[00:29] <Jak2000> in ubuntu howto:start restart mysql
[00:33] <patdk-lap> service mysql restart
[02:41] <twb> $customer has ordered me to make their tomcat servlet "work under linux", and given me an allegedly working box with the servlet running under XP to work from.
[02:42] <twb> Apparently the XP box has a "postgresql-9.0-801.jdbc4.jar" in its tomcat lib dir.  On Debian apt-file turns up a libpostgresql-jdbc-java: /usr/share/java/postgresql.jar, which sounds like what I want, but I can't find the same package under lucid -- why not?
[02:46] <ajmitch> twb: probably because it was added well after lucid released
[02:47] <ajmitch> twb: last release in debian changed the binary package name
[02:48] <ajmitch> it used to be libpg-java
[02:48] <twb> Ah, it's just a different package name?  Thanks.  I should probably just install apt-file on a lucid server.
[02:49] <twb> (I haven't yet because for some reason my local mirror has a broken Contents despite me telling debootstrap to sync it.  So that host would need non-mirror:// non-file:/// URLs in sources.list)
[05:10] <koolhead17> hi all
[05:11] <hallyn> stgraber: the lxc package at ppa:serge-hallyn/virt, merge with debian, seems to be working (passing lp:~serge-hallyn/+junk/lxc-test)
[05:31] <koolhead17> hallyn, can  you help me to understand this wiered kvm error log
[05:32] <koolhead17> http://paste.ubuntu.com/946826/
[05:33] <koolhead17> kvm              1:84+dfsg-0ubunt
[05:33] <koolhead17> my openstack setup seems to have gone 4 toss
[05:57] <hallyn> koolhead17: on precise?  (kvm version is not helpful)
[05:57] <koolhead17> hallyn, yes precise
[05:58] <hallyn> check the /var/log/libvirt/qemu/vmname.log , might have better info
[05:58] <koolhead17> k. i rebooted my system gimme few mins
[05:59] <hallyn> ok - i'll likely be in bed, pls feel free to open a bug, i'll look in the morning.  the monitor openign error looks weird
[05:59] <hallyn> almost like a dns error is causing it
[06:00] <koolhead17> k thanks
[06:41] <ruben23> mysql  Ver 14.14 Distrib 5.1.61, <------------------gusy i ahve this mysql version can ias k what si teh default engine for this..InnoDB or MyISAm..?
[06:44] <ivoks> kees: around, by any chance? :)
[06:46] <bluefrog> ruben23, http://dev.mysql.com/doc/refman/5.5/en/innodb-default-se.html
[06:48] <twb>  8 getaddrinfo(http://10.130.2.8,5900): Name or service not known
[06:48] <twb> Sounds like an obvious error to me
[06:49] <twb> Someone has put "http://1.2.3.4" where they should have put only "1.2.3.4"
[06:49] <twb> You do want to see the kvm log as hallyn suggested, though
[07:25] <ivoks> kees: please take a look at https://bugs.launchpad.net/ubuntu/+source/resource-agents/+bug/833368/comments/3 - can this be solved by just changing defaults in lvm.conf, instead of using that patch?
[07:35]  * koolhead17 is happy :)
[08:56] <koolhead17> hallyn, thanks :)
[08:56]  * koolhead17 has complete setup working with latest openstack pkg. :)
[11:23] <koolhead17> somebody let me know when i can hit the download button  :)
[11:59] <ogra_> Daviey, rbasak, does the server team update the pandaboard server install instructions ?
[11:59] <ogra_> (since they still point to oneiric)
[11:59] <rbasak> I don't know who maintains that, sorry.
[11:59] <ogra_> well, its a server image
[12:00] <Daviey> ogra_: linky?
[12:00] <ogra_> GrueMaster used to maintain it, but he isnt with us anymore
[12:00] <ogra_> https://wiki.ubuntu.com/ARM/Server/Install
[12:00] <rbasak> The instructions for a generic ARM server install should be based on armadaxp now.
[12:00] <rbasak> Did NCommander take an action on something related to this?
[12:01] <ogra_> i just updated https://wiki.ubuntu.com/ARM/OmapDesktopInstall in case you need a template
[12:01] <Daviey> rbasak: Is that something you can handle today?
[12:01] <NCommander> rbasak: it got sidetracked with release tasks
[12:02] <ogra_> rbasak, how do armadaxp instructions help the community to test server userspace on arm ? :)
[12:02] <rbasak> Daviey: rbasak or NCommander, and keep panda or replace with armadaxp?
[12:02] <NCommander> rbasak: keep the panda
[12:03]  * ogra_ would add armada instructions additionally on a separate page if you really expect that community people can buy these boards soon
[12:38] <Ztane> in 12.04 ec2 instance store images, can the kernel be updated within the instance?
[12:50] <Ztane> hmm
[12:55] <memoryleak> is it my fault or is the ubuntu site not available?
[12:56] <ogra_> its all your fault !
[12:56] <ogra_> (no, its actually known and being worked on)
[12:56] <memoryleak> ok
[12:56] <koolhead17> memoryleak, its working now :P
[13:07] <memoryleak> koolhead17: it still doesnt :)
[13:10] <memoryleak> how ever, the 12.04 server is final right?
[13:13] <koolhead17> memoryleak, i think so but Daviey or others be right folks
[13:15] <erichammond> The EC2 AMI ids for Ubuntu 12.04 Precise need to be updated on http://cloud.ubuntu.com/ami/
[13:15] <utlemming> erichammond: looking at that now
[13:15] <erichammond> That page is still showing the AMI ids for beta2 even though the query API is returning correct ids.
[13:17] <erichammond> Good job launching the AMIs with the rest of 12.04.  I've come to expect this level of excellence from Canonical, but now and then I remember how much work goes into it.
[13:18] <koolhead17> erichammond, are you the Alestic folk
[13:19] <erichammond> koolhead17: singular, yes.
[13:19] <koolhead17> erichammond, your doing good job
[13:19] <koolhead17> :)
[13:20] <erichammond> koolhead17: glad to help.
[13:23] <erichammond> http://www.canonical.com/ is not accepting connections.
[13:23] <erichammond> ah, nor is http://www.ubuntu.com/
[13:24] <smb> release day ...
[13:24] <daker> i guess this is due to the hight demand
[13:24] <lynxman-> the website is being hammered to the extreme
[13:25] <ogra_> known, being worked on
[13:25] <smb> lynxman-, I am sure you can see lots of hair being pulled :)
[13:25] <lynxman-> smb: not in the office today, otherwise I definitely would :D
[13:25] <smb> :)
[13:25] <smoser> erichammond, i'll poke on cloud.ubuntu.com/ami .
[13:25] <smoser> that actually reads /query, which is updated, so i'm not sure what is going wrong.
[13:26] <erichammond> smoser: utlemming may also be looking at it.
[13:26] <smoser> yeah. he pinged me too.
[13:27] <erichammond> smoser: yep, Alestic.com is picking up the correct AMI ids from /query
[13:27] <smoser> erichammond, note, though... i think that is kind of deprecated. :-(
[13:28] <smoser> i've tried to get a few minor things fixed with it
[13:28] <smoser> bug 907362
[13:29] <smoser> bug 904910, bug 898846
[13:29] <smoser> they were largely closed as invalid due to bug 920538
[13:43] <erichammond> smoser: I'm confused about how bug 920538 is marked fixed released, but cloud.u.c/ami does not redirect to cloud-images.u.c
[13:46] <smoser> erichammond, well, the link to that is gone (i think)
[13:46] <smoser> ie, you only know of it because you know of it.
[13:47] <SysTom> http://releases.ubuntu.com/12.04/ <- is that the "released" 12.04 LTS?
[13:52] <Pici> scubes13: yes
[13:52] <Pici> scubes13: oops
[13:52] <Pici> SysTom: yes
[13:53] <SysTom> Thanks
[13:55] <erichammond> smoser: The page itself should be replace with either a redirect or a hyperlink to the recommended replacement.  Just because you don't link to it doesn't mean there aren't other links to it.  In fact, it's the second result on Google for "ubuntu amis"
[13:57] <smoser> well, largely it continued to work.
[13:57] <smoser> and i think i convinced people that it at least should continue ot.
[14:04] <erichammond> smoser: I think it's fine to have multiple ways to get at the same information so people can choose what works best for them.
[14:05] <erichammond> but now we're back to getting the AMI ids to be correct on that page.
[14:06] <smoser> yes.
[14:07] <erichammond> precise uses the same AKI as oneiric.  Nice.
[14:08] <paws> Alternate Koolaid Interface?
[14:08] <koolhead17> nijaba_tab, //O.o\\
[14:09] <kgee> I have 3 web servers, ubuntu, ubuntu, and mac on my LAN. I can get to each of them through localhost, yet only the mac server is available over the LAN. This leads me to believe there is some network blocking/firewall rule on the default ubuntu install that is preventing access. Can anyone confirm/deny this?
[14:10] <EvilResistance> kgee:  what're you running on the ubuntu servers?
[14:10] <EvilResistance> by default, there's no firewall rules
[14:10] <EvilResistance> (but your network might be blocking it)
[14:11] <kgee> EvilResistance: yeah, I'm trying to narrow down the problem but I'm hitting nothing but air here. The catch is that the ubuntu servers are virtual machines, and the router I have is a piece of garbage. However, since I can get to the mac server remotely I'm led to believe the router is not the culprit
[14:11] <kgee> and the virtual machines have their own IP address over a bridged network adapter, so that usually lets things run smoothly
[14:12] <EvilResistance> emphasis on *usually* :P
[14:12]  * EvilResistance has similar setups with VBox on his active-development laptop, and sometimes things explode
[14:12] <frankban>  hi hallyn: when you have a minute, I'd like to talk about lxc-ip.
[14:13] <kgee> regardless, if my router has ip addresses in the dhcp lease table to match the output from ifconfig, I imagine the network is set up properly....
[14:29] <Apreche> Does anyone know when the official AMIs will be updated? Thanks.
[14:40] <erichammond> Apreche: Official AMIs for Ubuntu 12.04 LTS Precise have been released.
[14:40] <andygraybeal_> AMI ?
[14:40] <andygraybeal_> american megatrend ?
[14:40] <Apreche> erichammond: awesome, but I don't see them on http://cloud.ubuntu.com/ami/ Do you have the AMI-IDs?
[14:41] <wiherek> hi
[14:41] <erichammond> Apreche: That page is being updated.  In the meantime I list the AMI ids on the top right of http://Alestic.com
[14:41] <wiherek> i am getting a postfix-policyd related error while trying to install iredmail
[14:41] <Apreche> erichammond: sweet, thanks!
[14:41] <wiherek> on ubuntu 11.10
[14:41] <weblinux> does anybody know how to host torrents/make them
[14:42] <erichammond> Apreche: and you can get to them by clicking around here: http://cloud-images.ubuntu.com/
[14:42] <patdk-wk> weblinux, sure
[14:42] <weblinux> do u have a name of a program. thats all i need
[14:43] <hallyn> frankban: what's up?
[14:44] <frankban> hallyn: finally I had time to write the script, following your suggestions: it uses Python/ctypes to call setns and get_init_pid (from liblxc). You can find the relevant MP here: https://code.launchpad.net/~frankban/lpsetup/lp-lxc-ip/+merge/103502
[14:44] <hallyn> frankban: ooh, aweseome
[14:44] <frankban> hallyn: I'd really like you opinion about the script, and in particular a suggestion about line 209: get_init_pid(name[:85])
[14:44] <erichammond> weblinux: Here's how I host torrents (using AWS): http://alestic.com/2012/04/aws-s3-torrent
[14:44] <kirkland> jcastro: what happened to #ubuntu-cloud?  freenode is saying that it's invite-only now
[14:44] <frankban> hallyn: There I slice the container's name to avoid a buffer overflow error, and I'd like to know if there is a way to systematically identify the max length of the string (rather than using the number 85, found by trial and error).
[14:45] <erichammond> kirkland: I was told that #ubuntu-cloud is supposed to redirect here now.
[14:45] <kirkland> erichammond: that's what they all say....  all the good secret stuff is really happening there :-)
[14:45] <hallyn> frankban: looking
[14:45] <erichammond> :-)
[14:45] <kirkland> :-D
[14:46] <frankban> hallyn: thanks
[14:47] <hallyn> frankban: i should probably look at liblxc before asking - but is it PATH_MAX you're looking for?
[14:48] <hallyn> I think the library should be doing the sanity checking
[14:48] <frankban> hallyn: I've looked PATH_MAX, but I don't know how to obtain the number 85 starting from that
[14:49] <hallyn> frankban: no, i think lxc_command() should be using snprintf and checking its return value
[14:49] <hallyn> I've been meaning to do an audit of the lxc code ...
[14:50] <hallyn> frankban: which releases would you be using that on?  precise+?
[14:50] <frankban> hallyn: yes
[14:50] <hallyn> Does what i'm saying make sense to you, or does it sound like the problem you were addressing might be different?
[14:50] <doctormon> I'm interested in looking at JuJu charms, but I can't find a list of available charms.
[14:51] <jcastro> http://jujucharms.com/
[14:52] <hallyn> frankban: let me hand you a package (in a bit) to test on without your slicing the name
[14:53] <frankban> hallyn: it makes sense. I've already encountered buffer overflows when trying to call get_init_pid without the slicing
[14:54] <hallyn> kind of embarassing :(
[14:55] <frankban> hallyn: ah! and also lxc-info suffers from the same problem
[14:56] <hallyn> because of its call to lxc_init_pid()?
[14:56] <hallyn> i hope
[14:57] <frankban> I didn't checked that, but it's reasonable
[14:58] <frankban> hallyn: however, I am really interested in your opinions about my branch
[14:59] <hallyn> ah, sorry, got distracted :)  lemme look again
[14:59] <hallyn> (distracted with the security implications)
[15:01] <hallyn> note that you don't really need to re-set the netns at the end (cause you're going to exit), but it doesn't hurt
[15:02] <doctormon> thanks jcastro
[15:02] <frankban> hallyn: I do that basically to isolate tests
[15:03] <doctormon> bogger, no openldap charms.
[15:03] <hallyn> frankban: I've never done anything quite so awesome in python :)  looks great
[15:03] <wiherek> alrite, so.. I found the package here:
[15:03] <wiherek> http://www.ubuntuupdates.org/package/core/oneiric/universe/base/postfix-policyd
[15:04] <wiherek> but how can I install it
[15:04] <wiherek> from the terminal?
[15:04] <hallyn> frankban: what do you think about searching /sys/class/net in the container for an interface?  maybe if given --search-interface option?
[15:04] <wiherek> there is no direct link to the file
[15:04] <designbybeck> hello all, I'm trying to find a "whats new" in 12.04 server. where might be the best place to look?
[15:06] <frankban> hallyn: interesting, you mean, retrieving interface names from the directory names inside /sys/class/net and then return all ip addresses for each interface?
[15:12] <hallyn> frankban: not for lo, but the others, yes
[15:12] <frankban> hallyn: sounds great
[15:12] <hallyn> frankban: thanks for writing that.
[15:13] <frankban> hallyn: thank you for your hints. So i suppose I can remove the slicing once the bug you filed is fixed
[15:17] <SysTom> resolve.conf gets overwritten on reboot on 12.04 LTS, I'm not using DHCP on the eth0 interface
[15:18] <SysTom> I see the whole "do not edit this file by hand" warning, but am unsure on the supported method for manually supplying nameservers
[15:19] <smb> SysTom, There is a description in man resolvconf
[15:19] <smb> requires two lines in /etc/network/interfaces
[15:19] <smb> (which I usually forget the exact syntax off)
[15:20] <SysTom> smb: Thanks, I'll check it out
[15:28] <SysTom> smb: That worked
[15:28] <SysTom> (Obviously!)
[15:29] <smb> :)
[15:40] <cocoa117> how do you run program in different userid?
[15:51] <RichardRaseley> Congrats on the 12.04 release to all the contributors!
[15:56] <halvors> I'm trying to setup a PXE boot server.
[15:56] <halvors> https://help.ubuntu.com/community/PXEInstallServer
[15:57] <halvors> Reqarding to this, the root of the tftp server should be "/media/cdrom/install/netboot/"
[15:57] <halvors> But in 12.04 it seems like it don't exists...
[15:57] <halvors> Anyone knows why?
[16:12] <dob> why is a do-release-upgrade not possible with 10.04 to 12.04?
[16:15] <ogra_> dob, because for LTS thats only enabled when the .1 release shows up
[16:15] <ogra_> 10.04 -> 12.04.1
[16:37] <Karwin> Does anyone know if there is a fast upgrade path from 10.04 LTS directly to 12.04 LST?
[16:42] <de_> hello.. its there anyone using maas?
[16:43] <patdk-wk> Karwin, yes, that is a supported upgrade path
[16:43] <patdk-wk> do-release-upgrade :)
[16:43] <Karwin> Thank you
[16:57] <zrtsch> Hello
[16:57] <zrtsch> Does anyone know what causes the error mountall: Plymouth command failed, mountall: Disconnected from plymouth ? This is happening on first install on several machines lately
[16:57] <zrtsch> I've fixed it before by editing the grub line, but now I forget what it was that I changed
[16:57] <zrtsch> People online are referring to nvidia drivers, but I just installed ubuntu minimal. I don't have any nvidia drivers installed.
[16:59] <zrtsch> nomodeset, I think that was it.
[17:02] <Ztane> hmm ec2 images now handle authorized_keys on ubuntu differently, right?
[17:03] <Ztane> we've been building 10.04 images with ec2-bundle-vol and expected that the authorized_keys on ubuntu account stay the same, now with 12.04 cannot login anymore
[17:14] <Ztane> hmm so how to modify this?
[17:24] <Ztane> wtf really
[17:25] <Ztane> I specify the "webserver-keys" as key name for boto run_instances but am still unable to log in
[17:51] <jcastro> smoser: utlemming: erichammond: cloud.u.c ami browser fixed, sorry about that.
[17:53] <kennett> I need help with accessing my server  through a domain name.
[17:53] <zrtsch> 'through' a domain name?
[17:53] <smoser> utlemming, mosh is pretty cool. thanks for pointing me at that. thing that sucks though is that i only opne port 22 by default on instances.
[17:54] <zrtsch> why not open a non-default port?
[17:54] <kennett> like myserver.dyndns.com
[17:58] <kennett> I'm trying to access my server through the web can some1 help me?
[18:01] <kennett> I tried using dynamic dns but I can't get it to work.
[18:05] <stefg> kennett: but you know your servers IP, right?
[18:06] <kennett> IP on the network?
[18:06] <stefg> kennett: public IP... so this implies your server is nat'ted behind a router?
[18:07] <kennett> Yes I know the  IP.
[18:08] <stefg> kennett: so given that you have openssh-server installed it should as simple as ssh <server-ip> -p <port> (if its not 22)
[18:09] <stefg> !ssh
[18:10] <kennett> I have a site with pics and stuff that I'm trying view.
[18:11] <stefg> and your browser says what?
[18:11] <limpc> hi guys
[18:11] <limpc> i upgraded to 12.04 this morning (new partition, new system)
[18:12] <limpc> having trouble re-establishing some sshfs maps i had previously though
[18:12] <limpc> shows owner and gid as "???"
[18:12] <limpc> and access denied when i try to go into the folder i mapped it to
[18:12] <stefg> kennett: if DNS isn't working you can still just type the IP into your browsers address bar
[18:12] <kennett> It says theres an error when I'm outside the network.
[18:12] <limpc> I do specify the uid/gid in sshfs params, uid=1000,gid=1000 which are correct
[18:13] <stefg> kennett: so this "server" lives behind a router in a LAN, and not in a DMZ, right?
[18:13] <kennett> Right
[18:14] <stefg> kennett: so it isn't publicly accesible... you need to set up a DMZ, or use VPN to first connect to the remote LAN and reach the server from within this LAN.
[18:15] <limpc> nevermind i found the issue
[18:15] <limpc> looks like its a bug with the version that ships with ubuntu 12.04
[18:16] <stefg> kennett: so without gaining access to that machine you'll not be able to reach it from outside the LAN now
[18:16] <kennett> I can't use port forwarding?
[18:17] <stefg> kennett: you could.... if you use non-standard ports. You'll not like to set up port forwarding for 80 :-)
[18:17] <limpc> exit
[18:18] <kennett> So what do I need to do to get it to work?
[18:19] <stefg> kennett: first of all access to the server and to the router...
[18:20] <stefg> kennett: http://en.wikipedia.org/wiki/DMZ_%28computing%29
[18:29] <kennett> So if I turn on DMZ it should let me access the server?
[18:33] <stefg> kennett: you don't "turn on" a DMZ, you are creating a subnet separate from the 192.168.x.x LAN. so the server needs a new IP, and this DMZ subnet needs to be publicly accessible. (which is what what a typical home router tries to prevent....)
[18:35] <stefg> kennett: so it's a bad idea of mixing server and desktop workloads on the same machine.... the server should only have a minimum footprint, thus no X, nor unnecessary services
[18:36] <stefg> !ebox
[18:36] <kennett> So how can I set this up?
[18:37] <kennett> I use webmin.
[18:38] <stefg> see above ... first check your routers docu how to create a DMZ, then you'll probably have to physically change the network wiring and put your server in that DMZ
[18:39] <stefg> kennett: BTW webmin isn't officically sanctioned for ubuntu,
[18:39] <stefg> !webmin
[18:51] <kennett> stefg: I'm back.
[18:55] <kennett> stefg: Can you help me set it up?
[18:56] <stefg> kennett: i'm about to leave... and actually this is beyond the scope of ubuntu support, it's a course in LAN engineering :-)
[18:58] <Jak2000> how to install the ftp client on a ubuntu-server? (root@core /backup/mysql# ftp 192.168.0.87
[18:58] <Jak2000> -bash: ftp: command not found)
[18:58] <kennett> stefg: OK np.
[18:59] <stefg> !info ncftp
[19:00] <Jak2000> stefg: root@core /backup/mysql# ncftp 192.168.0.87
[19:00] <Jak2000> -bash: ncftp: command not found
[19:00] <Jak2000> same error
[19:00] <stefg> Jak2000: sudo apt-get install ncftp
[19:02] <Jak2000> http://pastebin.com/4JUT7rQX
[19:03] <Jak2000> cant install
[19:03] <stefg> Jak2000: ubuntu 12.04 just came out, there servers are swamped ...
[19:04] <Jak2000> i am on a old version of ubuntu?
[19:04] <stefg> Jak2000: and if you just want to d/l something... wget is you friend
[19:05] <stefg> Jak your on 10.04 LTS... perfect for a workhorse server
[19:05] <Jak2000> with wget how to get files form a Ftp server these are stored in /MysUbubtufiles/
[19:07] <stefg> Jak2000: man wget is much smarter than i am....
[19:19] <_ruben> bah .. dutch mirror changed ip .. breaking my sync .. now im 5GB of package behind :p
[19:22] <henkjan> _ruben: thats why people use dns ;)
[19:22] <_ruben> then again .. at 60Mbps it's going fairly quick
[19:22] <_ruben> henkjan: restrictive outbound firewall using ip addresses :/
[19:33] <alphanum> Hello. Pici referred me to your channel.
[19:33] <alphanum> He said you guys were keen on ec2.
[19:33] <alphanum> I am looking for an ubuntu vpn ami, but am unsure which to pick
[19:34] <alphanum> rather, how to set it up
[19:35]  * alphanum waits
[19:36] <_ruben> that's usualy the best approach indeed, wait :)
[19:36] <_ruben> (no ec2 experience here)
[19:36] <alphanum> :)
[19:37] <alphanum> It's like a hunting session, really--have you ever seen animal planet?
[19:37] <alphanum> The lioness waits hours on end for a lone zebra to drink from a lake?
[19:37] <_ruben> ;)
[19:39] <alphanum> Upon building a VPN w/ ec2, how easy is it to identify that you're connecting through it?
[19:40] <alphanum> Taftse, FACEFOX, are you keen on ec2?
[19:40] <Taftse> sorry ?
[19:40] <Taftse> ec2 ?
[19:41] <Taftse> alphanum, what you talking about ?
[19:41] <alphanum> amazon web services?
[19:41] <Taftse> ok
[19:50] <alphanum> So, is anyone keen on ec2?
[19:52] <alphanum> hi Jak2000
[19:52] <alphanum> are you inactive?
[19:53] <alphanum> hi marrusl
[19:54] <Pici> alphanum: This channel isn't as busy as #ubuntu is.
[19:54] <alphanum> I know :(
[19:54] <alphanum> I'm trying to figure out how to build a vpn via ams
[19:54] <Pici> alphanum: And to answer your question about the ec2 VPN, just use something to determine your ip address. If it points to the VPN, then your traffic is going through it.
[19:54] <alphanum> perhaps I should try google again
[19:55] <alphanum> I mean, I know that but I want to create a vpn via ec2
[19:55] <erichammond> alphanum: What do you mean "create a vpn via EC2"?
[19:56] <alphanum> launch a micro instance; ubuntu openvps ami
[19:56] <erichammond> Do you want to communicate with your EC2 instances via a VPN? communicate with the Internet using a VPN to EC2 as a gateway out of your local firewall? something else?
[19:56] <alphanum> run traffic from my home comp to aws ec2 vpn
[19:57] <alphanum> Does that clear things up?
[19:57] <erichammond> alphanum: The basic process is (1) start an EC2 instance running Ubuntu, (2) install and configure the VPN like you would for any Ubuntu server.
[19:58] <erichammond> There's no need to try to find an AMI that has VPN pre-installed.
[19:58] <alphanum> I have never set up a server in my life.
[19:58] <alphanum> Maybe windows instance suits me more closely?
[19:58] <erichammond> You have a fun road ahead of you :-)
[19:58] <erichammond> Probably the wrong channel to offer you that sort of advice.
[19:59] <alphanum> :(
[19:59] <alphanum> mind a pm?
[19:59] <erichammond> alphanum: Here's an old article I wrote providing instructions on how to set up openVPN with EC2 to escape a restricted network: http://alestic.com/2009/05/openvpn-ec2
[20:00] <erichammond> Karmic is past end-of-life, so you'd need to adapt to a more modern Ubuntu.
[20:01] <erichammond> There is a lot of documentation available on how to get things done with Ubuntu and Ubuntu servers.  Once you have an EC2 instance running, almost all of it applies.
[20:02]  * erichammond heads off to the office.
[20:04] <philipballew_> ubuntu server over windows anyday
[20:20] <pabelanger> well, that is kinda sucky:$ ls -la | grep glance
[20:20] <pabelanger> drwx------   2 glance   glance    4096 Apr 26 16:19 glance
[20:20] <pabelanger> not even rw permission for the group
[21:14]  * michael_tn waves, good day all
[21:14] <michael_tn> i'm spinning up a new maas build and having a bit of trouble
[21:15] <michael_tn> went through the config and it looks like dnsmasq is not getting configured, the config is all pounded out
[21:27] <robbiew> michael_tn: did you install maas-dhcp?
[21:27] <michael_tn> apt-get install maas-dhcp
[21:28] <michael_tn> answered the on screen prompts with my network info
[21:29] <michael_tn> but end up with fully commented out /etc/dnsmasq.conf
[21:29] <robbiew> hmm
[21:29] <michael_tn> nod, what i said too :-)
[21:30] <robbiew> RoAkSoAx: any ideas? ^
[21:31] <roaksoax> michael_tn: a fully commented out? can you pastebinit
[21:31] <michael_tn> sure
[21:31] <roaksoax> thanks
[21:33] <michael_tn> http://pastebin.com/aMxdKh70
[21:35] <roaksoax> michael_tn: ah
[21:35] <roaksoax> michael_tn: ok, I was confused
[21:36] <roaksoax> michael_tn: so dnsmasq gets configured but it does why using a template, not /etc/dnsmasq.conf
[21:37] <michael_tn> hmmm?
[21:37] <roaksoax> michael_tn: it uses /etc/cobbler/dnsmasq.template
[21:37] <michael_tn> ok
[21:37] <michael_tn> lemme look
[21:37] <michael_tn> ok
[21:38] <michael_tn> that looks correct :-)
[21:38] <robbiew> RoAkSoAx: but I thought cobbler generated /etc/dnsmasq.conf from the template
[21:38] <robbiew> does he need to run a `maas-import-iso`?
[21:39] <robbiew> maas-import-iso*s* :/
[21:39] <michael_tn> its running now
[21:39] <michael_tn> still running
[21:40] <pabelanger> Well, this is not good. Nova's logging system has DEBUG level logs enabled by default.  Not a good default level for a production box
[21:40] <pabelanger> looks like a bug in nova's log.py code
[21:40] <michael_tn> i'll let it finish and restart and see its state
[21:41] <michael_tn> thanks guys!
[21:41] <michael_tn> i'll check back in tomorrow :-)
[21:46] <roaksoax> robbiew: noep, dnsmasq ships a commented out config file per se
[21:47] <roaksoax> robbiew: maas-provision just tells dnsmasq to use the "template" config file
[21:47] <robbiew> oh
[21:47] <robbiew> command line?
[21:47] <robbiew> gotcha...so it passes the "template" file as the config file
[21:47] <robbiew> got it
[21:47] <roaksoax> yes
[22:01] <drecute> does sudo postconf -e 'home_mailbox = Maildir/' mean that Maildir is chosen as default mailbox format or it just specifies a directory?
[22:05] <roaksoax> robbiew: my bad, it does use the template to generate a new /etc/dnsmasq.conf, idk why I thought it just passed the template as a config file
[22:05] <robbiew> :)
[22:05]  * robbiew figured you might have changed it, as I recall it doing that when I was playing with it
[22:05] <robbiew> no worries
[22:06] <robbiew> again, I think you have to run maas-import-isos after installing maas-dhcp
[22:06] <robbiew> that triggers the cobbler reload...or whatever it's called
[22:06] <roaksoax> robbiew: yes, it has to be run, as itruns 'cobbler sync'
[22:07] <drecute> i want mails to be stored in a cifs mounted filesystem. How do I do this?
[22:09] <roaksoax> robbiew: what it does do though, is that it uses a cobbler hosts file (with dnsname, mac, ip) and the template sources from there
[22:09] <robbiew> ack
[22:10] <robbiew> drecute: I suggest trying http://askubuntu.com/questions/ask
[22:12] <robbiew> RoAkSoAx: did you catch my email about https://maas.ubuntu.com/images/ephemeral/releases/precise/
[22:12] <robbiew> who creates those?...as we need a newer one ;)
[22:21] <hallyn> stgraber: I'm about to push 0.8.0~rc1-4ubuntu1 to quantal.  I expect to have 0.8.0-rc2-based from debian soon, so it won't stay long...  but anyway, if you'd like to take a look first, pls let me know
[22:24] <hallyn> (it's at http://people.canonical.com/~serge/lxc_0.8.0~rc1-4ubuntu1.dsc, freshly rsync'ed)
[22:27] <flaccid> http://cloud-images.ubuntu.com/precise/ only seems to be daily, where is final?
[22:28] <flaccid> smoser: any idea when final precise AMIs available?
[22:29] <drecute> robbiew: ok
[22:29] <drecute> robbiew: hoping to get an answer
[22:31] <robbiew> utlemming: ^^
[22:32] <robbiew> flaccid: http://cloud-images.ubuntu.com/releases/precise/release/
[22:40] <flaccid> ah right thanks, bit weird how thats not linked to from http://cloud-images.ubuntu.com/precise/
[22:45] <erichammond> robbiew: Perhaps it's worth having a short UDS session or hallway discussion to talk about how to make AMI ids easier to find and use.
[22:45] <robbiew> erichammond: amen to that :)
[22:46] <robbiew> erichammond: I suggest just pulling utlemming and smoser aside and discussing it...b/c right now, we can't set the length of UDS sessions...all fixed :/
[22:47]  * robbiew notes this is a wishlist for UDS scheduling :)
[22:47] <erichammond> I'm going to plug hiding 32-bit and instance-store unless the user specifically clicks to request seeing them.  If a new user wants to run Ubuntu on EC2, there's really just one AMI id needed right now for each region (Ubuntu 12.04 Precise, 64-bit, EBS boot) and making them dig for it in a list of dozens is cruel.
[22:48] <robbiew> agreed
[22:56] <flaccid> did they just use the builds from the 24th for the final ? noticing the datestamps on these images
[22:59] <henkjan> flaccid: yep
[22:59] <flaccid> yeek i could of released MCIs already, ah wells i will just do today, thanks
[23:06] <SpamapS> flaccid: "MCI" ?
[23:06]  * SpamapS remembers when that was a phone company
[23:06] <flaccid> yeah http://support.rightscale.com/21-Community/RightScale_OSS#Ubuntu_MultiCloud_Images
[23:06] <flaccid> i'll update the alph2 ones to final
[23:18] <eftepede> Hi.
[23:18] <eftepede> I have ubuntu-server 11.10. Tried to do release-upgrade to 12.04, but since it's OpenVZ guest on host with 2.6.18, everything just broke down.
[23:19] <eftepede> Is there anything I can do?
[23:19] <eftepede> With most of my software I get:
[23:19] <eftepede> irssi: /lib/x86_64-linux-gnu/libc.so.6: version `GLIBC_2.14' not found (required by /lib/x86_64-linux-gnu/libglib-2.0.so.0)
[23:19] <doctormon> eftepede: Sounds like a binary package didn't get uninstalled correctly.
[23:21] <eftepede> doctormon: Which one?
[23:21] <eftepede> libc?
[23:21] <eftepede> WARNING: this version of the GNU libc requires kernel version^M
[23:21] <eftepede> 2.6.24 or later. Please upgrade your kernel before installing^M
[23:21] <eftepede> glibc.^M
[23:21] <eftepede> it's from screenlog
[23:21] <doctormon> eftepede: Not sure, hard to tell from here. what's uname -a say?
[23:22] <eftepede> Linux insomniac 2.6.18-238.9.1.el5.028stab089.1 #1 SMP Thu Apr 14 14:06:01 MSD 2011 x86_64 x86_64 x86_64 GNU/Linux
[23:22] <doctormon> eftepede: reboot server, use newer kernel.
[23:22] <eftepede> I can't.
[23:22] <eftepede> It's kernel on OpenVZ host.
[23:22] <eftepede> Not my local server.
[23:23] <eftepede> And I can't access OpenVZ host.
[23:23] <doctormon> eftepede: Then you're buggered I shouldn't wonder. you don't have ssh access?
[23:23] <eftepede> I have.
[23:23] <eftepede> Machine is not rebooted.
[23:24] <doctormon> The reboot command wouldn't reboot? or it would reboot in a way that would break?
[23:24] <eftepede> I suppose reboot is not a good idea now.
[23:24] <eftepede> Machine won't start properly, I guess.
[23:25] <doctormon> Do you have much invested in it? Upgrading without being able to reboot = bad idea.
[23:25] <eftepede> Not sure if I understand you correctly. I have the 'ability' to reboot.
[23:26] <eftepede> I'm just worried, that system is too broken to start properly.
[23:26]  * RoyK is tired and wants his servers and network to work a bit better
[23:27] <eftepede> My only idea now is to manualy change sources.
[23:28] <doctormon> RoyK: Have you tried putting the ethernet wires in diluted metal shavings? ;-)
[23:28] <eftepede> But will it perform a downgrade?
[23:28] <doctormon> eftepede: No, downgrade your machine manually if you have to. There isn't an automatic way to downgrade but if you spend all night on it, it can be done.
[23:29] <doctormon> eftepede: Or just accept that you've messed it up and try to recover a fresh install (somehow)
[23:29] <eftepede> I can't - it's shell server for ~50 users, the last thing I want is to reinstall it ;-)
[23:29] <eftepede> Or maybe: I can, but it's the last option, if everything fails.
[23:35] <qman__> eftepede, unfortunately, as far as I know, there is no way to roll back an upgrade
[23:35] <qman__> you must manually remove and reinstall the old versions
[23:35] <qman__> which is a huge task for an entire server that's been upgraded
[23:36] <qman__> you should never do a release upgrade on a server that 1) isn't backed up 2) is in production 3) you don't have physical access to
[23:36] <qman__> and virtualization always throws another wrench in
[23:37] <SpamapS> utlemming: hey, I think I can say this with confidence after a few days.. the S3 mirrors are *fast*
[23:38] <eftepede> qman__: I know the basics, really ;-) a) I have backed up most important things; b) it's my private machine on which I have some shell accounts for my friends, so it's just kind of 'production'; c) it's hard to have physical access to virtual machine ;-)
[23:39] <eftepede> Reinstall is still an option - but I want to avoid it.
[23:40] <qman__> IMO that will be easier
[23:41] <eftepede> http://pastebin.com/aJxx7GDw - it's when problems starred. Sorry for syntax, it's from screenlog :/
[23:42] <eftepede> So upgrading libc from 2.13 to 2.15 caused all this.
[23:43] <qman__> the problem is that running 12.04 means running a kernel much newer than the one you have
[23:43] <qman__> running your old kernel isn't an option with the right version of libc
[23:43] <qman__> and running an older version of libc means running lots of other new software won't work either
[23:43] <eftepede> Yup. Unfortunatelly as I'm not the owner of host machine, I can't upgrade.
[23:43] <qman__> so, in short, upgrading to 12.04 is not an option
[23:44] <eftepede> Now I know it ;-)
[23:44] <eftepede> I didn't know, that 2.6.18 will be too old and I need 2.6.25
[23:44] <eftepede> s/25/24/
[23:44] <qman__> the only way to restore your system is to remove all upgraded packages, and reinstall the packages they replaced
[23:44] <qman__> 12.04 actually runs 3.2
[23:45] <qman__> though that libc doesn't depend on such a new kernel, other things probably do
[23:45] <eftepede> Yes, I know. I have lot non-virtualized machines on 11.10 at work and I know, that they're using 3.x
[23:45] <qman__> 2.6.18 is _really_ old though
[23:46] <qman__> even ubuntu 8.04 is running 2.6.24
[23:46] <eftepede> I'm aware of that, too. Maybe I can ask OpenVZ-host owner to upgrade.
[23:46] <eftepede> But it will be followed by reboot of all virtual machines, mine included.
[23:46] <eftepede> And I'm really afraid to reboot mine ;-)
[23:47] <qman__> it's probably hosed either way
[23:47] <doctormon> qman__: Agreed
[23:47] <qman__> only way I can see of fixing it is to boot a live environment and chroot, then go through manual dependency hell to downgrade
[23:47] <qman__> hours of work at best
[23:47] <doctormon> Although you'll want to eat your own foot that do it again ;-)
[23:49] <eftepede> I need to choose, if I want to have hours of this hell or days of hell restoring users accounts ;-)
[23:49] <eftepede> As my /home was not backed up and users were aware of this, I don't care about data.
[23:50] <eftepede> But they'll want their accounts back
[23:50] <qman__> keep in mind also that it won't be what it was before
[23:51] <doctormon> eftepede: And you might have run out of coffee by the end.
[23:52] <eftepede> The worst thing is, I have last cigarette and it's 2:00 am here, so I can't but another box ;-)
[23:53] <qman__> unless you've got a comprehensive list of all packages that you upgraded from, and all packages that you upgraded to, you're probably going to spend days, not hours
[23:53] <qman__> if not weeks
[23:53] <eftepede> But: I'm not afraid of that manual dependency hell. I'm working as system administrator, so it's kinda 'funny' problem for me to solve. Reinstall is just too easy. It MUST be the way to get thins working again ;-)
[23:54] <eftepede> s/thins/things/
[23:54] <qman__> it's always possible, the question is if it's worth it
[23:54] <eftepede> And "I work as", excuse my bad grammar.
[23:54] <qman__> and in my opinion, and I'm guessing doctormon's as well, it's not nearly worth it
[23:54] <qman__> you've very seriously broken it with no easy way back
[23:55] <eftepede> Ok, I'll figure out what to do tomorrow. I'll consult with my ubuntu-sysadmins at work (I'm mostly involed with *BSD systems) and decide, which way to choose.
[23:56] <eftepede> Thanks your guys for your time and opinions, take care ;-)
[23:56] <doctormon> qman__: Agreed, I know I'd rather walk a 1,000 miles to the data center, put on burglar outfit, sky hook my way in to the building and get access to the machine.
[23:56] <doctormon> eftepede: good night and good luck!
[23:56] <eftepede> Tx, bye.