--- Day changed

--- Log opened Thu Aug 14 00:00:02 2003
00:07 -!- jax [~Nimaj@148.100.218.121] has quit [Quit: sleep]
00:46 -!- Aetherios [~jeff@rrcs-west-24-199-13-10.biz.rr.com] has joined #uml
01:21 -!- Umler [~anon@dsl-200-67-146-73.prodigy.net.mx] has quit [Quit: Client Exiting]
02:39 -!- jamiebecker [~jamie@68.89.163.233] has joined #uml
02:40 < jamiebecker> hello -- does anyone know why my umid=apache arg isn't getting accepted? it shows up in the "Kernel command line:  umid=apache ubd0=/uml/vms/apache/rootfs.d/" but then mconsole is started on a random umid...
02:41 -!- jamiebecker [~jamie@68.89.163.233] has quit [Client Quit]
02:41 -!- jamiebecker [~jamie@68.89.163.233] has joined #uml
02:41 < jamiebecker> yikes did i miss a response?
02:43 < coryb> negative
02:44 < jamiebecker> coryb: bummer ;-)
02:44 < coryb> sorry
02:44 < coryb> I have to sleep :)
02:56 -!- jamiebecker [~jamie@68.89.163.233] has quit [Read error: Connection reset by peer]
03:10 -!- dunc [~dunc@217.79.100.249.adsl.griffin.net.uk] has quit [Quit: leaving]
03:56 -!- ticallion [~ticallion@213.175.160.102] has joined #uml
04:27 -!- G2_ [~G2@193.195.148.66] has joined #uml
04:45 -!- wojci [~wojci@62.107.97.240] has joined #uml
05:14 -!- ticallion [~ticallion@213.175.160.102] has quit [Quit: later]
05:34 -!- Aetherios [~jeff@rrcs-west-24-199-13-10.biz.rr.com] has quit []
06:14 -!- cray [cray@dhcp-8-165.it.uu.se] has joined #uml
06:17 < cray> Got some problems with multiple UML:s and tuntap interfaces. I boot my uml:s by ./linux ubd0=rootfs eth0=tuntap,tap10x
06:17 < cray> before that I have done: tunctl -u root -ttap10x (where x is a number of course)
06:18 < cray> I setup the interfaces in the uml:s. ifconfig eth0 192.168.200.10x up
06:18 < cray> I setup the tap interfaces on the host machine: ifconfig tap101 192.168.200.1 and ifconfig tap102 192.168.200.2
06:19 < cray> I can ping the host machine from ONE (1) uml, but not the other one... How come_
06:19 < cray> ?
06:20 < cray> I can reach both 192.168.200.1 and .2 from the uml, and the other one doesn't reach anyone... (sending out ARP requests)
07:02 < BB> are you sure you havnt routed 192.168.200.x down tap10a ? you might have to bridge them to get it to work in the same subnet
07:04 -!- G2_ [~G2@193.195.148.66] has quit [Remote host closed the connection]
08:04 -!- Squik [~npa@ip-86-245.evc.net] has quit [Remote host closed the connection]
08:17 -!- arthur [~arthur@64.171.188.145] has joined #uml
08:28 -!- jax [~Nimaj@148.100.208.253] has joined #uml
08:29  * jax is away: @work
08:54 -!- Jon [~jon@194.196.110.14] has joined #uml
08:54 < Jon> hello
09:00 -!- jax-work [~computer_@148.100.210.81] has joined #uml
09:04 < Jon> there's something inherently appealing about 'killall linux'
09:08 < Jon>  /me fights with copy_from_user
09:20 -!- mistral [mistral@jstevenson.plus.com] has joined #uml
09:33 < cray> BB: yes... of course. I put them on different subnets and it works like a charm now.. thx
09:36 -!- mistral [mistral@jstevenson.plus.com] has quit [Ping timeout: 501 seconds]
09:42 -!- jdike [~jdike@dhcp89.ists.dartmouth.edu] has joined #uml
09:45 < jdike> hi guys
09:46 < jax-work> hey
09:49 < BB> cray :)
09:54 < green> Hi Jeff!
10:06 < jdike> Hey Oleg
10:06 < jdike> green: haven't checked out your task_struct leak theory yet, but it looks promising
10:43 < green> yes, I know it is ;)
11:22 -!- AquaJo [~torbofh@pD9559AE6.dip.t-dialin.net] has joined #uml
11:24 -!- probonic [~mr@host81-129-12-121.in-addr.btopenworld.com] has joined #uml
11:24 < probonic> hello
11:24 < probonic> Just wondering if someone could help me with a problem with ethertap...
11:25 < probonic> I followed instructions at http://www.geocrawler.com/mail/msg.php3?msg_id=6441084&list=597 but getting errors inside the UML when i bring eth0 up
11:26 < probonic> etap_tramp : uml_net failed
11:26 < probonic> uml_net : Too few arguments to ethertap_v4
11:26 < probonic> etap_tramp failed - errno = 22
11:26 < probonic> Why is it trying to use uml_net when I have specified no IP address for ethertap??
11:28 < probonic> anyone?
11:28 < jdike> probonic: what was the eth argument on the command line?
11:33 < probonic> same as on that page -  eth0=ethertap,tap0,fe:fd:f0:00:00:01
11:37 < probonic> any ideas?
11:37 < jdike> probonic: it needs uml_net because ethertap is somewhat broken
11:37 < jdike> probonic: this looks like a bug, what's the hos?
11:37 < jdike> host
11:38 < probonic> 2.4.20-19.8
11:38 < probonic> i'm pretty new to UML
11:38 < jdike> probonic: TUN/TAP is a better idea anyway
11:38 < probonic> it does work if I give an IP address
11:38 < probonic> will tun/tap work the way I want to with the bridge?
11:38 < probonic> i.e. the same way as ethertap?
11:42 -!- AquaJo [~torbofh@pD9559AE6.dip.t-dialin.net] has quit [Quit: Die Schwierigkeit, mit den meisten Leuten umzugehen, besteht darin, zu ihnen gleichzeitig ehrlich  und höflich zu sein - Heller, André]
11:45 < probonic> right I'm having another problem now
11:45 < probonic> "tunctl -u root" gives the error "Failed to open '/dev/net/tun' : No such device"
11:49 < jdike> probonic: TUN/TAP is in the kernel?
11:51 < probonic> well i think so - it definitely works when i let uml_net do it
11:51 < probonic> and /dev/net/tun definitely exists
11:52 < probonic> so when i do "eth0=tuntap,,,192.168.0.254" it works...
11:54 < jdike> probonic: well, I get the same thing, but it works after an 'insmod tun'
11:58 < probonic> "insmod: a module named tun already exists"
11:59 -!- miah [miah@199.201.145.61] has joined #uml
11:59 < miah> do i gain any additional security by running a uml as a user and not root?
11:59 < miah> i use skas
12:00 < jdike> miah: If UML is unbreakable, no
12:00 < jdike> miah: but if there is a way out of it, then you're better off running it as non-root
12:01 < jdike> miah: there's no reason I know of to run it as root anyway
12:01 < probonic> jdike: now i get "TUNSETIFF failed, errno = 16"
12:01 < probonic> jdike: "SIOCSIFFLAGS: Device or resource busy"
12:03 < jdike> probonic: do you have a UML using a tap device?
12:03 < probonic> jdike: not at the moment
12:03 < jdike> probonic: rmmod it and insmod it and see if that helps
12:04 < probonic> jdike: its ok its working now
12:04 < probonic> jdike: i think that "insmod tun" did it, even though it said it was already there
12:04 < probonic> jdike: i tried tunctl again and it assigned a new tap device
12:04 < jdike> probonic: strange
12:04 < probonic> jdike: just started the UML with that device and it works :/
12:06 < miah> jdike: right, im just trying to figure out the best way to load several uml's at startup
12:06 < miah> and right now, the way im setting things up, each uml is going to be run as a different user
12:36 < probonic> jdike: I've got it all working now
12:36 < probonic> jdike: Thx a lot for your help :-)
12:39 < jdike> probonic: Cool, that ethertap problem looks like a bug to me, I'll have a look at it
12:39 < probonic> jdike: yeah. tun/tap is working exactly the same way as ethertap with the bridge
12:42 < probonic> jdike: Is there any way of removing "dummy0" from my UML? Is that a standard UML device?
12:43 < jdike> probonic: it's a standard Linux device, look for it on the host
12:43 < probonic> jdike: Well I've never seen it anywhere else
12:44 < probonic> jdike: ohh yeah - it was showing up in phpSysInfo
12:45 < probonic> jdike: but this is the first time i've ever seen it listed in phpSysInfo
12:51 < miah> you dont need dummy
12:52 < miah> its really only usefull if you're doing slip or ppp
12:52 < miah> its a kernel option in networking devices
12:52 < miah> you can disable it, i did with my uml kernel
13:01 < probonic> well i didn't compile my own
13:01 < probonic> i'm new to it so i just used the RPM hehe
13:01 < probonic> and umlbuilder to build my UML
13:06 < miah> i haven't looked at umlbuilder
13:08 < miah> hrm, umlbuilder looks interesting
13:09 < miah> though, i hate that its an X interface
13:10 < jdike> miah: it has a command-line version, I think
13:10 < miah> oh?
13:10 < miah> ah yea
13:11 < jdike> OK, test3 builds
13:14 < jdike> Linux usermode 2.6.0-test3-1um #2 SMP Thu Aug 14 13:14:46 EDT 2003 i686 unknown
13:14 -!- Aetherios [~jeff@robocop.valueclick.com] has joined #uml
13:15 < miah> awesome
13:26 -!- pirlouit [~peter@64.162.195.202] has joined #uml
13:31 -!- Aetherios [~jeff@robocop.valueclick.com] has quit [Quit: Computer going to sleep...]
13:58 -!- mistral [mistral@212.159.71.212] has joined #uml
14:09 -!- mistral [mistral@212.159.71.212] has quit [Ping timeout: 490 seconds]
14:44 < jdike> CyBuX: hi, what's up?
14:50 -!- AquaJo [~torbofh@pD9559AE6.dip.t-dialin.net] has joined #uml
14:51 -!- cray [cray@dhcp-8-165.it.uu.se] has quit [Quit: ircII EPIC4-1.1.11 -- Are we there yet?]
15:17 -!- Aetherios [~jeff@64.70.54.15] has joined #uml
15:17 -!- Aetherios [~jeff@64.70.54.15] has quit [Client Quit]
15:17 -!- Aetherios [~jeff@64.70.54.15] has joined #uml
15:30 < miah> ok
15:30 < miah> strange problem..
15:31 < Aetherios> strange solution?
15:32 < BB> but did it fix it?
15:32 < miah> hrm, its not working
15:33 < miah> VFS: Cannot open root device "ubd0" or 62:00
15:33 < miah> Please append a correct "root=" boot option
15:33 < miah> Kernel panic: VFS: Unable to mount root fs on 62:00
15:33 < miah> whats 62:00 ?
15:33 < miah> hrm, i think i might know what the problem is..
15:34 < miah> im running uml via su -l user -c ..
15:34 -!- Fede [~Fede@OL6-77.fibertel.com.ar] has joined #uml
15:35 < miah> and its stripping my other variable...
15:35 < BB> have you tried it in "./linux ....."  ?
15:35 < miah> that would work
15:36 < miah> but it would totally circumvent the reasoning behind the script
15:36 < BB> I dont see how it would
15:36 < miah> i have a script that starts multiple uml's
15:36 < miah> all the uml users start with uml_
15:36 < jdike> miah: 2.6?
15:37 < miah> it scans the /home directory, and then starts a uml for each uml_ user that has a .uml_enabled file in their ~
15:37 < BB> and - su -l user -c "/pathto/linux ..." # woudlnt work?
15:37 < miah> no
15:37 < miah> BB: IT WOULD BUT IT WOULD BLOW AWAY THE SCRIPT
15:37 < BB> weird script is all i can say
15:37 < miah> no
15:37 < miah> not really
15:38 < miah> i am going to have 6 umls that need started during boot
15:38 < miah> i could easily just add './linux ....' to /etc/rc.local
15:38 < miah> or, i could have my script dynamically build the ./linux options and set everything else up, like the tap devices, routes, etc
15:40 < miah> which is what im doing now, and it works great, i need to fix a few things in it to make it a little smarted.. but obviously need to fix the showstopper first 
15:40 < miah> i could just run all the uml's as root.. but i'd like to get any extra security i can
15:40 < miah> su is definately not working the way i want it to..
15:45 < Aetherios> miah: is the problem with su the fact that it's launching a non-interactive shell and not defining variables?
15:46 < miah> well, its that its not preserving my variables, even though im using the -m option
15:47 < Aetherios> hmmm
15:47 < miah> i definately think there is a bug in su
15:48 < caker> try putting stuff in quotes, that worked for me
15:48 < miah> im using single quotes
15:48 < miah> right now, im testing with su --preserve-environment $UMLUSER -c 'env;set'
15:49 < miah> ah
15:49 < miah> i needed to export my variable
15:50 < miah> though, i was building the variable on the line before the su in the script..
15:50 < miah> so i figured it would be in the stack
15:54 < miah> i need to make my script clean up if the uml fails to start though, during testing everything was going fine up until running the uml, and so i was up to like tap23 =)
16:09 < miah> yea, love it, i reboot, and it stops working
16:09 < miah> grr
16:13 < wojci> miah: I have a perl script to read some config files, start/stop/setup UML, also works at boot time ;)
16:13 < miah> yea?  i've just about worked out the problems in my bash script though
16:13 < wojci> miah: I also control iptables from it etc.
16:14 < miah> nice
16:14 < wojci> miah: About 1000+ LOC of dirty perl .. I am going to reweite it using a real language, when I am finished with my current project.
16:14 < miah> blkmtd: error, missing `device' name
16:14 < miah> Initializing software serial port version 1
16:14 < miah> mconsole (version 2) initialized on /root/.uml/uml_smtp/mconsole
16:14 < miah> unable to open root_fs for validation
16:14 < miah> weird
16:14 < miah> damn
16:14 < miah> my script is like 15lines of bash
16:14 -!- ElectricElf [~david@elf.noc.oftc.net] has quit [Ping timeout: 240 seconds]
16:15 < wojci> I do check for a number of things - if files exist, if they are owned by the _right_ users, their permissions etc.
16:15 < jdike> miah: was there anything above that about failing to lock the file?
16:15 < wojci> oh .. and if the UML _really_ run.
16:15 < miah> \me looks
16:16 < miah> i dont see anything
16:16 < miah> Kernel command line: eth0=tuntap,tap15,,134.174.11.41 udb0=uml_smtp.cow,/usr
16:16 < miah> al/uml/root_fs.rh9 udb1=uml_smtp.swap umid=uml_smtp mem=128M con=none root=/
16:16 < miah> ubd0
16:17 < miah> that didnt paste well..
16:17 -!- david [~d@207.166.203.138] has quit [Ping timeout: 488 seconds]
16:17 < miah> oh
16:17 < miah> weird
16:18 < miah> im having a problem with my OPTIONS variable again
16:18 < miah> it seems to be working randomly
16:18 < miah> su --preserve-environment -l $UMLUSER -c echo $OPTIONS;/usr/local/uml/linux $OPTIONS
16:18 < miah> that echo doesn't print anything
16:18 < miah> but the kernel commandline is accurate
16:18 < wojci> miah: What about screen?
16:19 < miah> not using screen atm
16:19 < wojci> Why not?
16:19 -!- mistik1 [rasta@68.192.32.70] has quit [Ping timeout: 488 seconds]
16:19 < miah> dunno
16:19 < miah> ill set that up as soon as i get past this
16:19 -!- david-work [~david@impaler.ooi.net] has quit [Ping timeout: 501 seconds]
16:19 -!- jax [~Nimaj@148.100.208.253] has quit [Ping timeout: 490 seconds]
16:29 -!- probonic [~mr@host81-129-12-121.in-addr.btopenworld.com] has quit [Read error: Connection reset by peer]
16:29 < miah> this makes absolutely no sense
16:31 < miah> ok, such strangeness
16:32 < miah> my variable OPTIONS is exported, but my su command doesn't seem to pass it until i do 'env', so for example:
16:32 < miah> su -m -p --preserve-environment -l $UMLUSER -c env;$LINUXBIN $OPTIONS
16:32 < miah> works
16:32 < miah> (-m -p --preserve-environment all do the same thing btw)
16:33 < miah> zone(2): 0 pages.
16:33 < miah> Kernel command line: eth0=tuntap,tap53,,134.174.11.41 udb0=uml_smtp.cow,/usr/local/uml/root_fs.rh9 udb1=uml_smtp.swap umid=uml_smtp mem=128M con=none root=/dev/ubd0
16:33 < miah> Calibrating delay loop... 1168.77 BogoMIPS
16:33 < miah> ...
16:33 < miah>   Type:   Direct-Access                      ANSI SCSI revision: 03
16:33 < miah> blkmtd: error, missing `device' name
16:33 < miah> Initializing software serial port version 1
16:33 < miah> mconsole (version 2) initialized on /root/.uml/uml_smtp/mconsole
16:33 < miah> unable to open root_fs for validation
16:34 < miah> makes no sense
16:34 < miah> i know my .cow file doesnt exist, but that should be created on the fly..
16:34 < miah> and root_fs.rh9 is readable by everybody
16:34 < miah> -rw-r--r--    1 root     uml      4294967296 Aug 11 13:43 root_fs.rh9
16:35 < jdike> miah: ummm
16:35 < miah> yes?
16:35 < jdike> miah: udb0 != ubd0
16:35 < miah> OH SHIT
16:35 < miah> <-- been looking at this too long
16:36 < miah> AWESOME, works now
16:36 < miah> thanks for pointing out my mistake
16:48 -!- jax-work [~computer_@148.100.210.81] has quit [Ping timeout: 501 seconds]
17:00 -!- AquaJo [~torbofh@pD9559AE6.dip.t-dialin.net] has quit [Quit: Das Schicksal kommt, wenn's niemand merkt - Publius Syrus]
17:05 -!- Fede [~Fede@OL6-77.fibertel.com.ar] has quit [Quit: Client Exiting]
17:07 < miah> wojci: so, if i want to use screen, i need to change the con parameter to something like 'con=pts' ?
17:09 < caker> miah: con=null con0=fd:0,fd:1
17:10 < caker> miah: make sure a getty is listening on tty0 in your uml's /etc/inittab
17:10 < miah> well, its a rh filesystem, so by default mingetty is listening to tty0
17:10 < caker> miah: prob not
17:11 < Getty> i am not listing on tty0! is this slavery here or what? ;-)
17:11 < caker> Getty: geg
17:11 < caker> er heh even
17:11 < Getty> (running gag, i love him)
17:11 < miah> ah, default is 1-6
17:11 < Getty> caker: its a highlight for me, so i'm here, so i must make a joke ;)
17:11 < caker> miah: you could also do "con=null con1=fd:0,fd:1 console=tty0"
17:11 < caker> make that tty1
17:12 < miah> hrm, see i dont like specify static things like 'fd:0' =)
17:12 < miah> because im loading several umls at boot
17:32 -!- jdike [~jdike@dhcp89.ists.dartmouth.edu] has quit [Quit: Leaving]
17:42 < miah> hrm
18:18 -!- spenne [~spenneb@212.204.27.151] has joined #uml
18:20 < spenne> Hi, just wanted everybody to know that networking still works in 2.6.0-test2 when applying the little modification mentioned on the Devel mailinglist. Even IPsec using setkey, racoon and isakmpd works. FreeS/WAN still needs testing.
18:20 < spenne> Good Work!
18:20 -!- spenne [~spenneb@212.204.27.151] has quit [Client Quit]
18:29 -!- Aetherios [~jeff@64.70.54.15] has quit []
18:55 -!- com4 [~com4@12-209-152-183.client.attbi.com] has quit [Ping timeout: 501 seconds]
20:42 -!- Aetherios [~jeff@robocop.valueclick.com] has joined #uml
21:16 -!- probonic [~mr@81.129.12.121] has joined #uml
21:16 < probonic> Does anyone know if doing bandwidth accounting for UMLs using the "tap" devices is accurate?
21:46 -!- com4 [~com4@12-209-152-183.client.attbi.com] has joined #uml
22:47 < caker> uml.org down?
22:47 < caker> Anyone mirror those contributed redhat 9 images?
22:55 < Aetherios> heh, apparently uml.org is located on east-coast US, with no UPS??
22:55 < caker> hmm good point
22:56 < caker> mine (right near NYC) have remained online :-)
22:56 < Aetherios> you running on UPS?  or your power never went out?
22:57 < caker> Not sure -- that equipment is at XO.  They juts had a data-center wide power failure about two weeks ago, so maybe they bulked up
22:58 < Aetherios> yeah, all my important personal servers are in my employers cage at exodus :-)  They claim to be able to last 7 days without outside power :-)
22:58 < caker> in chicago?
22:58 < Aetherios> LA
22:58 < Aetherios> LA, CA, no LA Louisianna(sp)
22:58 < caker> I figured :-)
22:58 < Aetherios> :-)
23:12 -!- Umler [~anon@dsl-200-67-146-73.prodigy.net.mx] has joined #uml
23:19 -!- arthur [~arthur@64.171.188.145] has quit [Ping timeout: 490 seconds]
23:47 -!- Aetherios [~jeff@robocop.valueclick.com] has quit []
23:48 -!- Umler [~anon@dsl-200-67-146-73.prodigy.net.mx] has quit [Quit: Client Exiting]
23:48 -!- Umler [~anon@dsl-200-67-146-73.prodigy.net.mx] has joined #uml
--- Log closed Fri Aug 15 00:00:00 2003