[02:48] <VictoriaXOXO> Q: Hello! How do I FULLY disable ICMP timestamp requests and responses on my Ubuntu 14.04 box? :)
[03:10] <VictoriaXOXO> Q: Hello! How do I FULLY disable ICMP timestamp requests and responses on my Ubuntu 14.04 box? :)
[03:14] <k2gremlin> VictoriaXOXO, shut it down? lol j/k idk
[03:15] <VictoriaXOXO> k2gremlin: :|
[03:15] <OerHeks> old answer, still valid, http://blogs.reliablepenguin.com/2009/11/17/iptables-filter-icmp-timestamp-requests >> but why would you want that?
[03:30] <VictoriaXOXO> OerHeks: Dude. You didn't include my nick.
[03:30] <VictoriaXOXO> So I missed it.
[03:30] <VictoriaXOXO> OerHeks: Does this mean that it worked? http://pastebin.com/mmQtxuf8
[04:20] <trippeh_> root@mental:~# virsh start ato-pelle
[04:20] <trippeh_> error: Failed to start domain ato-pelle
[04:20] <trippeh_> error: unsupported configuration: numad is not available on this host
[04:20] <trippeh_> root      1134  0.0  0.0  19956   136 ?        Ssl  05:14   0:00 /usr/bin/numad -i 15
[04:20] <trippeh_> hrms ;)
[04:31] <trippeh_> not sure if numad is cool even, just trying things out
[04:32] <sarnold> interesting, I hadn't heard of numad before; is it libvirt-specific?
[04:32] <trippeh_> doesnt seem like it
[04:32] <sarnold> hmm, I wonder what it does that numactl doesn't already do
[04:32] <trippeh_> libvirt can query it for placement information
[04:32] <sarnold> there we go
[04:32] <trippeh_> well, recomended placement
[04:33] <sarnold> i've thought about building that myself, in part becuase I don't seem to get along with libvirt :)
[04:33] <trippeh_> libvirt tells it how many vcpus and how much memory, and it gives some recomendation back
[04:33] <sarnold> nice
[04:33] <trippeh_> "use these nodes"
[04:34] <trippeh_> of course now THP stopped working :P
[04:37] <trippeh_> hm, changed the config, it seems to do something
[04:37] <trippeh_> Wed Nov 25 05:37:32 2015: Advising pid 1969 (qemu-system-x86) move from nodes (0-1) to nodes (0)
[04:38] <trippeh_> and THP suddenly works
[04:39] <trippeh_> and numastat says all memory for the vm is on the correct node
[04:39] <trippeh_> huh
[04:39] <sarnold> woo :)
[04:40] <sarnold> now start ten more and see what happens :D
[04:41] <trippeh_> I'm trying to keep my huge Plex media VM on node 0 as all the SAS HBAs and 10Gbes are on it as well ;)
[04:41] <sarnold> your qpi link will thank you :)
[04:51] <trippeh_> could be kernel is smart enough out of the box these days tho
[04:51] <sarnold> i'm not sure about that
[04:51] <trippeh_> ok maybe not for hba/10g locality, but mem
[04:51] <sarnold> riel's been working on code to keep the memory and the processes aligned but I haven't heard of any efforts to keep the close tothe devices they need
[04:56] <trippeh_> hmm maybe thp stopped working because qemu is forgetting to madvise if you set static numa config or something
[04:56]  * trippeh_ tries thp = always instead
[05:00] <trippeh_> nopes
[05:02] <trippeh_> for some reason if numad tells qemu to use a node, thp works, but if node is set in libvirt config, it doesnt anymore
[05:03] <trippeh_> smells buggy ;)
[05:08] <trippeh_> just using a cpuset seems to do the right thing
[05:08] <sarnold> and hopefully riel's work will keep the memory local just out of convenience
[05:08] <trippeh_> numastat says all the memory is on the right node
[05:09] <trippeh_> 3190 (qemu-syste  49221      0 49221
[05:09] <sarnold> is that pid, comm, node1, node2, total?
[05:10] <trippeh_> ya
[05:10] <trippeh_> well node0, node1 ;)
[05:11] <trippeh_> altough some of the other vms seems to bleed over a little
[08:13] <Xat`> hi guys
[08:14] <Xat`> anyone knows why I'm getting "Commissionning to failed commissionning", and in maas.log I have "Marking node failed: Node operation 'Commissioning' tmed out after 0:20:00"
[08:14] <Xat`> I tried to debug this during about 5 hours, unsuccessfully
[08:19] <Xat`> I have maas 1.7
[12:50] <AtuM> I am searching for a way to integrate ubuntu to MS-AD. I know that login to ubuntu can be done with an AD account.. but would that also give me proper access to windows shares or not?
[12:51] <AtuM> I would like to use autofs to reach the necessary shares.. I am now just wondering if I'll have to set up "credentials" files or not..
[12:57] <AtuM> found it. i have to see if it works on ubuntu
[15:50] <eahmedshendy> Hi, I tried to uninstall mysql-server-5.5, then install mysql-server-5.6 but I got this error while installation: http://paste.ubuntu.com/13502312/
[15:50] <eahmedshendy> Ubuntu 14.04.3
[15:52] <ikonia> thats coming from the Mysql PPA
[15:59] <eahmedshendy> ikonia: yes, so?
[16:17] <Mik0z> Hello, I have a fairly simple question that I can't seem to find a consensus on what to do. I am trying to figure out how to setup permissions for ubuntu 14.04.1 to allow apache to host files out of /srv/www/site.com/public_html and allow a group called webmasters, to upload files, etc. Im struggling to figure out what permission set results in this configuration, while being secure
[16:35] <nat0> Can anyone tell my why preseeding a fresh install of 14.04 from a local mirror fails after searching for dists/precise-updates/Release, which doesn't even exist on the 14.04 installer DVD?
[16:36] <TJ-> nat0: something in the preseed file maybe
[16:36] <nat0> I've combed through it several times. :-\
[16:36] <nat0> Happy to put it online for other eyes to gleen though.
[16:52] <tpsilva> I'm trying to deploy Openstack with autopilot (Ubuntu 15.04), but it hangs at 82%... can anybody help me?
[18:00] <sarnold> tpsilva: try using fatrace or or something similar to see what it's doing? is there anything enlightening in any of the log files? (there ought to be enough to choose from..)
[18:20] <tpsilva> sarnold: I can see from cinder logs that rabbitmq is probably not running properly
[18:20] <sarnold> tpsilva: aha,a start.. :)
[18:21] <tpsilva> sarnold: but I don't even know which node is supposed to run rabbitmq... I'm used to single node devstack :)
[18:22] <sarnold> tpsilva: does the autopilot thing use juju to deploy the the nodes? you may be able to juju ssh rabbitmq/0 if it does..
[18:22] <tpsilva> it does use juju... let me check that
[18:25] <tpsilva> sarnold: it says that there is no environment specified, but when I try to get a list of environments (juju switch -l), it gives me no result
[18:26] <sarnold> aww :/
[18:26] <sarnold> juju switch is newer than the last time I used juju.. look around for ~/.juju directories?
[22:46] <keithzg> Huh, under 15.10 the sort order of `ls -la` is all wonky---by which I mean it ignores dots and case. Which, sure, for average users might be best, but really not what I'd want myself. What changed in that regard since 14.04?
[22:47] <keithzg> Things go back to how I expect if I export LC_ALL=C, but I can't seem to find an equivalent "incorrect" (from my perspective) setting anywhere in /etc.
[22:48] <keithzg> Or at least, the obvious culprit /etc/default/locale has the same value now as it did in 14.04 . . .
[22:49] <keithzg> ...wait, was it wonky then and I just didn't notice? Now I'm feeling like I'm a crazy person ;)
[23:10] <sarnold> keithzg: en_US and en_US.utf8 have had the annoying non-C sort for a dozen years or more
[23:11] <sarnold> keithzg: my guess is you've have LC_COLLATE or LC_ALL or LANG set to C or C.utf8 in your ~/.bashrc or something similar :)
[23:13] <keithzg> sarnold: Ah, interesting. Yeah, that's probably it then, LC_COLLATE sounds very familiar; I definitely remember monkeying around with that in the past, and most of the systems I deal with either don't have cluttered home directories or have ones longstanding and/or mounted from central network shares, so I probably "fixed" this for myself long ago and only now am noticing it isn't that way everywhere.
[23:13] <sarnold> :)
[23:14] <keithzg> sarnold: Thank you, as always :)
[23:15] <sarnold> it took me the longest time to admit defeat and accept en_US.utf8 sorting order..
[23:42] <Gnjurac> does minimal come with some partrition tool
[23:43] <Gnjurac> ?
[23:45] <tarpman> Gnjurac: fdisk(8), and I think cfdisk(8) too
[23:45] <sarnold> probably fdisk and maybe cfdisk or gdisk or .. ?
[23:45] <sarnold> gparted perhaps?
[23:45] <sarnold> parted?
[23:47] <Gnjurac> ok
[23:47] <Gnjurac> tarpman:  ty