one. currently, the path to images is hardcoded in cobbler, so we need to change this cobbler setting too.
14:44 evgeniyl ikalnitsky: oh, I got it, thanks, can we rename this ticket then?
14:45 adanin joined #fuel-dev
14:56 angdraug joined #fuel-dev
15:08 ikalnitsky evgeniyl: done
15:13 evgeniyl ikalnitsky: thanks!
15:26 dshulyak_ joined #fuel-dev
15:49 al_ex12 joined #fuel-dev
16:00 shauno__ joined #fuel-dev
16:01 e0ne joined #fuel-dev
16:09 rmoe joined #fuel-dev
16:14 akasatkin dpyzhov: what team is responsible for MOS components? There is a bug on ceilometer: https://bugs.launchpad.net/fuel/+bug/1350429
16:14 akasatkin *on running ceilometer
16:17 dpyzhov akasatkin: для начала library, вдруг там проблема в манифестах
16:18 akasatkin dpyzhov: ok
16:28 taj joined #fuel-dev
16:29 jaypipes joined #fuel-dev
16:55 warpc_ evgeniyl: ikalnitsky: yes, everything is exactly as ikalnitsky said
17:07 adanin joined #fuel-dev
17:11 adanin joined #fuel-dev
17:19 tatyana joined #fuel-dev
17:37 angdraug joined #fuel-dev
17:42 xarses joined #fuel-dev
17:44 brain461 joined #fuel-dev
17:48 Longgeek joined #fuel-dev
17:55 angdraug xarses: https://review.openstack.org/110168
17:56 tatyana_ joined #fuel-dev
18:23 xarses angdraug: +1
18:38 aglarendil hey guys. two important fixes: https://review.openstack.org/#/c/110710/4 and https://review.openstack.org/#/c/110745/2
18:38 aglarendil for parallel deployment - do not believe in galera state transfer mech too much ))
18:38 aglarendil angdraug: xarses rmoe ^^
18:41 aglarendil https://review.openstack.org/#/c/110699/ and this one also
18:42 aglarendil angdraug: xarses ^^
18:42 salmon_ aglarendil: https://review.openstack.org/#/c/110745/2 what a ugly python code...
18:42 salmon_ give me 10 minutes I will rewrite it in normal python :)
18:43 aglarendil salmon_: okay)
18:43 aglarendil salmon_: what is the best way? using iteritems or what?
18:43 aglarendil salmon_: I failed to find the better way
18:44 aglarendil salmon_: how can I pick each N elements from an array even if array length may be not dividable by N ?
18:44 aglarendil salmon_: and for each part I need to call a generator for priorities
18:47 Longgeek joined #fuel-dev
18:52 e0ne joined #fuel-dev
18:53 angdraug salmon_: +1, looking forward to a patch set from you that's readable :)
18:53 angdraug aglarendil: why do you have to do +0 in https://review.openstack.org/#/c/110710/6/deployment/puppet/galera/templates/wsrep.cnf.erb?
18:53 aglarendil for puppet's sake
18:53 salmon_ aglarendil: wait a moment
18:54 aglarendil angdraug: or it will consider it as a string
18:54 angdraug aglarendil: then you're doing it wrong
18:54 angdraug 'abc' * 2 in ruby returns 'abcabc'
18:54 bookwar joined #fuel-dev
18:54 angdraug '12' * 2 will be '1212'
18:54 aglarendil angdraug: nope. it will be number
18:55 aglarendil +0 makes it a number
18:55 aglarendil just run
18:55 aglarendil puppet  apply -e '$a=min(max($::physicalprocessorcount * 2 + 0, 4 + 0), 12 + 0) notify{"$a":}'
18:56 adanin joined #fuel-dev
18:56 salmon_ aglarendil: the unit test is failing for this code https://fuel-jenkins.mirantis.com/job/verify-fuel-web/1048/testReport/junit/nailgun.test.integration.test_orchestrator_serializer/TestNovaOrchestratorHASerializer/test_set_deployment_priorities/
18:57 salmon_ {'priority': 1000, 'role': 'other'} vs  expected {'priority': 1100, 'role': 'other'}
18:57 angdraug angdraug@x1:~$ puppet  apply -e '$a=min(max($::physicalprocessorcount * 2, 4), 12) notify{"$a":}'
18:57 angdraug Notice: /Stage[main]/Main/Notify[4]/message: defined 'message' as '4'
18:58 angdraug looks like *2 by itself is enough for puppet to coerce it to int
18:59 aglarendil salmon_: sorry, I was stupid and copy-pasted wrong string
18:59 aglarendil salmon_: fixed now
18:59 aglarendil salmon_: there is actualy no builtin function in python to split array into chunks of n elements
19:00 salmon_ aglarendil: I know
19:00 aglarendil salmon_: and what is your suggestion how to do it ?
19:01 salmon_ aglarendil: I need working tests first and patch 3 is still failing
19:03 aglarendil salmon_: where do you see it failing ?
19:03 salmon_ aglarendil: you can test is quickly with nosetests nailgun/nailgun/test/integration/test_orchestrator_serializer.py:TestNovaOrchestratorHASerializer.test_set_deployment_priorities_many_cntrls
19:03 salmon_ aglarendil: I test it localy
19:06 artem_panchenko joined #fuel-dev
19:10 aglarendil salmon_: my bad with arithmetics. I really feel ashamed
19:10 aglarendil patchset 4
19:11 salmon_ aglarendil: don't worry, it's late :)
19:11 xarses salmon_: did you have a better way to do 110745?
19:12 salmon_ xarses: I belive I have, testing now
19:13 xarses i think i have an option too (in my head)
19:13 salmon_ xarses: 5min
19:14 aglarendil salmon_: xarses it is not really urgent right now. we do not deploy more than 3 controllers. just in case someone firex big cluster deployment
19:16 xarses I know, Its a nice code challenge
19:16 salmon_ aglarendil: 4 is still failing
19:18 aglarendil salmon_: okay. gonna install fuel-web testing venv and test it by myself until it works :-)
19:23 aglarendil salmon_: angdraug do you know what the hell is this: UnicodeDecodeError: 'ascii' codec can't decode byte 0xd0 in position 19: ordinal not in range(128) ?
19:23 aglarendil I am just running nosetests
19:23 salmon_ aglarendil: it's python hell :)
19:23 aglarendil setting LANG to POSIX does not help
19:23 salmon_ aglarendil: xarses my solution http://pastebin.com/jumpAAfV
19:24 angdraug salmon_: you can just upload a new patchet into aglarendil's review
19:24 angdraug patchset even
19:24 xarses oh, I think mine might be simpler
19:24 salmon_ angdraug: I know, but I'm waiting for working tests to test it first
19:24 salmon_ xarses: show it :)
19:25 xarses working it out still
19:25 salmon_ ok
19:25 salmon_ aglarendil: where do you get this error?
19:26 aglarendil salmon_: the problem was old nose
19:26 aglarendil salmon_: I updated it
19:26 salmon_ ah, ok
19:26 aglarendil salmon_: and now it is trying to setup some stuff in DB which I do not need
19:27 salmon_ aglarendil: nose?
19:27 aglarendil salmon_:  nosetests
19:28 salmon_ I think integration tests are using db
19:33 aglarendil salmon_: yep. I have the fix
19:33 aglarendil salmon_: do you need it ?
19:34 aglarendil salmon_: I forgot that python does not catch the last index of slice
19:36 salmon_ aglarendil: push it, I will check it with my version
19:37 e0ne joined #fuel-dev
19:38 xarses its not workin in tests but heres mine http://paste.openstack.org/show/89229/
19:41 aglarendil it is working in tests
19:41 aglarendil okay. you are talking about yours
19:41 xarses aglarendil: yes
19:41 salmon_ xarses: I like yours more
19:41 salmon_ less code
19:42 aglarendil angdraug: why did not you say me that I am an idiot writing down puppet code in erb file ?
19:43 xarses ok, fixed mine it's passing the single test now http://paste.openstack.org/show/89230/
19:44 xarses oops left debug in the tests http://paste.openstack.org/show/89231/
19:52 aglarendil angdraug: I uploaded correct version of galera patch
19:52 aglarendil angdraug: merge it please after it passes CI
19:53 aglarendil and then we will need to rebuild mirrors
19:53 angdraug yup
19:53 aglarendil they will contain fix for neutron ml2 migration
20:03 xarses salmon_: I like where you went with the generator
20:04 salmon_ xarses: it's over kill for this problem
20:05 angdraug xarses: https://review.openstack.org/110602
20:05 e0ne joined #fuel-dev
20:21 xarses aglarendil: care if i push my version for 110745?
20:21 xarses angdraug: https://review.openstack.org/#/c/110710/
20:25 e0ne joined #fuel-dev
20:29 aglarendil xarses: I do not care.
20:30 aglarendil angdraug: xarses why did you merge this request while CI is still running ?
20:33 angdraug oops
20:33 angdraug confused it with another of your commits that did pass ci
20:34 angdraug no worries, I'll revert it if it fails ci
20:43 xarses aglarendil: angdraug https://review.openstack.org/#/c/110745
20:47 angdraug I don't like how sec_controller_priority initialization implicitly depends on first index returned by enumerate being 0
20:48 aglarendil it seems that galera fix is going to fail CI
20:48 aglarendil let's revert it and revert the revert to bring it back into review queue
20:49 aglarendil angdraug: ^^
20:50 angdraug let it fail first
20:52 angdraug I want to see the diagnoastics snapshot before I revert it
20:54 aglarendil okay. it failed
20:55 aglarendil angdraug: log is full of errors of galera not starting
20:58 angdraug yup
20:58 angdraug and of course we don't have galera config in the snapshot
21:01 aglarendil yep
21:01 aglarendil let's revert first
21:03 angdraug I don't see any evidence in the snapshot that it failed because wsrep.cnf was broken
21:03 angdraug can you log onto ci-slave06?
21:03 angdraug it won't let me in
21:04 aglarendil I cannot either
21:04 aglarendil but this is the only thing we changed
21:05 aglarendil oh! I am a stupid person!
21:06 aglarendil I wrote -%> in a erb instead if %>
21:06 aglarendil wsrep_slave_threads=4wsrep_sst_method=mysqldump
21:06 angdraug bah
21:06 angdraug so, revert and new patch?
21:07 aglarendil nope
21:07 aglarendil I will make a fix
21:07 angdraug ok
21:09 aglarendil https://review.openstack.org/#/c/110790/
21:09 aglarendil you all must be grateful to me
21:09 aglarendil I saved the world and millions of lives
21:10 aglarendil by not applying as a nuclear plant engineer
21:11 bookwar aglarendil: "I wrote -%> in a erb instead if %>i" is it syntax error?
21:12 bookwar shouldn't it be catched by puppet-lint check on CI?
21:12 aglarendil bookwar: nope. -%> does not return the line
21:12 aglarendil %> does
21:13 angdraug ok lets merge that before CI is done :p
21:13 aglarendil angdraug: why do you want this?
21:13 angdraug just kidding :)
21:13 aglarendil angdraug: well. I have just tested it. it works
21:13 angdraug nah, this time we wait, and won't merge it if it fails for unrelated reason :)
21:14 aglarendil as alex kornienko said fuel development is like gamedev
21:14 bookwar one day i'll just disable Fuel CI and check how many patches you'll merge before noticing that
21:14 aglarendil why don't you want to gamble a little bit
21:14 aglarendil bookwar: it was just a misclick, I think
21:15 angdraug yup. I looked at one commit, it had +1 from Fuel CI, then I got distracted, then I switched to the tab with that commit, and approved...
21:15 aglarendil let's write a song on queen melody: "Too much merge will kill you"
21:15 bookwar yes, this time it was, but i guess Ui and Python team wouldn't notice it for a week or two
21:16 angdraug joined #fuel-dev
21:23 adanin joined #fuel-dev
21:24 bookwar if i remember correctly, openstack infra says that at least 10% of all CI test failures are related to bugs already merged in openstack, and that they plan to limit the amount of tests just to make it possible for bug fixes to go through the ci process in less then 10 rechecks
21:25 aglarendil bookwar: they have elastic recheck for that
21:27 bookwar elastic recheck helps to run recheck if needed. But it happens too often. They want to decrease the amount of rechecks.
21:28 aglarendil I started mirrors rebuild
21:29 aglarendil it should finish right after CI finishes for my fix and right before master iso starts building
21:30 bookwar mirrors rebuild triggers iso rebuild immediately
21:31 bookwar we'll need to cancel this build
21:31 bookwar merge what needs to be merged and build iso afterwards
21:34 aglarendil it will be merged befor mirrors rebuild
21:40 mihgen xarses: aglarendil any idea why xenolog moved https://bugs.launchpad.net/fuel/+bug/1342009 to 6.0 ? take a look please..
21:41 aglarendil mihgen: it is still not confirmed
21:41 aglarendil mihgen: and it looks invalid
21:42 aglarendil mihgen: 1) different ip addresses for different dhcp agents is normal
21:42 mihgen then let's explain it there or get clarifications from aleksandr_null
21:44 aglarendil mihgen: commented and marked as incomplete
21:46 aglarendil angdraug: deployment of the fix finished succesfully. waiting for OSTF
21:46 aglarendil it will take ~ 5-10 minutes
21:54 aglarendil angdraug: success
21:54 aglarendil angdraug: I will merge it
22:08 aglarendil okay. new master ISO is building
22:08 aglarendil gonna sleep now. have good day, guys
22:24 xarses angdraug: https://review.openstack.org/#/c/109961/
22:31 mihgen rmoe: angdraug folks I think https://bugs.launchpad.net/fuel/+bug/1350286, https://bugs.launchpad.net/fuel/+bug/1348701 are duplicates
22:31 rmoe most liekly
22:31 mihgen let's pay attention to these, we can't simply close .. as it happens there and there..
22:32 rmoe I was never able to reproduce 1348701 and never got a log snapshot
22:34 angdraug looks like a swift bug to me
22:34 angdraug rmoe: were you trying to reproduce it with ceph?
22:34 rmoe nope
22:34 rmoe with swift
22:34 rmoe i'ts probably a swift bug
22:35 angdraug yep, definitely looks like a duplicate
22:36 angdraug or maybe not
22:36 angdraug 1348701 has 503 from swift
22:39 rmoe yeah, that's why I wanted to look at the logs :)
22:42 mihgen well there is at least one snapshot from our MLNX folks
22:57 xarses angdraug: https://review.openstack.org/#/c/110745
23:31 aleksandr_null aglarendil: when I address got assigned to the VM by one server and then IP address of server will be changed then this IP couldn't be prolonged by VM. dhcp client will try to reach dhcp server using old address to prolong IP lease till the end. After that it will send absolutely new request.
23:33 aleksandr_null *till the end of lease.
23:47 angdraug aleksandr_null: pretty sure he's finally asleep
23:59 aleksandr_null =)

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