[00:00] <Spiritus> Like Obiwan :)
[00:01] <Spiritus> Pondering Aragorn freeze.
[00:03] <Spiritus> Yompa: So, what do you do here ? Talk about why servers arent updated and breaches are made ?
[00:07] <Spiritus> The US military wanted some coding so i did that, but didnt even get a lousy t-shirt in return. Sure it was cool, but come on! I spent 3 days on those changes, day and night wo sleep.
[00:08] <Spiritus> And all i asked for was a t-shirt to hang on my wall.
[00:11] <Spiritus> Ooh, i see a lack of empathy.
[00:12] <Spiritus> Ah, theusgas
[00:14] <Spiritus> Report: The only sign of intelligent life is called "Yompa" :)
[00:18] <jeeves_moss> how can I script a way of moving a file from the input directory to an output directory, however, leave a "touched" file with the same file name in the input directory?
[00:20] <Spiritus> mv PATH/oldfile PATH/newfile |
[00:22] <Spiritus> mv PATH/oldfile PATH/newfile | echo @1 ... if this -> @$1 or $1 gives the path/filename you can use $2 etc
[00:22] <Spiritus> Depends on what shell youre scripting it in..
[00:23] <jeeves_moss> Spiritus, I'm guessing it's bash.  it's the standard one
[00:23] <Spiritus> #!/bin/bash or #!/bin/sh -> tcsh etc
[00:24] <jeeves_moss> Spiritus, #!/bin/sh
[00:24] <Spiritus> Always let the runtime shell know what file its trying to compile. At top of the file give the origin: #!/bin/bash (Newline here)
[00:25]  * RoyK packs up to thaw up a little in Reykjavík
[00:25] <Spiritus> Use bash, the docs arent good for sh
[00:25] <Spiritus> RoyK: Rockinvik! /greetz
[00:26] <RoyK> :)
[00:28] <Spiritus> jeeves_moss: stdout doesnt work well with #sh as with #bash if its for example a forked process in some cases.
[00:28] <Spiritus> On Gotland we have Rouks! :)
[00:28] <Spiritus> Im from the mainland though
[00:29] <Spiritus> Rauks
[00:30] <Spiritus> qbric ... thats a sucky network trying to keep the truth from the people, aka RealPlayer.
[00:30] <Spiritus> They also feed us with false information on a daily basis since 2003.
[00:31] <Spiritus> IE: The Swedish government tv garbage
[00:33] <Spiritus> 90% of what they show is disecting humans and or blowing them up (NCIS, some murering idiot and some doctor guy with a caine etc etc etc)
[00:34] <Spiritus> Or that fucking chef thats all about "JallaJalla/YallaYalla" or whatever stress they can put their young under.
[00:34] <Spiritus> the f-word idiot
[00:37] <Spiritus> I like Simpsons, Seinfeld, Absolutely Fabulous, SG and fishing programs. I hate opera, mtv and anything stressfull
[00:37] <Spiritus> -1l
[00:37] <Spiritus> Why hasnt mtv been cancelled yet ?
[00:37] <The_Tick> because it generates tons of revenue
[00:37] <Spiritus> Complete garbage.
[00:38] <Spiritus> So idiots watch that, really ?
[00:38] <The_Tick> there's some decent shows on it I suppose
[00:38] <The_Tick> but it doesn't matter
[00:38] <The_Tick> if the show can generate viewers
[00:38] <The_Tick> then the ads go from 40k to like 300k
[00:38] <The_Tick> for a single 30 second clip
[00:38] <The_Tick> or even higher
[00:38] <The_Tick> so if a show has what, 3 or 4 ads
[00:38] <The_Tick> per break
[00:38] <The_Tick> and 2-3 breaks
[00:39] <The_Tick> and gets rerun, what, 30 times a week?
[00:39] <The_Tick> tons of cash
[00:39] <The_Tick> you go tell mtv to shut the doors and they'd just laugh at you
[00:39] <The_Tick> because you don't get it :)
[00:40] <The_Tick> in the US when the whole late show problem occured with conan and jay leno
[00:40] <The_Tick> jay went to an earlier time slot
[00:40] <The_Tick> almost nobody watched it
[00:40] <Spiritus> I have watched it, couldnt find anything good on it "Oh i got a lipsucktion those balls" :)
[00:40] <Spiritus> MTV sucks arse!
[00:40] <Spiritus> I actually have to pay for it wether i want to or not. Thats the really odd part.
[00:40] <The_Tick> the ads STILL generated 40k per ad
[00:40] <The_Tick> per showing of the ad
[00:40] <The_Tick> with almost nobody watching
[00:41] <Spiritus> 90% of the money i pay for my tv license goes to needy people in the under developed countries. Why cant i just go there with some companies and build them some wells ?
[00:41] <The_Tick> you pay to watch tv?
[00:41] <Spiritus> MTV blows
[00:41] <The_Tick> woah, what?
[00:41] <The_Tick> you actually pay?
[00:41] <Spiritus> You do too, tv-license
[00:41] <The_Tick> that's hilarious
[00:41] <The_Tick> nope
[00:41] <The_Tick> there's no licensing in the US
[00:42] <The_Tick> and I live in the US
[00:42] <The_Tick> so I do not pay
[00:42] <Spiritus> Swedes pay billions each year to watch tv we ourselves made.
[00:43] <Doonz> hey does anyone use byobu here?
[00:43] <The_Tick> silly swedes
[00:43] <Spiritus> Then again, we are the richest country on earth right now... i doubt it.
[00:43] <The_Tick> richest because you pay dumb taxes like that one
[00:43] <Spiritus> The_Tick: Wanna play with the fjords of doom ? :) /You make me so happyfied :)
[00:44] <Spiritus> !Entail ^^
[00:44] <Spiritus> You have entered a database of doom, want to climb out ?
[00:45] <The_Tick> you hate mtv and yet consider this entertaining?
[00:45] <Spiritus> !Envicathiem est
[00:45] <Spiritus> You seem nice
[00:47] <Spiritus> The_Tick: Isnt chatting with people better then mindnumbing brainlessness ?
[00:47] <The_Tick> no
[00:47] <Spiritus> IF you think not, then why are you here ?
[00:47] <The_Tick> I think there's room for both
[00:47] <Spiritus> LOL
[00:47] <The_Tick> but I think bitching about it forever is annoying
[00:47] <Spiritus> Youre as fun as a sack of old potatoes
[00:48] <Spiritus> Youre rubber im glue... whatcha gonna do ? :)
[00:48] <Spiritus> We are fun to the onlooking crowd. Im liking it.
[00:49] <Spiritus> The_Tick: Dont you think IRC can also be fun, like it was before 2003 ?
[00:50] <Spiritus> I for one liked that betterwebs
[00:51] <The_Tick> Spiritus: I think it can't be like it was in 95, no
[00:53] <Spiritus> The_Tick: Ill settle for 2000, atleadt then crazy manicas didnt roam freely as ops (jokes where ok ans so where discussions). Then Idiots entered by governmet policy and removed all the fun, so we had to have fun with the evils.
[00:53] <Spiritus> And it was indeed fun.
[00:54] <Spiritus> So if i see anyone saying !ot again anywhere ill know its a microsoft employee amongst a few other tell tell signs.
[00:55] <The_Tick> Spiritus: but you're from sweden
[00:55] <Spiritus> We dont like those ;)
[00:55] <The_Tick> your opinions on irc do not matter
[00:55] <Spiritus> Explain ?
[00:55] <The_Tick> too close to norway
[00:55] <Spiritus> So then finland is crap as well ?
[00:55] <Doonz> does anyone know how to add more hard drives to the status line in byobu?
[00:56] <Spiritus> To you ?
[00:56] <Spiritus> The_Tick: So Mrs tick :) ... Do you live in US ?
[00:57] <The_Tick> Doonz: they don't have good docs?
[00:57] <Doonz> no actually
[00:57] <The_Tick> bla
[00:57] <lifeless> Doonz: have you checked out 'man byobu' ?
[00:57] <The_Tick> any config file?
[00:58] <Doonz> http://manpages.ubuntu.com/manpages/maverick/en/man1/byobu.1.html
[00:58] <Doonz> this is what im reading
[00:58] <Doonz> but what the man pages are saying the config file isnt really lining up
[00:58] <Spiritus> I miss the US inputs. I talk to my coders by phone instead.
[00:59] <Doonz> it doesnt really say how to add more disk monitoring in it
[00:59] <The_Tick> Doonz: what kind of monitoring does it provide anyhow?
[01:00] <Spiritus> But talking to people on irc tells if people are good or not so i prefer hireing them here granted good test results or very good credentials.
[01:00] <Doonz> basically on the status line it just shows free space of the drive
[01:00] <Doonz> 106GB,3%
[01:00] <Doonz> thats what it shows for my main drive but i have 3 others that i would like to monitor
[01:00] <Spiritus> df -hP /
[01:01] <The_Tick> Spiritus: unless you know what the hell he's asking
[01:01] <The_Tick> which you don't based on that answer
[01:01] <The_Tick> please be quiet
[01:01] <kirkland> Doonz: so you want to monitor multiple drives?
[01:01] <Doonz> Spiritus: yeah trying to avoid typing that
[01:01] <Doonz> kirkland: yea
[01:01] <Doonz> i have sd[a-d]
[01:01] <Spiritus> The_Tick: Yes sir! :) /rotfls
[01:01] <The_Tick> Doonz: why do you need to know that constantly? :)
[01:01] <kirkland> Doonz: okay, do this ... "mkdir ~/.byobu/bin"
[01:01] <Doonz> cause my wife has a habit of filling up the hdd's
[01:01] <The_Tick> heh
[01:02] <Doonz> she loves photography
[01:02] <The_Tick> Doonz: besides this I'd add a cronjob to run every 15 minutes, and if you're at 10% email yourself
[01:02] <Doonz> but 10Gb of pictures of our vase in the kitchen......
[01:02] <The_Tick> 10% left
[01:02] <Doonz> yeah i have that set up now
[01:02] <Doonz> :)
[01:02] <kirkland> Doonz: cp /usr/lib/byobu/disk ~/.byobu/bin/17_disk
[01:02] <kirkland> Doonz: then edit ~/.byobu/bin/17_disk
[01:02] <kirkland> Doonz: and have it monitor each of your disks
[01:02] <Doonz> ok let me try
[01:03] <kirkland> Doonz: that's how you create a "custom" byobu status script
[01:03] <kirkland> Doonz: you could make that script do whatever you want
[01:03] <Spiritus> The_Tick: http://en.wikipedia.org/wiki/Brigadier_General .. :)
[01:04] <Doonz> uuh
[01:04] <Doonz> hmm
[01:04] <Doonz> in the original if statment do i just change it to a 2
[01:05] <Doonz> and then set the mp to where ever that drive is mounted
[01:05] <Spiritus> The_Tick: By helping to specify and construct the GNU and in coding Linux i think i could possibly know what im doing sir :) .. It means Freedom!
[01:06] <The_Tick> Doonz: sounds like you'll have to play with it
[01:06] <Doonz> yeah wish me luck woot
[01:06] <Doonz> !!!
[01:06] <Doonz> bbl
[01:07] <Spiritus> I think freedom can be whiffed. The freedom to know.
[01:10] <Spiritus> I like this new bounce vpn via wirelsss
[01:12] <Spiritus> Packets embedded via subnet broadcast PEBCAK's bounced from lands to seas. Its fairly cool.
[01:14] <Spiritus> Im on wireless atm, but the interchanging HUBs pass the traffic on via SO_LINGER and reconnects under the timeouts. Im loving it
[01:15] <Spiritus> Has anyone tried headerpassing ?
[01:15] <Spiritus> portknocking, but thats bound to fail at some point..
[01:17] <Spiritus> Why... Consider a max amount of knocks=65535 (IANA registered port range) .. Hack the kernel and you have 0-infinite almost.
[01:17] <Spiritus> Hackers today are so confined
[01:19] <Spiritus> Same goes for subnetting and adressing. Otherwise youre playing ... on the fields of barley ... :)
[01:19] <Spiritus> Not to say Barley isnt cool, because it is my friends ;)
[01:20] <Spiritus> On the fields of barley! ... nananna!
[01:20] <fluvvell> Spiritus, move it to ubuntu-offtopic please
[01:21] <Spiritus> fluvvell: Would you say that barley is raw sugar, or stronger then raw sugar sir ?
[01:22] <Spiritus> Or did you wnt to ask a computer related question ?
[01:25] <Pici> Spiritus: Could you please try to stay on topic here.  If you want to chat about randomness, there is #ubuntu-offtopic
[01:26] <Spiritus> char c = 'x';
[01:26] <Spiritus>     while((c = cin.get()) != '\n' && c != '\0'); ?
[01:28] <sbeattie> Pici: thank you. I think you were the 4th person to ask Spiritus to respect the channel topic.
[01:56] <april__> i'm trying to move a bunch of files from one folder to it's parent folder. I tried 'for f in *; do mv /path/to/directory/"$f"; done' and i get the error "mv: missing destination file operand after '/path/to/directory/filename.zip' what am i missing?
[02:05] <qman__> april__, mv * ../ ought to work
[02:07] <qman__> and you're missing a source or destination file
[02:07] <qman__> you only have one file listed after the mv command
[02:15] <Daethz> That pici seems way gay.
[03:00] <mike01> has anyone configured honeyd on ubuntu server?  i'm testing it out via 127.0.0.1 - my virtual honeypots respond to pings, but not TCP connections
[03:14] <i0nic> trying to restart ssh but its not in /etc/init.d/ssh ?
[03:17] <dragoon123> Hi, I just installed Hibernate and rebooted my server know it is stuck @ the loading screen I tried recovery mode but it's the same. How do I get around this?
[03:18] <twb> i0nic: in 10.04, /etc/init.d/ssh.conf provides sshd normally, so run "restart ssh".
[03:18] <twb> Sorry, /etc/init/ssh.conf
[03:18] <twb> i0nic: the sysvinit /etc/init.d/ssh is provided by disabled by default, because the upstart version can't handle some ssh configurations.
[03:21] <hansin> i0nic: I am not sure if I am getting this right (twb, correct me if I am wrong), but with many "init scrips" moving from sysvinit to upstart, there are changes happening. And right now some scripts are native upstart style and some still sysvinit. Sounds like SSH uses an upstart script, but supplies the sysvinit (in /etc/init.d/) as backup.
[03:21] <i0nic> twb ahh thanks
[03:22] <twb> hansin: openssh-server is unusual
[03:22] <twb> NORMALLY when an upstart job is provided, a backwards compatibility symlink is placed in /etc/init.d
[03:22] <i0nic> so im trying to setup a reverse dns between a tablet and my server and I keep getting
[03:22] <hansin> I'm just sort of figuring this all out, so if what I say is obvious, just ignore. Okay, I think I get it.
[03:23] <i0nic> ssh_exchange_identification: Connection closed by remote host
[03:23] <i0nic> ssh_exchange_identification: Connection closed by remote host
[03:23] <i0nic> ssh_exchange_identification: Connection closed by remote host
[03:23] <i0nic> ssh_exchange_identification: Connection closed by remote host
[03:23] <i0nic> ssh_exchange_identification: connection closed by remote host
[03:23] <i0nic> ouch, sorry for spamming
[03:23] <i0nic> my keys look to be right.. would this be logged somewhere?
[03:24] <hansin> Most cases if there is a native upstart script, then /etc/init.d/ just contains a symlink to native upstart script. *But*, because SSH upstart script has limitations, a true native sysvinit script (not a symlink) is also included as well. This is what you are saying, right?
[03:27] <dragoon123> Anyknow any fast ways of troubleshooting a kernel loading prob?
[04:11] <Doonz> is there a bandwdith monitoring program for the cli?
[04:14] <twb> i0nic: ssh access issues are logged in auth.log *on the server side*.
[04:14] <twb> i0nic: by design, the client is not told WHY it was refused access.
[04:14] <twb> i0nic: I'm gonna guess that the problem is insecure permissions on ~/.ssh or ~/.ssh/authorized_keys
[04:34] <hansin> Doonz: I found this link; it might be close to what you are looking for. Both listed packages are in the Ubuntu repos: http://www.surlyjake.com/linux/linux-command-line-bandwidth-monitor/
[04:50] <twb> Hmm, free -m is reporting
[04:50] <twb> -/+ buffers/cache:        680       7283
[04:51] <twb> But nothing in top has mem% above 0.0
[04:51] <twb> So what's using all that 680MiB?
[05:25] <Cromulent> hmm what does the mysql_secure_installation command actually do? I forgot to run it and would like to take the hardening steps myself
[05:28] <The_Tick> http://dev.mysql.com/doc/refman/5.1/en/mysql-secure-installation.html
[05:29] <The_Tick> http://dev.mysql.com/doc/refman/5.5/en/mysql-secure-installation.html
[05:29] <The_Tick> and that's just by using google
[05:29] <Cromulent> ah thanks I assumed that was an ubuntu thing for some reason
[05:29] <The_Tick> most things are not
[05:30] <Cromulent> seems I have already done that
[05:30] <The_Tick> some things are, but bleh
[09:16] <kaushal> hi
[09:16] <kaushal> Can some one please suggest me about which Raid Controller Card is http://pastebin.ubuntu.com/541366/
[10:31] <garymc> Hi anyone know where my php error logs are stored? Im using PHP5.3
[10:37] <kaushal> Hi
[10:38] <kaushal> is there hpacucli available for 10.04 ?
[10:48] <jpds> kaushal: I believe you have to use the RPMs.
[10:48] <kaushal> jpds: oh ok
[10:48] <kaushal> jpds: Thanks
[10:49] <jpds> alien and all that jazz.
[10:49] <kaushal> jpds: sure
[10:49] <kaushal> will update you now
[10:53] <qman__> garymc, php errors are normally logged through the web server, which for apache would be /var/www/apache2/error.log by default
[10:53] <qman__> err
[10:54] <qman__>  /var/log/apache2/error.log
[10:54] <qman__> not www
[10:55] <kaushal> jpds: hi again
[10:55] <kaushal> i get http://pastebin.ubuntu.com/541402/
[10:55] <kaushal> Please suggest further
[11:15] <kaushal> jpds: got it now
[11:15] <kaushal> it worked
[11:15] <kaushal> fine
[13:17] <piquadrat> Hi! I just installed phpmyadmin on maverick, and it pulled in apache as a dependancy. But phpmyadmin works perfectly fine with other http servers like nginx or lighttpd. Is that dependancy really warranted?
[13:28] <patdk-wk> heh, it depends on mod-php :)
[13:29] <patdk-wk> that would be the bug I suppose
[13:29] <ilovegrolsc> my server keeps getting password attack ssh root login
[13:29] <ilovegrolsc> dozens of different ip's
[13:29] <patdk-wk> install fail2ban :)
[13:29] <ilovegrolsc> i have 10 login attemps per second
[13:31] <binBASH> change ssh port
[13:31] <ilovegrolsc> how
[13:31] <ilovegrolsc> at the moment im manually doing iptables -s mofo'sip -j DROP
[13:31] <patdk-wk> if you install fail2ban it will do the drops itself for you :)
[13:31] <patdk-wk> no config needed
[13:31] <ilovegrolsc> oh
[13:31] <ilovegrolsc> i'll google it thx
[13:31] <patdk-wk> unless you want to ban more than just craploads of failed ssh logins
[13:32] <ilovegrolsc> it would be nice if instead of just deny or drop there was like a 'fuckYou+deny' option
[13:32] <patdk-wk> tarpit :)
[13:33] <patdk-wk> that would require editing how fail2ban adds rules, and compiled the xtables modules
[13:34] <Daviey> patdk-lap: Best to raise a bug :)
[13:34] <Daviey> (re apache with phpmyadmin, via mod-php)
[13:34] <patdk-wk> daviey, heh, a quick look shows a crapload of packages require mod-php, some even mod-php4
[13:35] <Daviey> patdk-lap: eeeeeeek
[13:35] <ilovegrolsc> what does this mean
[13:35] <ilovegrolsc> Dec  9 08:34:14 vps1098 sshd[9525]: Did not receive identification string from UNKNOWN
[13:35] <ilovegrolsc> only entry after i blocked the hacker
[13:35] <ilovegrolsc> in my log
[13:36] <Daviey> ilovegrolsc: Looks like you were probed.
[13:36] <Daviey> (it's common)
[13:36] <ilovegrolsc> yea im not panicking or anything
[13:36] <ilovegrolsc> actually i dont care if ppl try to login they would never guess my password
[13:36] <ilovegrolsc> but i care if that 10 attemps per second slow down my server
[13:36] <ilovegrolsc> use resources
[13:37] <patdk-wk> I had it overload a t1 line
[13:37] <ilovegrolsc> must've been alot
[13:39] <ilovegrolsc> i'll install fail2ban sounds good
[13:40] <patdk-wk> iptables -L fail2ban-ssh -nv
[13:40] <jpds> ilovegrolsc: Remove the password and use SSH keys only?
[13:40] <ilovegrolsc> that sounds good but i dont know how to do it
[13:40] <patdk-wk> jpds, that still wouldn't help not overload a t1 with invalid requests
[13:41] <ilovegrolsc> i dont even know how to change the ssh service port
[13:41] <jpds> patdk-wk: Set up an SSH TARPIT.
[13:41] <patdk-wk> that is all editing /etc/ssh/sshd_config
[13:41] <jpds> patdk-wk: A la http://www.outflux.net/blog/archives/2010/11/10/tarpit-iptables-target/
[13:41] <patdk-wk> jpds, that is what I did, just fail2ban is esay and simple for the detection to manage the tarpit :)
[13:41] <ilovegrolsc> in 24 hour period i have like 20 different ip's doing 10 logins per second
[13:41] <ilovegrolsc> most from China
[13:42] <jpds> ilovegrolsc: Yeah, normal.
[13:44] <ilovegrolsc> seems simple if i only need edit sshd config
[13:45] <patdk-wk> ya, as much as changing the port helps and stuff, it's still only security through obscusion
[13:45] <patdk-wk> main reason I don't bother
[13:45] <jpds> ilovegrolsc: https://help.ubuntu.com/community/SSH/OpenSSH/Keys
[13:46] <patdk-wk> keys will make sure pretty much, you aren't brute forced via password
[13:46] <patdk-wk> but they can still chew up bandwidth
[13:46] <ilovegrolsc> yea im certain they would never guess my pass
[13:46] <ilovegrolsc> i'll get fail2ban and stop them from trying
[13:46] <patdk-wk> or login from a virus infected machine?
[13:47] <tdn> How do I make a bootable Ubuntu server installer from USB via ssh from another old ubuntu server?
[13:48] <ilovegrolsc> pki might be wort it if i was assange and the gov wanted into my server ;)
[13:48] <ilovegrolsc> worth*
[13:48] <jeremyA> with lucid 2.6.32-26-server on an Athlon64 X2 (AMD-based motherboard), I am experiencing frequent system hangs -- video goes dead, no kbd LEDs work, only hitting the restart button brings the machine back.  This is a server box without X running.  Nothing is logged.  Any tips for troubleshooting?  I have verified that RAM and cpu cooling are good, using memtest86+ and cpuburn.  No SMART errors are reported by the drives.
[13:49] <jeremyA> I have also tried the 2.6.34 mainline kernel, with little luck.
[13:49] <jeremyA> it lasted longer before hanging up -- 10 hours, when run with clock_source=hpet, but it still hung up.
[13:49] <patdk-wk> heh, I find pki logins nice, it's so fast :)
[13:51] <ScottK> Daviey: Next time you do a mass bug file for FTBFS, you might review the logs first to see if the build failures are actual package problems.
[13:51] <ilovegrolsc> just wondering what to type in console to download direct2ban
[13:51] <ScottK> I just checked all your failures that weren't also in http://udd.debian.org/cgi-bin/ubuntu_ftbfs.cgi and a large fraction of them were artifacts of how you did the rebuild.
[13:52] <Daviey> ScottK: If there is a next time i probably will... this was an experiment.
[13:52] <ScottK> OK.
[13:52] <jeremyA> ilovegrolsc:  I use pubkey authentication all over the place
[13:52] <Daviey> ScottK: Are you interested in being more involved the standard daily triage of the package set?
[13:52] <jeremyA> it's tons better than password, even for low-security things
[13:53] <ilovegrolsc> i'm already using it on my server for openvpn
[13:53] <ScottK> Daviey: Not for "fun" in my free time, no.
[13:53] <ScottK> I'm more interested in not getting bogus bugs landing in my inbox.
[13:53] <Daviey> ScottK: Do you find investigating Invalid FTBFS bugs fun?
[13:53] <Daviey> (seriously)
[13:53] <ScottK> No.
[13:53] <ilovegrolsc> with a beer on hand it might be fun
[13:54] <ScottK> Daviey: No.  I'm seriously annoyed you filed the bugs without even looking at the logs, thus wasting my time.
[13:54] <Daviey> ScottK: I did look at that log actually, i mean't to pluck it out.
[13:54] <Daviey> ScottK: However, i don't understand why you wasted time looking at them?
[13:54] <tdn> How do I make a bootable USB to isntall Ubuntu server?
[13:55] <ScottK> Daviey: The first one I looked at because it landed in my inbox, was Python related, and we've got a Python transition going on.
[13:55] <Daviey> ScottK: It would be really useful to know what part of the server you do find fun, it might be better use of the teams time as a whole - and allowing us to help present the stuff you do find fun.
[13:56] <ScottK> Daviey: The rest I looked at after I saw your email to ubuntu-server and I didn't want new contributors getting discouraged when asked to look into crap bugs.
[13:58] <patdk-wk> ilovegrolsc, apt-get install fail2ban
[13:58] <Daviey> ScottK: I really don't want you to waste your time on these, if you feel that it's of no benefit.
[13:59] <ScottK> I think it would have been a fine effort to involve new people if the bugs had represented actual failures.  IMO an unreasonable fraction of them didn't.
[14:00] <smoser> hallyn_, https://bugs.launchpad.net/ubuntu/+source/qemu-kvm/+bug/687997 is there a bug in qemu-kvm's apport-hooks ? there is no qemu-kvm package version there.
[14:00] <ScottK> Personally I use lists like http://udd.debian.org/cgi-bin/ubuntu_ftbfs.cgi and http://qa.ubuntuwire.com/ftbfs/ to work on FTBFS and not bugs in general to work on build failures, but I can see there value for others.
[14:01] <smoser> never mind, hallyn_ i see it.
[14:01] <hallyn_> smoser: ok
[14:02] <smoser> its not in the blobs in the summary
[14:02] <smoser> but in related pacakges
[14:02] <ilovegrolsc> 0.8.1 is latest version patdk?
[14:02] <Daviey> ScottK: Initially, i wasn't going to share the results.. but some other members of the team were interested... I nearly sent them to the mailing list... but i had concerns with noise on there.
[14:03] <patdk-wk> 0.8.4 on lucid
[14:03] <hallyn_> yeah, 'attach_related_packages'
[14:03] <Ninjix> anyone have any experience setting up 802.3ad bonding on Dell PowerConnect switches?
[14:03] <ScottK> My problem isn't with the list or the sharing, just the lack of QA before doing so.
[14:03] <Daviey> ScottK: Getting an extra 23 bugs in my Inbox, i didn't consider to be a big deal - perhaps i was wrong.
[14:03] <patdk-wk> ninjix, lots of netgear and cisco :)
[14:04] <ScottK> Daviey: If they are valid bugs it's not.
[14:04] <zertyu> hi
[14:04] <zertyu> is there any script to disable plymouth ?
[14:04] <Daviey> ScottK: This was purely a first cut.. i don't know if i'll do it again
[14:04] <Ninjix> patdk-wk: I'm only seeing one interface getting used
[14:05] <patdk-wk> how did you configure it?
[14:05] <patdk-wk> dynamic or static? I found dynamic isn't very reliable
[14:05] <patdk-wk> and you need to configure it on both ends
[14:05] <Ninjix> patdk-wk: I've setup LACP static
[14:05] <patdk-wk> on both switchs?
[14:06] <patdk-wk> or is this switch to computer?
[14:06] <Ninjix> switch to computer
[14:06] <patdk-wk> how do you know it's only using one?
[14:06] <Ninjix> I'm running bmon
[14:06] <patdk-wk> what is bmon?
[14:07] <Ninjix> handy interface monitoring tool for console
[14:07] <patdk-wk> ok, but that isn't a test
[14:07] <patdk-wk> I assume yo uattempt a file transfer or something while watching bmon?
[14:07] <consumerism> i have written some utility scripts i want to make available to users on several servers. i've never made a .deb before and the scripts are just a few files in a directory that have historically just been added to $PATH for people to use them. what's the best way to get these more easily packaged/installable?
[14:07] <Ninjix> for test I'm running bonnie++ against NFS on 10g NAS
[14:08] <patdk-wk> well, that will only use one interface, how 802.3ad works :)
[14:08] <patdk-wk> how if you did a bonnie++ against 4 different nfs servers
[14:08] <patdk-wk> then it should use both
[14:09] <Ninjix> ahh...
[14:09] <patdk-wk> 2 is probably enough, but just incase :)
[14:09] <patdk-wk> it's to solve the whole, tcp packets out of order thing
[14:09] <patdk-wk> one wire might be longer or slow than the other
[14:09] <ilovegrolsc> ok installed, in jail.conf i can put my static ip in the ignoreip list? just wanna make sure im not locked out
[14:09] <patdk-wk> so all ip -> ip traffic uses the same interface
[14:09] <Ninjix> gotcha
[14:10] <Ninjix> it's less magical than I had hoped
[14:10] <patdk-wk> well, it depends what you want
[14:10] <patdk-wk> to access one very large nfs, kind of useless
[14:10] <Ninjix> that's how obeys ethernet standards
[14:10] <patdk-wk> to serve to the internet, very useful :)
[14:11] <Ninjix> I see different hash methods L2,L3 and L4
[14:11] <ilovegrolsc> why does it want my email address in the jail.conf patdk?
[14:11] <ilovegrolsc> am i gonna get an email every time it blocks an ip?
[14:11] <ilovegrolsc> that would be bad
[14:11] <patdk-wk> Ninjix, I forget, the hash methods might need to match the other side (the switchs)
[14:11] <Ninjix> is that negotiated by the kernel driver?
[14:12] <patdk-wk> ilovegrolsc, it can email you each time it bans someone
[14:14] <Ninjix> patdk-wk: thanks for the quit explanation. Now I know what to look for in the IEEE docs and have found more information
[14:15] <Ninjix> now I just need to find better Linux docs covering the output of /proc/net/bonding/bond0
[14:16] <patdk-wk> heh? it looks pretty straight forward :)
[14:18] <ilovegrolsc> one last Q about it... i changed ssh-iptables to true, thats all? cuz i also see ssh-tcpwrapper
[14:18] <ilovegrolsc> and ssh-ipfw
[14:18] <Ninjix> should the Aggregator ID equal the same integer I've assigned to the LAG on the switch? or is it internal to the server?
[14:19] <patdk-wk> Ninjix, looks like the hash method doesn't matter much
[14:19] <patdk-wk> but you would have to change it on the switch to affect your incoming hash method from other stuff to your machine
[14:20] <patdk-wk> so using the later3+4 hash method would use more than one interface if you mounted the same nfs server more than once
[14:20] <Ninjix> ahh... so it has more to do with how the switch decides which line throw the packet down
[14:20] <patdk-wk> but I think packets from nfs to your computer will still only use one interface :)
[14:20] <Ninjix> that matches my test
[14:21] <patdk-wk> most switchs don't let you change it
[14:32] <ilovegrolsc> hmm
[14:32] <ilovegrolsc> i see fail2ban has added its own chain to iptables
[14:51] <raubvogel> ilovegrolsc, I thought that is what it always does, so it is easy to find the rules it adds
[14:54] <ilovegrolsc> its working
[14:54] <ilovegrolsc> kicking ass and adding iptables rules
[14:57] <cdubya> is there a package available for spam/virus filtering? I read about the suggestion on wiki and talk about possibly doing something with tasksel or something like that, but has anything come of it?
[14:57] <raubvogel> cdubya, you mean besides spamassassin and the likes?
[14:57] <cdubya> yeah
[14:58] <keyz182> Hi all, I've got a question about vmbuilder. When I run it, once it's finished doing it's thing I get the directory ubuntu-xen as expected, and inside are the filesystems. They seem to have names based on "tmp" followed by a random string. Is there a way to control the names of the files?
[15:05] <b0gatyr> anyone know of a way to make the terminal window "flash" when the audible bell goes off?
[15:11] <zertyu> hi
[15:11] <zertyu> is there any equivalent of rbash ?
[15:12] <soren> Just use rbash.
[15:12] <i0nic> anyone running lighttpd/php  ?
[15:12] <i0nic> I could use a howto on this =)
[15:12] <zertyu> rbash
[15:12] <zertyu> i can't understand how it works
[15:13] <soren> zertyu: Why do you think you need it, then?
[15:13] <zertyu> ok i try to explain my problem
[15:14] <zertyu> what i try to do is : i got a user called user1 i want to restrict access on my server to /var/www/sitex only
[15:14] <zertyu> rbash is the correct one ?
[15:15] <soren> Depends.
[15:16] <soren> But probably not.
[15:16] <zertyu> so what tool i have to use ?
[15:20] <zertyu> well
[15:22] <pmatulis> zertyu: what will that user be doing on the server?  based on the path you gave, probably uploading/downloading files.  in that case, consider a sftp chroot
[15:23] <zertyu> i simply want to allow user1 just only on that folder /var/www/sitex and access to read and write
[15:24] <pmatulis> zertyu: maybe a ssh chroot then
[15:25] <zul> JamesPage: the groovy looks good im just trying to reproduce and then ill merge
[15:25] <JamesPage> zul: OK; ping me if anything is weird....
[15:39] <zul> JamesPage: i just had to fix the changelog but it looks good anyways
[15:40] <JamesPage> zul: I should try to remember what release I'm working on :-)
[15:40] <JamesPage> zul: thanks
[15:40] <zul> JamesPage: no probs thanks
[15:47] <hallyn_> smoser: so, just wondering - why is bug 687997 triaged?  you know what it is?
[15:47] <hallyn_> bc, i've not seen that myself with several recent windows installs...
[15:49] <smoser> hallyn_, i do not know the problem, but i believe there is enough information to reproduce.
[15:49] <smoser> do you suggest a better state ?
[15:49] <smoser> there is a kvm command line there also.
[15:50]  * hallyn_ scratches his head
[15:50] <hallyn_> I guess I need to go re-read the bug status descriptions
[15:53] <smoser> hallyn_, i could be mistaken
[15:54] <smoser> or likely am. what would you have set it to ?
[15:54] <hallyn_> well, not sure - i guess i'd leave it at New until someone can confirm with the recipe
[15:55] <hallyn_> thing is, if it's Triaged I assume that means upstream knows about it and has an idea how to fix it
[15:55] <hallyn_> (or, we have an idea how to fix it)
[15:55] <hallyn_> which means - I'll ignore it when scanning bug lists
[15:55] <hallyn_> which is why i need to figure out whether i'm doing that wrong :)
[15:58] <\sh> ivoks: pingeling...lucid + drbd .. where in father xmas name is drbd module?
[15:58] <hallyn_> smoser: all right, well i guess i'm misinterpreting.
[15:59] <hallyn_> sorry :)
[16:00] <smoser> hallyn_, well, 'New' is definitely not righ.
[16:00] <smoser> but i dont really know what it should be there. ...
[16:00] <hallyn_> smoser: no, you were 100% right
[16:01] <hallyn_> Triaged means precisly that you think there's enough info there that someone can work on it, and it looks like a real bug
[16:01] <ivoks> \sh: in kernel
[16:01] <\sh> ivoks: it isn't
[16:01] <hallyn_> now, i think several of the qemu bugs are dups - i think there is a problem with the i386 emulator
[16:01] <\sh> http://packages.ubuntu.com/lucid-updates/amd64/linux-image-2.6.32-25-generic/filelist <- look for drbd
[16:01] <ivoks> 2.6.32?
[16:02] <ivoks> ah, not in 2.6.32
[16:02] <\sh> WHOOOT
[16:02] <ivoks> it got in in 2.6.33
[16:02] <ivoks> drbd8-source
[16:04] <\sh> this is not the truth?
[16:04] <\sh> lucid without drbd?
[16:04] <\sh> by default i mean
[16:04] <ivoks> drbd8-source package has dkmsed module
[16:04] <ivoks> i don't see where is the problem
[16:04] <\sh> no
[16:04] <\sh> module-assistent dkmsed the module now
[16:05] <SpamapS> hallyn_: re bug 687535 from yesterday.. I comented on it a bit.
[16:05] <\sh> it doesn't do it automagically
[16:05] <ivoks> \sh: it does, you need linux-headers-server (or -whatever server you use)
[16:05] <ivoks> \sh: after that, just installing drbd8-source will give you drbd module
[16:05] <ivoks> which will be compiled on every kernel upgrade
[16:06] <hallyn_> SpamapS: oh, thanks - i didn't get notes about that, must have forgotten to subsribe
[16:06] <ivoks> (-whatever kernel you use)
[16:06] <SpamapS> hallyn_: indeed you did
[16:06] <raubvogel> How can I quickly check if cron.d is being read? Would the log files show it?
[16:06] <hallyn_> SpamapS: have i mentioned how happy i am this morning to have found pentadactyl?  :)  i can 'p' and 'y' in firefox again
[16:06] <\sh> ivoks: why don't we have a dep on the linux headers when I install drbd8-source?
[16:07] <ivoks> \sh: we do: linux-headers-server | linux-headers-generic | linux-headers
[16:08] <RFleming> Greetings and salutations
[16:09] <RFleming> Could someone shed some light on how to calculate how much space a partition will take based on a raw disk's total capacity?
[16:10] <hallyn_> SpamapS: i agree with the guy - your 'invalid' comment was weird.  did you mean a dup of the other bug?
[16:10] <al> oh, great
[16:10] <al> exim is giving away root shells for free
[16:11] <ivoks> you deserved it
[16:11] <\sh> ivoks: hmmm....http://paste.ubuntu.com/541505/ <- have a look
[16:11] <al> http://www.exim.org/lurker/message/20101207.215955.bb32d4f2.en.html
[16:11] <ivoks> for playing with exim :)
[16:11] <ivoks> \sh: i know, i can't help you; you need to install headers first
[16:11] <\sh> ivoks: http://paste.ubuntu.com/541507/ <- the headers were already installed
[16:11] <SpamapS> hallyn_: Ugh, I meant Invalid in Upstart only.
[16:11] <\sh> that's my problem
[16:12] <ivoks> you are not looking right \sh
[16:12] <hallyn_> SpamapS: now that you're leading the upstart community :) what's your expert opinion on his pidfile suggestion?
[16:12] <ivoks> \sh: you don't have -server headers installed
[16:12] <\sh> ivoks: right, I'm having a -generic kernel
[16:14] <SpamapS> hallyn_: the trouble with pidfiles is race conditions
[16:14] <ivoks> Building for 2.6.32-25-server and 2.6.32-26-server
[16:14] <ivoks> doesn't look like
[16:15] <SpamapS> hallyn_: so lets say sshd forks on SIGHUP, then the new process writes its own pid to the pidfile, so upstart sees that and begins tracking the new pid great. But meanwhile, the old parent has exitted, and upstart has gone "uh oh, parent exitted" .. and started a new one, which failed..
[16:16] <\sh> ivoks: oh crap
[16:17] <SpamapS> hallyn_: I actually think pidfiles would be fine, but they've been rejected with aggressive hand waving multiple times by keybuk, and I'm sure he has a long winded good reason for it as well.
[16:18] <zul> SpamapS: not just keybuk as well
[16:19] <hallyn_> SpamapS: zul: sometimes a hammer doens't work right, but sometimes it's the only thing htat will
[16:19] <adamzap> hey guys im trying to do an ubuntu server install with only ubuntu-minimal packages, so ubuntu-standard. is there a trick to this? thanks
[16:19] <\sh> ivoks: gotcha
[16:19] <zul> hallyn_: hammer is not the right tool in this case...seldgehammer would be better
[16:19] <adamzap> *no ubuntu-standard
[16:20] <hallyn_> zul: do you have another suggestion for the sshd case?
[16:20] <zul> hallyn_: no unfortunately
[16:21] <hallyn_> zul: so, i marked it low prio, meaning i think it can wait for a design.  SpamapS marked it high priority, in which case i think waitint for the perfect design is irresponsible
[16:21] <SpamapS> cjwatson: re bug 687535 , I was wondering if you had any thoughts on why we chose to use 'expect fork' instead of sshd -D in openssh's upstart config.
[16:22] <RFleming> Does anyone know how to calculate the amount of usable disk space after a disk is partitions based on the raw disk size?
[16:22] <SpamapS> hallyn_: if sshd -D works, then thats the fix until we switch to proc connector / cgroups / whatever.. (and if its either of those, they can't be backported to lucid anyway AFAIK)
[16:22] <RFleming> IE: I have 20TB of raw space, and want to calculate what the total space is AFTER partitioning.
[16:22] <hallyn_> RFleming: i usually use df (or df -h)...
[16:23] <ivoks> hallyn_: usually? :)
[16:23] <RFleming> hallyn_, that works but I'm trying to calculate it before I purchase the hardware :)
[16:23] <SpamapS> RFleming: fdisk will tell you how much it was able to use before writing.
[16:23] <hallyn_> ivoks: other times i write a script to fill up the disk with '.', and then run wc to see how man '.'s there are :)
[16:23] <SpamapS> RFleming: ahh based on the geometry?
[16:23] <RFleming> yep
[16:23] <ivoks> my guess is that he thinks that partitioning reduces disk size
[16:24] <RFleming> ivoks, partitioning does reduce partition size :)
[16:24] <hallyn_> really it's the fs metadata that can punish you the most
[16:24] <ivoks> you do know that manufacturers lie about disk sizes?
[16:24] <SpamapS> RFleming: should be "damn close to 20TB"
[16:25] <ivoks> 120 GB isn't really 120GB... it's more like 100
[16:25] <RFleming> the whole GB/GiB thing
[16:25] <SpamapS> ivoks: they do not lie. They merely force reading every bit of microscopic fine print on their website and on the box the drive comes in to make sure it means 20*1024*1024*1024*1024 and not 20,000,000,000,000 ;)
[16:25] <hallyn_> SpamapS: so i'm still trying to figure out why your pidfile scenario makes sense in any case other than pathologically mis-behaving application
[16:25] <SpamapS> hallyn_: Me too actually
[16:26] <ivoks> SpamapS: right
[16:26] <hallyn_> i don't tthink cgroups work
[16:26] <SpamapS> hallyn_: why not?
[16:26] <hallyn_> well, then when you log in you'll be in that cgroup too :)
[16:27] <hallyn_> and really, it doesn't help any more than your pidfile objection - if sshd forksand then crashes, upstart again will just think the child is the running task
[16:27] <Psi-Jack> chkconfig doesn't work with Ubuntu 10.04.1 really, does it?
[16:27] <Psi-Jack> Everytime I try to use it with lsb scripts, it just ... throws a bunch of crap and doesn't even do anything.
[16:27] <hallyn_> SpamapS: now i suppose we can have PAM move you back out of the ssh cgroup, but then if that fails things will get exciting
[16:27] <RFleming> SpamapS, the reason I ask is I'm creating ~830TB volume.
[16:28] <hallyn_> can i have an account?
[16:28] <SpamapS> hallyn_: agreed, following forks is difficult. Letting the process tell you which pid to track is a good idea. I guess the question is, does a pidfile do a good job of that?
[16:28] <RFleming> sorry, we're not like Exim ;)
[16:28] <hallyn_> :)
[16:29] <RFleming> that 830TB is raw.  The partition table is going to be larger than a few gigs :)
[16:30] <hallyn_> SpamapS: well, the pidfile lets you know if there is a mismatch between what the service said and reality.  which you can flag as pathological, kill, and restart
[16:30] <hallyn_> now i dunno, maybe we should seriously follow up on the cgroup idea.  though i thought Keybuk hated that
[16:30] <SpamapS> hallyn_: I believe he does. ;)
[16:30] <hallyn_> SpamapS: oh, the ohter problem with cgroups:
[16:31] <hallyn_> SpamapS: they're new and fun enough that ppl are using them in funky ways, and their moutnts setup might interfere with what we want
[16:31] <hallyn_> though i guess we can just mount an empty cgroup named upstart
[16:34] <SpamapS> hallyn_: seems systemd uses cgroups
[16:36] <hallyn_> SpamapS: yeah, though i don't know exactly how
[16:36] <hallyn_> i'll go take a look i guess
[16:36] <SpamapS> RFleming: I'd just make sure its 20TiB and not 20 trillion bytes.
[16:36] <SpamapS> RFleming: the geometry loss should be minimal.
[16:37] <SpamapS> hallyn_: The way I understand the pidfile argument is simply that we shouldn't trust the daemon ever. I see the point in this, and I think I'd rather patch in --dont-daemonize cmdline options into each daemon than try to get pidfile creation working on all of them.
[16:38] <SpamapS> hallyn_: interestingly enough, many (most?) newer daemons have this built in because there are a lot of people running process trackers like upstart, runit, or daemontools
[16:38] <cjwatson> SpamapS: I think I was basically just trying to avoid changing too much at once
[16:39] <cjwatson> SpamapS: and I hadn't audited precisely what -D did
[16:39] <hallyn_> SpamapS: but i thought just doing -D and ditching expect fork was not suficient?
[16:40] <SpamapS> cjwatson: given that sshd does this fork->re-exec on SIGHUP, I don't think 'expect fork' will work, unless we can somehow tell upstart to expect fork every time we send SIGHUB
[16:40] <SpamapS> HUP
[16:40] <SpamapS> hallyn_: it is
[16:40] <cjwatson> SpamapS: if you're happy to test it to make sure everything still works, feel free to make that change
[16:40] <SpamapS> hallyn_: with -D, there's no fork->exec .. it just execs itself again.
[16:41] <hallyn_> SpamapS: i thought the bug poster had a case where it wasn't, looking
[16:41] <hallyn_> don't just tell me "it is"  :)
[16:41] <l3dx> if I have two disks in software raid, do I have to do something in order to detach them and use them separately?
[16:41] <cjwatson> RFleming: partition table sizes are typically measured in bytes or kilobytes, not gigabytes.  perhaps you're thinking of something else
[16:42] <hallyn_> SpamapS: oh, i see, he was talking about your suggestion about following the children
[16:42] <SpamapS> cjwatson: indeed, the sshd man page seems to suggest that -D is intended exactly for this purpose: "-D      When this option is specified, sshd will not detach and does not become a daemon.  This allows easy monitoring of sshd.
[16:43] <kirkland> smoser: yo
[16:43] <SpamapS> cjwatson: but its important enough to test thorougly.
[16:43] <smoser> oy
[16:43] <cjwatson> SpamapS: I think it's more likely that "monitoring" there refers to a sysadmin running it in a tty for testing purposes, actually
[16:43] <cjwatson> but sure
[16:43] <hallyn_> cjwatson: SpamapS: so any objections to taking that route?
[16:43] <cjwatson> hallyn_: as I said above, none from me provided that somebody else tests it
[16:44] <cjwatson> (to -D)
[16:44] <cjwatson> I definitely don't want to use pidfiles wiwth upstart
[16:46] <RFleming> cjwatson, then I am lost.
[16:47] <RFleming> I know drive manufacturers report space as XB not XiB, so a 1TB drive is really 0.9313TiB
[16:48] <RFleming> string 28 of those together and you get 26.077TiB, but after the disk is partitioned, the size is 25.4TiB.  Where did the ~690GiB go?
[16:48] <cjwatson> RFleming: that's just different units, nothing to do with partitioning.
[16:48] <cjwatson> er, wait, /me rereads
[16:49] <cjwatson> yes, it's units.  your disk manufacturer is lying to you more than you think.
[16:50] <Daviey> jdstrand: If you have a moment, i could really do with talking to you about apparmor :)
[16:50] <cjwatson> RFleming: 1000 * 1000 * 1000 * 1024 bytes is 0.9313TiB, yes.  However, 1000 * 1000 * 1000 * 1000 bytes is 0.9095TiB.  Multiply that by 28 and you get 25.47TiB.
[16:51] <RFleming> grr
[16:52] <RFleming> but then the unit is wrong, because it's not base 2
[16:52] <RFleming> we're back to base 10
[16:52] <RFleming> so it's not 25.47TiB but 25.47TB
[16:55] <jdstrand> Daviey: sure
[16:55] <cjwatson> RFleming: indeed.
[16:55] <cjwatson> RFleming: disk manufacturers do indeed typically quote base-10 sizes.
[16:56] <RFleming> cjwatson, ok, I just plugged in a single 1TB drive.  Windows reports it as 931.39GB
[16:56] <cjwatson> (this is why for example the Ubuntu installer's partitioner uses base-10 sizes throughout.)
[16:56] <RFleming> which really is 931.39 GiB
[16:56] <cjwatson> yes, you won't find the same names used consistently everywhere
[16:56] <cjwatson> sadly if you really care about the distinction you have to get things to show it to you in bytes
[16:56] <RFleming> cjwatson, but that brings back the original problem.
[16:57] <cjwatson> which original problem?
[16:57] <RFleming> if windows shows a 1TB drive as 931.39 GiB then 28 of those strung together is 25.47 TiB
[16:57] <RFleming> D'oh!
[16:58] <RFleming> my math sucks
[16:58] <resno> if i am looking to just add a drive to a machine, what mount point should i use?
[16:59] <resno> the drives main purpose is media stroage
[16:59] <resno> my current storage drive uses /home
[16:59] <RFleming> cjwatson, so does yours... curious how that happened
[17:02] <cjwatson> RFleming: hmm?
[17:03] <RFleming> cjwatson, your numbers all base 10 (909.5 GB * 28 / 1000) does equal 25.46 TB not TiB.  My NEW number of (931.39 GiB * 28 / 1024) gives 25.46 TiB
[17:03] <RFleming> units suck
[17:04] <cjwatson> No
[17:04] <cjwatson> 1TB base-10 == 0.9095TiB base-2; 28TB base-10 == 25.47TiB base-2
[17:06] <RFleming> 1TB base-10 == 0.93132 TiB base-2
[17:06] <cjwatson> that is incorrect, sorry
[17:06] <cjwatson> 1TB (base-10) == 1000 * 1000 * 1000 * 1000.  To convert that to TiB (base-2), divide by (1024 * 1024 * 1024 * 1024)
[17:06] <cjwatson> $ bc -lq
[17:06] <cjwatson> 1000 * 1000 * 1000 * 1000 / (1024 * 1024 * 1024 * 1024)
[17:06] <cjwatson> .90949470177292823791
[17:08] <cjwatson> the only way your calculation comes out the way you say is if you take 1TB == 1000 * 1000 * 1000KiB, i.e. 1000 * 1000 * 1000 * 1024, which is an odd hybrid unit that probably mainly serves to confus
[17:08] <cjwatson> *confuse
[17:08] <SpamapS> look at what you've done to cjwatson .. you've confus'd him
[17:08] <ilovegrolsc> any1 here able to ssl into a server?
[17:08] <ilovegrolsc> ssh
[17:09] <SpamapS> ilovegrolsc: I'd say most of us are able tos sh into a server. ;)
[17:09] <SpamapS> or to ssh either way ;)
[17:10] <RFleming> 1TB = 10^12, 1 TiB = 2^40
[17:11] <cjwatson> Yes.
[17:11] <cjwatson> $ bc -lq
[17:11] <cjwatson> 10^12/2^40
[17:11] <cjwatson> .90949470177292823791
[17:13] <RFleming> cjwatson, this is nuts.  why does a 1TB drive show up as .93139 GiB then?
[17:13] <RFleming> your math is sound
[17:13] <RFleming> but the reporting says otherwise
[17:15] <patdk-wk> heh, it should report ATLEAST .909gigs
[17:15] <patdk-wk> well, .909 terrabytes
[17:15] <cjwatson> RFleming: because units are hopelessly inconsistent
[17:16] <patdk-wk> drive makes use 1000 units
[17:16] <cjwatson> find out the size in bytes and don't worry about the "friendly" presentation
[17:16] <patdk-wk> computers use 1024 units :)
[17:16] <patdk-wk> now if everything used 500byte sectors :)
[17:18] <RFleming> .90949470177292823791 TiB = 1000000000000 B
[17:19] <cjwatson> I have no knowledge of how Windows computes disk sizes
[17:20] <RFleming> bizzare
[17:21] <RFleming> I think patdk-wk made a good point, it's at least .909, with some extra thrown in for 1024 units.
[17:21] <patdk-wk> na
[17:21] <patdk-wk> it normally depends on the lba mapping
[17:21] <patdk-wk> the drive might physically have 1.2tb of space
[17:22] <patdk-wk> but can only map it to the bios as 1.1tb, or 1.0tb depending on settings
[17:22] <patdk-wk> I dunno exactly how it works these days, I stopped caring at around 4gb drives :)
[17:22] <RFleming> 1000 * 1000 * 1000 * 1024 will make a nice 1,024,000,000,000 B or .93132 TiB
[17:22] <patdk-wk> like I can buy two 36gig sas drives, one is 36.7gigs, another is 37.2gigs
[17:23] <patdk-wk> RFleming, they might of done that, keep simple 4k or 512 sectors
[17:23] <patdk-wk> but the the firmware limits it to 1000*1000*1000 of them
[17:54] <zul> SpamapS: ping can you have a look at https://bugs.launchpad.net/ubuntu/+source/php5/+bug/316441 when you are around
[17:57] <dravekx_> what does it mean when your server says "There is 1 zombie process"
[18:08] <bluethundr> I am attempting to install sudo-ldap on a couple of machines (10.10 and 10.04LTS).. but I get the same error on each machine
[18:08] <bluethundr> http://pastebin.ca/2014942
[18:08] <bluethundr> does anyone know how I may get around this error?
[18:18] <dravekx_> rofl
[18:21] <Daviey> SpamapS / kirkland: Do either of you want to action the email to debian re bug #56679.. I see kirkland has been involved so far, but also SpamapS has a WI for that?
[18:26] <jdstrand> Daviey: whenever you file that bug, feel free to assign it to me
[18:34] <SpamapS> Daviey: seeing as Dustin is assigned I'll let him take it. I do think we need to find a way to work w/o that. I have some ideas beyond ARP btw. :)
[18:36] <kirkland> hmm, well i suppose i can;  i though cjwatson had asked Daviey to do so
[18:37] <SpamapS> round and round we go! :)
[18:37] <SpamapS> kirkland: can you confirm for me that cobbler does not have an API?
[18:37] <Daviey> kirkland: he did... i'll do it... i just noted that both of you were also involved and thought you might want to
[18:38] <kirkland> SpamapS: https://fedorahosted.org/cobbler/wiki/CobblerApi
[18:38] <SpamapS> err..
[18:38] <SpamapS> so..
[18:38] <Daviey> SpamapS: we did talk about cobbler's API last night.
[18:38] <SpamapS> Edison has what on Cobbler then?
[18:40] <Daviey> SpamapS: To me, the edison API looks easier to extend.  That is all...
[18:40] <SpamapS> hrm
[18:41] <SpamapS> my NIH avoidance alarm is going off
[18:42] <SpamapS> Daviey: readint his CobblerApi, it already does everything we want...
[18:43] <Daviey> SpamapS: I didn't realise there was a py interface to the cobblerapi... i thought it was pure xmlrpc
[18:43] <SpamapS> either way..
[18:43] <SpamapS> I'm polishing this proposal with references and examples.. and cobbler is kicking edison's ass
[18:43] <Daviey> SpamapS: Well the py interface is pretty significant
[18:44] <tonyyarusso> Anyone know why 'vmstat' consistently gives me incorrect CPU usage numbers?  ie, it always shows the CPU as idle, across multiple platforms I've tried it on.
[18:44] <Daviey> SpamapS: That seems fair...  I'm certain if Cobbler had the rot fixed, it would be fixed for good.
[18:45] <SpamapS> Daviey: ... crap.. here we go 'round agian.
[18:45] <Daviey> rot as in, Debian/Ubuntu host support and packaging
[18:45] <Daviey> SpamapS: ack :)
[18:45] <SpamapS> Daviey: I'm writing up what we'd have to do to Edison to make it work..
[18:45] <SpamapS> Daviey: btw, its not a great django app.. lots of assumptions and hard coded stuff. :-/
[18:46] <Daviey> SpamapS: It would be nice if Cobbler was fixed either way :)
[18:46] <SpamapS> Daviey: yeah, see, if we just fix cobbler, then we don't need to do anything else.
[18:46] <Daviey> SpamapS: and solves the upstream support concern.
[18:46] <SpamapS> But.. damn we seemed so excited about Edison yesterday. ;)
[18:46] <SpamapS> Daviey: heard back from them?
[18:47] <Daviey> SpamapS: I've not chased it today.
[18:52] <Daviey> jdstrand: raised that bug #688186 .. thanks for your help.
[18:53] <Daviey> SpamapS: Do you think we should commit some more time to getting Cobbler into shape?
[18:54] <RoAkSoAx>  Daviey http://www.threedrunkensysadsonthe.net/2010/07/installing-cobbler-on-ubuntu/
[18:54] <Daviey> RoAkSoAx: Thanks, aware of that :)
[18:54] <RoAkSoAx> ok :)
[18:54] <Daviey> he's the same author as edison fwiw
[18:54] <RoAkSoAx> oh I didn't know that :) thanks for enlightment
[18:55] <Daviey> RoAkSoAx: His fork still isn't quite there, but he has some good patches.
[18:55] <RoAkSoAx> Daviey:  better something than nothing ;)
[18:56] <Daviey> RoAkSoAx: Are you interested in helping out?
[18:56]  * Daviey imagines RoAkSoAx pondering.
[18:57] <RoAkSoAx> Daviey: I would, but first need to get other things done (PowerNap) :). I'm hoping to get it done in the next couple of weeks. After that I can help
[18:57] <Daviey> groovy
[18:58] <SpamapS> Daviey: I do actually think we should focus *most* of our effort on cobbler.
[18:58] <SpamapS> Daviey: for some reason I thought it was being rejected because it was built poorly or something.
[18:59] <SpamapS> But.. shoot.. whip it into Debian/Ubuntu shape.. and whats not to love?
[18:59] <Daviey> SpamapS: The main concern we had was the weight, and Deb'/Ubuntu support... and enrichment.
[18:59] <RoAkSoAx> SpamapS: cobbler has been something it's been wanted in Ubuntu for quite a long time
[19:02] <SpamapS> Daviey: yeah.. I don't care about the weight.. Deb/Ubuntu support should be doable in a series of bug fixes.. and I think what I see is that it doesn't need enrichment to achieve our goals for deploying UEC
[19:03] <Daviey> SpamapS: Aye.. i don't have a weighted opinion either way... :)
[19:03] <kirkland> SpamapS: Daviey: you guys duke it out and let me know what you decide
[19:03] <Daviey> both /will/ work.
[19:04] <SpamapS> I just see it as the standard currently for provisioning
[19:04] <Daviey> kirkland: Fancy throwing your package of cobbler into a ~ubuntu-virt bzr branch ?
[19:04] <SpamapS> we're going to have to make it work anyway
[19:04] <kirkland> Daviey: sure
[19:04] <kirkland> Daviey: one sec
[19:04] <Daviey> groovy
[19:05] <SpamapS> Yeah I think I'll take a look
[19:05] <jdstrand> Daviey: thanks, I should have an upload sometime today
[19:06] <Daviey> jdstrand: it's not a urgent for me, so don't prioritise it if you have other things.
[19:06] <kirkland> Daviey: SpamapS: pushing to lp:~ubuntu-virt/cobbler/ubuntu
[19:07] <Daviey> jdstrand: but rocking, thanks
[19:07] <Daviey> kirkland: great
[19:07] <kirkland> Daviey: SpamapS: we could talk to LP about getting cobbler's get autoimported to lp:cobbler
[19:07] <kirkland> Daviey: as I started from a bzr import of the git tree
[19:07] <kirkland> Daviey: done.
[19:07] <Daviey> \o/
[19:09] <kirkland> Daviey: i've been tracking patches in debian/patches
[19:09] <SpamapS> bzr: ERROR: Invalid url supplied to transport: "lp:~ubuntu-virt/cobbler/ubuntu": No such project: cobbler
[19:09] <kirkland> Daviey: nothing big there yet, but at some point, i'd assume we'd have a handful that we'll want to send upstream
[19:10] <kirkland> SpamapS: Daviey: whoops ...  lp:~ubuntu-virt/+junk/cobbler
[19:10] <Daviey> kirkland: I have a couple of patches
[19:11] <SpamapS> Its really hard for me to objectively test.. as the only other machine on my network is my wife's dell mini10.. :P
[19:11] <Daviey> SpamapS: kvm baby!
[19:11] <SpamapS> unless.. I could try to use it to install powerpc on my G5
[19:11] <Daviey> eeek... you are brave
[19:11] <SpamapS> Daviey: the networking on kvm scares me
[19:14] <Daviey> SpamapS: fair comment :)...  virtualbox :)
[19:18] <kirkland> cjwatson: so i have a preseed with "d-i partman-lvm/confirm boolean true"
[19:18] <kirkland> cjwatson: but I'm being held at that question anyway
[19:18] <kirkland> cjwatson: any hints?
[19:19] <kirkland> cjwatson: full preseed at http://pastebin.com/ME2CDpnx
[19:19] <SpamapS> liboobs ... one of the greatest bazingas ever
[19:20] <jdstrand> Daviey: well, it is fairly urgent as you can't use dhcpd at all atm :)
[19:20] <Daviey> jdstrand: true! :)
[19:22] <Daviey> kirkland: Interesting... just checked my preseed and that worked fine for me yesterday when i deployed natty :/
[19:22] <kirkland> Daviey: can you pastebin all or some of your preseed, then?
[19:23] <kirkland> Daviey: or tell me where mine differs?
[19:23] <i0nic> would you guys say its best practice to generate a gpg keypair on your "master server" and any nodes sending files to this server use its pub key to encrypt?
[19:23] <Daviey> kirkland: generated by uec-pro' http://pb.daviey.com/2sKc/raw/
[19:25] <kirkland> Daviey: weird ... nothing related in the diff
[19:25] <SpamapS> kirkland: impressive build-depends. :-P
[19:25] <Daviey> kirkland: Although... i did notice that my hosts in /etc/apt/sources.list were not the values i preseeded... so something could be skewed
[19:25] <kirkland> Daviey: one thing that's strange ... this machine's hard disk was detected as sdb (when it should probably be sda)
[19:26] <Daviey> (same preseed used on maverick DID give the correct hosts)
[19:26] <kirkland> Daviey: yeah, i have 3 microscopic fixes to uec-provisioning so far
[19:26] <kirkland> Daviey: i'm trying to get it back to the point where it "just works" out of the box
[19:26] <Daviey> kirkland: the itch i have is multi release support :)...  fancy tackling that? :)
[19:27] <Daviey> kirkland / SpamapS: Also, http://pb.daviey.com/qELJ/raw/ \o/
[19:27] <kirkland> Daviey: sure, if we go with uec-pro as our backend;  though i'm not going to invest any time in new feature dev if we agree upon some other technology though
[19:28] <kirkland> Daviey: under active development?
[19:28] <Daviey> kirkland: "lp:cobbler"
[19:28] <kirkland> Daviey: ah
[19:28] <kirkland> Daviey: cool
[19:28] <kirkland> Daviey: okay, my lvm-confirm issue is worked around by preseeding non-lvm disk partitioning :-P
[19:29] <kirkland> Daviey: i'll wait to hear back from cjwatson on that one
[19:32] <Daviey> yeah.. interesting :/
[19:33] <Daviey> kirkland: on your dell laptop?
[19:33] <kirkland> Daviey: yup
[19:39] <bobboau> I've suddenly lost my ability to use SSH into my server I'm not sure when it happened because I usually don't need to get into it, but when I enter my password it just hangs, everything else seems to be working fine
[19:58] <smt-mobil> since last apt-get upgrade on hardy server php isnt rendered anymore, any ideas?
[20:08] <dravekx> Anyone know how to enable tracking with phpmyadmin on 10.04LTS? I've done everything it says and it still fails.
[20:15] <jiboumans> i have the following line in my /etc/apt/sources.list: deb http://us-east-1.ec2.archive.ubuntu.com/ubuntu/ lucid restricted multiverse
[20:15] <jiboumans> this works just fine. another machine has it in sources.list.d/us-east-1-ec2.archive.ubuntu.com and it appears to be ignored
[20:15] <jiboumans> what am i missing?
[20:17] <jiboumans> .. and strace provides the answer; file must end in .list
[20:20] <SpamapS> jiboumans: how did we ever live without strace?
[20:20] <jiboumans> SpamapS: i dont know, but ever since i have to deal with Erlang and java, it's been my best friend
[20:20] <jiboumans> 'i cant read it or dont have the source..w tf is it doing???'
[20:22] <SpamapS> I don't know if they have it now, but it used to be the most annoying thing about admining *bsd
[20:22] <jiboumans> dtrace > strace though
[20:22] <jiboumans> but you can't have it all
[20:22] <SpamapS> strace named ... wtf.. cmd not found?!
[20:23] <SpamapS> dtrace is freebsd only isn't it?
[20:28] <zul> you have run strace as root now dont you?
[20:49] <SpamapS> zul: you don't have to , but you haev to run strace -p as root (or change a sysctl)
[20:50] <neopsyche> hi all.. quick question .. regarding squid.
[20:50] <neopsyche> question.. I noticed youtube recently made some changes to the way their videos are streamed.. making it difficult for 3rd party sites and programs to 'grab' their videos.. im just wondering if this effect's squids ability to cache youtube videos also.. for ISP's ?
[20:53] <fluvvell> dravekx, what do you mean tracking?
[20:54] <neopsyche> ?
[21:06] <psyferre_> hey folks.  I'm running ubuntu karmic with lvm and I'm trying to extend my root partition.  I've added unallocated space in vmware esxi, created a physical volume in the free space, added it to the volume group, and extended the logical volume to take up the free extents.  What I can't seem to figure out is how to get ubuntu to realize the volume is larger.  I'd like to use ext2online, but it appears to not be in the u
[21:07] <psyferre_> sorry about the wall of text =-/
[21:09] <h3sp4wn> resize2fs works
[21:09] <psyferre_> so boot from a desktop edition live cd and run that on the unmounted drive?
[21:10] <h3sp4wn> Anything will work (I would probably use grml dunno if there is a good ubuntu equivalent)
[21:11] <h3sp4wn> I take it you made a snapshot first ? (or a decent backup) if its important
[21:12] <psyferre_> not terribly important, but yes
[21:13] <steveng> ok i dont know if i should ask this here or not....I'm running ubuntu server 10.10 in a virtualbox vm....all of a sudden when I type ifconfig....my eth0 is not there...any ideas?
[21:14] <h3sp4wn> why not - ifconfig -a
[21:14] <Danawar2> Heyaa guys I have a server with apache2 and ftp I want to allow users to upload their websites by ftp and be able to view them via http is there an easy way I can do this?
[21:15] <h3sp4wn> steveng: Sometimes if you don't use -a then interfaces that are down won't be seen (No easy way for me to test it right this second though)
[21:15] <steveng> ok adding -a I see an eth2....?
[21:15] <steveng> no ip address though
[21:15] <steveng> is there a reason it would get changed from eth0 to eth2?
[21:16] <gholms> smoser: ping
[21:16] <psyferre_> h3sp4wn: thanks!  I appreciate it! :)
[21:16] <h3sp4wn> steveng: Did you change the interface ? (I think udev makes them unique)
[21:17] <h3sp4wn> steveng: also check the contents of /etc/iftab
[21:17] <h3sp4wn> (Might not exist)
[21:17] <h3sp4wn> steveng: Should be able to dhclient eth2
[21:17] <steveng> i did...in virtualbox....but I changed it back
[21:18] <h3sp4wn> (Thats if its in the default config)
[21:19] <h3sp4wn> steveng: Have a look in /etc/udev/rules.d
[21:19] <steveng> ok if I dhclient eth2 then I get an ip...but I want that set to static....I changed my settings in /etc/network/interfaces
[21:19] <dravekx> file permissions: If i have 'root' as the owner, and 'bar' as the group, and I make 'steve' part of bar, why cant i access the files with steve? :( Does steve need to be the owner?
[21:20] <h3sp4wn> group permissions
[21:20] <dravekx> oh right.
[21:20] <h3sp4wn> steveng: You cannot just do that with virtualbox
[21:21] <h3sp4wn> You need another interface type but I dunno what
[21:21] <dravekx> so 775 not 755.
[21:21] <h3sp4wn> yep or you can use acl's if you want
[21:26] <h3sp4wn> steveng: There is annoyances with the other 2 as well (like have to configure routing on the host) or bridged has some other quirks as well
[21:27] <steveng> eh
[21:27] <steveng> ill visit the vbox room to get them to help me with that
[21:27] <steveng> I got the eth2 up and back on static
[21:27] <steveng> I can access locall now
[21:40] <h3sp4wn> Is there a way to use a more minimal jeos (replace some stuff in ubuntu-minimal / standard with more minimal alternatives - its annoying it specifies certain packages not virtualks
[21:46] <patdk-lap> heh, I just have a large script I run, that does a more minimal system after I install the normal system
[21:46] <patdk-lap> apt-get purge ......
[21:46] <patdk-lap> mainly like ppp, pppoe, dhcp3-client, ...
[21:49] <FunnyLookinHat> Anyone here know how I could setup a cron job to be run by the apache2 user?
[22:00] <remix_tj> FunnyLookinHat: yeah, by setting it on /etc/cron.d/something
[22:00] <remix_tj> specifying just after the time and just before the command www-data
[22:01] <i0nic> my server just rebooted randomly, what can i check to see why?
[22:02] <FunnyLookinHat> remix_tj: cool thanks
[22:03] <remix_tj> i0nic: sure, take a look to the /var/log/syslog
[22:05] <FunnyLookinHat> remix_tj: if I am editing cron like this - sudo crontab -e ... that means I'm editing root's cron, right?
[22:05] <remix_tj> yeah FunnyLookinHat
[22:05] <FunnyLookinHat> kk
[22:13] <FunnyLookinHat> Wait - remix_tj  - couldn't I just sudo -u www-data crontab -e ?
[22:13] <remix_tj> FunnyLookinHat: is not a pretty way but you can,  i think
[22:14] <FunnyLookinHat> *crosses fingers as date approaches the next minute*
[22:49] <l3dx> I accidently delted /etc/apache2, how can I "reinstall" apache to get the default folder?
[22:58] <hallyn_> l3dx: you can apt-get purge apache2; apt-get install;  or you can grab the source (apt-get source apache2) and manually fetch them
[22:58] <l3dx> hallyn_: I tried
[22:58] <l3dx> but now I see that it's coming from apache2.2-common :)
[22:58] <l3dx> guess that will work better
[22:58] <hallyn_> oops
[22:59] <hallyn_> that trips me up quite often
[23:16] <ruben23> hi guys any thought how to create a file server on an ubuntu server on a hosted server-and teh client accessing it is using windwos..? any idea..?
[23:21] <DevoKun> ruben23: you need to install and configure Samba. You may want to check out TurnKey Linux, based on Ubuntu: http://www.turnkeylinux.org/fileserver
[23:22] <h3sp4wn> ruben23: You mean the hosted server is not on the same network right ?
[23:23] <ruben23> its on the vast internet not local conenction.
[23:24] <h3sp4wn> Openvpn and samba would work quite well
[23:24] <h3sp4wn> (openvpn has a windows client)
[23:24] <h3sp4wn> or openvpn and nfs would be ok if you use the windows built in nfs client (or SUA)
[23:26] <jhansonxi> ruben23: You should not access Samba directly over the Internet - use either SSH or a VPN.
[23:29] <ruben23>  jhansonxi: the problem is defining one person who can crate folder and give permision for others to access it. im planning SFTP and mapped it directory to windows
[23:31] <h3sp4wn> ruben23: You need to learn how to use ACL's
[23:32] <DevoKun> ruben23: I've had success routing SMB connections over stunnel to Windows clients before.
[23:36] <ruben23> guys its a bit hard.huhuhuhuh
[23:37] <ruben23> :'(
[23:37] <DevoKun> It will take a while to put together.
[23:37] <DevoKun> What specifically are you having problems with?
[23:39] <bmxer> hi. I'm on a AWS server running ubuntu and i've added the universe sources to sources.list but when i run update it doesn not retrieve list...it might be a network restriction?
[23:40] <jiboumans> bmxer: universe should be enabled by default. what AMI are you using?
[23:41] <bmxer> jiboumans, i've just added a prefix domain US and it worked deb http://us.archive.ubuntu.com/ubuntu/ lucid universe
[23:42] <jiboumans> there's an ubuntu mirror inside ec2 - faster and lower cost. but again, those should all be enabled by default
[23:47] <bmxer> jiboumans, now i see...just had to change the prefix to aws mirror. thx!
[23:51] <ruben23> DevoKun: my company weants me to setup a file server on a hosted ubuntu server then, some how creat direcftrey folders with access permissions fo user or define a single person who can do all and create folders and set permision on it. thats all and be mappend on each remote cleint windows users
[23:52] <twb> FWIW, I use "deb mirror://mirrors.ubuntu.com/mirrors.txt lucid main universe"
[23:53] <twb> Where mirrors.txt returns geoip-specific mirror list, and apt uses one (unfortunately, always the first).