[01:38] <pmatulis> thebwt: huh?
[02:12] <ScottK> thebwt: You mean MaaS on a public network?
[02:12] <ScottK> If so, the only documentation you need is "Don't do it".
[02:50] <engkur> hi all
[09:04] <Womkes> Is there a php5- package to install ioncube loader?
[09:50] <_johnny> hi, i'm running 12.04 and trying to add a redirect from port 8080 to 80 in iptables. the rule is saved but doesn't appear to have any effect. not being an expert in iptables, but isn't this "right" http://pastebin.com/nVms8wti ?
[09:54] <_ruben> PREROUTING isn't used for local packets
[09:54] <_ruben> do you want to redirect from anywhere, or just the local machine?
[09:55] <_johnny> it is just for my local machine
[09:55] <_ruben> try OUTPUT instead of PREROUTING
[09:56] <_johnny> works! thanks _ruben :)
[10:03] <Qu310> hi all, anyone using ataoe in production with zfs?
[10:15] <koolhead17> Daviey: ping
[14:23] <smb> stgraber, I think I found a fix for bug 1006937. I linked a branch with my fix to it.
[14:24] <stgraber> smb: thanks! I'll try to remember to have a look post-alpha1
[14:25] <smb> stgraber, Somehow its amazon how well the client works without finding its config file... :-P
[14:25] <smb> amazing I meant
[14:25] <smb> bah
[14:26] <stgraber> smb: ;)
[14:29] <smb> btw, what is the bzr equivalent to "git checkout -f"? (reset the working tree to what is checked in and dropping any other modifications)
[14:42] <stgraber> smb: bzr revert? (not sure exactly what git checkout -f does...)
[14:43] <stgraber> smb: did you manage to build isc-dhcp with that change? I remember trying CFLAGS=$(CFLAGS) before going with the really ugly hack of just hardcoding the one I needed and it failed for some weird reason
[14:43] <stgraber> though I didn't think of export ;) maybe export does something slightly different that makes it magically work
[14:44] <smb> stgraber, maybe. tried to explain above (get rid of any changes uncommited) I did build with sbuild in a quantal chroot. Yeah it seems odd that export is needed now but looking at the failures without it no CFLAGS defined in rules seemed to be passed on.
[14:45] <smb> And yes, tried the CFLAGS="$(CFLAGS)" road, too but unsuccessful
[14:46] <smb> Problem is passing all the quotes through shell ...
[14:48] <smb> stgraber, Ah, revert seems to be the command I was looking for
[15:03] <stgraber> smb: uploading isc-dhcp now as we'll be respinning everything and I'm not sure of all the possible side effects of not using dhclient.conf and the dhclient script
[15:06] <smb> stgraber, Ah ok. Yeah, I only noticed the not sending the hostname part because I grumbled about my quantal VMs not registering with my dhcp server.
[15:07] <stgraber> I'm pretty sure not using our dhclient.conf will break dhcpv6 at least
[15:07] <obelus> Is this for the new package in quantal?
[15:12] <smb> obelus, The discussion the isc-dhcp package, yes, quantal
[15:25] <Error404NotFound> I am running an Ubuntu server and its facing quite high load. http://pastebin.com/sqT9ZFNE As you se i have much of free ram, cpu is mostly idle, no i/o wait and very less % of cpu in used state. Any ideas?
[15:25] <Jeeves_> 92.0%id !
[15:27] <Error404NotFound> Jeeves_: yes, and how does that ring a bell? I would have assume an I/O problem but 0.0%wa. I can't comprehend the reason for high load when we have 7.0%us,  0.8%sy,  0.0%ni, 92.0%id,  0.0%wa,  0.0%hi,  0.1%si,  0.0%st, seems like stats of well behaving system.
[15:28] <patdk-wk> load has nothing to do with cpu usage
[15:29] <Error404NotFound> patdk-wk: true, but usually cpu util stats have some indication. In this case i fail to see that. Enough free ram, enough free swap, no misbehaving process, enough free cpu, no i/o issues....
[15:30] <patdk-wk> well, in this case, I see a load of 4
[15:30] <patdk-wk> cpu of 7%
[15:30] <patdk-wk> how many cores does that machine have?
[15:31] <Error404NotFound> 8
[15:31] <Error404NotFound> Intel(R) Xeon(R) CPU E5506  @ 2.13GHz
[15:33] <patdk-wk> single cpu? or dual?
[15:33] <patdk-wk> well, should say, single socket?
[15:33] <Error404NotFound> sorry?
[15:34] <patdk-wk> how many E5506's?
[15:34] <patdk-wk> oh, must be 2, no hyperthreading support
[15:35] <Error404NotFound> this one is actually c1.xlarge instance, /proc/cpuinfo = http://pastebin.com/fTuFMmQR
[15:35] <Error404NotFound> i am gonna be right back in 15 minutes, have to run to attend to something real quick before i get back to this stupid issue.
[15:36] <patdk-wk> hmm, atleast it's 0% st
[15:37] <hallyn> smb: re quantal kvm blueprint - the qemu-kvm merge hasn't been done yet :)
[15:37] <hallyn> smb: waiting for debian to merge qemu-kvm 1.1
[15:37] <smb> hallyn, Oops? I was thinking this is related to the ipxe update
[15:38] <smb> hallyn, Hm, maybe you are talking of something different...
[15:41] <smb> hallyn, No, probably it is me setting the workitem to done. So will qemu-kvm have an impact on ipxe/kvm-ipxe? Since that were the places which had a problem with the prom names.
[15:44] <thebwt> ScottK: re:maas on public network. I am trying to set up a test environment for my department.  Are there any diagrams that specify the network topologys? Our department 'owns' everything on our network except DHCP... So I need a clear grasp of things before I can start a meeting on the changes needed.
[15:44] <thebwt> onlive linux
[15:44] <thebwt> oops
[15:54] <Error404NotFound> back
[15:54] <Error404NotFound> so any thoughts on what would be hammering cpu load?
[15:55] <hallyn> smb: oh maybe you're right.  that was probably what we were worried about :)
[15:55] <hallyn> smb: thanks
[15:56] <smb> hallyn, Yeah I believe. :) Though who can be sure with so much multi-tasking. ;)
[16:41] <ahasenack> hi, is there an easy way to find out the ip address that an lxc container grabbed?
[16:42] <ahasenack> I used to check the lease file from dnsmasq, but it seems to keep it all in ram at least in the lxc case
[16:43] <ahasenack> ah, nm, found the lease file
[16:46] <Error404NotFound> http://pastebin.com/FLzP7Eqv contains bit more information. CPU, RAM, I/O everything seems fine.
[16:57] <Daviey> koolhead17: hey
[17:13] <ribo> so uhh, what's this:
[17:13] <ribo> whoopsie   744  0.0  0.2 187580  4024 ?        Ssl  17:07   0:00 whoopsie
[17:14] <ribo> creash reporter.
[17:46] <thebwt> how does one go about compiling libxxx:i386 on a x64 machine?
[17:53] <ribo> ok so this is annoying. I have scripts to automatically deploy ubuntu server VMs, but now, apt-get -y --force-yes dist-upgrade right after install upgrades grub-pc which prompts for a decision on using maintainer/keep/merge /etc/default/grub, pretty much making automated build outs impossible
[17:53] <ribo> how do I get debconf grub-pc to obey --force-yes or some other unattended mode?
[17:54] <veebull> having some problems with postgresql 9.1 on ubuntu 12.04...
[17:55] <veebull> basically I used 'GRANT ALL PRIVILEGES ON DATABASE bpsimple TO nuk;'
[17:55] <veebull> but when I log in as 'nuk' and try running 'SELECT * FROM items;'
[17:56] <veebull> I get: 'ERROR:  permission denied for relation customer'
[17:56] <veebull> what am I missing?
[18:22] <sveinse> Hi. I'm about to setup an older barebone PC as 4-disk NAS (with mdadm). I'm curios to a wise partitioning scheme. Boot in raid0 thats more or less given. a) One big RAID5 partition for the rest or b) One system partition with RAID0 perhaps, and a large data storage part with raid5
[18:22] <sveinse> Any thoughts, please
[18:24] <sveinse> bummer. s/raid0/raid1/...
[18:58] <ScottK> thebwt: What problem are you having that you think maas will solve?
[21:58] <sbeattie> adam_g: around? have you seen anything like http://paste.ubuntu.com/1025828/ before?
[22:00]  * sbeattie finds bug 981111
[22:02] <adam_g> sbeattie: yeah
[22:04] <adam_g> sbeattie: the handling of migrations in glance was broken at time of (our) release. i added a patch to require un-version controlled databases to be put under VC before running any migrations. so, new databases need 'glance-manage version_control 0 && glance-manage db_sync'.
[22:06] <adam_g> upgrades of un-VC'd diablo db's: glance-manage version_control 9 && glance-manage db_sync
[22:07] <sbeattie> thanks
[23:03] <koolhead17> Daviey, hi again :)