[00:53] <jerichowasahoax> Doow: "sudo -u gitolite-user gitolite args args args"
[01:11] <MannyLNJ> Hi I have a headless server running 16.04 I installed X2Go on it so I could get a GUI when working remotley but now I get no desktop how can I check that XFCE is running on it?
[07:40] <lordievader> Good morning
[07:41] <zioproto> good morning
[07:41] <zioproto> trying to compile packages for Openstack Neutron newton version
[07:42] <zioproto> jamespage: coreycb what is this weird tarball neutron_9.4.2~dev21.orig.tar.gz ?
[08:08] <zioproto> I am confused today: https://pastebin.com/QQ1yA2X9
[08:08] <zioproto> this used to work, to import the upstream tarballs
[08:15] <zioproto> are there some changes in these tools that are braking everyting ?
[08:17] <zioproto> I found the problem
[08:17] <zioproto> https://pastebin.com/Hcxm86cg
[08:18] <zioproto> it is hardcoded to use /usr/bin/pristine-tar
[08:18] <zioproto> but my tool is in /usr/local
[08:18] <zioproto> because I had to upgrade pristine-tar
[08:19] <zioproto> wow is getting hard to build stuff on Xenial
[08:19] <zioproto> tar: unrecognized option '--verbatim-files-from'
[08:20] <zioproto> what are you guys using to build the openstack ubuntu packages ?
[08:31] <zioproto> I upgraded tar by hand
[08:31] <zioproto> new problem :
[08:31] <zioproto> https://pastebin.com/uQdE0MBd
[08:32] <zioproto> sudo apt-get install xdelta3
[08:32] <zioproto> now it works !
[08:47] <zioproto> today is not my day
[08:47] <zioproto> E: neutron changes: bad-distribution-in-changes-file xenial-newton
[09:07] <jamespage> zioproto: I suspect that both coreycb and I will be running on the development release
[09:08] <zioproto> jamespage: hey there
[09:08] <jamespage> zioproto: that last error about xenial-newton is normal
[09:08] <zioproto> I am now building 9.4.1 with a patch backported from Ocata on top
[09:08] <jamespage> we use non-standard targets for the UCA
[09:08] <zioproto> jamespage: yes, I figured out I was looking for the dsc in the wrong path
[09:08] <zioproto> I fixed everything, update pristine-tar, tar, and xdelta
[09:09] <zioproto> I have to update my xenial provisioning script
[09:09] <zioproto> or I will not be able to build packages on Xenial anymore
[09:09] <zioproto> I did not understand if you are interested in the version bump to 9.4.1 or not
[09:09] <zioproto> what is the development release ?
[09:28] <jamespage> zioproto: that will happen yes - I suspect coreycb has it on his todo list
[09:38] <zioproto> ok, I hope to get to finish testing this today
[09:38] <zioproto> I will provide a MR
[11:20] <frickler> jamespage: if you happen to prepare a ceph-10.2.10 pkg, please be sure to include the patch from https://github.com/ceph/ceph/pull/15835 , we just ran into that. the patch was added after 10.2.10 into upstream
[11:21] <jamespage> frickler: ta
[13:29] <coreycb> jamespage: zioproto: hello, i started working on point releases last week. i'll be working on the rest in the background this week.
[13:30] <zioproto> coreycb: hey... I was trying to backport something from Ocata do Newton but looks too hard :(
[13:31] <zioproto> anyway importing just the latest newton tarball and compiling works fine
[13:31] <zioproto> the debian/patches for neutron is basically empty
[13:31] <zioproto> we are blocked by a performance problem, I guess we have to start the Ocata upgrade in production as soon as possible
[13:34] <coreycb> zioproto: ok. it's always tough balancing regression potential with retro-fitting a patch backport.
[13:35] <zioproto> We have this problem in Newton https://bugs.launchpad.net/neutron/+bug/1665215
[13:36] <zioproto> but looks like the only viable solution is to upgrade to Ocata
[15:17] <frickler> xnox: https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/1456789 seems fixed in newer systemd according to the debian bug, would be great to see some backport to xenial
[16:31] <rbasak> ahasenack: I just pushed https://anonscm.debian.org/cgit/pkg-mysql/mysql.git/log/?h=mysql-5.7/rbasak/ubuntu/devel which is my merge work in progress.
[16:58] <rbasak> ahasenack: git log --reverse -p 652783454c167bec03a291ba7bc347d5e9a8e864..pkg/ubuntu/devel -- debian/
[17:11] <rbasak> diff --git a/debian/patches/fix_mysql_config_flags.patch b/debian/patches/fix_mysql_config_flags.patch
[17:11] <rbasak> deleted file mode 100644
[17:11] <rbasak> index ee76415..0000000
[17:14] <rbasak> ahasenack: https://anonscm.debian.org/cgit/pkg-mysql/mysql.git/log/?h=mysql-5.7/rbasak/ubuntu/devel
[17:15] <rbasak> ahasenack: https://code.launchpad.net/~mysql-ubuntu/mysql-packaging/+git/mysql/+ref/mysql-5.7/rbasak/ubuntu/devel
[17:26] <rbasak> ahasenack: I forgot update-maintainer.
[17:26] <cpaelzer> rbasak: in case git ubuntu tag fails (blocked by it pretending the workign tree is not clean) - could I just tag it with git tag for the time being?
[17:26] <rbasak> cpaelzer: that should be OK I think. I can't remember if git-ubuntu uses annotated tags or not.
[17:26] <cpaelzer> rbasak: or is there a lot of extra magic behind "git ubtuntu tag" other than translating the changelog entry to a string that is tag compatible?
[17:26] <rbasak> No other magic.
[17:29] <cpaelzer> rbasak: it uses annotated tags
[17:29] <cpaelzer> rbasak: like http://paste.ubuntu.com/26113388/
[17:30] <ahasenack> rbasak: rubber duck review :)
[17:56] <rbasak> ahasenack: build test and dep8 passed against current bionic.
[18:00] <ahasenack> cool
[18:28] <trdillon1> I have a maas problem and the maas chanel is dead. Can anyone here help?
[18:30] <sarnold> "maybe"
[18:31] <dpb1> ask away
[18:32] <trdillon1> I can't commission som HPE BL460c G9 blades. They all fail with a segfault on the lshw script
[18:33] <trdillon1> I am running MAAS 2.2.2
[18:33] <trdillon1> Tried to commission with both 14.04 and 16.04 but they both fail with the same issue
[18:34] <sarnold> "Running lshw on a Dell Precision 5510 causes it to segfault unless I use `lshw -disable SCSI`."  https://bugs.launchpad.net/ubuntu/+source/lshw/+bug/1564517
[18:35] <trdillon1> Do I just set that as a global parameter?
[18:35] <sarnold> no
[18:35] <sarnold> I'm just surprised to find segfaults in the bug tracker :/
[18:35] <sarnold> I didn't expect lshw to do anything complicated enough to be that buggy
[18:37] <sarnold> trdillon1: I suggest trying to reproduce this with a minimal image, netboot image, installed system, etc., something that would let you run ubuntu-bug and report the issue to launchpad
[18:37] <trdillon1> Well this is a commission with MAAS not an install
[18:39] <sarnold> right
[18:39] <sarnold> but if it can be reproduced on a standard install it'll be way easier to report the bug :)
[18:41] <trdillon1> Alright. I understand now
[18:41] <trdillon1> Thanks
[19:11] <gun1x> guys, i just deployed kubernates on ubuntu 16.04 with conjure-up and i have no passwords. it deployed a lot of services, all ask for password and i don't know it
[19:22] <stokachu> gun1x: what all ask for password?
[19:33] <gun1x> stokachu: all the services deployed
[19:33] <stokachu> gun1x: how are you trying to connect to them
[19:33] <gun1x> stokachu: kubernates, heapster, grafana, influxdb
[19:33] <gun1x> https
[19:33] <stokachu> gun1x: wut..
[19:34] <gun1x> stokachu: https://bpaste.net/show/a43461bbec45
[19:34] <gun1x> i got the links at the end of the deployment
[19:34] <gun1x> but no password
[19:34] <stokachu> gun1x: odd
[19:34] <stokachu> gun1x: can you hop into #juju
[19:35] <gun1x> just did ...
[21:38] <stokachu> gun1x: just curious, what exactly are you trying ot accomplish with kubernetes?
[21:39] <gun1x> stokachu: my company is forcing me to use mirantis cloud platform for openstack, and that platform also has kubernetes deployment. also, we will have multiple internal projects which should have native cloud apps so kubernetes is a far better solution then openstack
[21:40] <gun1x> stokachu: so i have to learn kubernetes. fast. i hoped getting an automated build will speed up the process. i will go through manual install
[21:40] <stokachu> gun1x: k, well fwiw openshift will lock you into their custom kubernetes
[21:40] <stokachu> where are's is the closest you'll get to upstream
[21:40] <gun1x> stokachu: fwiw ?
[21:41] <stokachu> well the automated build via conjure-up does get you a full environment
[21:41] <stokachu> you use kubectl to manage your kubernetes after that
[21:41] <stokachu> so you'll need to do some additional reading on that
[21:41] <stokachu> i wouldnt base your experience on not being able to see metrics that doesn't actually exist because you havent deployed any workloads on it
[21:42] <gun1x> :D
[21:43] <stokachu> but, go ahead and try out other solutions , then come back to #juju when you need more help
[21:43] <gun1x> i saw juju stuff fail in the past. also deploying openstack with it is not delivering the same experience as openstack-ansible or kolla or trippleo ...
[21:43] <stokachu> wel lthey are different products by different companies
[21:43] <gun1x> stokachu: i don't want to use mirantis solutions. i would like to use openstack-ansible for openstack :D
[21:44] <stokachu> well we dont do openstack-ansible
[21:44] <stokachu> and when you saw juju stuff fail did you submit bugs?
[21:45] <gun1x> submiting bugs and getting involved is normally happening AFTER you decide to go for a solution. in the testing phase you just check out every solution to see which one you like most.
[21:46] <gun1x> for example, osa-aio, the all in one build from openstack-ansible, is working flawless on 3 distributions. think that has any influente on the decision of which tool to use for production deployment ? :D
[21:46] <gun1x> and regarding openshift, it's the last thing i want to try, since it provides bad support for ubuntu (or none)
[21:47] <gun1x> and centos/redhat have really old kernel atm.
[21:47] <stokachu> gun1x: so im guessing you upgrade flawlessly with openstack-ansible ?
[21:48] <gun1x> stokachu: upgrade what?
[21:48] <stokachu> openstack..
[21:48] <gun1x> stokachu: you mean from ocata to pike?
[21:48] <stokachu> sure
[21:48] <gun1x> stokachu: if you been upgrades, i didn't have a chance to try that. they have good docu for it, but never tried it out. i tried only prod builds and test builds. with cinder/swift and with ceph
[21:49] <stokachu> yea see, thats the thing, you may get flawless deploys with openstack-ansible
[21:49] <stokachu> but try to upgrade one
[21:49] <stokachu> https://docs.openstack.org/openstack-ansible/pike/user/index.html
[21:49] <stokachu> im guessing you have to do it manually
[21:50] <stokachu> anyway, good luck on your solutions search
[21:50] <gun1x> i don't think i will get to that anyway. company wants the "Strategical parnership" which i don't like anyway :D
[21:50] <gun1x> MCP doesn't provide pike yet, they are on ocata