1145750475 Q * ntrs_ Ping timeout: 480 seconds
1145750643 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com
1145751145 Q * ntrs Ping timeout: 480 seconds
1145751375 J * ntrs ~ntrs@68-188-51-87.dhcp.stls.mo.charter.com
1145751429 J * FireEgl Atlantica@Atlantica.Tcldrop.US
1145752210 N * Bertl_oO Bertl
1145752220 M * Bertl evening folks!
1145752241 M * daniel_hozac evening Bertl!
1145752268 M * insomniac hi daniel_hozac and Bertl :)
1145752284 M * Bertl hey insomniac!
1145752298 M * insomniac we came back at the same time
1145752303 M * _coocoon_ hey bertl
1145752306 M * insomniac is that a sign?
1145752307 M * insomniac :>
1145752317 M * Bertl daniel_hozac: could you give the network test scripts a spin and let me know what you think?
1145752333 M * Bertl insomniac: it must be a sign! :)
1145752339 M * Bertl hey _coocoon_!
1145752341 M * daniel_hozac sure.
1145752362 M * daniel_hozac nc_{client,server}.sh?
1145752372 M * Bertl yep, precisely, I updated them in place
1145752390 M * insomniac Bertl: does it mean that even if I recompiled a brand new kernel, and brand new utils, and (believe me? :) a brand new installation of a debian
1145752394 M * insomniac nothing works!
1145752398 M * insomniac (ta-daan)
1145752417 M * Bertl insomniac: hmm?
1145752455 M * insomniac you remember The Error of the Month[TM] ?
1145752457 J * matta ~matta@c-68-32-202-140.hsd1.pa.comcast.net
1145752462 M * insomniac ( save_ctxinfo: execv(): No such file or directory )
1145752464 M * derjohn insomniac, this may be a brand new error, too 
1145752493 M * Bertl yeah, but unlikely, I still opt for some 'strange' mount/tag/permission issue
1145752512 M * insomniac on the vservers dir?
1145752524 M * Bertl insomniac: but I think we can try to check certain things in this direction
1145752541 M * insomniac Bertl: you made me think the same yesterday
1145752553 M * insomniac but I'm following your light
1145752553 M * Bertl one thing to verify would be that you can reach the vserver guest as user (not root)
1145752584 M * Bertl that basically verifies many permission issues (not the vserver related ones though)
1145752606 M * daniel_hozac save_ctxinfo tries to exec env though...
1145752609 M * Bertl so after you logon (or become via su) some user, can you ls and chdir into the guest?
1145752612 M * derjohn Bertl, wouldnt help a strace and looking what file it tries to exec?
1145752636 M * insomniac yes, I can
1145752650 M * Bertl derjohn: yes, but the strace is not allowed to cross context boundaries (for security reasons) and it seems to happen outside the guest anyways
1145752677 M * Bertl insomniac: okay, now try with chcontext --xid 42 --secure -- bash
1145752688 M * derjohn insomniac, did you copy the guests root dir or did you build on the machine itself?
1145752692 M * daniel_hozac /usr/lib/util-vserver/save_ctxinfo /etc/vservers/<name> true works, right? (i seem to recall that...)
1145752721 M * insomniac derjohn: both of them
1145752727 M * insomniac Bertl: changed context
1145752757 M * insomniac and gracefully exited
1145752782 M * derjohn insomniac, did you apt-get util-vserver as well as compiling youself such to .../local /... ?
1145752794 M * insomniac compiled by myself
1145752802 M * insomniac it's a slackware host
1145752818 M * insomniac and a debian guest
1145752823 M * Bertl try to change the context as above, and then enter your guest
1145752835 M * Bertl (i.e. cd into the guest root, and execute chroot)
1145752849 M * insomniac can not change context: migrate kernel feature missing and 'compat' API disabled: Function not implemented
1145752854 M * insomniac ah ok
1145752863 M * Bertl aha?
1145752878 M * insomniac # ls -l /vservers/
1145752878 M * insomniac ls: /vservers/: Permission denied
1145752998 M * Bertl how did you work around the API message above?
1145753025 M * insomniac i just launched vserver debian start
1145753066 M * daniel_hozac Bertl: hmm, my netcat doesn't have the -q option.
1145753074 M * Bertl ah, I'm using 1.1
1145753101 M * Bertl 1.10 sorry
1145753111 M * daniel_hozac nc-1.84-3.2
1145753144 M * Bertl hmm ... they seem to have strange numbers
1145753151 M * Bertl the official version is 0.72
1145753167 M * Bertl the version I found was 1.10, now you have 1.84?
1145753239 M * daniel_hozac hehe.
1145753292 M * Bertl I added the -q because otherwise the tcp messages got lost
1145753307 M * daniel_hozac URL         : http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin/nc/
1145753309 M * Bertl -q seconds   after EOF is detected, wait the specified number of seconds and then quit.
1145753341 M * daniel_hozac hmm, strange that it would detect EOF before getting the message. especially for TCP.
1145753366 M * Bertl well, imho EOF is generated when the socket closes, which probably causes a signal
1145753387 M * Bertl depending on the implementation, this _could_ of course mean that the message is lost in the buffer
1145753394 M * Bertl (as it was for doener and me)
1145753394 M * daniel_hozac hmm, yeah, true.
1145753406 M * Bertl the -q 1 on server side fixed it
1145753418 M * Bertl but just remove it and try yourself
1145753455 M * daniel_hozac hmm, my nc still doesn't like the usage.
1145753494 M * Bertl okay, no problem, we can ship the nc with it, it's not supposed to work on every setup (i.e. no generic test suite stuff)
1145753497 M * daniel_hozac removing -s and -p seems to have gotten it working.
1145753514 M * Bertl hum, that is somehow besides the point, no?
1145753557 M * daniel_hozac hmm?
1145753596 M * Bertl well, you remove the -s (source address) and -p (port) where?
1145753623 M * Bertl server side?
1145753641 M * daniel_hozac yeah.
1145753644 M * Bertl (IMHO that might be okay with a change in syntax)
1145753653 M * Bertl but client side you definitely need them
1145753660 M * Bertl the scripts are used this way:
1145753670 M * daniel_hozac does the client side just output the problematic results?
1145753670 M * Bertl - start the server side, wait for the 15sec server ready msg
1145753686 M * Bertl - start the client side, wait for both to complete
1145753751 M * daniel_hozac my server side never seems to complete.
1145753763 M * Bertl it takes 60 seconds
1145753777 M * Bertl well, with the nc and -w we use :)
1145753815 M * daniel_hozac it has been 5 minutes now though.
1145753837 M * Bertl probably the result of the 'incompatible' netcat
1145753875 M * Bertl the client is supposed to give something like this: http://vserver.13thfloor.at/Stuff/SCRIPT/client_24.out
1145753885 M * Bertl and similar for the server (server_24.out)
1145753920 M * daniel_hozac i guess my nc is really incompatible, because my results look nothing like that :)
1145753938 M * Bertl okay, np, what distro is that?
1145753948 M * daniel_hozac FC5.
1145753989 M * Bertl so the mandrake nc should work for you too, I guess
1145754007 M * daniel_hozac most likely.
1145754044 M * Bertl interestingly the last netcat rpmfind finds is for RH 5.2 :)
1145754089 M * Bertl I'll upload a simple src rpm ... sec
1145754096 M * daniel_hozac http://rpmfind.net/linux/rpm2html/search.php?query=nc seems to give quite a bit of results.
1145754164 M * Bertl right, fc3 also used the same one
1145754192 M * Bertl http://rpm.pbone.net/index.php3/stat/26/dist/41/size/89141/name/nc-1.10-22.src.rpm
1145754214 M * Bertl (that's the one for fc3) so I guess they drastically improved the tool :)
1145754230 M * daniel_hozac i think they just changed the upstream.
1145754348 M * Bertl anyway, I move it into a different subdir now, before it gets indexed
1145754371 M * Bertl /Stuff/TEST/
1145754399 M * daniel_hozac hmm, that one still says invalid option for -q.
1145754486 M * Bertl http://vserver.13thfloor.at/Stuff/TEST/
1145754496 M * Bertl here is the one I used (the other one was from FC3)
1145754500 Q * FireEgl Ping timeout: 480 seconds
1145754510 M * Bertl it should compile fine on FC5 I'd say
1145754761 M * daniel_hozac wow, that was quick.
1145754850 M * Bertl derjohn: hmm @ LT-ML?
1145754945 M * derjohn Bertl, yes. Orga ... I had to tell everything multiple times via chat. We have 12+ members to coordinate, inkl. access to the testing machines for builting guests etc...
1145754969 J * Venomous Venom-One@host122-185.pool8255.interbusiness.it
1145754977 M * Venomous hi all
1145755007 M * Bertl derjohn: okay, but why in the hell yahoo?
1145755012 M * derjohn Bertl, but you may be an extra case, as you are some kinda super-VIP.
1145755042 M * Bertl AFAIK the only purpose of yahoo is to collect email addresses for spam :)
1145755115 M * derjohn Bertl, I asked all if they run some mailman or so. No one. So I transferred the task to  member soewono (a.k.a. zeng here) ... he setup the yahoo-foo. The list will vanish in two weeks anyway.
1145755127 M * Venomous Bertl
1145755146 M * Venomous on one guest
1145755146 M * derjohn Bertl, I get so much Spam, I dont care anymore. You can use a throwway addy as well ..
1145755155 M * Venomous i need all the network stuff
1145755163 M * Venomous ping tracerout socket etc..
1145755167 M * Bertl derjohn: well, somebody already used my email there, so the damage is already done :/
1145755168 M * derjohn Bertl, BTW: yahoo does harvest ?
1145755171 M * Venomous what i have to do?
1145755220 M * Venomous i have created a file bcapabilities and inside i have written CAP_NET_RAW
1145755228 M * Venomous but doesn't work
1145755229 M * derjohn Bertl, well, you have to filter anyway ... :)
1145755232 M * Bertl Venomous: ping and tracerout should work without CAP_NET_RAW with recent kernels
1145755241 M * Venomous bertl
1145755246 M * Venomous i have 2.4........
1145755257 M * Venomous RHE
1145755268 M * Bertl derjohn: it's no problem for me to kill yahoo (spam filter) but that doesn't buy me anything if they sell the addresses, no?
1145755290 M * Venomous does you know some work ariund?
1145755310 M * Bertl Venomous: well, 2.4 does not support that
1145755323 M * Venomous so no ping?
1145755331 M * Bertl not in a secure way
1145755347 M * Venomous wich is the unsecure way?
1145755362 M * Bertl giving CAP_NET_RAW and maybe CAP_NET_ADMIN
1145755368 M * derjohn Bertl, your filter doenst adapt itself, nor? (Spamassassin, Bogofilter ???)
1145755378 M * Venomous in the file bcapabilities?
1145755409 M * Venomous let's try
1145755421 M * derjohn bcapabilities in 2.4 ?
1145755428 M * Bertl derjohn: trust me, it adapts, but I just don't like folks 'subscribing' or 'inviting' me somewhere without even asking ...
1145755450 M * Bertl Venomous: what tools do you use
1145755457 M * Venomous ping google.it
1145755457 M * Venomous ping: icmp open socket: Operation not permitted
1145755492 M * Venomous 0.30.209
1145755493 M * derjohn Bertl, sry, can't be undone. But you may choose not to join the ML. 
1145755576 M * Venomous guys yuo are opening a new mailing list?
1145755585 M * Venomous for linux-vserver?
1145755603 M * daniel_hozac Bertl: ok, it works now.
1145755624 M * Bertl derjohn: yeah, please just let Soewono know that he should be more careful in those regards in the future, otherwise I might feel inclined to subscribe him to lkml on a regular basis :)
1145755637 M * daniel_hozac lol
1145755679 M * Bertl daniel_hozac: okay, you get similar (identical?) results?
1145755690 M * Bertl Venomous: no, we already have two mailing lists
1145755712 M * Venomous Bertl so no fix for me?
1145755731 M * Bertl derjohn: imho a mail alias would have been more than sufficient for that purpose, but let's move on now ...
1145755740 M * daniel_hozac http://daniel.hozac.com/vserver/nc_test/nc_client.out http://daniel.hozac.com/vserver/nc_test/nc_server.out
1145755751 M * Bertl Venomous: check the caps inside the guest please with
1145755751 M * daniel_hozac not identical, quite a few more passed for the client.
1145755779 M * Bertl Venomous: grep Cap /proc/self/status
1145755800 M * Venomous wait
1145755811 M * daniel_hozac we have two mailing lists? which is the second one?
1145755813 M * Bertl daniel_hozac: ah, I left out an important test detail :)
1145755822 M * Bertl daniel_hozac: the wiki mailing list
1145755829 M * daniel_hozac ah.
1145755863 M * Venomous # grep Cap /proc/self/status
1145755864 M * Venomous CapInh: 0000000000000000
1145755864 M * Venomous CapPrm: 00000000144c04ff
1145755864 M * Venomous CapEff: 00000000144c04ff
1145755864 M * Venomous CapBset:        00000000144c04ff
1145755881 M * Bertl you need to specify two IPs, one which exists, and another which does not exist but would be on the network
1145755889 M * Bertl (that was for daniel_hozac)
1145755897 M * daniel_hozac ah, ok.
1145755924 M * Bertl the scripts 'assume' 192.168.0.1 to be local (existing)
1145755932 M * Bertl and 192.168.0.9 to be non existant
1145755948 M * derjohn Bertl, I think I will leave office now. I send you an invitiation to the ML from my private account some hour ago, so you can choose to subscribe with an alias you like.
1145755986 M * thorkild I am using the helper scripts, but some of them sometimes complains about the lack of a /etc/vservers/<name-of-vserver>.conf file. Am I correct when I guess that was the old way of setting the config, which is now replaced by the /etc/vservers/<name-of-vserver>/ hiearchy? (I am on ubuntu dapper, using something which looks like scripts from the debian distribution)
1145755997 M * thorkild or is the .conf for other settings?
1145756010 M * daniel_hozac the .conf is the legacy configuration format.
1145756019 M * Bertl thorkild: which helper script does use the old config?
1145756023 M * daniel_hozac uninstall vserver-debiantools now and save yourself from a great deal of pain :)
1145756030 M * Bertl ah :)
1145756042 M * thorkild Bertl: vserver-copy, but I guess it is me using old software. :)
1145756049 M * thorkild daniel_hozac: I'll do that right away :)
1145756059 M * Bertl Venomous: you are missing the capabilities
1145756072 M * daniel_hozac vserver-copy is indeed part of util-vserver, but it's considered legacy too.
1145756075 Q * mountie Read error: No route to host
1145756084 M * Bertl CAP_NET_RAW is 13 which means 0x2000 hex
1145756104 M * Bertl you should have that in CapBset at least
1145756135 M * thorkild daniel_hozac: are there any replacements? (I am aware of how simple it is to copy a server, so it isn't really an issue)
1145756158 M * Bertl daniel_hozac: has a patch to add a 'clone' feature
1145756169 M * daniel_hozac well, there's a patch for vserver ... build -m clone, but's it's just a cp -a.
1145756180 M * thorkild heh
1145756184 M * derjohn Bertl, didnt I ever invite you for joining openbc ?
1145756191 M * micah derjohn: do you have a source copy of the bash completion (the wiki code is weird from formatting)
1145756209 M * derjohn micah, yes, I mail you .. hold a sec
1145756212 M * micah thanks
1145756218 M * Bertl derjohn: openbc? not that I'd remember
1145756242 M * daniel_hozac micah: can't you choose to edit that page and copy from the textbox?
1145756258 M * Bertl yes, but wouldn't that be cheating?
1145756271 M * daniel_hozac lol
1145756278 M * Venomous So Bertl, what i have to do?
1145756299 M * Bertl I'd assume you are hitting a bug in 0.30.209
1145756309 M * Venomous great
1145756312 M * micah daniel_hozac: I did, but it contains html formatting characters, such as &lt for '<', most of all I've fixed, but i want to make sure its not broken by looking at an actual one
1145756323 M * Bertl Venomous: I remember something with secure being too secure
1145756340 M * micah daniel_hozac: and the end of the script with the } } on the same line makes me suspicious
1145756340 M * derjohn strange with firefox i was able to cut & paste (middle mouse key) 
1145756345 M * Bertl daniel_hozac: when was that fixed so that not so secure (bcaps) could work?
1145756349 M * daniel_hozac micah: hmm? are you sure you didn't view source?
1145756358 M * micah yes
1145756376 M * daniel_hozac Bertl: not fixed yet, but that only affects chcontext direct usage.
1145756393 M * Bertl daniel_hozac: and for legacy/2.4?
1145756434 M * daniel_hozac i have no idea, i don't think the chcontext script is even used in that case.
1145756476 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com
1145756479 M * derjohn micah, I mailed you the file as txt attachment. It is called 'vserver' (I hope I didnt click on my root-pw-list)
1145756484 M * Bertl hmm, Venomous: please do a guest startup with --debug and upload the output to a pastebin service (or somewhere else)
1145756506 M * Venomous can i remove interface from guest server just remivng the conf folder?
1145756510 M * micah derjohn: hehe, thanks,  I just need to compare
1145756542 M * Bertl Venomous: yes, deleting a subdir in the interfaces will remove that 
1145756548 M * derjohn micah, thanks for you incredibly fast reaction to my wishlist-bugs ... 
1145756549 M * Venomous Bertl thanks for your help, but i'm really tired, 4:00 am
1145756551 M * Venomous LOL
1145756566 M * derjohn 4 am ? WTF , I leave now
1145756570 M * derjohn bye all ....
1145756575 M * Bertl well, it's 3:40 here, probably my clock is out of sync then :)
1145756578 M * micah derjohn: I want to make the package better, so anything you have for suggestions
1145756589 M * Venomous Bertl where are you?
1145756594 M * Venomous europe?
1145756603 M * Bertl yup, austria
1145756612 M * Venomous close to me!
1145756616 M * Venomous i'm italian
1145756637 M * Bertl so, the night is young and so are we .. ? :)
1145756640 M * Venomous in fact here 3:45 am
1145756671 M * Bertl anyway, np to continue another time
1145756687 M * Venomous and so we are young i think
1145756696 M * Venomous correct?
1145756769 Q * mountie Quit: 
1145756801 M * Bertl Venomous: nevermind (was from Dance with Me)
1145756810 M * Venomous LOL sorry
1145756814 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com
1145756849 Q * mountie Quit: 
1145756910 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com
1145756917 M * Bertl wb mountie!
1145756931 M * mountie Bertl: Thanks....
1145756938 A * mountie is playing kernel pingpong...
1145756958 M * Bertl with whom?
1145757167 M * thorkild why is there both a name and a uts/nodename file in the current configuration system? 
1145757189 M * thorkild humm, I guess it might be so the vserver name could be something else than the hostname
1145757261 M * Bertl precisely
1145757385 M * thorkild it is funny sometimes. You first realise the answer to something just after you've asked the question. 
1145757396 M * Bertl yep, I know that
1145757457 M * Venomous Bye all guys
1145757461 M * Venomous good night
1145757464 M * Bertl Venomous: good night!
1145757476 Q * Venomous Quit: KVIrc 3.2.0 'Realia'
1145757513 Q * mountie Remote host closed the connection
1145757540 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com
1145757626 Q * mountie Remote host closed the connection
1145757647 M * thorkild is there any way to see what IP-address a vserver has from the inside of the vserver? (ifconfig just gives me UP BROADCAST RUNNING NOARP, even though I know the ip adress is correctly set)
1145757676 M * Bertl try 'ip addr ls'
1145757689 M * thorkild ah, iproute2 of course.
1145757695 M * thorkild should have thought about that. Sorry
1145757699 M * Bertl np
1145757717 M * thorkild worked perfectly. Now I'll go make my vservers multiply..
1145757731 M * Bertl proliferate?
1145757737 M * Bertl :)
1145757744 M * thorkild hihi
1145757851 J * matt1 ~matta@c-68-32-202-140.hsd1.pa.comcast.net
1145757867 M * thorkild my users like to install php applications as part of the domains I am hosting for them, and some of them gives me a Bad Feeling when it comes to security (now, I am not bashing php here, you can write insecure programs in pretty much everything), so I am stuffing their apaches in vservers to reduce the impact of vulnerabilities. Seems to work great.
1145757925 M * Bertl good to hear, btw, php is pretty safe when you use the security feature (safe mode) but it requires properly written programs too
1145757977 M * thorkild Bertl: yeah, I know. My personal favourites are those who need register_globals on.. that is a baaad sign.
1145757990 M * Bertl indeed
1145758002 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com
1145758021 J * coocoon3 ~coocoon@p54A055AD.dip.t-dialin.net
1145758038 Q * _coocoon_ Ping timeout: 480 seconds
1145758053 Q * coocoon3 Quit: 
1145758150 J * _coocoon_ ~coocoon@p54A055AD.dip.t-dialin.net
1145758153 M * Bertl I guess I will call it a day too ... will be back tomorrow, and then we can test on the lo patches ...
1145758192 M * Bertl daniel_hozac: please let me know if you see significant differences on your system or have any suggestions to improve the test scripts
1145758226 Q * matta Ping timeout: 480 seconds
1145758241 M * daniel_hozac it looks a lot more like your results now.
1145758340 M * Bertl okay, good ...
1145758692 Q * matt1 Ping timeout: 480 seconds
1145758873 J * matta ~matta@c-68-32-202-140.hsd1.pa.comcast.net
1145759035 M * Bertl okay, finally off now ... have a nice one everyone!
1145759041 N * Bertl Bertl_zZ
1145759066 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1145760163 Q * mef Ping timeout: 480 seconds
1145760315 J * kilgur_ user@p50813758.dip0.t-ipconnect.de
1145760733 Q * kilgur Ping timeout: 480 seconds
1145761915 J * FireEgl Atlantica@Atlantica.IRCNut.Com
1145763932 M * thorkild Does xid-marking protect against rogue setuid binaries? I.e: if a user makes a binary inside a vserver, and then accesses that binary from outside the vserver later on.. does it still work (and run setuid)? 
1145765187 M * thorkild hmm, no it doesn't
1145767224 Q * cryo Ping timeout: 480 seconds
1145769230 J * eyck eyck@ghost.anime.pl
1145769824 J * Guest395 ~Guest395@ppp-124.121.93.101.revip2.asianet.co.th
1145770370 Q * Guest395 Quit: Bye
1145775649 M * Hollow thorkild: how would a vserver user access the binary outside the guest? (i.e. from the host)
1145775857 M * cehteh he usually can not
1145775868 M * cehteh mhm nm
1145778192 J * Viper0482 ~Viper0482@p5497779B.dip.t-dialin.net
1145781185 J * Aiken ~james@tooax6-079.dialup.optusnet.com.au
1145785383 J * bonbons ~bonbons@83.222.37.206
1145788294 M * derjohn cehteh, ping:p-msg
1145788695 J * _coocoon_ ~coocoon@p54A06EC3.dip.t-dialin.net
1145788704 M * _coocoon_ hello
1145789243 Q * _coocoon_ Ping timeout: 480 seconds
1145789868 Q * softi42 Ping timeout: 480 seconds
1145790290 M * Ben___ hello
1145790410 M * derjohn are there any plans to include SUM(TIME+) in the vserver-stat? Would be nice to see the CPU consumption of all processes of a guest.
1145790492 J * softi42 ~softi@p549D5A0F.dip.t-dialin.net
1145792153 J * Pater_John ~Patrick09@194.112.144.69
1145792193 M * Pater_John hello, I have a question
1145792201 M * Pater_John maybe someone can help me
1145792217 M * Pater_John i installed fedora vore 5
1145792250 M * Pater_John and want to run a vserver, but there is an error
1145792318 M * Pater_John rpm-fake-resolver: vc_new_s_context(): Funktion not implemented
1145792324 M * Pater_John what can i do?
1145792367 M * derjohn Pater_John, pray! Or install a vserver patched kernel. That it IMHO not default in fc5.
1145792438 M * Pater_John german?
1145792463 M * derjohn you may also pray in latin !
1145792465 M * derjohn :)
1145792485 M * derjohn yes, but this is usually not recommended there and will be severely punished.
1145792488 M * Pater_John i used this install instructions: http://linux-vserver.org/VServer+installation+Fedora+Core+5
1145792519 M * derjohn Ich bin kein Fedora-Mann. Warte mal ab ob Daniel_H Dir antwortet, der macht Fefora.
1145792546 M * Pater_John hab doch gewusst dass du deutsch sprichst
1145792551 M * derjohn 1.) Do you have an vserver kernel installed? 2.) what does testme.sh -v tell you ?
1145792619 M * Pater_John wo bekomme ich die testme.sh her?
1145792939 M * derjohn http://linux-vserver.org/ hat eine Suchen-Funktion: http://linux-vserver.org/Linux-Vserver+FAQ -> recht weit oben
1145792970 M * derjohn testme must show all "OK" or you wont be able to run a guest.
1145793148 M * Pater_John sind zwei fehler
1145793168 M * derjohn mit -v oder --verbose siehst du auch, which one.
1145793217 M * Pater_John chcontext und chbind
1145793241 M * Pater_John hast du irgendwelche erfahrungen darin?
1145793264 M * derjohn do you have a vserver patchd kernel? form where? self-compiled?
1145793414 M * Pater_John Daniels RPMs have been proven stable for a longer period
1145793425 M * Pater_John wie in der anleitung beschrieben habe ich diesen benutzt
1145793445 M * Ben___ ihr macht da ja ein denglish *g
1145793455 M * Pater_John ^^
1145793470 M * derjohn well, you know, the "johns" :)
1145793478 M * Pater_John warum english wenns auch deutsch geht
1145793485 N * Ben___ ben_john
1145793488 M * ben_john :D
1145793490 M * Pater_John I'm not the best in english
1145793499 N * ben_john ben_
1145793518 M * derjohn because here are chatlogs which save the conversation for ever in order to make the information available for all other (and save us lots of support time)
1145793569 M * derjohn Pater_John, what does uname -a tell you ?
1145793583 M * Pater_John one moment please
1145793596 M * derjohn I still feel you booted an standard fedora kernel ...
1145793667 M * derjohn I *guess* daniel_hozac names them linux-..2.6.16....vs2.1.1..  or similar
1145793705 M * ben_ unstable *g
1145793751 M * ben_ stable  Linux 2.6.14.3-vs2.0.1 #2 SMP PREEMPT Sat Apr 22 20:42:43 CEST 2006 i686 GNU/Linux
1145793793 M * derjohn ben_, debian does not name them like this
1145793825 M * ben_ self compiled from kernel.org, i hate the pre compiled shit...
1145793924 M * derjohn ben_ should upgrade to 2.6.16/vs2.1.1rc17 (and rc18 coming soon!)
1145793969 M * Pater_John how can i use the mozilla irc chat?
1145793990 M * ben_ its non stable version atm, next week i ll be testing the new 2.6.16 
1145793994 M * derjohn Pater_John, eh, well that's the wrong channel ;)
1145794006 M * derjohn ben_, take rc17
1145794023 M * Pater_John no, i have to switch every time between the two displays
1145794055 M * derjohn ben_, trust me, there is no reason atm not to take it. besides that it knows memoty limits better (didnt you ask?)
1145794055 M * Pater_John Linux test 2.6.15-1.2054_FC5smp #1 SMP Tue Mar 14 16:05:46 EST 2006 i686 i686 i386 GNU/Linux
1145794076 M * Pater_John that outputs uname -a
1145794089 M * ben_ derjohn, on the online server with customers i cant test anthing, must be stable next week i can test the 2.6.16 with the adaptec aar 2410 @ home
1145794115 M * derjohn Pater_John, I dont think thats a vserver kernel. maybe the machine booted the standard one (grub config?)
1145794128 M * Pater_John yes grub
1145794146 M * derjohn ben_, /me slaps himself. yes, next week
1145794225 M * ben_ Pater_John, the first kernl on top of the list there will be boot *g
1145794291 M * ben_ voting for a german support channel, my english is :(
1145794299 M * Pater_John i rebootet my server and saw, that the false kernel is activated
1145794318 M * Pater_John lets try with the other kernel
1145794347 M * derjohn maybe I should open 0900-vserver :)
1145794452 M * ben_ i am the first on #0900-vserver *g
1145794493 M * derjohn :/
1145794815 A * derjohn afk
1145794888 Q * Aiken Ping timeout: 480 seconds
1145796080 Q * phedny Ping timeout: 480 seconds
1145796572 J * phedny ~mark@volcano.p-bierman.nl
1145796825 J * Oli ~skycode@212.224.239.33
1145797607 J * Oli_ ~skycode@212.224.239.33
1145797680 Q * Oli Ping timeout: 480 seconds
1145798149 Q * mountie Ping timeout: 480 seconds
1145798403 Q * Loki|muh Ping timeout: 480 seconds
1145798587 J * Loki|muh loki@satanix.de
1145799190 J * mountie ~mountie@CPEdeaddeaddead-CM000a739acaa4.cpe.net.cable.rogers.com
1145800469 M * daniel_hozac Pater_John: are you sure you installed the SMP kernel?
1145800469 J * Oli ~skycode@212.224.239.33
1145800722 Q * Oli_ Ping timeout: 480 seconds
1145801551 M * Pater_John i installed the normal kernel
1145801561 M * Pater_John for single prozessor system
1145801563 M * Pater_John s
1145801703 M * thorkild Hollow: I was thinking about the case where a user has access to both the host os and the vserver. I was trying to figure out what the file system taggging was for.
1145801723 J * doener ~doener@i5387DADC.versanet.de
1145801797 M * Pater_John @ daniel_hozac: I managed the problem, it was so simple - the machine started with the false kernel!!
1145801845 M * daniel_hozac Pater_John: so you don't have an SMP machine? (i.e. no hyperthreading, dual core, multiprocessor, etc.)
1145801866 M * Pater_John hmm, one moment please
1145801886 M * daniel_hozac thorkild: for accurate accounting of per-context files, otherwise disk limits become very hard to implement.
1145801938 M * Pater_John model name : Intel(R) Pentium(R) 4 CPU 2.80GHz
1145801974 M * thorkild daniel_hozac: ok. So it isn't used by any access limitation.
1145801983 M * daniel_hozac thorkild: well, it's used for that too.
1145801999 M * daniel_hozac thorkild: one guest cannot touch another guest's files.
1145802015 M * thorkild daniel_hozac: ah. But the host os doesn't look at it in any way?
1145802117 M * Pater_John (daniel_hozac): its peculiar, i thought that it is a hyperthreading machine but the cpuinfo shows only one cpu
1145802122 M * daniel_hozac thorkild: the host sees all.
1145802136 M * daniel_hozac Pater_John: if you boot a non-SMP kernel, you'll only see one CPU.
1145802160 M * Pater_John ah, you are right
1145802168 M * Pater_John sorry my mistake
1145803582 J * _coocoon_ ~coocoon@p54A07AA7.dip.t-dialin.net
1145804068 Q * _coocoon_ Ping timeout: 480 seconds
1145804251 A * micah pokes h01ger 
1145804429 A * h01ger tickles micah - its #3 on (the remaining of) todays lists
1145804445 A * micah giggles
1145804844 Q * Viper0482 Ping timeout: 480 seconds
1145805099 M * gdm ooh, is that the new munin plugin upcoming?
1145805409 J * Viper0482 ~Viper0482@p54976175.dip.t-dialin.net
1145805999 J * _coocoon_ ~coocoon@p54A0754E.dip.t-dialin.net
1145807016 Q * nebuchadnezzar Quit: ERC Version 5.1.2 $Revision: 1.796.2.4 $ (IRC client for Emacs)
1145807125 J * nebuchadnezzar ~nebu@zion.asgardr.info
1145807356 J * doener_ ~doener@i5387EB56.versanet.de
1145807523 M * micah gdm: yeah, a few of them
1145807586 Q * doener Ping timeout: 480 seconds
1145807793 J * cryo ~say@psoft.user.matrix.farlep.net
1145808053 M * gdm kool!
1145808974 M * micah Bertl_zZ: looks like upstream dietlibc picked up the parisc fix06 patch
1145809440 M * Wonka "upstream dietlibc" would be fefe?
1145810071 M * h01ger micah, gdm: you plugin is gpl'ed or what? or gplv2 or?
1145811294 M * gdm h01ger: dunno, but GPL i think: http://dev.riseup.net/
1145811465 A * h01ger nods
1145812728 Q * _coocoon_ Ping timeout: 480 seconds
1145813386 J * _coocoon_ ~coocoon@p54A0660D.dip.t-dialin.net
1145813437 J * Venomous Venom-One@host65-34.pool871.interbusiness.it
1145813444 M * Venomous hi all guys
1145814122 M * micah h01ger: yeah, free as can be
1145815780 M * insomniac daniel_hozac: you believe? I tried to install a "virtual slackware", and it works. Debian doesn't.
1145815789 M * insomniac (greetings)
1145815911 J * bogus ~bogusano@fengor.net
1145818063 Q * Venomous Quit: KVIrc 3.2.0 'Realia'
1145819116 M * ray6 ah, a possibly silly question: can I add a /dev/ttyS0 (or others) into a vserver (from the host) and this way run for example a faxserver inside an vserver?
1145819373 M * derjohn ray6, why not? if the device node is available, the guest can access it (/me hates the capi4hylafax stuff, which seems to need more than only the dev)
1145819709 J * ScoobyD00 ~foo@80-195-186-201.cable.ubr08.newm.blueyonder.co.uk
1145819754 M * ScoobyD00 hi - does anybody know how i might capture how much CPU time is used by a vserver, say, a month?
1145819803 M * ScoobyD00 both kernel and userland
1145820225 M * derjohn ScoobyD00, a host or a guest?
1145820255 M * ScoobyD00 guest
1145820262 M * derjohn h01ger, create a plugin for monitoring the host via munin, which may grapg the host
1145820267 M * derjohn *graph
1145820291 M * ScoobyD00 so i can charge for CPU time - I would need pretty accurate usage figures
1145820295 M * ScoobyD00 for each guest
1145820345 M * derjohn ScoobyD00, forget that. I would not know how to do that in a fair way.  Even in the non-vserver world, e.g. usage of a mysql-db ...
1145820365 M * derjohn it's probably better to gurantee certain timings ...
1145820387 M * derjohn more money, more prio. the scheduler in devel can do lots of things
1145820405 M * derjohn wel ...
1145820413 M * derjohn hm ... can you code perl or bash?
1145820498 M * derjohn ScoobyD00, look on the host into /proc/virtual/401/cvirt . there you could snapshot the load avg.
1145820509 M * derjohn s/401/<ctx>/
1145821484 J * virtuoso ~s0t0na@80.253.205.251
1145821585 J * matt1 ~matta@68.32.202.140
1145821820 Q * mountie Ping timeout: 480 seconds
1145821938 M * ScoobyD00 yeah - but i don't want to charge for more priority - that is still charging for "potential" use of the CPU - i want to charge for actual CPU time (like Sun Grid)
1145821940 Q * matta Ping timeout: 480 seconds
1145821969 M * ScoobyD00 i was hoping that the cpu token bucket might be recorded?
1145821985 M * ScoobyD00 if i understand correctly (is that what vserver uses to limit cpu time to guests)
1145822161 M * ScoobyD00 http://pastebin.com/677457    <-- i dont suppose that last line is total tokens used?
1145822295 M * daniel_hozac ScoobyD00: /proc/virtual/.../sched
1145822309 M * ScoobyD00 yeah i pasted the output from that
1145822317 M * ScoobyD00 is that last line is total tokens used?
1145822320 M * daniel_hozac yes.
1145822332 M * daniel_hozac well, devel will give you a line per CPU.
1145822344 M * ScoobyD00 ok - may i ask what a token is exactly? :o)
1145822351 M * ScoobyD00 (thanks daniel btw)
1145822366 M * daniel_hozac one tick, IIRC.
1145822391 M * ScoobyD00 so it is pretty much a reliable way to charge money for actual CPU resource?
1145822398 M * daniel_hozac yes.
1145822427 M * ScoobyD00 if i was to run a benchmark, i could in theory work out how many ticks is equivelent to, say a 3.0GHZ p4?
1145822441 M * ScoobyD00 and then no matter what hardware, compare results
1145822517 M * daniel_hozac i think the length of a tick depends on the kernel, and the kernel configuration.
1145822525 M * ScoobyD00 oh ok - so like the timer in the kernel
1145822585 M * ScoobyD00 so you think it might be different for different kernel versions (even if the config had the same value)?
1145822660 M * daniel_hozac it depends on the kernels, if they supported setting it in the configuration ;)
1145822731 M * ScoobyD00 :)
1145822786 M * ScoobyD00 i'm quite excited to be able to charge a vserver by CPU time (and in reality all the machines will have the same kernel version and config)
1145822818 M * ray6 derjohm: I'd somehow have to assign the /dev/ttyS0 the right "virtual uid" or something I guess?
1145822840 M * ray6 derjohn: won't need CAPI4Linux staff, I'll be using a USR Courier I-Modem :)
1145822840 M * daniel_hozac a tick's efficiency varies depending on the processor, it's a set amount of time.
1145822867 M * daniel_hozac ray6: shouldn't have to do anything like that, just copy it into the guest's /dev.
1145822979 M * ScoobyD00 i think if i test each new platform to get a "performance to tokens" indicator it should be ok
1145823107 M * daniel_hozac ScoobyD00: you may want to wait for Bertl_zZ to verify what i said, i'm still rather fuzzy on the scheduler stuff.
1145823480 M * ScoobyD00 :o)
1145823506 M * ScoobyD00 what times is bertl awake?
1145823600 Q * FireEgl Ping timeout: 481 seconds
1145823605 Q * Viper0482 Remote host closed the connection
1145823824 M * daniel_hozac every once in a while ;)
1145823992 M * ScoobyD00 reading: http://linux-vserver.org/vsched+explained   i think you are right
1145824063 J * matta ~matta@c-68-32-202-140.hsd1.pa.comcast.net
1145824415 Q * ben_ Quit: 
1145824468 Q * matt1 Ping timeout: 480 seconds
1145825323 Q * ScoobyD00 Quit: 
1145825770 Q * Pater_John Quit: get satisfied! • :: ««« (Gamers.IRC) »»» www.gamersirc.net ::
1145825881 J * FireEgl Atlantica@Atlantica.DollarDNS.Net
1145827053 Q * nebuchadnezzar Ping timeout: 480 seconds
1145827663 J * mugwump ~samv@watts.utsl.gen.nz
1145827705 J * nebuchadnezzar ~nebu@zion.asgardr.info
1145828326 Q * _coocoon_ Quit: KVIrc 3.2.0 'Realia'
1145828930 J * Aiken ~james@tooax7-247.dialup.optusnet.com.au
1145829144 Q * Oli Quit: Oli
1145829389 Q * bonbons Quit: Leaving
1145830557 M * rmoriz hm
1145830570 M * rmoriz where do i need to set raw_icmp with vs2.1?
1145830689 M * rmoriz aaah
1145831319 J * kilgur user@p50813758.dip0.t-ipconnect.de
1145831733 Q * kilgur_ Ping timeout: 480 seconds
1145832297 Q * matta Ping timeout: 480 seconds
1145833286 J * matta ~matta@c-68-32-239-173.hsd1.pa.comcast.net
1145833310 Q * pflanze Quit: [x]chat
1145833950 Q * matta Ping timeout: 480 seconds
1145834253 M * thorkild Groan. Wasted a lot of time getting a nfs-server on the host and on the client to work together, when all I really needed was to fix /etc/vserver/<name>/fstab . I feel stupid now.
1145834977 J * kilgur_ user@p50813758.dip0.t-ipconnect.de
1145835393 Q * kilgur Ping timeout: 480 seconds
1145836427 J * doener ~doener@i5387CE6B.versanet.de
1145836592 Q * doener_ Ping timeout: 480 seconds