[00:08] <rostam> HI I would like upon ssh to my ubuntu 12.04, the /etc/profile to be sourced, is there a way to configured this on any login? thx
[00:11] <Daviey> zul: hey
[00:11] <sarnold> rostam: if you're using bash, that should just happen automatically...
[00:13] <rostam> sarnold: that is what I thought but it does not,  I have to manually do . /etc/profile so the changes to take effect??
[00:14] <sarnold> rostam: are you changes in /etc/profile over-ridden by settings in ~/.bashrc or ~/.*profile files?
[00:20] <rostam> sarnold, it does not override that I am sure...
[00:22] <rostam> sarnold there is a specific profile which I have placed it in the /etc/profile.d once the /etc/profile is source that file also gets source
[01:00] <zul> Daviey:  still around?
[07:39] <frojnd> Hi there.
[07:39] <cfhowlett> frojnd, greetings
[07:39] <frojnd> I have a huge problem. Somehow my IP is in /etc/hosts.deny
[07:39] <cfhowlett> d'oh!
[07:39] <frojnd> even if I remove that entry it keeps coming back :) so I can't ssh to it with another temrinal
[07:40] <frojnd> luckily I'm still logged in by one terminal
[07:40] <frojnd> would it helped if I put my Ip in hosts.allow?
[07:41] <frojnd> but would like to know how my home ip ended up in hosts.deny in the first place
[07:44] <frojnd> cfhowlett: any ideas? :=
[07:44] <frojnd> :)
[07:45] <cfhowlett> frojnd, sorry, no.  I'm silent on areas outside my range of knowledge --- i.e. most server problems.  As we seem to be the only ones awake here, perhaps moving the query to #ubuntu might be in order.
[07:52] <frojnd> nice idea
[07:57] <frojnd> noone there either
[07:57] <frojnd> :)
[08:32] <frojnd> It was denyhosts
[08:32] <frojnd> interestingly
[08:33] <cfhowlett> frojnd, so how ??
[08:35] <frojnd> it was the configuraton file. I wasn't careful when I was updating and in /etc/denyhosts wasn't set PURGE_DENY
[09:05] <jamespage> zul, you needed a allow-stderr restriction in tests/control
[09:06] <jamespage> otherwise stderr output == test failure
[10:48] <yolanda> jamespage, smoser, zul, what do you think about that? https://code.launchpad.net/~yolanda.robla/keystone/icehouse_fix-distribution/+merge/198275
[12:15] <jamespage> yolanda, how would you feel about targetting that to upstream first?  I'd like to know that there is appetite to accept this type of header data as normal before we start adding patches to all of the packages.
[12:16] <yolanda> jamespage, ok, what's the process to send some patch upstream? just submit like a code change to gerrit?
[12:16] <jamespage> yolanda, yes
[12:16] <yolanda> should i create some bug first?
[12:17] <jamespage> yolanda, sounds good
[12:17] <yolanda> ok
[12:17] <jamespage> people can then comment on the bug
[12:17] <jamespage> yolanda, that approach means it can be overridden by configuration right?
[12:17] <yolanda> jamespage, yes, if you add this section to config, it will be overriden
[12:18] <yolanda> it only shows platform.dist() on default
[12:18] <jamespage> if so it would also be nice if there was a way to disable adding the header altogether
[12:19] <jamespage> Distribution = None
[12:19] <jamespage> or something similar
[12:23] <yolanda> ok
[12:23] <yolanda> i'll do some tests with it and then send upstream
[12:28] <jamespage> zul: https://code.launchpad.net/~james-page/cinder/python-six-1.4.1/+merge/198378
[12:34] <jamespage> zul, working through some neutron bugs today - might do an 0ubuntu2 later
[12:35] <zul> jamespage:  can you look at that horizon merge from yesterday as well
[12:35] <jamespage> zul, oh - yes of course
[13:07] <Daviey> zul: you wanted me?
[13:08] <zul> Daviey:  yeah when you get a sec can you do a source new review for me?
[13:10] <Daviey> zul: oslo.rootwrap or swift-bench?
[13:28] <zul> Daviey:  both
[13:30] <Daviey> zul: Ok, should be able to do it in about 1.5 hours.  Tied up in meetings until then.
[13:30] <zul> Daviey:  cool thanks..
[13:45] <zul> jamespage:  horizon is ok?
[14:29] <smoser> yolanda, that looks good
[14:30] <yolanda> sent upstream
[14:30] <mdeslaur> does anyone know how to create an explicit deny ACE in samba?
[14:30] <smoser> is there any reason you wouldnt want to default to something like '/'.join(platform.dist)
[14:30] <smoser> maybe not.
[14:32] <smoser> yolanda, the only othe rcomment i have is why force 'X-' prefix ?
[14:32] <smoser> as opposed to letting that be configured.
[14:33] <yolanda> smoser, normally in http all extra headers are prepended with X- , right?
[14:33] <yolanda> i just tried to send X-Distribution in config file but it was transformed to X_Distribution
[14:36] <smoser> well that sucks.
[14:36] <smoser> i understand all "extra headers" being normally prefixed with 'X-'
[14:37] <smoser> but don't know who's definition of 'extra' is right.  if you allowed thatin the config, then the provider can choose to put 'X-' or not put 'X-'.
[14:38] <yolanda> smoser, but then i'll need to replace the _ with - in headers, or they won't look right. And if some people just wanted to send a header with _ ?
[14:38] <yolanda> maybe just look for starting X_ and transform to X-
[14:58] <yolanda> smoser, jamespage, what do you think about the X- headers then?
[14:58] <smoser> yolanda, i'm fine to let upstream discuss it and take it either way.
[14:58] <smoser> if it were my choice, i'd not want to force any restrictions on the headers.
[14:58] <smoser> but i dont think its a very big deal.
[14:58] <smoser> annoying that config format squashes your '-'
[14:59] <yolanda> smoser, yes, my first try was to don't force the X- and give freedom
[14:59] <yolanda> let's see if people add comments there
[14:59] <smoser> someone (me) might do it :)
[15:00] <gyre007> can someone tell me if Precise LTS has ipvs support in kernel ?
[15:04] <gyre007> basically lvs support is what Im after
[15:09] <jrwren> gyre007: signs point to no: http://pastebin.ubuntu.com/6551563/
[15:09] <gyre007> oh no
[15:10] <jrwren> you can do lxc.
[15:10] <jrwren> https://www.stgraber.org/2012/05/04/lxc-in-ubuntu-12-04-lts/
[15:10] <gyre007> not really...can I ?
[15:11] <gyre007> but I will need an image which has kernel with ipvs support compiled in
[15:11] <jrwren> afaik, lxc does not use ipvs or lvs kernel mods
[15:11] <gyre007> I just actually installed keepalived on KXC
[15:11] <jrwren> instead it uses namespaces and cgroups, which are newer and better.
[15:11] <gyre007> *LXC
[15:12] <gyre007> and I get the http://askubuntu.com/questions/388768/lvs-support-lts-12-04
[15:12] <jrwren> but I am not an expert in this area.
[15:12] <gyre007> LXC runs of the host's KErnel AFAIK
[15:12] <jrwren> yes, there is only one kernel with lxc
[15:12] <gyre007> if host's Kernel' doesnt have lvs support you're screwed
[15:12] <jrwren> so you really want to run keepalived ?
[15:12] <gyre007> so am I :-)
[15:12] <gyre007> well yeah
[15:13] <gyre007> whats wrong with that ? :)
[15:13] <jrwren> nothing, i'm just trying to understand what you want :)
[15:13] <gyre007> basically I need something which will give me vrrp
[15:13] <gyre007> and that's keepalived
[15:13] <gyre007> but it needs lvs
[15:13] <jrwren> you running routing protocols?
[15:14] <gyre007> Im trying to set up HA load balancer...
[15:14] <gyre007> with auto failover
[15:14] <gyre007> and keepalived looks like a pretty good candidate
[15:14] <jrwren> cool.
[15:15] <gyre007> well not really :-) if the kernel doesn't support it then....
[15:15] <jrwren> vrrpd or ucarp?
[15:15] <gyre007> I dont think I need vrrpd
[15:15] <gyre007> keepalived is actually filling the role of vrrpd right ?
[15:16] <jrwren> i'm suggesting either of those instead of ipvs
[15:16] <jrwren> i don't know keepalived or ipvs very well.
[15:16] <gyre007> but can keepalived work with those ?
[15:16] <jrwren> I think they would be used instead of keepalived
[15:16] <gyre007> mm
[15:17] <jrwren> i wish I knew someone esperienced with them to ask which to use, but I don't.
[15:33] <arie_kiyoshi> hello guys
[15:38] <zul> jamespage:  ping...horizon review?
[15:45] <stgraber> hallyn_: hey, did you ever get sarnold to re-check lxc-user-nic? I'm getting tired of manually setting the setuid bit everytime I update LXC :)
[16:04] <jamespage> yolanda, how about taking a look at active/active rabbitmq across the openstack charms
[16:04] <jamespage> its on the blueprint and is an important delivery this cycle
[16:04] <jamespage> https://blueprints.launchpad.net/ubuntu/+spec/servercloud-1311-openstack-charms
[16:05] <yolanda> wow, sounds challenging!
[16:05] <jamespage> yolanda, oh - and merges
[16:05] <yolanda> yes, i was taking a look at merges right now :)
[16:05] <jamespage> I'm sure there are still some on the list - stuff lands in debian all the time!
[16:07] <hallyn_> stgraber: not since I pulled some code out
[16:07] <beatstreet> I have an ubuntu server that keeps locking up and requires reboot to get it back online
[16:07] <hallyn_> sarnold: ^
[16:08] <beatstreet> what's the best log to look up
[16:13] <zul> jamespage:  waitress and bs4 have been promoted fyi
[16:16] <hallyn_> waitress?
[16:20] <zul> hallyn_:  python-waitress -  dependency for python-webtest - dependency of keystone
[16:21] <jamespage> zul, marvellous
[16:23] <zul> jamespage:  ill wait for it to get out of proposed and run boom
[16:24] <jamespage> zul, how are we on the rest of icehouse-1?
[16:25] <jamespage> glance - dep wait
[16:25] <zul> checking glance
[16:25] <zul> jamespage:  argh..
[16:26] <jamespage> zul, ceilometer - rebuilding now - bust on webtest
[16:27] <zul> jamespage:  glance rebuilding now
[16:27] <caribou> smoser: I saw that you mentionned merges during the meeting, any server specific list of merges or merge-o-matic is the source ?
[16:27] <caribou> smoser: I started to look at those & a few ftbs
[16:28] <zul> jamespage:  btw https://code.launchpad.net/~zulcss/horizon/2014.1.b1/+merge/197957
[16:28] <jamespage> zul, doing it now
[16:28] <zul> jamespage:  thanks
[16:30] <smoser> caribou, i didn't think anyone would call me on that :)
[16:30]  * smoser goes to look for some stuff
[16:30] <smoser> grep-merges is helpful.
[16:31] <smoser> caribou, but yeah, generally merge-o-matic is your friend there.
[16:31] <smoser> i could look for something that looked "easy", but i dont hink you need something terribly easy.
[16:34] <caribou> smoser: I've done a couple already
[16:34] <caribou> smoser: I helped with crash & libpng only to find out that the original person responsible had done it already
[16:34] <jamespage> zul, it was still missing static assets, so I re-did them and pushed, and uploaded
[16:34]  * smoser should probably do mine.
[16:34] <Daviey> zul: oslo.rootwrap, where did the orig tarball come from?
[16:34] <smoser> caribou, yeah, its always best to try to ping the last touch'd
[16:35] <caribou> smoser: I'll look around and see what I can find
[16:35] <Daviey> jamespage: binNEW on ceilometer?
[16:35] <jamespage> ?
[16:35] <Daviey> jamespage: binNEW on ceilometer-agent-notification ?
[16:36] <jamespage> yeah - thats OK
[16:36] <zul> Daviey:  pypy
[16:36] <jamespage> its new this cycle
[16:36] <zul> Daviey:  er...pypi
[16:37] <zul> Daviey:  we know the upstream author pretty well https://pypi.python.org/pypi/oslo.rootwrap/1.0.0
[16:38] <Daviey> zul: right, but it makes it quite hard to verify the source when the only hint i get is "Source: https://github.com/openstack/oslo-incubator"
[16:39] <zul> Daviey:  argh i need to fix that up
[16:39] <Daviey> zul: Fany adding a get-orig-source?
[16:39] <Daviey> fancy*
[16:39] <zul> Daviey:  sure i can do that
[16:41] <Daviey> zul: ok, rejected that.. with that change (& copyright entry) happy to accept
[16:41] <zul> Daviey:  ok gimme a sec dont go anywhere
[16:44] <beatstreet> nsk
[16:44] <zul> Daviey:  just uploaded
[16:45] <zul> jamespage:  https://launchpad.net/ubuntu/+source/glance/1:2014.1~b1-0ubuntu1/+build/5317701
[16:52] <hallyn_> win 31
[16:53] <jamespage> zul, rebuilding neutron as well - webtest dependency issues
[16:54] <jamespage> zul, so i think
[16:54] <jamespage> cinder: done
[16:54] <jamespage> ceilometer: done
[16:54] <jamespage> heat: done
[16:54] <jamespage> glance: done
[16:54] <jamespage> nova: done
[16:55] <jamespage> neutron: inprogress
[16:55] <jamespage> horizon: done
[16:55] <jamespage> keystone: done
[16:55] <zul> ack
[16:55] <jamespage> swift: nothing todo right?
[16:55] <jamespage> zul, ^^
[16:56] <zul> jamespage:  well they just did a rc this morning
[16:56] <jamespage> urgh
[16:56] <zul> i noticed
[16:56] <zul> so gimme a sec
[16:56] <jamespage> ack
[16:56] <hallyn_> stgraber: bug 1254338   can you see any downside to detecting kernel version in /etc/init/lxc-net.conf and donig the iptables rule for veth checksum offloading if needed?
[16:58] <stgraber> hallyn_: in theory there will be a small CPU overhead in doing that when it's not needed (recent dhcp client) but no, I'm not opposed to it. If we do it, I think it'd probably be easier to just always add that rule (like I believe libvirt does).
[16:58] <zul> jamespage:  actually its milestone proposed...so i would rather wait til thursday for that one
[16:58] <jamespage> zul, after christmas we should get trove out of -proposed and into CI as well
[16:58] <jamespage> we don't need to MIR it yet
[16:58] <zul> jamespage:  or before
[16:58] <jamespage> but it should at least work :-)
[16:58] <zul> in theory yes :)
[16:58] <hallyn_> smoser: bug 1090223, maybe it's time to request vmbuilder be pulled from the archive
[16:59] <zul> hallyn_:  yes please :)
[16:59] <hallyn_> stgraber: well id on't really want to end up with 200 copies of that rule though
[16:59] <hallyn_> hm.  i wish iptables was a little more useful :)
[16:59] <hallyn_> zul: how do we go about that?
[17:00] <zul> hallyn_:  open up a bug and ask the archive admin to do it
[17:00] <stgraber> hallyn_: isn't libvirt adding an interface-specific rule?
[17:00] <zul> that reminds me we should remove xcp as well
[17:00] <stgraber> hallyn_: yeah, it's:
[17:00] <stgraber>     5  1656 CHECKSUM   udp  --  *      virbr0  0.0.0.0/0            0.0.0.0/0            udp dpt:68 CHECKSUM fill
[17:01] <hallyn_> stgraber: ok, cool.  so i guess i'll add that in trust, but i don't wanna sru
[17:01] <hallyn_> zul: open bug against vmbulider itself, or against ubuntu?
[17:01] <stgraber> hallyn_: yeah, that's fine. I expect it'll mostly hit people upgrading from 12.04 anyway
[17:01] <hallyn_> smoser: any objection to that (removing vmbuilder from archive)
[17:02] <zul> vmbuilder itself
[17:02] <hallyn_> stgraber: and those cloud-archive people zul likes to cultivate
[17:02] <zul> hallyn_:  meh :)
[17:02] <hallyn_> zul: <sigh> i wonder if we should firrst see if any users want to fork+maintain it
[17:02] <hallyn_> nah, we can always add it back i guess
[17:03] <zul> hallyn_:  i dont think anyone wants to maintain it
[17:03] <zul> too buggy, too old
[17:03] <Daviey> hallyn_: I'd like to see it proposed on ubuntu-server / ubuntu-devel first.
[17:04] <Daviey> Some crazy people use it.. they should get the chance to step up and maintain it :)
[17:04] <Daviey> (unlikely)
[17:04] <hallyn_> jodh: btw, github pull requests for cgmanager and lxc get misfiled in my mail.  until i fix that, if anything urgent looks like it's being ignored pls just yell at me
[17:05] <hallyn_> Daviey: poor slobs like https://bugs.launchpad.net/ubuntu/+source/vm-builder/+bug/1090223 are losing data
[17:05] <hallyn_> but, ok.
[17:08] <Daviey> hallyn_: you make a good point.. but it seems this will continue to happen for the foreseeable future if this is removed anyway.  I guess someone should fix it.. /me hides. :)
[17:08]  * hallyn_ throws the broken code and a copy of vim under the blanket Daviey is hiding under
[17:08] <hallyn_> "you'll need those" :)
[17:09]  * Daviey wonders where soren is hiding.
[17:10] <smoser> hallyn_, i have no such objection.
[17:12] <hallyn_> best to have it out before trusty is released, i think
[17:15] <Daviey> hallyn_: your right.  How about raise a bug for removal, lets do it.. then state the fact on the ML's and suggest it could be re-introduced with a reliable offer to maintain it.
[17:17] <hallyn_> Daviey: too late, sent the email.
[17:17] <jamespage> zul, urgh - testing from the trunk testing ppa; nova can't connect to libvirt
[17:17] <hallyn_> 'll open the bug tomorrow
[17:18] <jodh> hallyn_: ack, thanks.
[17:19] <zul> jamespage:  log?
[17:19] <jamespage> zul, http://paste.ubuntu.com/6552155/
[17:19] <zul> jamespage:  why permission denied?
[17:30] <hallyn_> stgraber: did i do anything stupid here?  http://people.canonical.com/~serge/lxc-dhcp.debdiff
[17:30]  * hallyn_ fires up a precise host to test on
[17:34] <hallyn_> jodh: oh hey, we now have the 'official' github tree at github.com/cgmanager/cgmanager.  probably best to start doing merge requests against it
[17:35] <hallyn_> github.com/hallyn/cgmanager will become my devel archive, so i can push prettier commits to the real tree :)
[17:36] <jodh> right. I have got one o/s pull request (https://github.com/hallyn/cgmanager/pull/3). Not sure why that's showing up against your branch though as I had thought I'd used the new official branch.
[17:37] <hallyn_> jodh: also, the realpath_escapes() fixup, i think it's better to do it the other way
[17:37] <hallyn_> so always have the actual raising of errors be in the top level function
[17:37] <hallyn_> so we don't have to guess whether the fn we called already did it
[17:37] <stgraber> hallyn_: you probably want a matching -D rule on stop?
[17:38] <hallyn_> stgraber: not sure, i was wondering if it would jsut go away when we delete the device
[17:38] <jodh> hallyn_: yeah, I did wonder about that.
[17:38] <hallyn_> stgraber: oh but i'm using the network not the device :(
[17:39] <stgraber> hallyn_: oh yeah, -o should be passed the interface name
[17:39] <stgraber> hallyn_: and I just checked, iptables doesn't care whether an interface exists or not
[17:39] <stgraber> hallyn_: so you'll need to clear it in stop
[17:39] <hallyn_> drat
[17:39] <hallyn_> well then i don't need the -o
[17:40] <stgraber> you don't need it but you should have it
[17:40] <stgraber> otherwise iptables will try to fill the checksum on all interfaces
[17:41] <stgraber> what we want is: iptables $use_iptables_lock -t mangle -A POSTROUTING -o ${LXC_BRIDGE} -p udp -m udp --dport 68 -j CHECKSUM --checksum-fill
[17:41] <stgraber> and the same but -D in stop
[17:41] <hallyn_> stgraber: and the device should be the bridge tiself?
[17:41] <hallyn_> ok
[17:41] <stgraber> hallyn_: right. We can't filter using -d and -s since the DHCP traffic is broadcast, so we need to use -o with LXC_BRIDGE
[17:44] <hallyn_> jodh: so that's the downside of these pull reqeusts :)  I see a whole bunch of commits there.
[17:44] <hallyn_> lemme try and apply them 1-by-1
[17:44] <hallyn_> actually no,
[17:45] <hallyn_> let's see if we can get the pll request aimed at cgmanager/cgmanager :)
[17:47] <hallyn_> jodh: changed my mind again :)  pulling, will move the errors out, and push to the other tree.  thanks!
[17:54] <yolanda> jamespage, what's that neutron plugin doc?
[17:54] <jamespage> yolanda, ideas for how to manage neutron plugins via packaging a little easier
[17:55] <jamespage> zul, b-o-m running
[17:55] <zul> ack
[18:01] <jodh> hallyn_: thank you!
[18:02] <hallyn_> no, thank you :)
[18:37] <jamespage> zul, CA staging should be all up-to-date now
[18:39] <zul> jamespage:
[18:39] <zul> jamespage:  ack
[18:47] <jamespage> zul, OK - all done
[18:47] <jamespage> going to the openstack meeting for beer now
[18:47] <zul> lucky bastard ;)
[18:53] <w0rmie> whats the fonctional difference between dhcp3-server and isc-dhcp-server?
[18:54] <Nomad_> Can anyone help with making my 13.10 .iso into a valid usb bootable image?  I have used YUMI from the pendrivelinux team to make an image but every time I try to install the server just reboots.  If I try to do a rescue or memcheck it tells me /linux/* is missing.  I get the same thing with 12.10 and 13.10 desktop as well.
[18:54] <jpds> w0rmie: $ apt-cache show dhcp3-server
[18:55] <jpds> w0rmie: N: Can't select versions from package 'dhcp3-server' as it is purely virtual
[18:56] <sarnold> Nomad_: just use dd: dd if=/path/to/ubuntu.iso of=/dev/sdg  (or whatever your usb device is, check using dmesg to see what address it is assigned..)
[18:57] <w0rmie> jpds: the problem is that i need to modify /etc/default/dhcp3-server so i don't know really what's its equivalent by using isc-dhcp-server
[18:58] <Nomad_> sarnold: dd the iso to the usb stick?
[18:59] <Nomad_> didn't think that worked straight up
[18:59] <sarnold> Nomad_: it does because the ubuntu ISOs are 'hybrid isos', designed to work that way. it's convenient, especially now that the isos are too large to fit on CDs. hehe.
[19:00] <Nomad_> oh, nice to know
[19:00] <Nomad_> thank you
[19:01] <jpds> w0rmie: It'll be under /etc/dhcp/*
[19:05] <hallyn_> stgraber: uh.  say.  in lxc blueprint, what does this mean?  Update the upstart job to setup the LXC profile as a namespace
[19:06] <stgraber> hallyn_: I think there's an "apparmor" missing in there
[19:07] <stgraber> hallyn_: from what I remember, jjohansen said we'd need to pass some extra flag to apparmor when loading the lxc profile so that it'd allow profile stacking by setting up a new namespace
[19:07] <hallyn_> stgraber: ah, thanks
[19:09] <syfhvbgos7> WARNING       WARNING      WARNING,                       WARNING
[19:09] <syfhvbgos7> WARNING             WARNING              WARNING,     WARNING         WARNING
[19:09] <syfhvbgos7>  YOU MAY BE WATCHED
[19:09] <syfhvbgos7> YOU MAY BE WATCHED
[19:09] <syfhvbgos7>                 YOU MAY BE WATCHED
[19:09] <syfhvbgos7> Do usa&israel use chat&social communication prog(facebook&twitter) to collect informations,,,,can we call that spying!!!!
[19:09] <syfhvbgos7> Do usa&Israel use chat &facebook 2 spy?!?!?!?
[19:56] <NBeeuwsaert> Hey, would this be an appropriate place to ask about downloading the messages from one server to another?
[20:13] <err-or> NBeeuwsaert: what?
[20:14] <NBeeuwsaert> I need to have one server fetch mail from a user on another server, I was wondering if there was an efficient way to do that without having to set up imap or pop
[20:15] <NBeeuwsaert> (Also if this is the appropriate place for asking that?)
[20:15] <err-or> NBeeuwsaert: did you have a look on fetchmail already?
[20:15] <err-or> NBeeuwsaert: don't ask to ask!
[20:16] <err-or> NBeeuwsaert: but i'm not sure it works without mda
[20:16] <NBeeuwsaert> yeah, but then I'd have to set up dovecot or courier on the server, and that seems excessive
[20:16] <err-or> NBeeuwsaert: its not that much :P and you have lots of more experience you could add for your next job request ;)
[20:17] <NBeeuwsaert> Yeah that's true :P
[20:17] <NBeeuwsaert> I Was thinking setting up postfix to use Maildir, and then rsyncing down from the server
[20:17] <err-or> i run fetchmail - postfix - dovecot - thunderbird (or any other imap capable mail-client)
[20:20] <err-or> NBeeuwsaert: this one is a good one ;) https://help.ubuntu.com/community/PostfixVirtualMailBoxClamSmtpHow
[20:20] <jamespage> zul: everything flushed through to proposed
[20:20] <jamespage> I'll re-run the deploy test I did with those packages
[20:21] <zul> jamespage:  ack
[20:23] <err-or> NBeeuwsaert: https://help.ubuntu.com/community/PostfixVirtualMailBoxClamSmtpHowTo sry :P
[20:24] <err-or> NBeeuwsaert: arg... https://help.ubuntu.com/community/PostfixVirtualMailBoxClamSmtpHowto
[20:26] <NBeeuwsaert> lol did you type those by hand?
[20:28] <err-or> NBeeuwsaert: nope first i forgot to copy the 'to' and entered 'To' manually. don't just copy & paste !! :D
[20:28] <NBeeuwsaert> ok ok :P
[21:12] <NBeeuwsaert> ack, hit a snag
[21:12] <NBeeuwsaert> no permission to put fetchmail on the server
[21:16] <d1n0> i have a ipmipower process consuming 100% cpu, i got a feeling its from a node i didn't delete properly
[21:17] <NBeeuwsaert> just kidding, fetchmail was always there, and nobody can say otherwise >:)
[21:19] <beatstreet> I have an ubuntu server that keeps locking up and requires reboot to get it back online
[21:19] <beatstreet> what's the best log to look up
[21:20] <guntbert> beatstreet: /var/log/syslog has lmost everything
[21:20] <beatstreet> ok
[21:20] <guntbert> *almost
[21:20] <beatstreet> thanks
[21:21] <w0rmie> nfs-commons is installed but no way to symlink /etc/network/if-up.d/mountnfs within /etc/rcS.d/S35mountnfs
[21:21] <w0rmie> i can't find mountnfs in the /etc/network/if-up.d
[21:24] <FunnyLookinHat> Hey - got a quick question for you guys - I'm trying to migrate a ton of files from one server to another ( slow transfer rate, long story ) and I want to run my rsync job only between midnight and 5am... can you think of a clever way to run an rsync for only that specified time period other than cronning a start and kill script ?
[21:28] <NBeeuwsaert> FunnyLookinHat, cron?
[21:28] <NBeeuwsaert> OH wait, I read that weird
[21:29] <NBeeuwsaert> I think rsync has a --time-limit option
[21:30] <sarnold> FunnyLookinHat: if you've got a spare few days to learn about tc, you could keep it running continuously but change the bandwidth allotted to the transfer during 'work' hours so it won't eat so much bandwidth...
[21:33] <FunnyLookinHat> sarnold, already using --bwlimit
[21:33] <FunnyLookinHat> Ooh I see what you mean
[21:40] <w0rmie> nfs-commons is installed but no way to symlink /etc/network/if-up.d/mountnfs within /etc/rcS.d/S35mountnfs
[21:40] <w0rmie> i can't find mountnfs in the /etc/network/if-up.d/ folder
[21:41] <d1n0> Can anyone help with this error: could not access file '334b532c-c71f-483e-8f36-beea1905536d-provider-state': gomaasapi: got error back from server: 401 UNAUTHORIZED
[21:42] <d1n0> I get that when trying to bootstrap the maas
[21:46] <NBeeuwsaert> FunnyLookinHat, Looks like rsync has a "--time-limit" and "--stop-at" option in version 3.1 (with a patch) so if yours feeling upto compiling rsync you could give that a go
[21:46] <FunnyLookinHat> Yeah but it's not built into the rsync I have
[21:46] <FunnyLookinHat> so I just used timeout
[21:46] <FunnyLookinHat> :)
[21:47] <FunnyLookinHat> NBeeuwsaert, timeout <seconds> <command>
[21:47] <FunnyLookinHat> And it just send a kill to the process after the seconds have elapsed :)
[21:47] <NBeeuwsaert> huh I thought that was just for individual transfers, and would fail it if it took too long
[21:52] <igalic> I. Hate. Preseed.
[21:57] <FunnyLookinHat> NBeeuwsaert, Yeah that's what I thought too - but tested and it works  :)
[21:57] <NBeeuwsaert> it won't cut a file off halfway?
[22:39] <w0rmie> is there any way to capture current data on RAM and save it into recoverable image from HDD?
[22:51] <Patrickdk> w0rmie, hibernate?
[23:03] <w0rmie> Patrickdk: thank you for your answer, so can i recuperate the RAM image as a file to store on the HDD and "des-hibernate" it to make it readable?
[23:06] <w0rmie> i'm installing a cluster to execute too much data on my 72GB of RAM so i would to save the current RAM work in order to reload other calculations
[23:08] <sarnold> yikes, that does not sound fun, and not exactly something to try to get the hang of in the middle of an important job..