[00:20] <axisys> i installed latest lucid 10.04.2 on my refurbised ibm box and I am only getting bug soft lockup cpu#0 stuck for 61s! on the screen a reboot after the fresh install
[00:21] <twb> axisys: so piss about with notsc, noapic, acpi=off, etc?
[00:22] <andygraybeal_> with kvm, can i use a real physical nic dedicated to a virtual machine?
[00:23] <andygraybeal_> so i can have a virtual machine connected directly to a hardware network?
[00:23] <twb> andygraybeal_: almost certainly
[00:23] <twb> WOrst case: you bridge it but don't assign an IP on the dom0 for that iface
[00:23] <andygraybeal_> awesome thank you twb.  i nee to play with this.
[00:24] <andygraybeal_> twb, *nods*
[00:24] <andygraybeal_> thank you
[01:08] <axisys> twb: don't follow
[01:08] <axisys> so want me to try acpi=off ?
[01:08] <axisys> but i do not see the grub
[01:08] <axisys> how do I catch the grub during boot
[01:11] <Jasonn> Hold shift while bootin
[01:11] <Jasonn> booting*
[01:11] <axisys> Jasonn: ok.. let me try
[01:11] <Jasonn> Okies
[01:11] <Jasonn> Its shift on all the newer versions
[01:11] <Jasonn> escape, or delete on the older ones
[01:14] <axisys> ok shift gave me the grub
[01:14] <axisys> Jasonn: thanks
[01:14] <Jasonn> No problem :)
[01:15] <twb> axisys: *I* don't care
[01:15] <axisys> i installed 10.04.2 server 32bit.. but the kernel shows 2.6.32-28-generic-pae instead of server
[01:15] <axisys> so append acpi=off in the kernel line may be?
[01:15] <Jasonn> nah, it doesnt matter
[01:16] <Jasonn> i suppose
[01:16] <Jasonn> Yer probably gonna get better help if you ask in #ubuntu for that one mate
[01:16] <axisys> Jasonn: ok
[01:17] <twb> crond tries to reduce the "nice" and "nproc" ulimits of its child processes.  How do I prevent this?
[01:17] <twb> Hmm, it mentions pam_limits, so maybe it's pam not crond at fault...
[01:19] <twb> Fuck it, I'm just going to ignore the complains in the log
[01:19] <twb> I'm even getting... pam_limits(cron:session): Could not set limit for 'nice' to soft=20, hard=20: Operation not permitted; uid=0,euid=0
[01:20] <twb> Which *can't* be widening a limit, so it's clearly just not allowed to change limits at all
[01:30] <axisys> twb: thanks for your help.. acpi=off worked.. system finally up
[02:09] <twb> OK, WTF
[02:09] <twb> I have a new CUPS server, and it ONLY works if I refer to it as its unqualified hostname.
[02:11] <twb> http://paste.debian.net/112372/
[04:59] <twb> Where is /etc/cups/mime.convs supposed to come from?
[05:55] <dschuett> not really a big deal, but does anyone know how to change the order of the interfaces in the systen information that comes up during login?
[06:06] <juzzy__> hi
[06:15] <twb> What "system information" ?
[06:20] <YankDownUnder> dschuett, Um...the stats?
[08:17] <kaushal> hi
[08:17] <kaushal> when i run dmidecode i get Type: Unknown Speed: Unknown in http://pastebin.ubuntu.com/587238/ Any clue ?
[08:21] <twb> any clue about what?
[08:23] <kaushal> twb: about Type: Unknown Speed: Unknown
[08:24] <kaushal> is there anyother way to know the RAM Speed and Type ?
[08:24] <twb> Try lshw
[08:24] <twb> or open the box
[08:25] <kaushal> twb: ok
[08:26] <_ruben> or your hardware inventory documents ;)
[08:27] <twb> haha
[08:27] <twb> "ask your hardware vendor"
[08:29] <kaushal> twb: its colocated
[08:44] <shauno> I don't suppose it's a VM that's giving you empty results because there's technically no hardware?
[08:55] <BbluE> Hello all.
[08:57] <twb> Grmph
[08:57] <twb> Why does nagios3 use /usr/bin/mail but not recommend on it
[09:02] <twb> Apparently because bsd-mailx meets its deps, but provides a slightly different binary
[09:03] <BbluE> Are you running Nagios on your server?
[09:05] <twb> I am.
[09:06] <kaushal> shauno: its a physical hardware
[09:06] <BbluE> Nice monitoring solution.
[09:07] <twb> Personally I think it could do with a massive cleanup
[09:07] <twb> I mean step #1 would be to use execvp instead of system() FFS
[09:07] <BbluE> I have a question for anyone who might be able to answer it...
[09:08] <BbluE> I used apt-get to install php5... is there a format or anything I can put into the php.ini, to comment out extensions I know I'm not going to use?
[09:08] <twb> That's a question for #php, I think
[09:09] <BbluE> Ah,I figured someone here might have ran into that issue... installing it on Ubuntu Server 10.10.. but, you're right. I should go check there. Thank you.
[09:40] <nick_name> hello
[09:45] <nick_name> is there a way to disable ttys temporarily?
[09:50] <twb> stop tty2
[10:05] <katmandoudou> Hello, I have configured my server to run JBoss as service. It works almost except that I have some exceptions when I try to stop it
[10:05] <katmandoudou> http://pastebin.com/y0at57Mv
[10:05] <katmandoudou> Caused by: java.net.ConnectException: Connection refused
[10:05] <katmandoudou> Could someone help me ?
[10:05] <rcconf> try with /etc/init.d/
[10:06] <katmandoudou> rcconf, ok
[10:12] <katmandoudou> rcconf, it has changed nothing ...
[10:12] <rcconf> k
[10:18] <katmandoudou> rcconf, would you want to see my /etc/init.d/jboss file ?
[10:19] <rcconf> im not geek
[10:19] <katmandoudou> rcconf, ah ...
[10:23] <jamespage> katmandoudou: I'm a little rusty on Jboss but from memory the shutdown command should be connecting to a control port to initiate the shutdown
[10:24] <jamespage> katmandoudou: /etc/init.d/jboss might be helpful to diagnose
[10:24] <jamespage> katmandoudou: netstat -a would also handy with JBoss up and running
[10:25] <katmandoudou> jamespage, ok at first I paste my /etc/init.d/jboss script to pastebin
[10:25] <jamespage> katmandoudou: great
[10:27] <katmandoudou> jamespage, ok http://pastebin.com/5ikL1Erp for the jboss script
[10:30] <jamespage> katmandoudou: well it looks like it still works the same way as it did in 4.3 and 5
[10:30] <katmandoudou> jamespage, yes but it doesn't work for the shutdown :)
[10:31] <jamespage> katmandoudou: yeah - so we need to figure out why the shutdown code can't connect to the JBoss instance
[10:32] <jamespage> katmandoudou: some log data from jboss would also be helpful as I think the containers output what ports they are listening on
[10:32] <katmandoudou> jamespage, the server.log ?
[10:33] <jamespage> katmandoudou: yep -that should contain the right info
[10:50] <katmandoudou> jamespage, here my boot.log http://pastebin.com/eST0qhku
[10:56] <jamespage> katmandoudou: looking now
[10:56] <katmandoudou> jamespage, and after having cleaned my server.log, I don't have a lot of interesting things inside
[10:57] <katmandoudou> jamespage, here my server.log http://pastebin.com/NFRLFHLS
[11:00] <katmandoudou> jamespage, I must leave for a while now. See you after, just the time to go to home
[11:02] <brontosaurusrex_> in karmic server, what would be a working way to setup internet time adjustement once per day?
[11:03] <Caribou> brontosaurusrex_: why would you want to do that, instead of running a ntp daemon ?
[11:04] <Caribou> This will cause sudden offset in your time reference (i.e. changing once a day)
[11:04] <brontosaurusrex_> Caribou, ok, that makes sense, so: sudo apt-get install htp ?
[11:04] <guampa> brontosaurusrex_: add something like * * * * * ntpdate time.nist.gov to crontab, mind that some servers, notably dovecot, will die if time moves backwards too much
[11:05] <Caribou> well, afaik, ntp is installed by default.
[11:05] <Caribou> lemme check
[11:07] <brontosaurusrex_> Caribou, ok, that makes sense, so: sudo apt-get install ntp ? > related: can i use ntp on desktop as well?
[11:09] <Caribou> as long as you have a internet connection, it should work fine either on desktop or server
[11:09] <brontosaurusrex_> also, can i assume that ntp comes with some sane defaults? and can i assume i dont need to play with crontab with ntp?
[11:11]  * Caribou is looking at it right now
[11:12] <guampa> brontosaurusrex_: ntp is the ntp daemon, not ntpdate which is a different program. ntpd doesn't need crontabs
[11:12] <brontosaurusrex_> guampa, i understand
[11:12] <guampa> it will run at startup and stay in background syncing time while the system is up
[11:12] <brontosaurusrex_> i understand that as well
[11:13] <Caribou> I just installed it. Seems to configure without a glitch and sync with ntp.ubuntu.com
[11:26] <tigreton> hi
[11:26] <tigreton> could you recommend an admin panel like webmin? i didn't find nothing
[11:26] <brontosaurusrex_> Caribou, ok, thanks for help
[11:26] <Caribou> np
[11:29] <tigreton> i have problems win openldap too, if someone can recommend some tutorials too, thanks
[11:54] <katmandoudou> jamespage, I'm back
[12:21] <katmandoudou> jamespage, are you still here ?
[12:21] <katmandoudou> jamespage, it should comes from my run.conf !
[12:28] <jf_> hi, I'm having a really irritating problem installing ubuntu server on my server machine
[12:28] <jf_> after installing and rebooting multiple times, everything is okay with the network
[12:29] <jf_> but as soon as I install dhcp 3 and reboot, every last bit of networking fails to work properly (it's fine before restarting)
[12:29] <jf_> *dhcpd
[12:30] <jf_> ah it is actually dhcp3-server
[12:30] <pmatulis> jf_: "every last bit of networking" meaning other systems affected?
[12:31] <jamespage> katmandoudou: sorry - just popped out for a bit
[12:32] <jamespage> katmandoudou: I tried to reproduce your issue; with no success (i.e. I was able to startup and shutdown as the init script does)
[12:33] <jamespage> katmandoudou: some of the init script looks a bit suspect when compared to the shutdown.sh that comes as part of JBoss
[12:37] <katmandoudou> jamespage, sorry wait a minute I've almost solved the problem
[12:39] <jf_> pmatulis: no, just the server
[12:41] <etrisnanto> hello
[12:43] <pmatulis> jf_: check the logs
[12:44] <jf_> i looked at dmesg, everything's fine there
[12:44] <pmatulis> jf_: /var/log/daemon.log
[12:47] <jf_> okay, from dhcpd there are only warnings there (host declarations are global, not limited to scope declared in)
[12:55] <pmatulis> jf_: you need to discover what is actually happening.  do you lose your IP address assignment?
[12:58] <jf_> what do you mean? the server itself has a static ip (which it retains) and all other network devices eventually don't get a new dhcp lease
[13:05] <jf_> but since the server has a static ip and the router does too, it should at least have internet
[13:10] <hallyn> Daviey: (haven't updated yet) did you enable modpython in your ppa's new znc yesterday?  (if you ever got around to updating)
[13:25] <jf_> pmatulis: I also just noticed that nothing in ifconfig changes if I remove the network cable
[13:25] <jf_> or put it into another rj45 slot
[13:25] <pmatulis> jf_: and when dhcpd is not running that behaviour changes?
[13:25] <tigreton> well bye
[13:27] <jf_> no it does not
[13:28] <jf_> everything breaks after installing dhcpd
[13:28] <jf_> *dhcp3-server
[13:28] <pmatulis> jf_: so it's the installation and not the running of it
[13:28] <jf_> even if I remove the package, it's still broken
[13:28] <pmatulis> jf_: remove package and reboot mabye.  sounds very odd
[13:32] <pmatulis> jf_: i'm starting to think that dhcpd has nothing to do with your problem
[13:33] <jf_> it does seem rather weird
[13:35] <Daviey> hallyn, no... but i did email the debian maintainer asking why it wasn't enabled
[13:36] <Daviey> it could have very good reason... but doesn't seem so
[13:36] <Daviey> hallyn, if you are bored, and have a diff.... :)
[13:37] <hallyn> Daviey: no to both
[13:37] <hallyn> thanks :)
[13:37] <hallyn> Daviey: today i hope to get testing done on vmware
[13:38] <hallyn> first need to update the hallyn labs infrastructure :)
[13:38] <Daviey> hallyn, Yeah... the latest bug update wasn't clear if it was good or bad.
[13:39] <hallyn> oh, it was bad
[13:39] <Daviey> I think it indicated that it tried to modprobe before it had been dkms built
[13:39] <Daviey> which isn't necessarily a problem, is it?
[13:44] <zul> morning
[13:49] <hallyn> Daviey: oh, no I thought it was saying that the debian packaging was trying to install two modules which no longer exist.
[13:49] <hallyn> but his second msg said that clipboar doesn't work
[13:49] <hallyn> anyway i just need to test it
[13:50] <hallyn> biab, time for a reboot
[13:52] <katmandoudou> jamespage, still here ?
[13:52] <jamespage> katmandoudou: sure am
[13:52] <jamespage> did you manage to fix your issue?
[13:52] <katmandoudou> jamespage, ok so after some tests I can say that the problem is from the run.conf
[13:53] <katmandoudou> when I put the option -Djboss.service.binding.set=ports-02 into the JAVA_OPTS variable
[13:54] <Disconnect> is there a way to use preseeding to manually create a partition before partman runs? (if i can do that, i can get a reasonable number of inodes, then i can just tell partman to keep the now-existing partition..)
[13:54] <katmandoudou> without it, I can stop jboss normally
[13:54] <katmandoudou> with I have the exceptions you have seen before
[13:55] <katmandoudou> this option is to change the ports of JBoss, to add 200 to each ports
[13:55] <jamespage> katmandoudou: OK - that was a new feature in JBoss 5
[13:56] <katmandoudou> jamespage, and ?
[13:58] <jamespage> katmandoudou: so the way the init script calls the Shutdown code needs to be modified - see http://community.jboss.org/wiki/configureports (right at the bottom)
[13:58] <jamespage> it does not observer the settings in run.conf
[14:02] <katmandoudou> jamespage, thanks I try to start and stop jboss and I tell you the results
[14:13] <talntid> I'm trying to solve an issue. I have an app that sends many emails to the same addresses every day. Recently, it stopped "working". Seems the emails are going out, from what I can see.. but none of the addresses are receiving them. The mail HOST is gmail. Unlikely that I am being blocked. It responds and says mail accepted for delivery -- just, it's not happening. Any ideas on troubleshooting?
[14:14] <ScottK> If your logs say gmail accepted them, then it's an issue internal to gmail.  You'll have to talk to them.
[14:15] <Disconnect> not to point out the obvious, but check spam. gmail sometimes decides that automated mail like that is spam..
[14:16] <talntid> I did check the spam mail.
[14:16] <talntid> It's not in there :)
[14:17] <talntid> I am under the impression that it's internal to GMail too, ScottK. Just wanted to make sure I wasn't missing something dumb...
[14:17] <pmatulis> talntid: you might want to pastebin some mail logs
[14:18] <Disconnect> and its not under 'all mail'?
[14:18] <pmatulis> fg
[14:18] <katmandoudou> jamespage, it should be not exactly the same values to pass
[14:19] <katmandoudou> jamespage, I did this : JBOSS_CMD_STOP=${JBOSS_CMD_STOP:-"java -classpath $JBOSSCP org.jboss.Shutdown --shutdown -s jnp://localhost:11299"}
[14:20] <katmandoudou> jamespage, and I get this exception : Exception in thread "main" java.net.MalformedURLException: Service URL must start with service:jmx:
[14:20] <katmandoudou> 	at javax.management.remote.JMXServiceURL.<init>(JMXServiceURL.java:143)
[14:20] <katmandoudou> 	at org.jboss.Shutdown.main(Shutdown.java:234)
[14:20] <talntid> correct, Disconnect
[14:21] <Disconnect> dunno then. we had that problem with large mailer reports, just sucked it up and moved them to the internal server
[14:21] <talntid> I want to convince my office staff that the crap they are having me auto-email me them, is crap ;)
[14:22] <talntid> They can view the info live, any time, using the intranet page. Instead, they want it emailed to them every day. weak. :P
[14:22] <jamespage> katmandoudou: I don't like the way the init script calls the java code directly; it would be better if it called shutdown.sh with appropriate parameters '-p 1290 -o localhost' should work
[14:23] <katmandoudou> jamespage, yes let's try
[14:28] <katmandoudou> jamespage, I changed my JBOSS_CMD_STOP like this : JBOSS_CMD_STOP="cd $JBOSS_HOME/bin; ./shutdown.sh -p 1290 -o localhost"
[14:28] <katmandoudou> but the problem is the same
[14:28] <katmandoudou> I have still the ConnectException
[14:35] <jamespage> katmandoudou: so are you still using the -Djboss.service.binding.set=ports-02 setting?
[14:36] <katmandoudou> yes normally
[14:36] <katmandoudou> yes I have just checked
[14:37] <jamespage> katmandoudou: sorry - thats my fault the parameter for the port should be '-r 1290' not -p
[14:37] <RoAkSoAx> morning
[14:38] <jamespage> ./shutdown.sh with no args give you all of the options
[14:38] <jamespage> hey RoAkSoAx
[14:39] <RoAkSoAx> jamespage: howdy! how's it going today?
[14:39] <jamespage> RoAkSoAx: good thanks! hows yours?
[14:39] <RoAkSoAx> jamespage: good good
[14:48] <katmandoudou> jamespage, ok it was the -r
[14:48] <katmandoudou> jamespage, thanks a lot !!
[14:48] <jamespage> katmandoudou: no problem - might be worth reporting that as a bug back to jboss as its pretty broken how it is today.
[14:48] <zul> SpamapS: ping when you around....is squid upstart stuff fixed in lucid?
[14:55] <SpamapS> zul: I think so.. I think the reports we got were largely busted conffiles from people who had never re-started their squid.
[14:56] <zul> SpamapS: so we should be ok to make the changes in maverick?
[15:07] <SpamapS> zul: yeah should be fine unless I missed more bug reports coming in
[15:07] <AdvoWork> Hi there, i have paths set like: mirror_lenny = http://cdn.debian.net/debian  for something im using, how would i specify the path for the latest server version?
[15:11] <zul> SpamapS: cool
[15:13] <shauno> AdvoWork: archive.ubuntu.com/archive would be the equivalent of that.
[15:13] <shauno> err, /ubuntu rather
[17:26] <Error404NotFound> i am following http://www.linuxjournal.com/content/centralized-logging-web-interface to send my syslog to a center server. How would i also send other logs such as apache as well?
[17:36] <airtonix> Error404NotFound: well apache has some custom errorloging modules
[17:37] <Error404NotFound> airtonix: so, the only option is to send those messages to syslog as well?
[17:37] <airtonix> it's been a while since i looked at it, but i think you can "stream" it to a custom handler
[17:42] <SpamapS> Error404NotFound: apache can log to syslog very easily
[17:44] <SpamapS> Error404NotFound: the simplest way is just   CustomLog |logger -t apachelogs -p user.info
[17:45] <SpamapS> actually you need the full path to logger, so   CustomLog | /usr/bin/logger -t whatever -p something.somethingelse common
[17:46] <evilsushi> does syslog.conf come installed on the ubuntu 10.04 lts distro?
[17:47] <evilsushi> root@ironman:/etc# find / -name syslog.conf
[17:47] <evilsushi> root@ironman:/etc#
[17:48] <evilsushi> erm
[17:48] <evilsushi> rsyslog?
[17:48] <evilsushi> what is rsyslog?
[17:49] <pmatulis> evilsushi: the logging framework since 9.10
[17:50] <evilsushi> cool so ill just install ng then
[17:50] <pmatulis> evilsushi: why?
[17:51] <evilsushi> pmatulis: because its the norm and works great?
[17:51] <pmatulis> evilsushi: rsyslog works great too
[17:51] <evilsushi> do you have config examples?
[17:52] <pmatulis> evilsushi: to do what?
[18:24] <Slyboots_> Mm.. Anyone know how to get wget to name files it downloads correctly
[18:24] <Slyboots_> Say I .. give it wget http://fake.com/download?12345 , Wget will save the file as download?12345 instead of filename.zip
[18:39] <genii-around> Slyboots_: You can try something like wget -O filename.zip http://wherever/something?12345
[18:41] <Slyboots_> genii-around: Yea, but it would be preferable for wget to follow the link as it should and obtain the correct filename
[18:47] <genii-around> Slyboots_: It does follow the link, but the file is being named differently by some server side script.
[18:54] <fosterdv> wget http://longname -O short_name
[18:57] <Anon275> nick/starwarsfan
[19:37] <semiosis> Slyboots_: according to the wget man page, a new exprimental option called "--content-disposition" will do what you want
[19:38] <Slyboots_> Mmm..
[19:39] <semiosis> Slyboots_: idk what version introduced that option, i'm looking at maverick's man pages
[19:39] <Guest98493> hi room
[19:39] <Slyboots_> I'll gve it a try, see how it goes
[19:40] <semiosis> Slyboots_: cool, let us know how it goes
[19:41] <smoser> SpamapS, ping
[19:41] <smoser> bug 745930
[19:41] <smoser> request upstart job help. jhunt__ would also be helpful possibly.
[19:42] <smoser> hm... that wasnt the bug that i meant to say, but your input might be useful there also. i meant to ask about bug 745946
[19:44] <SpamapS> smoser: looking now
[19:45] <SpamapS> smoser: the udev one, btw, I totally stalled out on. Its really puzzling.
[19:46] <smoser> i'm not convinced that the first bug above is not related to udev.
[19:46] <smoser> ie, it would fail to find the metadata service if the job ran, but the network interface wasn't really up.
[19:46] <SpamapS> Agreed.
[19:47] <SpamapS> smoser: but in that case, udev appears to have started normally
[19:49] <SpamapS> smoser: is it possible cloud-config never stopped?
[19:51] <Guest98493> Can anyone help me to setup a web server?
[19:51] <smoser> SpamapS, i can't be sure, but i dont think its likely.
[19:51] <smoser> SpamapS, we are moving to a different test suite, and so i dont have some logs that i would have had :-(.
[19:53] <Guest98493> I have ubuntu 10.10 installed with LAMP
[19:53] <Guest98493> where to go from now?
[19:53] <Pici> Guest98493: Then you already have a webserver setup.
[19:53] <SpamapS> Guest98493: well what do you want your web server to do?
[19:53]  * SpamapS hopes the answer is "PROFIT"
[19:53] <Guest98493> none profit
[19:54] <Guest98493> :D
[19:54] <Guest98493> just hosting myself and my buddies
[19:54] <Guest98493> I want to have hosting panel something close to cpanel
[19:55] <Guest98493> so the idea is just setup a web server and forget about it
[19:55] <SpamapS> udevd[59]: worker [66] did not accept message -1 (Connection refused), kill it
[19:55] <SpamapS> smoser: have to figure that one out. So weird.
[19:56] <smoser> but SpamapS regarding the 'start on' for those two jobs
[19:56] <smoser> does that seem sane ?
[19:56] <SpamapS> totally
[19:56] <smoser> could any race condition cause cloud-final to not start ?
[19:56] <smoser> ok.
[19:56] <SpamapS> Those two things, rc and cloud-config, are decoupled from one another
[19:56] <SpamapS> so its a safe and
[19:57] <slim_> hello all, is there a sip server to use it for chat (text/audio/video) and can communicate with MS OCS server ?
[19:57] <SpamapS> one thing that I don't like about that tho, is that it will block running 'runlevel 1'
[19:58] <SpamapS> actually thats not true.. runlevel does not wait
[19:58] <SpamapS> and.. single user on a cloud instance is.. a bit silly
[19:58] <smoser> you have a better suggestion? i need "rc.local"
[19:58] <smoser> hm...
[19:58] <smoser> hm... maybe there should be an upstart job called rc.local
[19:58] <SpamapS> smoser: stopped rc RUNLEVEL=[2345] is about as close to rc.local as it gets
[19:59] <smoser> right.
[19:59] <smoser> i was suggesting there should be a job though
[19:59] <smoser> so that people dont have to write "stopped rc RUNLEVEL=[2345]" many times
[19:59] <SpamapS> no
[19:59] <SpamapS> ;)
[20:00] <SpamapS> rc.local is still in /etc/init.d IIRC
[20:00] <semiosis> Guest98493: check out webmin for server control panel
[20:01] <smoser> 59 jobs in /etc/init on the system i'm looking at
[20:01] <smoser> 5 of them (almost 10%) have "stopped rc RUNLEVEL=[2345]"
[20:01] <smoser> that would seem a reasonable thing to abstract
[20:01] <smoser> and then run rc.local via it.
[20:02] <Guest98493> semiosis: thanks
[20:02] <SpamapS> smoser: start on boot-finished ?
[20:02] <SpamapS> or maybe 'post-boot'
[20:02] <smoser> basically, yes, that is what rc.local starts on
[20:02] <smoser> :)
[20:03] <SpamapS> nice thing is you can just dump that in rc.conf's post-stop .. 'initctl emit post-boot'
[20:03] <smoser> but scott will tell you that boot has never finished
[20:03] <SpamapS> post-automatic-boot
[20:03] <SpamapS> post-system-boot
[20:04] <SpamapS> whatever we call it.. it means that the stuff we run at the beginning has completed.
[20:04] <lder> hi.  i am running a server on a closed network and want to deliver updates via CDROM once a week.  is there a special day of the week that security updates are released on the repo?
[20:05] <SpamapS> lder: no.
[20:05] <SpamapS> lder: they're released "as fast as possible"
[20:06] <lder> SpamapS: ok, thanks.
[20:34] <RoAkSoAx> kirkland: I NQA preseed + SSD rockS!!
[20:34] <kirkland> RoAkSoAx: :-)
[20:34] <kirkland> RoAkSoAx: + TestDrive for extra speedy options
[20:37] <RoAkSoAx> kirkland: indeed, can we pass the nqa arguments by default though? This way I can just make an option available in testdrive (In the preferences "Enable NQA" "Enter Pressed file location"), so that's even faster?
[20:37] <kirkland> RoAkSoAx: nope
[20:37] <kirkland> RoAkSoAx: however, in a few months, it'll get easier
[20:38] <kirkland> RoAkSoAx: you can actually add kernel command line arguments on the kvm line
[20:38] <kirkland> RoAkSoAx: however, you have to provide both the kernel and initrd parameters too
[20:39] <kirkland> RoAkSoAx: aliguori (upstream QEMU) has a brand new, super cool hack for qemu that will let you address the kernel/initrd in the -cdrom argument by a url style cdrom://path/to/vmlinuz
[20:39] <kirkland> RoAkSoAx: once we have that in qemu, then, yes, you will be able to add such support to TestDrive ;-)
[20:39] <RoAkSoAx> kirkland: cool!! that;s great then. I guess that for now I can have a customized ISO to "Install from Pressed" defaulting to your preseed
[20:40] <RoAkSoAx> so that I don't have to enter it manually
[20:45] <kirkland> RoAkSoAx: cool
[20:46] <RoAkSoAx> kirkland: have you though on having this included in the official ISO?
[20:47] <RoAkSoAx> (the pressed file)
[20:47] <RoAkSoAx> or a hidden option to not have to enter it manually
[20:48] <kirkland> RoAkSoAx: i've discussed the concept of having a "no-questions-asked" seed file in the /preseed directory of the Ubuntu Server before with cjwatson
[20:49] <RoAkSoAx> kirkland: that'd we great
[20:49] <kirkland> RoAkSoAx: as i recall, cjwatson didn't like the idea because a true "no questions asked" install would probably need to blow away the whole disk (as my sample preseed does)
[20:49] <kirkland> RoAkSoAx: and that can be dangerously destructive
[20:50] <kirkland> RoAkSoAx: however, I still think providing just such a file in /preseed would be extremely high value, low cost effort
[20:51] <RoAkSoAx> kirkland: indeed! but with proper documentation I don't think it would be dangerous because the administrator would know what they are duing if they use the preseed. And I also agree with you on having it shipped in /preseed (at least) would be awesome
[20:51] <kirkland> RoAkSoAx: moreover, i'd love to have that as an option in the F4 Modes menu of syslinux, and additionally append the "priority=critical locale=en_US" to the kernel line when selected
[20:52] <kirkland> RoAkSoAx: so I suggest your toss this out there as an Oneiric-Topic on the mailing list
[20:52] <kirkland> RoAkSoAx: perhaps CC cjwatson so that he can comment on it
[20:52] <RoAkSoAx> kirkland: cool, I will
[21:07] <hallyn> hey dn
[21:07] <hallyn> hey d'nalrick, how's it going
[21:11] <cjwatson> kirkland: a preseed file might be one thing, but I refuse to put it in the F4 Modes menu - it's far too dangerous for exposure there
[21:11] <cjwatson> absolutely no way
[21:11] <cjwatson> a preseed file can have comments and such
[21:11] <kirkland> cjwatson: fair enough
[21:12] <kirkland> cjwatson: any suggestions as to avoiding the need for the locale=en_US additional parameter?
[21:12] <kirkland> cjwatson: perhaps making an assuming from the language chosen on the first language prompt in syslinux?
[21:13] <cjwatson> just document it in the preseed file
[21:13] <kirkland> cjwatson: and I'm pretty sure that I could avoid the priority=critical with a more complete preseed, right?
[21:13] <cjwatson> sure, though many preseeders use priority=critical anyway
[21:13] <cjwatson> saves hassle
[21:14] <cjwatson> you could use Kickstart, that's able to set locale parameters from a file
[21:14] <kirkland> cjwatson: right, so ideally, I would just add file=/preseed/no-questions-asked.seed
[21:14] <cjwatson> (with a disgusting hack)
[21:14] <kirkland> cjwatson: and if that's the case, then we assume a locale and priority=critical
[21:15] <kirkland> and "no-questions-asked" should probably be something shorter, save keystrokes
[21:15] <cjwatson> I wish there were less focus on this and more on server team members fixing bugs in the server installer :-/
[21:15] <cjwatson> TBPH
[21:15] <cjwatson> https://help.ubuntu.com/10.10/installation-guide/i386/automatic-install.html documents using Kickstart, FWIW
[21:16] <kirkland> hallyn: howdy :-)
[21:16] <kirkland> cjwatson: thanks
[21:18] <bastidrazor> i used the 64bit install CD and now i notice Linux servitude 2.6.32-30-generic-pae #59-Ubuntu SMP Tue Mar 1 23:01:33 UTC 2011 i686 GNU/Linux ..
[21:19] <kirkland> cjwatson: TBPH, I wish we had more leverage to change the server installer in positive ways
[21:19] <bastidrazor> since i don't have the 64bit kernel nor mroe than 3.2GB of RAM, can i safely install the -server kernel?
[21:20] <Guest98493> hi guys
[21:21] <Guest98493> I'm quite stuck trying to install shorewall following this guide http://net.tutsplus.com/tutorials/php/how-to-setup-a-dedicated-web-server-for-free/
[21:21] <Guest98493> when I type "sudo cp /usr/share/doc/shorewall-common/examples/one-interface/* /etc/shorewall/"
[21:22] <pixel13> let's try this again... ( I should've known there was an ubuntu-server channel!) ;)
[21:22] <Guest98493> it gave me cp: cannot stat ......
[21:22] <smw> Guest98493, why do you need shorwall?
[21:22] <pixel13> anyone here successfully used DRBD 8 / OCFS2 on Maverick?
[21:22] <pixel13> ...specifically _with_ Pacemaker?
[21:22] <smw> Guest98493, please get a real nick
[21:22] <Guest98493> How do I get real nick?
[21:22] <Guest98493> :)
[21:23] <RoyK> Guest98493: type /nick mynick
[21:23] <smw> Guest98493, type /nick aname
[21:23] <RoyK> what a nick...
[21:23] <yessir> hehhe
[21:23] <yessir> smw: I don't know anything
[21:24] <pixel13> For those of you familiar with DRBD, I'm referring specifically to this guide:
[21:24] <pixel13> https://wiki.edubuntu.org/ClusterStack/LucidTesting#Pacemaker,%20drbd8%20and%20OCFS2%20or%20GFS2
[21:24] <smw> yessir, ok, I suggest against shorewall
[21:24] <yessir> I was just trying to follow th guide
[21:24] <yessir> smw: so I don't need shorewall at all?
[21:24] <smw> yessir, it is not worth the time.
[21:24] <smw> yessir, no
[21:25] <pixel13> I know it's for Lucid, and DRBD + OCFS2 is not a problem, that's been really easy to setup...it's just I have not been able to get Pacemaker work with O2CB for the life of me, really hoping someone on this board can help point me in the right direction
[21:25] <smw> yessir, shorewall is more of a pain than anything else when used on a local server.
[21:25] <smw> yessir, (local server == not on the internet)
[21:25] <yessir> oh
[21:25] <yessir> I'm trying to setup a web server
[21:26] <yessir> can you have a look at this guide: http://net.tutsplus.com/tutorials/php/how-to-setup-a-dedicated-web-server-for-free/
[21:26] <smw> yessir, one that will be available to others?
[21:26] <yessir> Yes.
[21:26] <smw> yessir, run from home?
[21:26] <yessir> Yup.
[21:27] <smw> yessir, you don't need a fw because your router will block ports for you.
[21:27] <yessir> I'm planning also to install a hosting panel
[21:27] <yessir> Oh, I see.
[21:27] <cjwatson> kirkland: I can think of lots of ways we could do really useful things for server users in the installer
[21:27] <yessir> I should the skip to step 7?
[21:27] <smw> yessir, but, just a warning, most ISPs do not allow you to host a website on port 80
[21:28] <cjwatson> kirkland: for example, I intend to propose that I spend time rearranging the preseeding format so that LVM and RAID preseeding is much more straightforward
[21:28] <cjwatson> *that's* the sort of server installer work I want us to be doing
[21:28] <yessir> how can I test am I allowed or not?
[21:28] <smw> of course, if you have a "business account"...
[21:28] <pixel13> hmm..no one here using DRBD for Ubuntu clustering??
[21:29] <smw> cjwatson, what?
[21:29] <cjwatson> smw: wasn't talking to you?
[21:30] <pmatulis> pixel13: try the ubuntu-server ML
[21:31] <pixel13> thx @pmatulis, good call. I may just give that a try...
[21:32] <kirkland> cjwatson: sounds useful;  we'll have quite a bit of work around OpenStack next cycle, and we'll need a good experience installing it;  I'm hoping we explore ways to make this work well with less invasive changes required to d-i than we did with UEC
[21:33] <cjwatson> kirkland: right, that sort of thing I also think is useful.  What I'm worried about is spinning wheels on nice-to-haves
[21:34] <yessir> "ECONNREFUSED - Connection refused by server".
[21:35] <yessir> I tried to connect to the server using Filezilla
[21:35] <kirkland> cjwatson: such as a functional no-questions-asked seed?  that's just file in rootskel, right?
[21:35] <smw> yessir, maybe shorewall is messing with you? lol
[21:35] <pixel13> @pmatulis ...dang it, seems most recent talk of OCFS2 on ubuntu-server ML is over a year ago!
[21:35] <kirkland> cjwatson: don't worry about my request for the priority/locale selections, i can see that as being more of a pain, and more of a nice-to-have
[21:35] <pixel13> how can I ask a fresh question on the ML?
[21:36] <yessir> Oh no..
[21:36] <cjwatson> kirkland: changes to several different udebs, all of which require testing, QA, bug-fixes, ongoing merges, etc.
[21:36] <yessir> Can you guide me how to uninstall it?
[21:36] <cjwatson> preseeding aliases live in preseed, for instance
[21:36] <cjwatson> preseed files live in cdimage, not rootskel
[21:37] <smw> yessir, sudo apt-get remove shorewall
[21:37] <smw> !who | yessir
[21:37] <kirkland> cjwatson: interesting, i had no idea;  i think i'd be satisified with shipping a single file somewhere in the server cd filesystem
[21:37] <cjwatson> if that's all it takes, I don't care; but scope creep is a problem
[21:37] <yessir> ubottu: Yes sir! :)
[21:39] <cjwatson> what I'm trying to say isn't "I'll obstruct any useful change you want to make", but more "I wish server effort went slightly more in the direction of the giant stack of bugs and feature requests we already have from users, rather than adding new ones"
[21:40] <SpamapS> cjwatson: ++
[21:40] <cjwatson> I mean server installer effort, sorry - I don't see the rest of it
[21:41] <SpamapS> Whether intentional or not, You captured my exact stance on the server product quite well there.
[21:42] <yessir> How do I check if I have installed OpenSSH  or not
[21:43] <yessir> my memory did not serve me well :o
[21:43] <SpamapS> which ssh
[21:43] <SpamapS> ssh --version might also work
[21:44] <semiosis> dpkg -l | grep ssh
[21:44] <semiosis> TMTOWTDI (tm)
[21:45] <yessir> semiosis: dpkg: unknown option -1
[21:46] <SpamapS> yessir: thats an l not a 1
[21:46] <semiosis> l as in list
[21:46] <yessir> semiosis: do you mean l letter?
[21:46] <semiosis> no i mean l as in list
[21:47] <yessir> I got it
[21:47] <yessir> :)
[21:51] <yessir> Anyone know how can I connect to my server using Filezilla via SFTP?
[21:53] <semiosis> yessir: thats actually pretty easy, if you can SSH into your server, you should be able to SFTP in as well... make sure you use SFTP (SSH File Transfer Protocol) and not the other choice FTPS (which is something different)
[21:55] <yessir> Sounds too simple but yet too difficult for me
[21:55] <yessir> :(
[21:55] <yessir> How can I make sure it's SFTP not FTPS
[21:56] <semiosis> it's the Server Type option in Filezilla, when you are setting up the connection options
[21:59] <yessir> semiosis: Should I use HTTP/1.1 using CONNECT method or SOCKS 5
[21:59] <yessir> also which port should I use?
[22:00] <SpamapS> yessir: thats for proxy!
[22:00] <semiosis> yessir: just leave that stuff alone unless the defaults dont work
[22:00] <semiosis> yessir: you should only need hostname, server type, username and password
[22:07] <yessir> hostname is my ip, right?
[22:07] <yessir> as you could tell I'm quite noob
[22:07] <yessir> :)
[22:08] <semiosis> yessir: yes sir
[22:38] <SpamapS> kirkland: your new challenge.. do for mutt what byobu did for screen. :)
[22:40] <kirkland> SpamapS: :-)
[22:41] <kirkland> SpamapS: i spent about 6 months using sup
[22:41] <yessir> I think I messup something
[22:41] <SpamapS> seriously.. mutt is awesome but you have to read 9 pages of docs to figure out how to make it do the basics.
[22:41] <kirkland> SpamapS: i liked it, but it was buggy, and being in ruby, i didn't hack on it much
[22:41] <kirkland> SpamapS: i've heard "notmuch" is pretty cool
[22:41] <yessir> I typed "yes" and pressed entered... now I see on y letters running
[22:41] <SpamapS> I've always been an Evolution user..
[22:41] <kirkland> SpamapS: but i like your challenge :-)
[22:41] <yessir> how can I make it stop
[22:41] <kirkland> SpamapS: i may well take a stab at that
[22:41] <SpamapS> But I try to do mutt about once every 9 months
[22:41] <semiosis> yessir: ctrl+c
[22:42] <yessir> semiosis: thanks
[22:42] <yessir> what is apt-get update
[22:45] <semiosis> yessir: software is distributed in "packages" which can be managed with apt-get, the update command (of apt-get) will refresh the list of packages available from the package repositories.  type "man apt-get" to see the manual for apt-get, it will explain more about it
[22:46] <yessir> semiosis: 13: Permission denied
[22:46] <yessir> Unable to lock the administration directory ......, are you root?
[22:47] <KB1JWQ> yessir: sudo that.
[22:47] <yessir> KB1JWQ: Can you give me presise command for that
[22:47] <yessir> :)
[22:50] <KB1JWQ> yessir: Have you read the fine documentation? :-)
[22:51] <yessir> KB1JWQ: I have not
[22:52] <yessir> *shame on me*
[22:52] <yessir> getting quite late now.. I should be in bed....
[22:52] <yessir> Thanks for everyone that helped me this far
[22:52] <semiosis> yw
[22:58] <jibel> hi all, there is only 1 mandatory testcase left for natty beta 1 iso testing : Install (default + RAID1) http://iso.qa.ubuntu.com/qatracker/result/5241/286 , any volunteer ?
[23:02] <hggdh> Daviey: if you are there... Dan seems to have found the issue
[23:02] <hggdh> which is sort of weird
[23:15] <SpamapS> kirkland: hmmm.. notmuch seems pretty good
[23:19] <jibel> hggdh, smoser, are you testing ubuntu server ec2 images ?
[23:19] <skaet> hggdh, around?
[23:19] <jibel> :-)
[23:19] <skaet> :)
[23:20] <hggdh> skaet, jibel: sorry, got involved in raid1 tests and my critically unloved bug
[23:23] <skaet> hggdh,  we're just wondering if the ec2 tests are close to be finished off?
[23:29] <osmosis> how do I stop a service from starting at boot?
[23:30] <hggdh> skaet: I did not have time yet to look at them, but Jenkins ran a series of tests. I will collate them now to the tracker
[23:30] <semiosis> osmosis: which service?
[23:30] <skaet> hggdh,  cool.    they're the biggest unknown at this point in time.
[23:31] <osmosis> any service
[23:31]  * skaet will be glad when jenkins talks to iso tracker directly ;)
[23:31]  * hggdh also
[23:31] <skaet> jibel, ^^
[23:31] <osmosis> my friend is telling me that ubuntu has no chkconfig and its problematic.
[23:31] <hggdh> skaet: I intend to have at least one session on UDS on this
[23:31] <skaet> hggdh,  please sign me up.  :)
[23:32] <hggdh> skaet: roger wilco
[23:37] <ejv> hi guys, partitioning new disks to be assembled into a new mdadm software raid, because the disks are 2TB, 4k format, I used GPT this time around. Is setting: `parted /dev/<device> set <partition number> raid on` akin to setting fd for Linux raid autodetect? or are they fundamentally different? should I not set any 'flags' at all? thank you! :)
[23:39] <osmosis> semiosis, ^
[23:44] <cjwatson> ejv: kind of similar, I guess; it makes the GUID of that partition be one that means "this is a RAID physical volume"
[23:45] <ejv> cjwatson: ok
[23:45] <cjwatson> you should set that flag, yes
[23:45] <ejv> heheh ok, i want to be able to (in the future) call `mdadm --auto-detect` and have it pick up these GPT partitions and roll them together back into the array
[23:49] <osmosis> im being told that upstart is inferior to sysV because it doesnt factor in USR1, USR2 signals that some daemons use for log rotation and graceful reloading
[23:50] <SpamapS> osmosis: in that case, sysv's init daemon stays out of that game entirely. The scripts implement those signals.
[23:50] <SpamapS> osmosis: I'd say that in this regard, upstart is deficient when compared to LSB-init ..not sysv
[23:51] <cjwatson> wouldn't you send those signals directly to the daemon anyway?
[23:51] <SpamapS> osmosis: anyway, custom reload actions is an open bug in upstart.
[23:51] <cjwatson> oh, I guess you mean how reload is hardcoded to be SIGHUP
[23:51] <SpamapS> cjwatson: the init script allowed aggregating all of the things relevant to that daemon in one place
[23:52] <cjwatson> yeah.  it wouldn't be hard to implement custom reload actions, I suspect
[23:52] <SpamapS> it was sort of the operational clearing house for the daemon.. upstart is more "do it our way or don't do it"
[23:52] <SpamapS> 'tis an old bug .. bug 94873
[23:53] <SpamapS> Most daemons that implement such things have their own control program too.... apachectl .. mysqladmin.. pg_ctl ..
[23:55] <osmosis> good info. thanks