[03:16] <baako> hi guys why am i getting http://kopy.io/OTo8V when i try to install sudo apt-get install nginx
[07:51] <spm_draget> In my log I gett apparmor messing with cups: kernel: [476909.591863] type=1400 audit(1424677626.641:8866): apparmor="DENIED" operation="connect" profile="/usr/sbin/cupsd" name="/run/s pid=30683 comm="cupsd" requested_mask="w" denied_mask="w" fsuid=0 ouid=0
[07:51] <spm_draget> Can anyone help me fixing this?
[09:01] <lordievader> Good morning.
[11:15] <jamespage> coreycb, promoting the juno updates to the proposed PPA - I think they are all ready right?
[12:49] <coreycb> jamespage, they're all set to upload
[12:50] <coreycb> jamespage, I'll get it if you haen't
[13:50] <jamespage> zul, coreycb: hey eventlet just release 0.17
[13:50] <jamespage> including ipv6 support - we might want to ffe that
[13:50] <jamespage> its pretty important IMHO
[13:50] <zul> jamespage: good idea
[13:54] <coreycb> zul, you're all over that right? :)
[13:56] <zul> coreycb:  sure
[13:57] <coreycb> zul, thanks
[14:17] <zul> jamespage/coreycb: FFE filed, packaged and built
[14:17] <jamespage> bug ref?
[14:17] <jamespage> can we link it to the blueprint please
[14:18] <zul> https://bugs.launchpad.net/ubuntu/+source/python-eventlet/+bug/1424639
[14:18] <danwest>  /join #ubuntu-devel
[14:18] <zul> jamespage:  done
[14:18] <zul> heh
[16:30] <sneezewort> Is there a way to see the changes that are made to my system in regards to networking when I run "apt-get install linux-generic-lts-trusty linux-image-generic-lts-trusty"?
[16:31] <bekks> sneezewort: That will install a new kernel.
[16:31] <sneezewort> Right. But can I see the changes in the kernel in regards to networking?
[16:32] <teward> sneezewort: you might want to look at the changelogs for the packages those install
[16:32] <bekks> sneezewort: You'd have to investigate the kernel changelog then. But whats the actual issue behind your questions?
[16:34] <rberg_> sneezewort: for the high level you could look at kernel newbies changelog digest, for the details look at the changelogs on kernel.org..
[16:35] <sneezewort> I posted in here last week. I am on a University campus. Using VMWare ESXi when I install Ubuntu 12.04 I can download large files no problem. When upgrade with the above command and change nothing else, large files stall and become very slow.
[16:35] <sneezewort> On campus network is still fast.
[16:36] <sneezewort> I am trying to figure out what is happening so I can ask campus networks to help.
[16:36] <pmatulis> sneezewort: what command?
[16:36] <sneezewort> "apt-get install linux-generic-lts-trusty linux-image-generic-lts-trusty"
[16:37] <sneezewort> Or do-release-upgrade. basically anything greater than the initial install of 12.04.
[16:48] <pmatulis> sneezewort: what mirror are you using?
[16:48] <sneezewort> http://us.archive.ubuntu.com/ubuntu/
[16:49] <pmatulis> sneezewort: see if esx can emulate a different network adapter
[16:53] <sneezewort> It can I have tried all 3. No dice. I have tried several different versions of VMWare tools. Thats what make this so hard to track down. It is a combination of VMWare Ubuntu and my campus network that is the problem. I am just trying to get a grasp on what is happening.
[16:54] <pmatulis> sneezewort: try using iperf to see if it's systematic or not.  also: http://goo.gl/qadgoJ
[17:34] <smoser> arges, http://paste.ubuntu.com/10374983/
[17:34] <smoser> is that related to https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1329434
[17:34] <smoser> $ uname -r
[17:34] <smoser> 3.13.0-43-generic
[17:34] <smoser> ie, see it in a guest running that kernel, but no way o fkonwing what the host is running.
[17:34] <smoser> this is on digital ocean
[17:36] <smoser> does the host have to possibly update their kernel ?
[17:57] <smoser> smb, so above.. woudlthe host kernel have to have been updated also ?
[17:58] <smb> smoser, I remember it only vaguely but if its the same issue then it might require an update on the host (if that is 3.13 based)
[17:58] <smoser> cool. thanks.
[18:07] <sneezewort> Just FYI gang, I tried the newest Debain using 3.2.0-4-amd64 kernel and the problem does not occur.
[18:15] <pmatulis> sneezewort: would be good to know if issue was observed with Ubuntu 14.04
[18:47] <sneezewort> pmatulis, yep fresh install of 14.04, same behavior.
[18:48] <pmatulis> sneezewort: does vmware web site show 12.04 and 14.04 as supported for your version of esx?
[18:49] <sarnold> sneezewort: is your campus doing any goofy transparent proxy things to your network traffic?
[18:49] <pmatulis> http://www.vmware.com/resources/compatibility/search.php?action=base&deviceCategory=software <= sneezewort
[18:51] <sarnold> (Ubuntu is pages 5-... under "Canonical", heh)
[18:51] <sneezewort> sarnold, probably. I only have control of the network in our server room. Outside of that I have no idea.
[18:52] <sneezewort> pmatulis, It looks like 14.04 is not supported, but 12.04 is.
[19:06] <anubhav> Hi, I had some question regarding deployment and if ubuntuserver + openstack fits my needs, is someone familiar with all this?
[19:29] <sarnold> anubhav: welcome; irc tends to work best if you can ask specific questions..
[19:30] <SchrodingersScat> !ask
[19:44] <anubhav> okay thanks... here is my question.... I have a few systems(4-5) in my institute and that have 16cpu and 32gb ram each, we are planning to organize the architecture to be used as a super computer and i am looking into the open source solutions. Rocs cluster is one option but i think ubuntu cloud would be better. So how does it work. Will using IaaS allow better queuing technique? By this i mean that in rocks cluster, you submit a job and it schedu
[19:44] <anubhav> les it to one of the compute node. Instead can the cloud check on the fly how many resources are available and use them effectively?
[19:45] <anubhav> So will it essentially become a single system with (4*16) cores and 4*32 gb ram?
[19:45] <anubhav> Also i guess adding new nodes will be easier in such a system
[19:52] <sarnold> anubhav: you'll have to figure out how you really want to use your cluster; if you want to use it like AWS, where it's cheap and easy to make virtual machines for individual users, then openstack is a fine approach
[19:53] <sarnold> anubhav: if you want to run large, involved simulations that require the combined resources of all the machines to complete, that requires a different kind of option, something more like beowulf or gridengine or torque or similar..
[19:53] <anubhav> sarnold, mostly it is going to be used for academic purpose only, running simulations etc. So will it incur a performance hit if I use openstack
[19:54] <sarnold> anubhav: openstack involves virtualization, which tends to introduce some overhead; how much overhead depends upon workload, etc., you'd have to measure it to see if you care..
[19:59] <anubhav> okay, the reason i am considering this is because if i have it virtualized, then I can allocate a smaller instance as well to someone who needs only decent power. There can be one large instance reserved for simulations and other tasks to be run
[20:01] <anubhav> is the performance difference between a beowulf cluster and a virtualized cluster with same specs significant, assuming that at a given moment only a few users are online and submitting jobs
[20:06] <sarnold> anubhav: you'd probably be writing custom code to use the features of a beowulf cluster; is your team writing MPI-based calculations? that might help..
[20:07] <anubhav> sarnold, not my team. but there is another and they are using libraries that i am pretty sure have good mpi implementation
[20:08] <anubhav> sarnold, I am involved in managing the cluster and figuring out which might be the best path
[20:09] <sarnold> anubhav: then you might want something more like beowulf rather than openstack. of course, if you've got mixed loads, openstack might still be useful, you could probably build a beowulf out of virtualized systems... but .. I'm scared to suggest it. :)
[20:09] <anubhav> sarnold, I understand. I was also sceptical is virtualizing would be a good thing
[20:10] <sarnold> anubhav :)
[20:10] <anubhav> sarnold, thanks for the suggestions. I will look deeper into beowulf
[20:12] <sarnold> anubhav: have fun :)
[20:24] <nch> hello can someone help me with, grub, mdadm (raid 5 with 8x2tb) and a boot error: "device with uuid not found"?
[20:25] <Prezident> did you try boot grub without uid ?
[20:25] <nch> how can I do that?
[20:26] <Prezident> In your grub file
[20:26] <Prezident> GRUB_DISABLE_LINUX_UUID=true
[20:26] <Prezident> Have you installed/loaded kernel to /boot?
[20:27] <Prezident> This is three errors i see, this is for the boot error.
[20:28] <nch> yep all worked well put I reconneted some harddrives to other sata ports and then I got this error
[20:28] <nch> you sure that I can disable the uuids, because all drives in the grub config file are defined by uuids
[20:29] <Prezident> Aha, so this happend after you installed some drives?
[20:30] <nch> no the drives are all the same as before I only changed some cables and I think I reconnected some drives to other sata ports
[20:30] <Prezident> Aha, bios can see them all?
[20:30] <nch> I read that the uuids wont affect the raid because the raid information is stored directly on the drives, am I right?
[20:30] <nch> yes all drives are available in bios
[20:32] <nch> http://sourceforge.net/projects/boot-repair-cd/
[20:33] <nch> is this compatible with raid and gpt?
[20:48] <nch> well boot repair tells me to create a partition with the bios_grub flag, but one of my drives have it -.-