Perl 6 - the future is here, just unevenly distributed

IRC log for #fuel-dev, 2014-07-11

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

All times shown according to UTC.

Time Nick Message
00:09 xarses joined #fuel-dev
00:21 angdraug intLabs: yes, all code in fuel is ASL 2.0
00:21 angdraug you're welcome to use it in your project, as long as you keep the relevant copyright notices in place in the source code it should be ok
00:22 angdraug we would however appreciate if you contribute back your bugfixes
00:23 angdraug it will help you track your own changes, too
00:24 angdraug if you can, please post a writeup about what you're doing with fuel
00:24 angdraug we're always looking for all kinds of feedback
00:38 prmtl joined #fuel-dev
00:45 intLabs angdruag: Thanks, I’ll certainly report back any findings; I’m in the process of producing a series of blog articles atm.
00:46 intLabs I’ll offer any bug reports and fixes up through the appropriate channels - though you lot have built a wonderful project so far - I’ve not hit many yet I can’t attribute to my own clumsyness :)
01:01 teran joined #fuel-dev
01:13 teran joined #fuel-dev
01:15 teran_ joined #fuel-dev
01:56 bdudko joined #fuel-dev
01:56 agordeev joined #fuel-dev
01:56 ykotko joined #fuel-dev
01:56 enikanorov_ joined #fuel-dev
01:56 mrasskazov joined #fuel-dev
01:56 scroiset joined #fuel-dev
01:56 warpc_ joined #fuel-dev
01:56 bas_ joined #fuel-dev
01:56 aignatov joined #fuel-dev
01:56 intLabs joined #fuel-dev
01:56 rshon_ joined #fuel-dev
01:56 saibarspeis joined #fuel-dev
01:56 evgeniyl joined #fuel-dev
01:56 asledzinskiy joined #fuel-dev
01:56 dteselkin joined #fuel-dev
01:56 jeremydei joined #fuel-dev
01:56 xdeller joined #fuel-dev
01:56 ogelbukh joined #fuel-dev
01:56 dilyin joined #fuel-dev
01:56 aglarendil joined #fuel-dev
01:56 ikalnitsky joined #fuel-dev
01:56 sbanka joined #fuel-dev
01:56 dshulyak1 joined #fuel-dev
01:56 katyafervent joined #fuel-dev
01:56 zynzel joined #fuel-dev
01:56 asyriy joined #fuel-dev
01:56 mirrorbox joined #fuel-dev
01:56 anteaya joined #fuel-dev
01:56 amaretskiy joined #fuel-dev
01:56 igajsin joined #fuel-dev
01:56 msemenov joined #fuel-dev
01:58 nurla joined #fuel-dev
01:58 sirmax joined #fuel-dev
01:58 apalkina joined #fuel-dev
01:58 vkramskikh joined #fuel-dev
01:58 DinaBelova joined #fuel-dev
01:58 alex_didenko joined #fuel-dev
01:58 holser joined #fuel-dev
01:58 izinovik joined #fuel-dev
01:58 sbog_ joined #fuel-dev
01:58 meow-nofer joined #fuel-dev
01:58 rvyalov joined #fuel-dev
01:58 MiroslavAnashkin joined #fuel-dev
01:58 akislitsky joined #fuel-dev
01:58 monester_ joined #fuel-dev
01:58 christop1eraedo joined #fuel-dev
01:58 evg_ joined #fuel-dev
01:58 dburmistrov joined #fuel-dev
01:58 bookwar joined #fuel-dev
01:58 openstackstatus joined #fuel-dev
01:58 higgins joined #fuel-dev
01:58 MorAle joined #fuel-dev
01:58 mihgen joined #fuel-dev
01:58 wrale joined #fuel-dev
01:58 eshumakher joined #fuel-dev
01:58 IgorYozhikov joined #fuel-dev
01:58 akuznetsova joined #fuel-dev
01:58 apodrepniy_ joined #fuel-dev
01:58 enikanorov__ joined #fuel-dev
01:58 omelchek joined #fuel-dev
01:58 Tatyanka_Leontov joined #fuel-dev
01:58 Bomfunk joined #fuel-dev
01:58 tnurlygayanov joined #fuel-dev
01:58 jkirnosova joined #fuel-dev
01:58 kat_pimenova joined #fuel-dev
02:03 ruhe joined #fuel-dev
02:04 xarses joined #fuel-dev
02:04 SergeyLukjanov joined #fuel-dev
02:06 dmitryme joined #fuel-dev
02:06 artem_panchenko joined #fuel-dev
02:06 akasatkin joined #fuel-dev
02:06 justif joined #fuel-dev
02:06 a_teem joined #fuel-dev
02:06 aedocw joined #fuel-dev
02:06 mattymo joined #fuel-dev
02:06 dpyzhov joined #fuel-dev
02:08 sanek joined #fuel-dev
02:08 accela-dev joined #fuel-dev
02:14 teran joined #fuel-dev
02:20 ChanServ joined #fuel-dev
02:20 EeeBotu joined #fuel-dev
02:21 EeeBotu New bug #1340426 in fuel: "Empty section "System Tests" included in development docs" [Undecided, New] https://bugs.launchpad.net/bugs/1340426
02:21 EeeBotu New bug #1340414 in fuel: "Large number of disks could fail ubuntu installation" [Undecided, New] https://bugs.launchpad.net/bugs/1340414
02:21 EeeBotu New bug #1340397 in fuel: "vcenter_hash is not a hash or array when accessing it with host_ip" [Critical, Fix Committed] https://bugs.launchpad.net/bugs/1340397
02:21 EeeBotu New bug #1340353 in fuel: "[UX] as UI pages get larger, save | apply workflow becomes bothersome" [Undecided, New] https://bugs.launchpad.net/bugs/1340353
02:21 EeeBotu New bug #1340336 in fuel: "fuel-library puppet-lint tests stop on first failure" [High, Confirmed] https://bugs.launchpad.net/bugs/1340336
02:21 EeeBotu New bug #1340301 in fuel: "MySQL server package installation failure in HA mode on CentOS" [High, Fix Committed] https://bugs.launchpad.net/bugs/1340301
02:21 EeeBotu New bug #1340288 in fuel: "Need documentation of MTU setting in YAML files" [Undecided, New] https://bugs.launchpad.net/bugs/1340288
02:21 EeeBotu New bug #1340263 in fuel: "controller failover leads to lags with dead memcache server" [High, Confirmed] https://bugs.launchpad.net/bugs/1340263
02:21 EeeBotu New bug #1340260 in fuel: "VIPs do not failover in assymetric cluster" [Critical, In Progress] https://bugs.launchpad.net/bugs/1340260
02:21 EeeBotu New bug #1340237 in fuel: " Execution of '/usr/bin/apt-get -q -y -o DPkg::Options::=--force-confold --force-yes install iptables-persistent=0.5.3ubuntu2' returned 100" [High, Incomplete] https://bugs.launchpad.net/bugs/1340237
02:21 EeeBotu New bug #1340213 in fuel: "Error messages in anaconda.log file regarding nonexistent files on FUEL master-node  during node setup" [Low, New] https://bugs.launchpad.net/bugs/1340213
02:21 EeeBotu New bug #1340207 in fuel: "Booting bootstrap without any Disc Fuel 5.0.1" [Medium, New] https://bugs.launchpad.net/bugs/1340207
02:21 EeeBotu New bug #1340172 in fuel: "Add mamange/unmanage blocks to corosync init scripts" [Medium, Confirmed] https://bugs.launchpad.net/bugs/1340172
02:21 EeeBotu New bug #1340153 in fuel: "[Access to master node] Empty password should be validated when try to change password from UI" [Medium, Fix Committed] https://bugs.launchpad.net/bugs/1340153
02:21 EeeBotu New bug #1340141 in fuel: "[Access master node]Get message 'No handlers could be found for logger "keystoneclient.httpclient" when using fuel cli'" [Medium, In Progress] https://bugs.launchpad.net/bugs/1340141
02:21 EeeBotu New bug #1340098 in fuel: "Dependent settings incorrect behaviour" [Medium, Invalid] https://bugs.launchpad.net/bugs/1340098
02:32 EeeBotu joined #fuel-dev
02:33 EeeBotu New bug #1340426 in fuel: "Empty section "System Tests" included in development docs" [Undecided, New] https://bugs.launchpad.net/bugs/1340426
02:33 EeeBotu New bug #1340414 in fuel: "Large number of disks could fail ubuntu installation" [Undecided, New] https://bugs.launchpad.net/bugs/1340414
02:33 EeeBotu New bug #1340397 in fuel: "vcenter_hash is not a hash or array when accessing it with host_ip" [Critical, Fix Committed] https://bugs.launchpad.net/bugs/1340397
02:33 EeeBotu New bug #1340353 in fuel: "[UX] as UI pages get larger, save | apply workflow becomes bothersome" [Undecided, New] https://bugs.launchpad.net/bugs/1340353
02:33 EeeBotu New bug #1340336 in fuel: "fuel-library puppet-lint tests stop on first failure" [High, Confirmed] https://bugs.launchpad.net/bugs/1340336
02:33 EeeBotu New bug #1340301 in fuel: "MySQL server package installation failure in HA mode on CentOS" [High, Fix Committed] https://bugs.launchpad.net/bugs/1340301
02:33 EeeBotu New bug #1340288 in fuel: "Need documentation of MTU setting in YAML files" [Undecided, New] https://bugs.launchpad.net/bugs/1340288
02:33 EeeBotu New bug #1340263 in fuel: "controller failover leads to lags with dead memcache server" [High, Confirmed] https://bugs.launchpad.net/bugs/1340263
02:33 EeeBotu New bug #1340260 in fuel: "VIPs do not failover in assymetric cluster" [Critical, In Progress] https://bugs.launchpad.net/bugs/1340260
02:33 EeeBotu New bug #1340237 in fuel: " Execution of '/usr/bin/apt-get -q -y -o DPkg::Options::=--force-confold --force-yes install iptables-persistent=0.5.3ubuntu2' returned 100" [High, Incomplete] https://bugs.launchpad.net/bugs/1340237
02:33 EeeBotu New bug #1340213 in fuel: "Error messages in anaconda.log file regarding nonexistent files on FUEL master-node  during node setup" [Low, New] https://bugs.launchpad.net/bugs/1340213
02:33 EeeBotu New bug #1340207 in fuel: "Booting bootstrap without any Disc Fuel 5.0.1" [Medium, New] https://bugs.launchpad.net/bugs/1340207
02:33 EeeBotu New bug #1340172 in fuel: "Add mamange/unmanage blocks to corosync init scripts" [Medium, Confirmed] https://bugs.launchpad.net/bugs/1340172
02:33 EeeBotu New bug #1340153 in fuel: "[Access to master node] Empty password should be validated when try to change password from UI" [Medium, Fix Committed] https://bugs.launchpad.net/bugs/1340153
02:39 prmtl joined #fuel-dev
03:13 teran joined #fuel-dev
04:13 teran joined #fuel-dev
04:41 prmtl joined #fuel-dev
05:13 teran joined #fuel-dev
05:34 dshulyak joined #fuel-dev
05:42 prmtl joined #fuel-dev
06:13 prmtl joined #fuel-dev
06:13 teran joined #fuel-dev
06:26 al_ex11 joined #fuel-dev
06:36 al_ex11 joined #fuel-dev
06:39 sambork joined #fuel-dev
06:42 teran joined #fuel-dev
07:01 pasquier-s joined #fuel-dev
07:07 eshumakher joined #fuel-dev
07:11 mihgen xarses: hi
07:11 mihgen xarses: Does haproxy/vip issue exists in master right now or it's related to your ML2 patchset only?
07:12 * mihgen received an update on ML2 in mailing list from xarses
07:16 artem_panchenko joined #fuel-dev
07:22 msemenov1 joined #fuel-dev
07:22 mihgen mattymo: hey, how https://review.openstack.org/#/c/103897/ can break CI ?
07:23 meow-nofer__ joined #fuel-dev
07:25 EeeBotu New bug #1340555 in fuel: "astute will stop receive new messages if queue is redeclared" [High, In Progress] https://bugs.launchpad.net/bugs/1340555
07:30 Longgeek joined #fuel-dev
07:52 Longgeek_ joined #fuel-dev
07:53 Longgeek_ joined #fuel-dev
07:55 ikalnitsky aglarendil: could you please merge these? https://review.openstack.org/#/c/105750/ https://review.openstack.org/#/c/105739/
08:12 sambork joined #fuel-dev
08:23 ddmitriev joined #fuel-dev
08:24 e0ne joined #fuel-dev
08:25 ikalnitsky joined #fuel-dev
08:27 teran joined #fuel-dev
08:28 gilmeir joined #fuel-dev
08:31 EeeBotu New bug #1340588 in fuel: "ubuntu control node gro on on phystical interface" [Undecided, New] https://bugs.launchpad.net/bugs/1340588
08:31 EeeBotu New bug #1340559 in fuel: "Lack of firmware in the initramfs.img file of bootstrap mode." [Undecided, New] https://bugs.launchpad.net/bugs/1340559
08:33 IvanBerezovskiy joined #fuel-dev
08:33 mattymo mihgen, it can't. we don't even test master node in ci
08:33 mattymo it means HA on ubuntu is still not stable
08:37 EeeBotu New bug #1340593 in fuel: "Upload Cirros TestVM image failed with  - import pkg_resources Error communicating with http://10.108.6.2:9292 [Errno 32] Broken pipe" [High, New] https://bugs.launchpad.net/bugs/1340593
08:38 AndreyDanin joined #fuel-dev
08:41 nurla akasatkin: evgeniyl guys, what about merge https://review.openstack.org/#/c/105495?
08:42 evgeniyl nurla: it's a refactoring and doesn't fix any issues, we decided to merge it after 5.0.1 release.
08:42 nurla evgeniyl: thnx
08:43 akasatkin evgeniyl: I'll not add it to 5.0.1 tarball builds then. I've added it to all the builds till now.
08:44 evgeniyl akasatkin: yeah, just skip it for now
08:44 salmon_ joined #fuel-dev
08:46 mihgen mattymo: so why it broke CI?
08:47 brain461 joined #fuel-dev
08:47 teran joined #fuel-dev
08:49 mattymo mihgen, cluster deployed but failed ostf
08:50 mihgen mattymo: let's troubleshoot. we need our CI to be stable like hell
08:51 mattymo actually that was reading centos and it's fine
08:51 mattymo ubuntu failed and it looks like we have an issue with snapshots still not capturing /var/log/remote
08:52 mattymo oh it's here
08:52 mattymo nevermind
08:53 mattymo galera failed on ubuntu in CI
08:53 mattymo that's why it didn't pass CI
09:03 jaranovich joined #fuel-dev
09:07 teran joined #fuel-dev
09:07 pasquier-s joined #fuel-dev
09:08 ddmitriev joined #fuel-dev
09:18 vrovachev joined #fuel-dev
09:24 asheplyakov joined #fuel-dev
09:38 EeeBotu New bug #1340615 in fuel: "diagnostic snapshot should have /var/log/remote" [Medium, Confirmed] https://bugs.launchpad.net/bugs/1340615
09:53 meow-nofer__ joined #fuel-dev
09:54 mattymo ikalnitsky, I can't find anything in the "get" function for fabric about handling links
09:54 mattymo it would also be nice if fabric could preserve timestamps
09:57 ikalnitsky mattymo: I'm not a Fabric maintainer :) actually, I'll take a look in free time.
09:57 mattymo ok)
09:59 Longgeek joined #fuel-dev
10:02 teran joined #fuel-dev
10:08 skolekonov joined #fuel-dev
10:14 brain461 joined #fuel-dev
10:14 prmtl hey guys, could someone look at https://review.openstack.org/#/c/104903/ and https://review.openstack.org/#/c/105113/ ? it's ready to review and (hopefully) merge
10:17 guillaume__1 joined #fuel-dev
10:36 vkramskikh +1, but i think we need akasatkin or evgeniyl to +2 it
10:38 evgeniyl meow-nofer_: meow-nofer  meow-nofer__  take a look at https://review.openstack.org/#/c/104903/ please
10:38 EeeBotu New bug #1340662 in fuel: "UI: extra scroll bar on 'create environment' page" [Low, Confirmed] https://bugs.launchpad.net/bugs/1340662
10:38 EeeBotu New bug #1340657 in fuel: "'keystone' delays lead to unstable MOS operations" [Undecided, New] https://bugs.launchpad.net/bugs/1340657
10:42 meow-nofer__ roger that
10:57 prmtl vkramskikh: thanks
11:02 teran_ joined #fuel-dev
11:11 teran joined #fuel-dev
11:15 vkramskikh hi guys
11:15 vkramskikh as for this bug: https://bugs.launchpad.net/fuel/+bug/1338573
11:15 vkramskikh we can provide a fix for UI part, but this won't fix the bug finally
11:16 vkramskikh look here: https://github.com/stackforge/fuel-web/blob/master/nailgun/static/js/views/cluster_page.js#L133-L139
11:16 vkramskikh let me explain the code
11:16 vkramskikh after the task is started, we send 3 requests to get new data from nailgun
11:16 vkramskikh the cluster and its nodes and tasks
11:17 vkramskikh and then if all these request succeed, we schedule polling of the tasks and nodes
11:18 vkramskikh to fix the bug we will schedule polling always, not only if all these 3 requests succeed
11:18 vkramskikh the bug was because one of these requests sometime fails
11:18 vkramskikh usually is it request for nodes
11:18 vkramskikh and by "fails" I mean this
11:19 vkramskikh https://launchpadlibrarian.net/179740992/2014-07-11%2014_10_27-Fuel%20Dashboard%20-%20screen.png
11:19 vkramskikh response of nodes handler is malformed
11:19 vkramskikh the browser is unable to parse the response becuase of "No Content" before the beginning of valid HTTP response
11:20 vkramskikh this seem to happen sometimes in case of parallel http requests like these ones
11:20 vkramskikh so this is the real issue
11:20 vkramskikh some weird s&%t is going on
11:20 vkramskikh if we don't fix this, some strange "bugs" may happen here and there
11:21 meow-nofer__ holy guacamole
11:22 meow-nofer__ is it reproducible on development server or this started after switching to uWSGI?
11:22 vkramskikh this is reproducible on nailgun in fake mode
11:23 meow-nofer__ so, this happens for a long time? nd we just found it now?
11:24 vkramskikh well, i never faced this issue
11:25 vkramskikh it is also known that it reproduces rarely if nginx is used
11:26 vkramskikh it could have reproduced in other places, but it could be less visible
11:26 vkramskikh it seems this is the only places in code where we considered requests to be always successful
11:26 meow-nofer__ so it maybe some web.py internal issue
11:28 vkramskikh probably
11:28 vkramskikh kat_pimenova says it mostly happens with the second request here:
11:28 vkramskikh this.model.fetch(), this.model.fetchRelated('nodes'), this.model.fetchRelated('tasks')
11:28 vkramskikh second here is request for node list
11:28 vkramskikh so it seems to be kind of multithreading issue
11:29 vkramskikh because these 3 requests run in parallel
11:30 meow-nofer__ but stil I can’t imagine what in our code may cause bulshitting of the socket
11:38 EeeBotu New bug #1340692 in fuel: "Can't change hypervisor after enabling Ceph options" [Medium, Confirmed] https://bugs.launchpad.net/bugs/1340692
11:41 vkramskikh meow-nofer__: i've debugged it a little bit more
11:41 vkramskikh https://github.com/stackforge/fuel-web/blob/master/nailgun/nailgun/api/v1/handlers/tasks.py#L60 this message gets inserted
11:42 vkramskikh i changed it and got same error
11:42 vkramskikh the error seems to be reproducible only if we start deployment after stop/reset
11:42 vkramskikh because we delete the old task
11:43 vkramskikh and right after that start the new one
11:47 vkramskikh it seems this is the issue
11:47 pasquier-s joined #fuel-dev
11:47 vkramskikh the culprit i mean
11:47 vkramskikh it seems you return 204 with content
11:47 vkramskikh with content "No Content"
11:48 vkramskikh but browser/uwsgi/webpy/whatever doesn't expect it to be there
11:48 e0ne joined #fuel-dev
11:48 vkramskikh and in case if the connections are kept alive
11:48 vkramskikh "No Content" is treated as a part of the next response
11:48 guillaume__1 joined #fuel-dev
11:49 meow-nofer__ yeah, you may be right
11:49 vkramskikh http://media.melty.fr/article-1289213-ajust_930/sherlock-holmes.jpg
11:49 vkramskikh ok, i'll created the requests for both UI and python fixes
11:49 vkramskikh *i'll create
11:49 meow-nofer__ https://i.imgur.com/y4Lufun.jpg
11:56 meow-nofer__ vkramskikh: thanks
11:57 vkramskikh this bug exists since 2014-03-11 according to git blame
11:57 evgeniyl mattymo: what should I set for keystone/admin_token? I need some default value for upgrade from 5.0 versions.
11:59 mattymo it should be random
11:59 mattymo and distributed via astute.yaml in the normal way
11:59 mattymo can you generate a random 8 char string of letters and digits?
12:00 evgeniyl mattymo: sure
12:00 evgeniyl mattymo: thanks
12:01 mattymo it's a brand new key/value so why not
12:02 AndreyDanin joined #fuel-dev
12:13 teran joined #fuel-dev
12:14 akasatkin warpc_: can it be useful someday or should be abandoned https://review.openstack.org/#/c/88556/ ?
12:15 warpc_ akasatkin: i believe it will be useful
12:15 warpc_ akasatkin: at now moment
12:16 akasatkin warpc_: why -1 then ?
12:17 warpc_ akasatkin: this should work and test with other changes (upgrade fuel). To prevent merge set -1.
12:17 akasatkin warpc_: it is dated "April 29"
12:17 warpc_ akasatkin: i know. But really we have upgrade script only for 5.1 and work as i know, still in active stage
12:21 alex_didenko https://review.openstack.org/#/c/99150/ needs some reviews
12:22 akasatkin warpc_: afaic, we forgot about this patch. patching of openstack is under test now. as I know, this patch is not in test.
12:23 warpc_ akasatkin: yes. I know what we possible need thus in future, but can work with it in 5.1
12:23 warpc_ akasatkin: *without it i mean
12:24 akasatkin warpc_: okey
12:24 sirmax Hi Team.
12:27 YorikSar joined #fuel-dev
12:31 akasatkin dilyin: do we need https://review.openstack.org/90190 anymore? It's dated April 25.
12:33 akasatkin evgeniyl: ^^^ it's about openstack patching
12:35 evgeniyl akasatkin: hmm, I'm not sure, do you remember something about it ikalnitsky ?
12:39 EeeBotu New bug #1340711 in fuel: "building oslo-messaging for rpm and deb" [Critical, Confirmed] https://bugs.launchpad.net/bugs/1340711
12:39 EeeBotu New bug #1340707 in fuel: "[Upgrades] Cannot start nailgun container due to error in puppet manifests" [High, In Progress] https://bugs.launchpad.net/bugs/1340707
12:39 EeeBotu New bug #1340704 in fuel: "[418] Error running provisioning: <type 'exceptions.OSError'>:[Errno 2] No such file or directory: '/var/www/cobbler/images/centos-x86_64/initrd.img', trace: /usr/lib64/ruby/2.1.0/xmlrpc/client.rb:271:in `call'" [High, New] https://bugs.launchpad.net/bugs/1340704
12:39 EeeBotu New bug #1340703 in fuel: "TransactionRollbackError: deadlock detected" [High, New] https://bugs.launchpad.net/bugs/1340703
12:40 YorikSar Hi, #fuel-dev. Does Fuel setup Keystone with PKI or UUID tokens?
12:41 mihgen YorikSar: mattymo or sambork should know about that..
12:42 mattymo YorikSar, just a sec let me look at our manifests
12:42 mihgen oh wait, do you mean for master node purposes or you talk about openstack controller?
12:42 YorikSar OpenStack controller
12:44 ikalnitsky evgeniyl: akasatkin: wow. i know nothing about this patch.
12:47 EeeBotu New bug #1340725 in fuel: "Nailgun agent execution expired" [High, New] https://bugs.launchpad.net/bugs/1340725
12:47 evgeniyl ikalnitsky: akasatkin I think we don't need this patch since centos/ubuntu versions generated by make system and information about repos is generated by upgrade system.
12:51 akasatkin evgeniyl: it provided repos information on clean master node. but library has defaults and uses it for clean master.
12:52 evgeniyl akasatkin: ikalnitsky lets discuss it later, after demo or on monday
13:08 AndreyDanin joined #fuel-dev
13:14 YorikSar mattymo: Any luck?
13:14 mattymo YorikSar, sorry.. it's UUID
13:15 YorikSar mattymo: Ok... Does anyone know why it's not PKI?
13:16 mattymo I know why. it's the default from upstream keystone puppet manifest
13:16 mattymo that's all
13:16 mattymo we can change it any time
13:17 YorikSar mattymo: I remember there was some problems with it in Fuel... Did they resolve themselves?
13:18 mattymo YorikSar, I'm not aware with any issues with it
13:18 mattymo aware of*
13:19 mihgen YorikSar: have not heard about that too
13:20 msemenov1 joined #fuel-dev
13:20 YorikSar mihgen, mattymo: I think it's time to finally switch to PKI then :)
13:20 mattymo YorikSar, can you give a justification?
13:20 YorikSar mattymo: UUID tokens don't fit well in SQL and using Memcached to store them is a shitty option :)
13:20 mihgen YorikSar: well, now you will need to explain it then and create at least bug about it for sure :)
13:21 YorikSar It started with this bug 1340657
13:21 YorikSar Oh, we don't have that bot here...
13:21 YorikSar https://bugs.launchpad.net/mos/+bug/1340657
13:22 YorikSar Basically you can read that bug as "We use cache and storage and it doesn't work so well"
13:22 YorikSar *as storage
13:22 YorikSar With PKI tokens you don't rely on storage and don't care about memcached failures.
13:24 YorikSar btw, do we use caching in Keystone?
13:27 mihgen this is question to aglarendil and holser I think… ^^
13:28 aglarendil YorikSar: nope, we have it in roadmap, but for production usage we need to implement
13:28 aglarendil YorikSar: 1) deployment of CA for PKI or 2) usage of user's CA
13:29 aglarendil YorikSar: it is currently under development by our french team
13:29 YorikSar aglarendil: Great.
13:29 YorikSar aglarendil: Will it be the default option?
13:29 aglarendil YorikSar: but it will not be merged into 6.0
13:29 aglarendil YorikSar: of course, I am dreaming of it
13:29 mihgen why not 6.0?
13:29 mihgen not 5.1 for sure
13:29 aglarendil sorry
13:30 aglarendil YorikSar: mihgen: I meant until 6.0
13:30 YorikSar aglarendil: And what about caching? Is it on in keystone?
13:30 mihgen aglarendil: are you talking about this email thread http://www.mail-archive.com/fuel-dev@lists.launchpad.net/msg01206.html ?
13:30 aglarendil YorikSar:  yep, we have memcached configured as dogpile backend
13:30 YorikSar aglarendil: And does it use all controllers?
13:31 mihgen YorikSar: btw did you see our bug with memcached?
13:31 YorikSar mihgen: You mean 17:21 < YorikSar> https://bugs.launchpad.net/mos/+bug/1340657 ?
13:31 jaypipes joined #fuel-dev
13:31 mihgen YorikSar: nope
13:31 mihgen holser: what's the bug you filed to keystone prj?
13:31 YorikSar mihgen: Ok.. Which one?
13:32 YorikSar aglarendil: So... Does it use memcached servers on all controllers for caching?
13:32 holser I filed a bug regarding delays when memcache backend is down
13:32 mihgen YorikSar: oh actually yes, it's about the same: https://bugs.launchpad.net/keystone/+bug/1332058
13:32 mihgen that mos refers to https://bugs.launchpad.net/keystone/+bug/1332058
13:32 YorikSar mihgen: Yep.
13:33 YorikSar mihgen: That's the one I want to solve without coding :)
13:33 mihgen =)
13:33 holser dogpile uses simple roundrobin, if backend is down, it doesn’t turn off it for a while
13:33 mihgen YorikSar: lazy programmer)
13:33 holser that adds extra delay when one controller is down
13:34 holser when you have 2 controllers down, many OSTF tests will fail
13:34 YorikSar holser: There are two places where Memcached is used in Keystone. One is for token storage (stupid thing). The second one is for general caching.
13:35 holser YorikSar, that’s true
13:35 YorikSar For tokens we can switch to PKI and it should fix issues with missing controllers.
13:35 YorikSar For caching we should use only local memcached.
13:39 EeeBotu New bug #1340743 in fuel: "backport: neutron security-group-rule-list fails with URI too long" [Medium, Confirmed] https://bugs.launchpad.net/bugs/1340743
13:42 msemenov1 joined #fuel-dev
13:44 mihgen Folks, don't forget about demo today: https://etherpad.openstack.org/p/fuel-demo-agenda
13:48 EeeBotu New bug #1340757 in fuel: "add mellanox bootstrap image to fuel iso" [Undecided, New] https://bugs.launchpad.net/bugs/1340757
13:48 EeeBotu New bug #1340744 in fuel: "ml2 doens't take configuration from astut.yaml" [High, New] https://bugs.launchpad.net/bugs/1340744
13:50 sambork joined #fuel-dev
14:04 guillaume__1 joined #fuel-dev
14:04 pasquier-s joined #fuel-dev
14:07 jobewan joined #fuel-dev
14:28 jaranovich joined #fuel-dev
14:34 mihgen meeting started https://etherpad.openstack.org/p/fuel-demo-agenda
15:00 prmtl joined #fuel-dev
15:30 IvanBerezovskiy left #fuel-dev
15:31 teran joined #fuel-dev
15:41 EeeBotu New bug #1340796 in fuel: "Random JS errors on the settings tab" [High, Confirmed] https://bugs.launchpad.net/bugs/1340796
16:08 intLabs joined #fuel-dev
16:09 asledzinskiy left #fuel-dev
16:09 e0ne joined #fuel-dev
16:12 evgeniyl mattymo: https://review.openstack.org/#/c/106359/ in the patch akislitsky suggested to use stronger token for keystone, what do you think? And is there any constraints on length of token?
16:14 mattymo I'm ok with stronger tokens
16:14 mattymo I don't know the constraints, but I can find it
16:19 rmoe joined #fuel-dev
16:26 e0ne joined #fuel-dev
16:30 intLabs joined #fuel-dev
16:50 angdraug joined #fuel-dev
17:21 tatyana joined #fuel-dev
17:21 AndreyDanin joined #fuel-dev
17:21 teran joined #fuel-dev
17:28 teran joined #fuel-dev
17:28 dshulyak joined #fuel-dev
17:35 ogelbukh1 joined #fuel-dev
17:35 xdeller_ joined #fuel-dev
17:35 guillaume__1 left #fuel-dev
17:51 e0ne joined #fuel-dev
17:56 jeremydei joined #fuel-dev
17:57 dilyin joined #fuel-dev
18:01 nurla joined #fuel-dev
18:04 tatyana joined #fuel-dev
18:07 xarses joined #fuel-dev
18:20 aglarendil joined #fuel-dev
18:22 nurla joined #fuel-dev
18:25 intLabs joined #fuel-dev
18:32 prmtl joined #fuel-dev
18:44 nurla joined #fuel-dev
18:50 artem_panchenko joined #fuel-dev
18:52 Longgeek joined #fuel-dev
18:54 prmtl joined #fuel-dev
18:56 prmtl_ joined #fuel-dev
18:57 EeeBotu New bug #1340886 in fuel: "FUEL web UI incorrectly validate vcenter cluster names that contain space chars" [Undecided, New] https://bugs.launchpad.net/bugs/1340886
19:03 BillTheKat joined #fuel-dev
19:05 e0ne joined #fuel-dev
19:11 prmtl joined #fuel-dev
19:22 r0mikiam joined #fuel-dev
19:23 prmtl joined #fuel-dev
19:30 prmtl joined #fuel-dev
19:37 prmtl joined #fuel-dev
19:39 prmtl joined #fuel-dev
20:00 prmtl joined #fuel-dev
20:05 prmtl joined #fuel-dev
20:06 prmtl joined #fuel-dev
20:08 prmtl joined #fuel-dev
20:09 prmtl joined #fuel-dev
20:28 prmtl joined #fuel-dev
20:29 prmtl_ joined #fuel-dev
20:37 prmtl joined #fuel-dev
21:00 prmtl joined #fuel-dev
21:01 prmtl_ joined #fuel-dev
21:04 EeeBotu New bug #1340968 in fuel: "network namespaces require net.ipv4.ip_forward" [Medium, Triaged] https://bugs.launchpad.net/bugs/1340968
21:28 tatyana joined #fuel-dev
21:47 saibarspeis joined #fuel-dev
21:52 aglarendil guys
21:52 aglarendil It seems I have finally finished polishing of rabbitmq OCF script
21:52 aglarendil https://review.openstack.org/#/q/status:open+project:stackforge/fuel-library+branch:master+topic:bug1339080,n,z
21:53 aglarendil there are two requests
21:53 wrale can anyone tell me which interface to which the fuel node's rsync service should be binding?  is the the admin network interface?
21:53 wrale *is it
21:53 aglarendil 1st one modifies monitor logic to make node join to the "master" node
21:54 aglarendil 2nd one stores all the attributes and makes demote action stop all the rabbit servers thus making this operation idempotent
21:55 xarses mihgen: only on my change set, and its fixed now
21:56 xarses aglarendil: I noticed that after deployment, crm thinks that the monitor script is failed for rabbit, running a cleanup fixes this, might have something to do with the assembly issue
21:57 aglarendil yep, that's what these requests should fix
21:57 aglarendil I have rabbitmq and galera clusters reassembling after full hard reset of the cluster
21:57 mihgen wrale: well ideally to admin, yes
21:57 aglarendil nova-manage service list is full of smiles
21:57 mihgen I don't know what's the actual one
21:58 aglarendil the only problem I see that l3-agent does not create its namespace
21:58 mihgen wrale: I would consider it is a bug if it's listening on all interfaces
21:58 aglarendil but if I restart it - everything works
21:58 aglarendil it may be an oslo.messaging bug
21:58 wrale mihgen: thanks.  from the containers view, there is only one interface .. (in 5.0.1)
21:58 wrale mihgen: just learned this
21:59 aglarendil or performance problem
21:59 mihgen wrale: well but it should be mapped to one of the host interfaces I assume
21:59 aglarendil as when rabbitmq cluster reassembles all the services start reconnecting attempts and it takes a lot of time for rabbitmq to reassemble on 1 vcpu environment
21:59 aglarendil I would really appreciate if you tested it
21:59 wrale mihgen: cool.. in this way a bind with xinetd is redundant, i guess.. thank you
22:00 aglarendil I did not test it on the deployment, but only on already deployed cluster
22:00 xarses aglarendil: keep in mind nova-manage is deprecated. use nova service-list
22:00 aglarendil I do not care right now)
22:00 mihgen wrale: np
22:00 aglarendil xarses: I have ostf passing after hard cluster reset
22:00 xarses yay!
22:00 * xarses cheers
22:00 aglarendil but only after poking l3-agent
22:01 aglarendil we will need to investigate this issue
22:01 aglarendil I hope, deployment is much less destructive than hard reset and cluster will assemble easily
22:01 aglarendil just fetch these requests, as they only modify OCF script
22:11 aglarendil the neturon bug: https://bugs.launchpad.net/fuel/+bug/1340989
23:06 EeeBotu New bug #1341009 in fuel: "obsolete ceph package in fuel-repository for 5.0.1" [Critical, New] https://bugs.launchpad.net/bugs/1341009
23:06 EeeBotu New bug #1340989 in fuel: "l3-agent does not create namespace after full cluster hard reset" [High, New] https://bugs.launchpad.net/bugs/1340989
23:56 rmoe joined #fuel-dev

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