[01:52] <seijirou> bigjools: can you help with https://bugs.launchpad.net/maas/+bug/1038295
[01:52] <bigjools> yes
[01:53] <seijirou> Outstanding, what do you suggest?
[01:54] <bigjools> what is "ubuntu-mirror"?
[01:55] <seijirou> I have no idea
[01:55] <seijirou> I was hoping you could help me with that.
[01:55] <bigjools> 403 is permission denied, so it's trying to connect to something
[01:55] <bigjools> I suspect it's a default config in the preseed
[01:55] <bigjools> can you grep the preseeds
[01:56] <seijirou> Sure I'm bringing up the initial box now.
[01:56] <bigjools> preseeds are on the maas box
[01:56] <seijirou> Yes, the maas box is what I mean.
[01:56] <bigjools> ok
[01:57] <bigjools> sorry I'm not helpful enough at the moment, I'm trying to re-sync my head with the old package since we've been concentrating on a new architecture without cobbler
[01:57] <seijirou> Understandable.
[01:57] <seijirou> Box is up, i'm trying to remember where that preseed is.
[01:58] <bigjools> /etc/cobbler/snippets and /etc/cobbler/kickstarts
[01:58] <bigjools> iirc
[01:59] <seijirou> I do have /etc/cobbler/ubuntu-server.preseed
[02:00] <bigjools> just starting up my box too, I'll hacve a look
[02:01] <seijirou_> I might need to get a real client, freenode webchat is acting up..
[02:01] <seijirou_> I found /var/lib/cobbler/snippets and /var/lib/cobbler/kickstarts
[02:04] <bigjools> ah that's it
[02:05] <bigjools> seijirou_: I think it's set here /var/lib/cobbler/snippets/maas_proxy
[02:06] <bigjools> http/proxy
[02:06] <bigjools> is your server called ubuntu-server? :)
[02:06] <seijirou_> http://tny.cz/1f664b2b
[02:07] <seijirou_> The hostname?   It's "Galaxy1"
[02:08] <bigjools> I genuinely can't fathom how ubuntu-server  gets in there then
[02:08] <bigjools> let me look at your cloud-init log again
[02:08] <seijirou_> Alright, thank you.
[02:09] <bigjools>  Aug 17 17:40:16 Galaxy2 [CLOUDINIT] cc_apt_update_upgrade.py[DEBUG]: selected mirror at: http://ubuntu-mirror/ubuntu
[02:09] <bigjools> but why ...
[02:10] <bigjools> something has gone wrong in the preseeds somewhere
[02:10] <bigjools> did you customise anything?
[02:10] <seijirou_> Not a thing, I followed that guide directly
[02:10] <seijirou_> HAha and at this point I've started from the beginning several times.
[02:11]  * bigjools scratches head
[02:11] <bigjools> nobody else had this problem afaik
[02:12] <bigjools> can you try commenting out that http/proxy line in the preseed and try again (or set it to a fixed value if you have a local mirror that works)
[02:12] <seijirou> I'll comment it out.
[02:12] <bigjools> can you connect to port 8000 on Galaxy2?
[02:13] <bigjools> squid-deb-proxy might be on there
[02:13] <bigjools> sorry not galaxy2
[02:13] <bigjools> your maas box
[02:13] <seijirou> Sure
[02:14] <seijirou> Looks like yes
[02:16] <seijirou> Would you prefer I change @@server@@ to the maas' IP ?
[02:16] <bigjools> worth a shot
[02:16] <bigjools> just to eliminate this if nothing else
[02:17] <seijirou> Done and bringing up 2nd box on PXE now.
[02:21] <bigjools> if you kept that snippet I suggested last week you should have a password set on the ubuntu user
[02:21] <bigjools> (you did customise something :) )
[02:21] <seijirou> That and the NTP thing, you caught me
[02:22] <bigjools> heh
[02:22] <seijirou> But, I wasn't trying to inject any of my own good ideas :)
[02:25] <seijirou> Same rogue "ubuntu-mirror" domain.
[02:25] <seijirou> W: Failed to fetch http://ubuntu-mirror/ubuntu/dists/precise-updates/main/binary-amd64/Packages  403  Forbidden
[02:29] <seijirou> Reading suggestion "c" in /etc/apt/sources.list in the remote node, it suggests changing /etc/cloud/templates/sources.list.tmpl
[02:30] <seijirou> I tried that in both the i386 and amd64 images but it didn't have an effect.
[02:32] <bigjools> :/
[02:32] <bigjools> this is bizarre
[02:32] <bigjools> grep -r ubuntu-mirror /var/lib/cobbler
[02:33] <seijirou> No return
[02:34] <bigjools> I am out of ideas.
[02:35] <bigjools> we need roaksoax or smoser, they are more familiar with cloud-init than I
[02:35] <UndiFineD> seijirou, is "ubuntu-mirror" a local lan host or an alias for an online mirror
[02:35] <seijirou> UndiFineD:  I don't know where "ubuntu-mirror" came from.  I have no boxes with that name.
[02:35] <bigjools> it gets a 403, so it must be connecting to something
[02:36] <UndiFineD> well, it certainly is nof fqdn
[02:36] <UndiFineD> bigjools, no it is a false name
[02:36] <UndiFineD> check /etc/apt/sources.list
[02:36] <seijirou> In the maas box or the remote node?
[02:37] <seijirou> well i can tell you the maas box has us.archive.ubuntu.com
[02:37] <seijirou> the remote node has ubuntu-mirror
[02:37] <bigjools> yeah, cloud-init sets that up
[02:37] <UndiFineD> then the remote should be corrected
[02:37] <bigjools> no
[02:37] <UndiFineD> hmm
[02:37] <seijirou> I could manually correct the remote, but i don't know what other things teh script does after that.
[02:38] <seijirou> It's a pre-requisite to a lot of other stuff I expect.
[02:38] <bigjools> cloud-init writes that file based on preseed data
[02:38] <bigjools> and other things, I expecty
[02:38] <bigjools> something is weird in its mirror selection
[02:38] <UndiFineD> well, you could work around it
[02:39] <UndiFineD> by setting a name in /etc/hosts
[02:39] <bigjools> won't work
[02:39] <UndiFineD> it should
[02:39] <bigjools> the machine is getting installed at this stage
[02:40] <bigjools> seijirou: http://askubuntu.com/questions/144393/how-to-let-maas-cloud-init-client-select-internal-mirror
[02:41] <bigjools> try that
[02:41]  * bigjools afk for food
[02:41] <seijirou> strange, mine does that archive.ubuntu.com as the mirror/http/hostname
[02:42] <seijirou> I'll try it
[02:43] <UndiFineD> michael's answer is funny, i was right on /etc/hosts
[02:46] <seijirou> I added the line, resetting.
[02:46] <seijirou> If this doesn't work I'll try /etc/hosts.  It won't hurt anything.  However, i'm unsure how i'll get /etc/hosts set.
[02:51] <seijirou> That worked.... everything was archive.ubuntu.com
[02:51] <seijirou> Sort of. I anticipated it would use the proxy IP I set
[02:54] <seijirou> Well that's progress.  http://tny.cz/e433b33a
[03:00] <bigjools> seijirou: so it will take a while to install the node
[03:03] <seijirou> bigjools:  I think that install went well.  I'm attempting 'juju deploy mysql' now to a new box.
[03:03] <bigjools> excellent
[03:04] <bigjools> this will fail if your host lookups are not working :)
[03:04] <seijirou> oy vey
[03:44] <seijirou> So far so good, no new surprises
[03:54] <bigjools> excellent
[08:39] <_val_> Hi there. Can someone tell me what svn version is availiable on Ubuntu 12.04LTS ? I mean the latest version on the debian/ubuntu repo's?
[08:42] <maxb> !info subversion precise
[08:52] <_val_> maxb: thanks.
[08:52] <_val_> THe github version of svn is 1.7.6. When do you think this will be released?
[08:56] <_val_> s/released/availiable on ubuntu repo's/
[08:58] <xnox> !info subversion quantal
[08:59] <xnox> _val_: above is a lie, since  1.7.5-1ubuntu1 is in Quantal.
[09:00] <_val_> xnox: so ubottu needs an ubbdatu? :>
[11:01] <Jake232> I',m having trouble connecting to memcached from outsie localhost, but I think the issue is probably my usage of ufw to lock down the server
[11:02] <Jake232> This should allow me to connect to port 12111, right? https://gist.github.com/8aa3c0c6f5e0e16cf741
[11:36] <henkjan> Jake232: you can't connect from outside to localhost
[11:36] <Jake232> henkjan: I thought the rules I set should allow that?
[11:37] <henkjan> Jake232: you have to edit your memcached configuration to listen on a public ip to be able to connect
[11:38] <Jake232> henkjan: I got rid of the "-l 127.0.0.1"
[11:38] <Jake232> so it should listen from everywhere now, right?
[11:38] <henkjan> i don't know if memcached can listen on all availabe ips
[11:39] <henkjan> man memcached would give you more info about that
[11:39] <Pici> -l 0.0.0.0 might let you get around that if it doesn't seem to do it by default.
[11:55] <zul> good morning
[12:17] <phillw> Hi folks, is this edit okay to do? http://pastebin.com/SkkC1HkH
[12:20] <xranby> phillw: the ubuntu documentatin is correct check your /etc/sudoers file
[12:20] <xranby> # Members of the admin group may gain root privileges
[12:20] <xranby> %admin ALL=(ALL) ALL
[12:21] <xranby> this means that if your user is part of the admin group then sudo allows your user to execute all commands using sudo
[12:26] <phillw> xranby: thanks, I'll reply to the email.
[12:27] <phillw> But, possibly it could be better explained? If someone takes the time to report an issue it should be 'taken on board'.
[12:30] <rbasak> xranby: the sudo group is the correct and more consistent method in 12.04: https://wiki.ubuntu.com/PrecisePangolin/ReleaseNotes/UbuntuDesktop#PrecisePangolin.2BAC8-ReleaseNotes.2BAC8-CommonInfrastructure.Common_Infrastructure
[12:31] <rbasak> xranby: for consistency I think it would make sense for documentation intended for 12.04 users to refer to the sudo group instead of the admin group. Just my humble opinion though.
[12:33] <phillw> rbasak: xranby the e-mail I received simply pointed out that the instructions as given, do not work.
[12:35] <rbasak> phillw: that's odd. Both the admin and sudo groups should work.
[12:35] <phillw> does -server have an email address? I can forward the email to you so you can better liase with the OP?
[12:36] <phillw> It came to me via the -docs email address.
[12:37] <rbasak> There is ubuntu-server@lists.ubuntu.com. Or the full explanation of community support channels is at http://www.ubuntu.com/support/community
[12:38] <phillw> rbasak: thanks, I'll forward the original email to you with a link back for the OP
[12:39] <rbasak> phillw: I can't investigate personally, sorry. I've got tons to do before the 12.10 release.
[12:40] <phillw> rbasak: I meant to the mailing list - not you in person!
[12:40] <rbasak> Oh OK. Sure, go ahead :)
[12:41] <rbasak> If this does turn out to be reproducible it does sound like something that should really be fixed
[12:46] <roaksoax> morning
[13:04] <smoser> seijirou, your problem with cloud-init is bug 974509
[13:04] <smoser> you have broken dns.
[13:12] <seijirou> smoser:  Ahh that explains why I always get IP addresses for bogus domains.
[13:12] <smoser> yes.
[13:13] <smoser> use opendns or google dns.
[13:13] <seijirou> smoser:  But how then does "ubuntu-mirror" become the domain in source.list ?
[13:13] <smoser> this was a feature added to cloud-init so clouds could easily "advertise" a local ubuntu mirror, and cloud-init would pick it up.
[13:13] <seijirou> Oh I didn't read far enough
[13:14] <seijirou> Sorry I just woke up
[13:14] <smoser> if cloud-init got a response for "ubuntu-mirror", it assumed that there was a mirror there.
[13:14] <smoser> no worries.
[13:14] <smoser> but your ISP says "sure, we have 'ubuntu-mirror', and 'ubuntu-asdfjkl' also!"
[13:14] <seijirou> Yes exactly
[13:15] <smoser> (note if you choose to use opendns, you have to actually get an account and configure it to turn that behavior off.  by default they do the same garbage)
[13:15] <seijirou> I think I'll point my router to google and level3 dns
[13:16] <xnox> maybe it should check for /Release file for example, before trusting it....
[13:17] <xnox> and or do network check against start page thing
[13:34] <zul> Daviey/adam_g: have you guys figured out what you are going to do with lessc?
[13:36] <mala> hello
[13:37] <Daviey> zul: i think adam_g is driving that.. but i think it can be done sanely at source package build time.
[13:38] <zul> Daviey: ok because i rather not have the MIR team cringe
[13:40] <mala> i'm using ubuntu server at Amazon Web Services, but i can't understand why a regular ubuntu server use use 200mb of memory ram and the AWS instance use 380mb of memory
[13:40] <mala> can someone help me?
[13:41] <xnox> mala: use top, sort by memory usage, figure out.
[13:41] <mala> xnonx, i did it
[13:41] <xnox> mala: please note on Amazon it is using Ubuntu Server Cloud Image, which is different from a "normal" server image
[13:42] <xnox> (not much, but different)
[13:42] <RoyK> mala: also, see the 'RES' column in top, not 'VIRT'
[13:42] <RoyK> type M to sort by memory usage
[13:43] <stgraber> hallyn: hmm, just noticed that your lxc-wait changes actually never made it to Ubuntu as debian/local/lxc-wait overwrites the C one
[13:44] <mala> RoyK, inside top type shift M yes?
[13:46] <Daviey> zul: There seems to be no need for it to happen at package build or run time.
[13:46] <RoyK> yes
[13:47] <zul> Daviey: cool...or we can just download all the of js and package it that way ;)
[13:49] <Daviey> zul: we'll call that plan b
[13:49] <zul> Daviey: can we call it plan z?
[13:51] <Daviey> zul: having a "debian/rules prepare-media" would be awesome :)
[13:51] <Daviey> although that could be upstreamed
[13:54] <mala> Royk, always i change the ssh config file, when i restart the server, the ssh file come back to normal. Why?
[14:36] <CharlieSu> Hi all.  I have a SFTP server that constantly gets locked up because the 'console-kit-daemon' process is taking a lot of CPU usage..  Any ideas why?
[14:46] <Daviey> adam_g: hey, can you triage - bug 1038146 ?
[14:46] <Daviey> I am pretty sure you need it :)
[14:50] <SpamapS> Daviey: mind if I take that one?
[14:52] <Daviey> SpamapS: Oh, no.. that would be perfect
[14:53] <Daviey> koolhead17: also looks like you are using the PPA... which will need net access.
[14:54] <koolhead17> Daviey, ya i been told to use that :(
[14:55] <koolhead17> SpamapS, sir. so whats the final verdict i should or should not use the PPA
[15:02] <SpamapS> koolhead17: see response
[15:06] <koolhead17> SpamapS, in my case i have option 1. swift using another auth
[15:07] <koolhead17> SpamapS, i would not like to have my juju dependant of swift at all
[15:07] <koolhead17> i can use some other object storage someday
[15:09] <koolhead17> juju should be independent of all that. better there should be a provider charm and it can be of swift/ceph or something else
[15:10] <koolhead17> and if particular charm in my openstack environment needs object storage it should/could use it
[15:13] <koolhead17> SpamapS, https://bugs.launchpad.net/juju/+bug/1030897 also this hack
[15:14] <CharlieSu> Hi all.  I have a SFTP server that constantly gets locked up because the 'console-kit-daemon' process is taking a lot of CPU usage..  Any ideas why?
[15:15] <koolhead17> SpamapS, also what configuration in which file i need to make so it runs without stopping at failed swift creds?
[15:17] <njin> Hallo there's someont able to run maas on virtualbox ?
[15:17] <njin> someont/someone
[15:17] <koolhead17> njin, do a google i guess i saw some one writing blog on it
[15:18] <njin> I've tried google suggestion, but nothing
[15:20] <SpamapS> koolhead17: there's only one config file (~/.juju/environments.yaml)
[15:21] <koolhead17> SpamapS, and what configuration i have to pass so it does not ask for swift sir
[15:21] <SpamapS> koolhead17: for the "cloud without object storage" ... I see the use case, but juju needs a way to distribute files to new instances, so its a chicken-and-egg problem.. how do you deploy a charm whose purpose is distributing charms ;)
[15:21] <SpamapS> koolhead17: Not entirely sure what you're asking. I'm in a bit of a rush, so please forgive me
[15:22] <koolhead17> SpamapS, i dont want swift or object storage in my openstack infra
[15:22] <koolhead17> does that mean i cannot use Juju
[15:22] <koolhead17> ?
[15:22] <koolhead17> SpamapS, ok. later
[15:22] <koolhead17> just ping me once u have time i will test it for you sir
[15:23] <SpamapS> koolhead17: no, it means you will have to use some other S3 service tho
[15:23] <SpamapS> koolhead17: plain and simple, juju cannot operate without an object store.
[15:24] <koolhead17> SpamapS, but nowhere it says that OS cannot run without Object storage
[15:24] <koolhead17> :)
[15:25] <koolhead17> its a component if one wants uses it :D
[15:25] <koolhead17> seems like only solution would be to create some fake S3 env <-- which is not there :)
[15:27] <SpamapS> koolhead17: its fundamental to the design of juju. This has nothing to do with openstack reqs.
[15:27] <SpamapS> koolhead17: we need a place where we can push bundles to and have new instances find them.
[15:28] <koolhead17> SpamapS, either am confused or am not able to clarify my condition
[15:28] <SpamapS> koolhead17: yes we could run that on the machine 0.. no that is not on the priority list.
[15:28] <koolhead17> SpamapS, then i need a way out for this
[15:28] <koolhead17> i can load images kept inside glance
[15:28] <koolhead17> authenticated via
[15:28] <koolhead17> keystone
[15:28] <koolhead17> :)
[15:29] <smoser> utlemming, your review at https://code.launchpad.net/~smoser/vmbuilder/automated-ec2-builds.lp1035279/+merge/120410 would be appreciated.
[15:29] <SpamapS> koolhead17: yeah, thats great. Not at all related to juju, but thats great. :)
[15:30] <SpamapS> koolhead17: you need a globally accessible object store. Swift and "something that talks S3" are supported for openstack...
[15:30] <utlemming> smoser: looking
[15:30] <koolhead17> SpamapS, in my current enviornment i just dont use swift at all cos my images are taken care by glance
[15:31] <koolhead17> and i dont see its need in future
[15:31] <koolhead17> so cant we come with some fake auth for S3
[15:31] <koolhead17> to get out/away with this issue
[15:35]  * koolhead17 is out. ttl
[15:37] <smoser> utlemming, i can do the merge. i just wanted your eyeballs on it.
[15:37] <smoser> as a double check of the logic.
[15:37] <utlemming> smoser: yup, it looks fine
[15:38] <smoser> go ahad and comment there, and then i'll merge
[15:38] <utlemming> smoser: I kicked off a build against it to smoke check, just in case
[15:38] <utlemming> but it looks fine for merging
[15:38] <smoser> utlemming, well, i'll wait on the result then. thats great. thank you.
[15:39] <utlemming> smoser: one of the "features" of our new builder is the ability to test branches with out containating the main builds
[15:40] <Daviey> smoser: did you work out how to fix that damn python-keyring issue? it hit me ages ago, and i got around it.. but can't remember how.. i DON'T want to use a creds file.
[15:41] <smoser> Daviey, http://paste.ubuntu.com/1086257/
[15:41] <smoser> put that in ~/keyringrc.cfg
[15:42] <smoser> better link: http://bit.ly/nova-no-keyring
[15:43] <Daviey> smoser: that second link is MUCH better, thanks
[15:43] <smoser> :)
[15:43] <smoser> well, it means i can find it in my irc logs better.
[15:43] <Daviey> heh
[15:44] <Daviey> smoser: and great.. doesn't seem to work with w3m anymore?
[15:44] <Daviey> Grr
[15:45] <Daviey> ah no, does work
[16:05] <cyphermox> adam_g: ping
[16:07] <cyphermox> adam_g: I'd like to test one new release of dnsmasq that got released last week; I asked Daviey and he suggested I ask you about adding it (https://launchpad.net/~mathieu-tl/+archive/nm/+sourcepub/2613904/+listing-archive-extra) to the openstack CI to smoke test before uploading, just to make sure it doesn't break openstack
[16:07] <cyphermox> or you could just take the package from Debian unstable, since it's there now.
[16:09] <Daviey> cyphermox: hey, i'm sure we can do that.. adam_g Should be around shortly to confirm
[16:13] <Daviey> zul: http://people.ubuntu.com/~davewalker/cloud-archive-status.txt
[16:13] <Daviey> zul: that is updated hourly
[16:13] <zul> what am i looking at?
[16:14] <Daviey> zul: that is the list of things on the queue to migrate.. staging -> proposed.. proposed -> updates
[16:14] <zul> Daviey: gotcha
[16:23] <RoyK> how can I have ubuntu keep back certain packages? I've built my own ffmpeg package, and I don't want it to be upgraded automatically
[16:26] <xnox> RoyK: google for apt-pinning
[16:28] <ogra_> or just use synaptic it has an UI element that allows you to lock packages
[16:30] <Daviey> !pinning | RoyK
[16:30] <RoyK> thanks
[16:35] <xnox> Daviey: and his fancy ! commands
[16:35] <xnox> =)
[16:37] <RoyK> hrmf
[16:37] <RoyK> seems linux is unable to kick out a hanging drive from a raid
[16:39] <Daviey> well that isn't normally true
[16:39] <RoyK> http://paste.ubuntu.com/1157472/ got this almost half an hour ago
[16:39] <RoyK> started smartctl -t short /dev/sdd, and even that seems to be hanging
[16:40] <RoyK> trying to do something on /raid (where the filesystem is mounted) just blocks
[16:40] <RoyK> root@smilla:~# touch /raid/tmp/wtf
[16:40] <RoyK> ^C
[16:40] <Daviey> RoyK: what raid type?
[16:40] <RoyK> md raid-6
[16:41] <RoyK> http://paste.ubuntu.com/1157475/
[16:41] <Daviey> RoyK: well, something does indeed seem wookey
[16:41] <Daviey> can't really help atm
[16:42] <RoyK> the drive hung the other day as well
[16:42] <RoyK> and a reboot 'fixed' it, but I want to try to see if linux can sort this out without using debugging methods normally reserved for windows machines
[16:42] <xnox> RoyK: do you know which drive is failed?
[16:43] <RoyK> xnox: see the pastebin above - ata4 times out, the raid is hanging
[16:43] <RoyK> and even the smart test seems to be hanging
[16:43] <xnox> $ mdadm --fail /dev/sdX
[16:44] <xnox> then you can remove it.
[16:44] <xnox> for example.
[16:45] <RoyK> hm... raid stil hanging. trying to touch a file in there blocks the process
[16:46] <RoyK>  /proc/mdstat tells me recovery is in progress, though
[16:48]  * RoyK has seen drives 'die' like that earlier, and has found that a quick yank to the side while the disk is spinning, guaratees trouble-free replacement under warranty, while just returning the disk, often gives problems like 'hey, the disk works'
[16:52] <adam_g> cyphermox: i can test that dnsmasq, but im currently also trying to resolve 1035172
[16:53] <adam_g> cyphermox: so in the meantime, i can at least install an older precise kernel to avoid that bug, and confirm that the dnsmasq bits are working there. ...if that helps
[16:56] <cyphermox> Ah, yes, it would. Mostly as long as we can make sure the small changes in dnsmasq don't break openstack use cases
[17:03] <hallyn> stgraber: bleh
[17:03] <RoyK> hrmf! just ran smartctl -t short /dev/sdd and smartctl -a, which tells me
[17:03] <RoyK> Num  Test_Description    Status                  Remaining  LifeTime(hours)  LBA_of_first_error
[17:04] <RoyK> # 1  Extended offline    Completed: read failure       90%      6897         1514732911
[17:04] <RoyK> still, smartctl -H tells me the drive is healthy
[17:46] <adam_g> cyphermox: those dnsmasq packages seem to work fine, at least using FlatDHCPManager, which launches dnsmasq with http://paste.ubuntu.com/1157571
[17:47] <cyphermox> great. is there a different launch command you know of?
[17:48] <cyphermox> what I'm curious about is mostly whether there is anything that uses dhcp-range="interface:blah" or something like that
[17:54] <adam_g> cyphermox: i can check the source to see what the other managers use
[17:54] <cyphermox> adam_g: I guess I can do that
[17:55] <cyphermox> adam_g: is it all together in a source or are there multiple sources?
[17:56] <adam_g> cyphermox: i believe its all in nova/network/linux_net.py
[18:00] <cyphermox> adam_g: looks fine
[18:00] <cyphermox> adam_g: Daviey: mind if I go ahead and sync dnsmasq to 2.63-1?
[18:01] <adam_g> cyphermox: looks okay to me, have you checked with people working on MAAS? i dont know what their current use of dnsmasq is in the work thats being done to supercede cobbler
[18:01] <cyphermox> I had no clue it was used there too
[18:01] <adam_g> Daviey or roaksoax would know
[18:02] <cyphermox> fwiw, lxc and libvirt look okay already too
[18:04] <Daviey> cyphermox: sounds good
[18:04] <Daviey> cyphermox / adam_g: MAAS is dropping dnsmasq
[18:51] <Daviey> adam_g: did you look at compressing horizon's media at source package creation time?
[18:52] <adam_g> Daviey: not any more than we discussed last time
[18:53] <adam_g> Daviey: FYI 'add-apt-repository "http://ubuntu-cloud.archive.canonical.com/ubuntu precise-proposed/folsom"' results in an invalid entry to sources.list (two releases) 'deb http://ubuntu-cloud.archive.canonical.com/ubuntu precise precise-proposed/folsom'
[18:55] <Daviey> adam_g: ok
[19:08] <peeps[work]> i'm trying to set up a couple samba shares on my server and having some difficulty.  when I try to add a samba user with smbpasswd or pdbedit it says "Failed to add entry for user blah."   but gives not other information about why it failed.  does each samba user have to be an existing linux userr also?
[21:30] <sveinse> How do I configure kernel commandline to print *everything* during boot, rather than having this plymouth theme which hides it all?
[21:31] <sveinse> I have plymouth-theme-ubuntu-text installed, and I get text dots as progress. But that is not what I want
[21:40] <chilicuil> sveinse: press <Esc> while booting to change to the text mode (where all msgs are shown) or modify the grub entry for the kernel to remove 'splash': https://wiki.ubuntu.com/Plymouth#Debugging
[21:41] <chilicuil> you can also see the boot logs, if you run dmesg & check /var/log/boot.log out
[21:45] <sveinse> chilicuil: Well. I have no "splash" in the kernel command line at all. The only install theme is plymouth-theme-ubuntu-text, and it seems to hide everything between grub and getty.
[21:46] <iToast> apache2: bad user name ${APACHE_RUN_USER}
[21:46] <iToast> WHat do I do to fix that
[22:00] <xr1rr> I want to download some biggish files from a third party websites and store them on amazon s3. Is there anyway to do that without directly downloading them first to my local computer and then upload them back to amazon s3
[22:01] <ikonia> download them directly from your amazon machine /
[22:01] <xr1rr> but I am only using amazon s3
[22:03] <xr1rr> can it be done using amazon s3
[22:03] <ikonia> so you are just using amazon storage, not running a machine on amazon
[22:04] <xr1rr> yes
[22:04] <ikonia> then no, it has to go through your machine
[22:05] <xr1rr> but i need to upload files around 800mb/1gb in size and there is going to create alot of wasted bandwidth?
[22:06] <xr1rr> my ISP will hate me lol
[22:06] <ikonia> well, it has to hit your machine
[22:06] <Troy^> lol
[22:07] <Troy^> it is only 1gb
[22:07] <jernst> Needing a postfix guru … attempting to do smtp_sender_dependent_authentication and Amazon/SES and having no luck.
[22:07] <xr1rr> Troy^: i have about 30/40 different files tho
[22:08] <ikonia> jernst: just explain the issue
[22:08] <Troy^> xr1rr: i transfer about 1tb a month isp never says anything
[22:08] <jernst> I just wrote it up on ServerFault: http://serverfault.com/questions/419440/whats-wrong-with-my-postfix-setup-routing-through-amazon-ses-smtp-sender-depe
[22:09] <ikonia> sorry, I'm not reading through that
[22:10] <smw_> what is wrong with the following sudoers line? myuser ALL = (root) NOPASSWD: god terminate, /etc/init.d/god start
[22:11] <smw_> I would like myuser to have the power to run those two commands with a password as root
[22:11] <ikonia> smw_: root ?
[22:12] <jernst> ikonia: just trying to be specific.
[22:12] <smw_> ikonia, what?
[22:12] <ikonia> smw_: sorry, miss-read
[22:12] <ikonia> smw_: why have you got "god terminate"
[22:13] <jernst> ikonia: more specifically: what is the correct content of a smtp_sasl_password_map file when smtp_sender_dependent_authentication = yes? keys are sender email addresses? sender domains? relay hosts?
[22:13] <smw_> ikonia, I want someone to be able to run god terminate
[22:13] <smw_> ikonia, although, I should probably give the path to god...
[22:13] <ikonia> jernst: so you want postfix speciality in #postfix really, I thought it pulled that info from the DV rather than plain in the map file in your situation
[22:13] <smw_> ah!
[22:14] <smw_> ikonia, giving the complete path to god fixed it
[22:14] <ikonia> should do
[22:14] <ikonia> complete paths only
[22:14] <jernst> ikonia: "DV" = ?
[22:14] <smw_> ikonia, yeah, thanks for pointing out that "god" was wrong :-)
[22:14] <ikonia> ?
[22:15] <ikonia> smw_: I'd not idea what "god" was
[22:15] <smw_> ikonia, a command :-)
[22:15] <ikonia> I assumed you where doing some sort of comment or alias
[22:15] <smw_> ikonia, apparently, sudo did not know what god was either ;-)
[22:15] <ikonia> smw_: I hate to point out the obvious, but the default ubuntu sudo rule allows you to run all commands, was that not enough ?
[22:16] <smw_> ikonia, this is a separate user who I want to be able to run those two automated commands
[22:16] <smw_> ikonia, this allows my deploy user to restart a service
[22:16] <ikonia> smw_: fair enough, just checking
[23:11] <arrrghhh> hey all, i have a case of the dumb.  i'm trying to get a bash script to zip all files with a particular name... is this possible?
[23:11] <arrrghhh> like "zip -m ConnectionManager*.log ConnectionManager$(date +%Y%m%d_%H%M%S).zip"
[23:13] <arrrghhh> hrm.  seems i figured it out.  derp.