[00:00] <Patrickdk> yes
[00:00] <Patrickdk> when you change the dns, you would also adjust the redirect
[00:00] <Patrickdk> I am assuming each webserver is a dns server though
[00:00] <Patrickdk> for routing issues
[00:00] <weeb1e> In certain cases, such as my own, when streaming audio over HTTP, not all clients support multiple (or in rare cases, any) redirects, the only option is DNS based
[00:00] <Patrickdk> that way some of the internet can use one, and some of the internet hits the other
[00:01] <xperia> yes actually the one is a master and the other a slave but i can turn them both as masters
[00:01] <weeb1e> Patrickdk: I gave DNS as the first option, a web server based redirect is the second
[00:01] <Patrickdk> yes, but no reason you can't use both together :)
[00:01] <weeb1e> A third web server redirecting instead of using DNS changes will emilinate any downtime
[00:01] <weeb1e> But is not always suitable
[00:01] <Patrickdk> and then add bgp anycast on top :)
[00:01] <Patrickdk> personally I normally do bgp anycast -> loadbalancers -> bankends
[00:02] <weeb1e> Not everyone is in a position to use anycast :)
[00:02] <xperia> bgp anycast hmm never hear till yet. thing is i dont want to balance i want to cascade aka use allways server 1 in case server 1 shutdown use server 2
[00:02] <weeb1e> I'd love to use multicast, but my country not so much
[00:02] <Patrickdk> then your stuck doing dns
[00:02] <weeb1e> So YMMV
[00:03] <RoyK> IPv6 FTW!
[00:03] <Patrickdk> what? mobileip works? :)
[00:03] <Patrickdk> mobileip could so solve this :)
[00:03] <RoyK> IPv6 should be mandatory for all services
[00:03] <xperia> yeah dns is then the thing to go. thanks a lot for your suggestions.
[00:04] <weeb1e> xperia: Good luck
[00:04] <weeb1e> 2am, good time for a breakfast break.. then back to work
[00:19] <bananapie> hi. I have Ubuntu 10.04 LTS server and need to use debootstrap to build a machine with Trusty 14.04 LTS. i tried soft linking gutsy to trusty in /usr/share/debootstrap/scripts but it says it's not a valid suite
[01:14] <jak2000> hi all how to remove all java stuff? http://pastie.org/9432985  cant :(
[01:16] <sarnold> jak2000: try this: sudo dpkg --purge openjdk-6-jdk openjdk-6-jre openjdk-6-jre-headless openjdk-7-jre openjdk-7-jre-headless openjdk-7-jre-lib sun-java5-jdk sun-java6-jdk
[01:17] <jak2000> sarnold ok
[01:28] <jak2000> sarnold: http://pastie.org/9433006 not worked
[01:29] <jak2000> continue allmy javastuff on my ubuntu server
[01:30] <sarnold> jak2000: great -- re-run the command with "icedtea-6-jre-cacao openjdk-6-jre-lib default-jre-headless icedtea-6-jre-jamvm icedtea-7-jre-jamvm" added to the end
[01:32] <jak2000> ok, rerun...
[01:35] <jak2000> sarnold: http://pastie.org/9433017 same or similar result :(
[01:36] <sarnold> jak2000: this is normal :) now add "ca-certificates-java libcommons-dbcp-java tomcat7-common" to the command :)
[01:39] <jak2000> http://pastie.org/9433020 :(
[01:40] <sarnold> jak2000: add openjdk-6-jre-headless libtomcat7-java   :)
[01:41] <eatingthenight> hello I am having an odd issue. I have gone through every issue I can find on google regarding networking for ubuntu server 12.04 but I can't ping from this server to an IP or even the gateway. I am able to ssh into it though and access the webserver it runs.
[01:42] <eatingthenight> I have tried using static and dhcp to no avail.
[01:43] <sarnold> eatingthenight: do you get error messages when you try to ping the gateway's IP?
[01:43] <eatingthenight> no it will just hang I have left it for about 10 minutes and it just hangs.
[01:44] <jak2000> sarnold: http://pastie.org/9433029
[01:45] <sarnold> jak2000: openjdk-6-jre-lib  :)
[01:47] <sarnold> eatingthenight: hunh, that's odd.
[01:47] <eatingthenight> yeah i know it's very strange.
[01:48] <eatingthenight> I have never had this before. Server has been running great for 2 years.
[02:01] <xeno_> Say, my friend is trying to set up a remote desktop experience on an Ubuntu Server.  He is trying to provide a remote desktop experience, but apparently there are access problems that allow shutdown of the server which he wants to preclude and cannot.  Any suggestions?
[02:02] <xeno_> I know from experience that Gnome is less than comfortable on server for my tastes, but that may not be pertinent to this.
[02:04] <xeno_> Typically my friend provides remote user access to servers for shell, but not root access at all.
[02:04] <jak2000> sarnold: http://pastie.org/9433049
[02:07] <sarnold> jak2000: nice. that's probably all of them :)
[02:07] <sarnold> xeno_: I think what you need is in /usr/share/polkit-1/actions/org.freedesktop.consolekit.policy
[02:10] <sarnold> xeno_: oh. this looks more useful: http://askubuntu.com/a/251942/33812
[02:11] <xeno_> Yes, but he wants to make shutdown require an admin password.  Hmm.
[02:49] <MavKen> any recommended open source control panel for 14.04?
[02:52] <eatingthenight> what do you want to control?
[02:52] <eatingthenight> for the entire server?
[02:55] <MavKen> yes
[02:55] <MavKen> cpanel ..not zpanel
[02:56] <MavKen> cpanel replacement
[03:17] <eatingthenight> MavKen so you just want something for your database? or do you have a server that is running everything?
[03:18] <MavKen> I have a small hosting business.. I use cpanel to add new clients and them have access to a cp
[03:18] <MavKen> I just didnt like zpanel
[03:18] <MavKen> and don't want to pay for cpanel anymore
[03:19] <MavKen> since most of them use wordpress, a bash script to create a new user, create db user, install wp would be sweet...but I don't have experience with that
[03:20] <sarnold> MavKen: the juju wordpress charm may have some ideas you can mimic
[03:43] <jak2000> sarnold hi friend are you there?
[03:44] <sarnold> hey jak2000 :) it's almost time for me to quit for the night..
[03:44] <jak2000> waita moment please
[03:46] <jak2000> see please: http://pastie.org/9433133
[03:47] <sarnold> jak2000: yes?
[03:47] <jak2000> this files, deleted? manually?
[03:47] <jak2000> or not problem?
[03:48] <sarnold> jak2000: everything in that directory is cached downloaded packages -- keep that directory in mind, it can be very very helpful :)
[03:49] <jak2000> ok
[03:49] <sarnold> jak2000: if you need disk space, you can delete them. but they don't bother anything
[03:50] <jak2000> now i can install: apt-get install openjdk-7-jdk   right?
[03:51] <jak2000> sarnold can you give me a favour? paste me the last link, before i am discnnected,thanks(i want documment my problem) thanks
[03:52] <sarnold> jak2000: http://pastie.org/9432985 http://pastie.org/9433006 http://pastie.org/9433017 http://pastie.org/9433020 http://pastie.org/9433029 http://pastie.org/9433049 http://pastie.org/9433133
[03:52] <jak2000> thx
[03:52] <jak2000> goodnight
[03:52] <sarnold> nn :)
[03:53] <jak2000> ok installing now openjdk-7
[04:17] <jak2000> sarnold beeautiful,,,  with your help
[04:18] <jak2000> see please: http://pastie.org/9433163
[04:18] <jak2000> no errors now
[04:18] <sarnold> jak2000: nice :)
[04:18] <sarnold> jak2000: scorched earth never looked so good
[04:19] <jak2000> before your help i cant shutdown the tomcat service....
[06:17] <pds|2> hi guys wrote a preseed file, getting from utc for some reason
[07:53] <MavKen> for 1gb ram, any reason to get 64bit on vps?
[08:07] <gnuoy> I'm trying to understand why the  alembic package in ubuntu isn't a straight sync and what change Ubuntu has atop of the upstream version.  http://paste.ubuntu.com/7912454/
[08:08] <gnuoy> rbasak, I know it's not patch Friday but if you have a sec at some point ^
[08:51] <jak2000> hi all, how to open a port on ubuntu-server?
[08:52] <TJ-> jak2000: define "open" ?
[08:52] <jak2000> telnet 162.252.243.81 8080   connection refused...
[08:54] <TJ-> jak2000: Do you control 162.252.243.81 ?
[08:54] <jak2000> i want thee users can access: http://mydomain.com:8080  understand me TJ-?
[08:54] <jak2000> yes
[08:55] <TJ-> jak2000: OK, what firewall control are you using on that server?
[08:55] <jak2000> the default i have 14.04 version
[08:55] <TJ-> jak2000: the manual way would be to add a rule using iptables, but that won't stick across reboots unless you have some form of firewall manager, such as "ufw" or "shorewall" installed
[08:56] <jak2000> ithink ufw
[08:57] <jak2000> sudo ufw allow 8080/tcp   and   sudo ufw reload  ?
[08:57] <TJ-> jak2000: Then you'd need to add a rule using UFW to allow port 8080 TCP
[08:58] <jak2000> mmm not opened :(
[09:00] <TJ-> jak2000: what process is listening on port 8080 ?
[09:01] <TJ-> jak2000: if no process is listening, having the port allowed in the firewall won't change anything
[09:01] <jak2000> ok
[09:01] <jak2000> now opened :)
[11:14] <pmatulis> morning
[11:54] <jamespage> ok that's enough triage of really old bugs for one day...
[11:54] <jamespage> pm-utils I love you
[11:59] <pmatulis> me?  ;)
[12:15] <RoyK> patdk-wk: btw, your theory that enterprise drives don't have these smart counters doesn't seem right. the two root drives connected to local sata ports are also seagate constellation, enterprise drives
[12:43] <patdk-wk> my definition of enterprise disks was, ones that talk scsi, and non-enterprise talk ata
[12:43] <patdk-wk> generally that means sata=ata and sas=scsi
[12:43] <patdk-wk> now, if a manufacturer, say it's enterprise, but still talks ata, heh
[12:45] <patdk-wk> if the disk talks sata, you get attributes and thresholds
[12:45] <patdk-wk> if the disk talks scsi, you don't
[13:18] <JediMaster> hi guys, I'm upgrading to VMs from 13.10 to 14.04.1 LTS with do-release-upgrade, and both have been sitting at "Installing new version of config file /etc/pam.d/sshd ..." for the last half hour, I can still SSH into both, any ideas?
[13:19] <Abhijit> hi
[13:21] <pmatulis> JediMaster: woof, try http://serverfault.com/questions/321520/ssh-connection-lost-during-server-upgrade-how-to-reconnect-to-process
[13:21] <JediMaster> pmatulis, it's not lost the conenction, but thanks
[13:21] <pmatulis> having an ssh pam change during an upgrade is nasty
[13:22] <pmatulis> JediMaster: i know, try it anyways, the upgrade may be continuing and you don't know it
[13:22] <JediMaster> pmatulis, it's still connected to the ssh & do-release-upgrade, I can type and see characters come up
[13:22] <pmatulis> JediMaster: the key is to try reconnecting to the screen/byobu session
[13:23] <JediMaster> that never works for me
[13:23] <JediMaster> screen -list returns: "No Sockets found in /var/run/screen/S-root."
[13:23] <JediMaster> screen -x and byobu do not reconenct to it either
[13:23] <JediMaster> I am really still conencted anyhow
[13:24] <pmatulis> may be using tmux?  dunno.  need to try it myself
[13:24] <pmatulis> tmux ls
[13:24] <JediMaster> failed to connect to server
[13:25] <pmatulis> can you view logs?
[13:25] <JediMaster> I can see this running however:  screen -e \0\0 -L -c screenrc -S ubuntu-release-upgrade-screen-window /tmp/ubuntu-release-upgrader-n_ottj/trusty --mode=server --frontend=DistUpgradeViewText
[13:26] <pmatulis> maybe use some byobu command?
[13:26] <pmatulis> i don't use byobu, but tmux directly
[13:26] <pmatulis> byobu -r ?
[13:27] <pmatulis> you must do it with the root account i think
[13:27] <JediMaster> I am
[13:31] <StolenToast> Hello
[13:31] <StolenToast> I am attempting to use OpenVPN with my server but it will not function is the firewall is enabled, even though port 1194 is allowed
[13:31] <StolenToast> *if
[13:43] <weeb1e_> When trying to start cpufrequtils, I get "disabled, governor not available...", googling this issue just returns tons of questions with no solutions. The most informative thread I found says that someone was able to fix it by downgrading to an older kernel, and that the newer kernel is bugged and the devs do not want to fix it
[13:44] <weeb1e_> But that was in 2011, I've used newer kernels on other boxes just fine
[13:44] <weeb1e_> So does anyone know how I can check if the reason cpufrequtils is due to hardware, such as an incorrect bios setting?
[13:45] <weeb1e_> *the reason cpufrequtils is failing
[13:53] <ed8> hi, does the cron's string @reboot work on Ubuntu-server?
[13:53] <Pici> ed8: yes
[13:59] <ed8> Pici: are you aware of http://unix.stackexchange.com/a/109805/17362
[14:00] <Pici> ed8: curious... I recall it working at some point.
[14:01] <Pici> I don't have an Ubuntu system I can reboot here to test with though...
[14:02] <patdk-wk> at some point?
[14:02] <patdk-wk> it has always worked
[14:03] <ed8> patdk-wk: did you read the thread I linked to?
[14:03] <patdk-wk> yes, but nothing in that thread matters
[14:04] <patdk-wk> there is a lot of speculation and references to other people
[14:04] <patdk-wk> nothing referencing any bug reports, issues, or problems
[14:04] <pmatulis> ed8: why not simply test it?
[14:04] <ed8> pmatulis: that's why I'm here
[14:04] <ed8> I tested, it doesn't work
[14:04] <patdk-wk> what version of ubuntu?
[14:04] <ed8> here is my crontab content: https://gist.github.com/edouard-lopez/b45c524c26b8ffecc576
[14:04] <ed8> 14.04
[14:05] <ed8> + cron 3.0pl1-124ubuntu2
[14:06] <patdk-wk> works fine here on 14.04
[14:06] <ed8> my bad, my path to the scripts is incorrect
[14:08] <pmatulis> works fine here too (14.04)
[14:08] <pmatulis> ed8--
[14:09] <arges> smb: utlemming : we can discuss bug 1350522 here when you get a chance.
[14:10] <ed8> thanks :)
[14:14] <RoyK> hm... I beleive I asked here earlier, but I can't find it in the logs. Is there a simple way to enable disk LEDs for broken disks? IIRC there was something under /sys somewhere
[14:42] <dasjoe> RoyK: that should be possible, if your disk is in an enclosure with a lsi expander you get "/sys/class/enclosure/SOME_ID/Slot\ 01/" and can echo to "fault" or "locate", "echo 1 > /sys/class/enclosure/1\:0\:21\:0/Slot\ 01/fault" turns on a red light for me
[14:51] <RoyK> dasjoe: any idea how I can map device names to a slot without a manual lookup?
[14:56] <bitfury> hi, how would I upgrade security packages only?
[14:56] <bitfury> sorry I worded that wrong, I'm looking to just do security updates.
[15:00] <dasjoe> bitfury: I'd use unattended-upgrade from the unattended-upgrades package. By default it runs security upgrades daily, which you can disable if you want to run it manually
[15:01] <bitfury> yeah I would like to run it manually, I'll look into it thank you
[15:04] <pmatulis> bitfury: use the -security repo only?
[15:09] <bitfury> pmatulis: guess that could be an option as well
[15:09] <bitfury> :)
[15:10] <rberg_> Hi all.. I think I found a kernel regression in ubuntu 12.04. I was tracking down why in kvm and virtualbox the console resolution is 640x480 when I set it to 800x600 in /etc/default/grub with kernel 3.2.0-67-generic.. when I booted 3.2.0-48-generic the console was the correct resolution.
[15:32] <robot_z> How can I check total disk size versus total disk usage/availability as with df -H?
[15:32] <robot_z> Or, does dh -H give that info and I just am failing to read it correctly
[15:32] <robot_z> df -H*
[15:38] <rberg_> #24 on https://bugs.launchpad.net/ubuntu/+source/grub2/+bug/699802 seems to do the trick for me.. huzzaaa!
[16:03] <nyc-h0st> hi all, having a problem with Kerberos + LDAP, i can obtain krb tickets just fine and my user exists in ldap, however trying to login over ssh fails, logs are visible under; http://pastebin.com/BDCDFS2J
[16:04] <nyc-h0st> kerberos is embedded with pam and nss is configured to use ldap
[16:23] <lykwydchykyn> HI, I'm having a problem with my 12.04 ltsp server.
[16:23] <lykwydchykyn> HTOP shows the cores are constantly between 80 and 100%, but no processes are showing high cpu usage
[16:24] <lykwydchykyn> I've posted a screenshot on Ubuntu forums here: http://ubuntuforums.org/showthread.php?t=2237048
[16:24] <lykwydchykyn> Anyone seen this sort of behavior before?  It's just started recently, this server's been running for years without problems.
[16:31] <arrrghhh> lykwydchykyn, ick.  have you tried 'atop'?
[16:31] <arrrghhh> also, are you running with sudo
[16:31] <lykwydchykyn> I've tried sudo, but not atop
[16:31] <lykwydchykyn> I'll try it presently
[16:32] <sarnold> lykwydchykyn: "The administrator may have required you to register before you can view this page."
[16:32] <arrrghhh> sarnold, sign in
[16:33] <arrrghhh> lykwydchykyn, atop will show stuff like IO so if you have a disk going nuts
[16:33] <sarnold> hooray, the single-sign-on actually worked :)
[16:35] <lykwydchykyn> I'm trying to parse through the output of atop
[16:35] <sarnold> lykwydchykyn: can you pastebin ps auxw output?
[16:36] <lykwydchykyn> http://pastebin.com/8c9fFgwb
[16:37] <arrrghhh> lykwydchykyn, screenie of atop too if you have time :)
[16:37] <arrrghhh> it should highlight any "issues" if youw ill
[16:39] <lykwydchykyn> http://imgur.com/TUvO1sM
[16:40] <lykwydchykyn> The disk usage doesn't look high, but I'm not sure what normal would be
[16:40] <arrrghhh> hm ya it all seems processor bound
[16:41] <arrrghhh> ran with root on atop as well yes?
[16:41] <lykwydchykyn> Yes
[16:41] <sarnold> the easiest way to chew up that kind of CPU time without it showing up in ps or top or similar is to have a few forkbombs running where the parent immediately exits after spawning a child; no process lasts more than a few miliseconds but there's thousands of them born and killed each second
[16:42] <lykwydchykyn> My ltsp users keep running ps all the time, is that normal?
[16:42] <arrrghhh> lykwydchykyn, the other thing I can think of is back in htop, have you tried sorting by "time"?  see if there is a process that's staying up?
[16:43] <sarnold> lykwydchykyn: seems unlikely...
[16:43] <lykwydchykyn> the top process for time is a python process running twistd, seems to be related to epoptes
[16:44] <lykwydchykyn> It has 9:06:53 for TIME
[16:44] <arrrghhh> :/
[16:44] <lykwydchykyn> Not sure what killing it would do
[16:45] <sarnold> nine hours of CPU time for a heavily used service isn't too surprising..
[16:45] <arrrghhh> ya I don't think that's it
[16:48] <lykwydchykyn> wish I knew what was running ps all the time.  Is there a way to see parent pids in something like top?  I couldn't find it in the man page
[16:49] <lykwydchykyn> The processes disappear to quick to inspect their /proc folders
[16:50] <sarnold> lykwydchykyn: hit 'f', arrow down to ppid, hit space, then q
[16:52] <lykwydchykyn> Hrm.  It's my xsession script
[16:53] <sarnold> lykwydchykyn: if you're up for some work, you could install the auditd package, use auditctl to load syscall auditing rules and audit the 'clone' system call. if my guess is right, the logging will be horrendously quick, so be prepared to turn off the logging again before you run the command :)
[16:53] <sarnold> lykwydchykyn: oooo
[16:53] <lykwydchykyn> I have a line in there that checks to see if the browser is already running, and if not, it runs it.
[16:53] <lykwydchykyn> But it should only run if the browser is closed, which shouldn't happen really at all...
[16:54] <sarnold> lykwydchykyn: can you pastebin your xsession?
[16:56] <lykwydchykyn> http://pastebin.com/nSa7XpCW
[16:56] <lykwydchykyn> It's the same one that's been used for a while
[16:58] <sarnold> lykwydchykyn: if you run wcgbrowser in a terminal, do you get your prompt back quickly? or do you have to kill the browser to get your prompt back?
[16:59] <lykwydchykyn> You have to kill the browser
[16:59] <lykwydchykyn> It should block
[17:00] <sarnold> lykwydchykyn: dang.
[17:00] <sarnold> lykwydchykyn: well, I mean, that's good, but it means I don't spot the easy answer :)
[17:00] <lykwydchykyn> :) yeah
[17:01] <jak2000> hi sarnold thanks for your help
[17:01] <lykwydchykyn> I can't remember why I had to do the ps thing, it's been a while (I should comment these things).  Maybe I can just take that bit out.
[17:02] <lykwydchykyn> I have to run for a bit; thank you sarnold and arrrghhh
[17:02] <lykwydchykyn> I'll check back if anyone has any ideas...
[17:03] <sarnold> lykwydchykyn: hrm. good point. take out the ps | grep ..
[17:04] <sarnold> lykwydchykyn: another option is to take out the "keep it running" thing entirely -- let xsession end when the browser dies or is closed, and restart the session some other way
[17:07] <jak2000> sarnold you develop?
[17:08] <sarnold> jak2000: not much; I audit, complain, and sometimes fix :)
[17:10] <jak2000> oo interesting then your specialization is SYSADMIN?
[17:11] <sarnold> jak2000: no, security :)
[18:24] <lykwydchykyn> Hey all
[18:32] <lykwydchykyn> I think I may have figured something out
[18:32] <lykwydchykyn> My LTSP users seem to have two .xsessions running
[18:53] <lykwydchykyn> The clients get rebooted every night, but the old sessions stay running; since they can't run the browser without the client attached, they just loop ps forever
[18:53] <lykwydchykyn> I'll just add something to kill all the client sessions after the reboot
[18:54] <sarnold> lykwydchykyn: I really think you'd do better to take out the while true .. bit
[19:01] <lykwydchykyn> Yeah, but I need these to respawn if the browser gets closed for some reason
[19:01] <sarnold> lykwydchykyn: can't you just let the xsession die and then have the session manager restart the whole thing?
[19:01] <lykwydchykyn> I suppose so; it just takes that much longer
[19:02] <lykwydchykyn> But all in all, it's probably a better appraoch
[19:02] <lykwydchykyn> I'll test that on my dev system
[19:03] <lykwydchykyn> I've been killing off the superfluous .xsessions, the effect on the cpu usage is dramatic
[19:23] <lykwydchykyn> OK; infinite loop is axed.  I think this will be more stable now.  We'll find out in the morning!
[19:23] <lykwydchykyn> Thanks again for all the help!
[22:25] <WACOMalt> Hey folks. I just had my dedicated server host install ubuntu-server for me. I didnt specify a regular user, and it seems that ubuntu doesnt allow my root login by default
[22:25] <WACOMalt> is there any way for me to get in without re-loading ubuntu and specifying a normal user this time?
[22:27] <WACOMalt> apparently I can use the "ILO" to login as root, but I have no idea what that is
[22:30] <Patrickdk> ILO is HP for remote console
[22:36] <WACOMalt> Ok, I got that working. thanks.
[22:53] <jiffe98> anyone know any good cloud server providers besides digitalocean?
[22:53] <jiffe98> good/cheap
[22:54] <sarnold> jiffe98: hetzner.de is popular, I'm not sure if I'd call them "cloud" though..
[22:58] <Patrickdk> hetzner has been good to me
[22:58] <Patrickdk> I need to do a refresh sometime
[22:58] <Patrickdk> server has been running there for 5 years without ever being offline
[22:59] <sarnold> wow
[22:59] <Patrickdk> it runs kindof like tbp :)
[22:59] <Patrickdk> it's just a proxy server, no disk usage or anything
[22:59] <Patrickdk> just ipsec back to the real server
[23:00] <sarnold> tbp?
[23:01] <Patrickdk> tpb :)
[23:03] <sarnold> ahhh :)
[23:54] <WACOMalt> anyone in here used Virtualmin/Webmin?