1148429770 J * FireEgl FireEgl@adsl-220-219-103.bhm.bellsouth.net
1148429955 M * hardwire anybody used copyfs with a vrootdir?
1148431536 J * doener_ ~doener@i5387C3EC.versanet.de
1148431936 Q * doener Ping timeout: 480 seconds
1148432302 J * Methos ~mjoconr@hacker.pineview.net
1148434192 N * sarnold sars
1148434296 Q * hardwire Remote host closed the connection
1148437094 J * Aiken_ ~james@tooax8-160.dialup.optusnet.com.au
1148437301 M * Skram "does this Starbucks make me look fascist"?
1148437449 Q * Aiken Ping timeout: 480 seconds
1148437632 Q * MrX Ping timeout: 480 seconds
1148437746 Q * softi42 Ping timeout: 480 seconds
1148438372 J * softi42 cbxbcbp@p549D7B62.dip.t-dialin.net
1148439493 M * s0undt3ch is this last entry normal?
1148439493 M * s0undt3ch vserver-stat
1148439494 M * s0undt3ch CTX   PROC    VSZ    RSS  userTIME   sysTIME    UPTIME NAME
1148439494 M * s0undt3ch 0      101 397.5M 231.9M   4h39m16   1h29m52   1d07h46 root server
1148439494 M * s0undt3ch 3       31  96.3M  52.3M  30m14s13   1m12s19   1d07h39 mail
1148439496 M * s0undt3ch 4       72   219M   3.9G   4m45s36   0m47s64   1d07h39 web
1148439507 M * s0undt3ch 3.9G???
1148439514 M * Skram thats not right
1148439531 M * s0undt3ch it's running apache with mpm_worker support
1148439613 M * s0undt3ch Skram: what could be wrong?
1148439620 M * Skram the RSS/RAM
1148439625 M * Skram its kind of confusing
1148439644 M * s0undt3ch that is saying that it's using 3.9G of ram?
1148439661 M * Skram yeah
1148439668 M * Skram do you even have 3.9G of RAM?
1148439689 M * s0undt3ch well, with only 3*250Mb of phisical ram that seems impossible
1148439698 M * Skram VERY impossible
1148439700 M * Skram its not true
1148439739 A * s0undt3ch pings hollow to look at these ram stats
1148439747 M * s0undt3ch I'm running it on gentoo
1148440391 J * hardwire ~spencersr@209-112-218-196-cdsl-rb1.nwc.acsalaska.net
1148442731 Q * hardwire Quit: Ex-Chat
1148442768 J * stefani ~stefani@c-24-19-46-211.hsd1.wa.comcast.net
1148443778 Q * stefani Quit: going away now...
1148444081 N * otaku42_away otaku42
1148444574 M * eyck what is wrong with 3.9G RSS?
1148444662 M * s0undt3ch is it normal?
1148445315 M * eyck yeah, why not?
1148445366 M * eyck it may be waay higher then your physical potential, it counts shared usage as separate for example
1148445880 Q * mire Ping timeout: 480 seconds
1148448806 J * MrX ~urk@219.95.0.128
1148449157 M * Hollow morning folks!
1148449306 M * eyck morning
1148449978 Q * Loki|muh Remote host closed the connection
1148450008 J * Loki|muh loki@satanix.de
1148450194 J * mire ~mire@170-167-222-85.COOL.ADSL.VLine.verat.net
1148450194 Q * Loki|muh Read error: Connection reset by peer
1148450308 J * Loki|muh loki@satanix.de
1148450998 J * NetAsh ~NetAsh@62-212-199-5.elsis.lt
1148451017 M * NetAsh hello all
1148451614 M * Skram hey
1148454289 Q * Aiken_ Ping timeout: 480 seconds
1148454390 M * NetAsh the question out of topic:
1148454423 M * NetAsh how do you fight with image based spam?
1148454481 M * NetAsh my server resently is atacked with loads of html based spam with only one image containing the spam "message" in it
1148454555 M * Hollow NetAsh: i generally delete all html mails :)
1148454572 M * Hollow using procmail as mailfilter
1148454594 M * NetAsh whis would kill half of legit mail I receive
1148454622 M * Hollow too bad, i don't consider html mails legit :)
1148454643 M * NetAsh I would like too
1148454690 M * NetAsh but unfortunately moust customers/partners send html based emails
1148454700 M * NetAsh this sucks but this is true
1148454700 M * Hollow yeah... thanks outlook
1148454724 M * NetAsh unfortunatly thunderbirds default too
1148454742 M * Hollow well, thunderbird is a windows app like outlook, it was just ported to linux ;)
1148454750 M * Hollow same for mozilla
1148454752 M * Hollow :P
1148454758 M * Hollow *firefox
1148454766 M * Hollow *scnr*
1148454794 J * s0undt3c1 ~s0undt3ch@bl7-247-101.dsl.telepac.pt
1148454842 M * NetAsh I considered filtering all html mail with embedded images, but this is bad too
1148454862 M * Hollow NetAsh: not sure, but maybe you can use one of the checksum databases (razor & co iirc)
1148454891 M * Hollow and dcc
1148454907 M * NetAsh trust me, these images are fresh as hell
1148454912 M * Hollow hm
1148454958 M * NetAsh gues I could be the one who submits whese cheksums, but not use them becous to me its alreadi too late :)
1148454981 M * NetAsh ah crappy my english :(
1148455033 M * Hollow heh yeah... nasty spam
1148455202 Q * s0undt3ch Ping timeout: 480 seconds
1148455236 J * tassieboy ~s4074789@g110-9706.students.itee.uq.edu.au
1148455240 M * tassieboy :q
1148455247 M * tassieboy whoops
1148455255 M * tassieboy vi and mirc have different commands hey :P
1148455257 Q * tassieboy Quit: 
1148455776 Q * derjohn Ping timeout: 480 seconds
1148455805 J * ||Cobra|| ~cob@146.50.22.204
1148455899 J * cdrx ~legoater@cimai.net4.nerim.net
1148455997 Q * cdrx Quit: 
1148456201 J * hardwire ~spencersr@209-112-218-196-cdsl-rb1.nwc.acsalaska.net
1148456205 M * hardwire sup ladies
1148456254 J * Aiken ~james@tooax6-164.dialup.optusnet.com.au
1148457048 J * MrX` ~urk@60.49.41.149
1148457105 J * dna ~naucki@p54BCD20B.dip.t-dialin.net
1148457132 Q * MrX Ping timeout: 480 seconds
1148457615 J * dsoul_ darksoul@vice.ii.uj.edu.pl
1148457947 J * bonbons ~bonbons@83.222.38.81
1148457973 Q * dsoul Ping timeout: 480 seconds
1148458725 J * Milf ~Miranda@ipsio382.ipsi.fraunhofer.de
1148458733 M * Milf Good morning
1148459459 N * dsoul_ dsoul
1148459939 Q * Methos Quit: Methos
1148460153 J * derjohn ~derjohn@80.69.37.19
1148460391 M * hardwire morning?
1148460395 M * hardwire yeh.. for 47 minutes
1148460399 A * hardwire goes to sleep
1148460665 Q * shedi Quit: Leaving
1148460854 J * Methos ~mjoconr@hacker.pineview.net
1148461158 Q * hardwire Quit: Ex-Chat
1148461272 J * hardwire ~spencersr@209-112-218-196-cdsl-rb1.nwc.acsalaska.net
1148461498 Q * lonewolff Remote host closed the connection
1148463629 J * kir ~kir@swsoft-mipt-nat.sw.ru
1148463990 M * Milf Yes, it's good morning. This channel lives in Bertltime, a timezone where it's morning the whole day long :)
1148464029 M * Hollow i love BUT :)
1148464054 M * Hollow i will adapt it the next 4 months too *g*
1148464192 M * tokkee All times are Bertl? ;-)
1148464216 M * tokkee All your times belong to Bertl ;-)
1148464253 A * Milf add an 'are' into the correct position in that last statement.
1148464278 M * Hollow :)
1148464332 M * tokkee Um... yeah... d'oh ;-)
1148464382 M * eyck good morning then
1148464405 M * SNy GBTG?
1148464411 M * SNy generic Bertl time greeting?
1148464414 M * SNy ;p
1148464569 M * sid3windr Hollow: you love butt? :p
1148464713 M * Milf Success: I just ported an Ubuntu into a Vserver hosted on a 2.4 Kernel
1148464732 M * Milf No he loves BUT: Bertl Universal Time
1148464745 M * Wonka Milf: is that more than debootstrap and stuff?
1148464782 M * Milf The thing was already installed with a gforge running inside it and I had to make a VServer out of it.
1148464801 M * Milf Also the original has a 2.6 Kernel and mit Host is running 2.4
1148464818 M * Milf Quote: 'I hope to god it worked.'
1148464835 A * Wonka doesn't really understand why one would switch from 2.6 back to 2.4
1148464867 M * Milf I havn't switched to 2.6. I'm too fscking lazy to update my host servers.
1148465062 M * Hollow BUT = GMT +/- NaN
1148465063 M * Hollow ;)
1148465252 J * shedi ~siggi@dsl-og-108-50.du.vortex.is
1148466022 J * DarthVader ~Aniken@203.177.212.163
1148466781 Q * cdlu Ping timeout: 480 seconds
1148466981 M * Milf What's NaN?
1148467262 M * eyck Non-Arithmetic Number
1148467275 M * eyck usually used for 0/0, or sth like that
1148467281 M * eyck 1/0 is also a NaN
1148467288 M * eyck same goes for -1/0 etc..
1148467723 M * Milf Oh, so BUT is a complex Timezone.
1148468411 Q * waldi Ping timeout: 480 seconds
1148468413 J * waldi ~waldi@bblank.thinkmo.de
1148468850 Q * derjohn Remote host closed the connection
1148468921 Q * NetAsh Ping timeout: 480 seconds
1148469478 J * lonewolff ~lonewolff@adleman.lonewolff.info
1148470360 N * s0undt3c1 s0undt3ch
1148470726 J * sladen paul@starsky.19inch.net
1148471012 J * complexmind ~mark@162.84.2.81.in-addr.arpa
1148471031 J * Dr4g ~Dr4g@tor-irc.dnsbl.oftc.net
1148471250 M * complexmind can anyone tell me... I am using the dhozak rpms for the first time - I use a custom path for my vserver config and file systems but the rpms are configured to the standard path...  Can I just edit the util-vserver-vars file to point everything to the right place or are the paths hard coded anywhere else?
1148471401 N * Bertl_zZ Bertl
1148471409 M * Bertl morning folks!
1148471423 M * s0undt3ch morning
1148471445 M * complexmind morning bertl!
1148471541 Q * Hunger Ping timeout: 480 seconds
1148471701 J * alamar ~alamar@sol.diddens.de
1148471864 M * Bertl welcome alamar!
1148471981 M * SNy Milf: Hehe.
1148472022 M * Milf good morning Bertl
1148472144 M * Bertl complexmind: I would get the source rpm and adjust the pathes there
1148472181 M * complexmind great thanks, that's what I just did :)
1148472265 Q * kir Quit: Leaving
1148472326 Q * Aiken Ping timeout: 480 seconds
1148472663 J * Hunger Hunger.hu@Hunger.hu
1148472675 M * Bertl wb Hunger!
1148472804 J * derjohn ~derjohn@80.69.37.19
1148472828 M * Bertl hey derjohn!
1148472899 M * derjohn Bertl, himself :) I had to reboot ... nor rc21 and 2.6.16.18 ... man, the updates come faster than my ability to compile ;)
1148472926 M * Bertl yes, .18 was the snmp fix, IIRC
1148473084 M * derjohn well, I did even overlook .17 ;) 
1148473116 M * derjohn do you think rc21 will be the next release? (Assuming there wont be errors reported ...)
1148473167 M * Bertl I'm still working on some minor cleanups, but the functionality should be the same
1148473300 M * derjohn ok. then I will wait until i upgrade by production boxes
1148473313 M * derjohn (to avaiod downtime)
1148473317 M * derjohn *avoid
1148473622 M * Wonka i got a problem here
1148473642 M * Wonka i cannot remove /vservers/$vserver/proc, device is busy
1148473651 M * Wonka but i cannot identify what busies it
1148473799 M * Bertl lsof should help here, maybe in the guest context?
1148473873 M * Wonka lsof: status error on /vservers/$vserver/proc: Permission denied
1148473922 M * Wonka in context 0, just nothing
1148473952 M * Wonka /proc/virtual/$xid doesn't exist...
1148473970 M * Bertl ah, then it isn't a guest process keeping it busy
1148474085 M * Wonka what might it be else?
1148474101 M * Wonka i suspected a procfs mounted there
1148474116 M * Wonka but i cannot find out how to find out or how to umount
1148474228 M * Wonka is there a possibility to get something in /proc in contexts 0 or 1 which just shows _all_ namespaces, _all_ xids, and stuff?
1148474244 Q * sladen Remote host closed the connection
1148474271 M * Wonka so that one could find out such things by "chcontext --xid 1 cat /proc/mounts" or something?
1148474373 M * Bertl Wonka: nope, the namespaces are per process
1148474388 M * Bertl Wonka: so there (currently) is no way to list all of them
1148474402 M * Bertl you can only enter namespaces assigned to contexts
1148474449 A * Wonka sees a necessity for /proc/$pid/namespace there...
1148474484 M * Wonka i would like to be able to see anything there is :)
1148474490 M * Wonka mh
1148474496 M * Wonka still, why is it busy?
1148474504 J * the_hydra ~a_mulyadi@202.59.168.5
1148474658 M * Bertl welcome the_hydra!
1148474689 M * the_hydra hi bert
1148474697 M * the_hydra we must prepare to bomb malaysia...
1148474700 M * Bertl Wonka: for proc it is unusual, typically only a process 'reading' a file there or sitting in a directory can dirty it
1148474703 M * the_hydra wait...i want to prepare C4
1148474725 M * Bertl the_hydra: hmm?
1148474740 J * sladen paul@starsky.19inch.net
1148474757 M * the_hydra Bertl: yeap, badawi plans to claim another Indonesian Island...fun isn't it?
1148475774 J * harti ~Hartmut@85-124-100-166.dynamic.xdsl-line.inode.at
1148475789 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net
1148476329 J * Snow-Man ~sfrost@kenobi.snowman.net
1148476793 M * tokkee Wonka: Was there a vserver running before?
1148476806 M * Wonka tokkee: it was, yes
1148476821 M * tokkee Wonka: You just might wanna try "umount /vservers/$vserver/proc"
1148476830 M * Wonka "not mounted"
1148476889 M * tokkee Wonka: I guess you deleted that vserver?
1148476893 M * Wonka yes
1148476901 M * Wonka config dir is still there
1148476921 M * Wonka but in /vservers/$name, there's only proc left 
1148476931 M * tokkee I had that very same problem a couple of weeks ago. Bertl told me how to solve it, but I can't remember exactly right now :-/
1148476946 M * Wonka maybe i'll find it in the logs...
1148476965 M * tokkee You problably should.
1148477005 M * tokkee Wonka: You either have to enter that context or the namespace and execute umount /proc... something like that.
1148477022 M * doener_ exactly :)
1148477044 M * doener_ vnamespace -e foo umount /vservers/foo/proc
1148477052 M * Wonka chcontext --xid $xid umount $path didnt help
1148477084 M * tokkee chcontext actually does not make much sense ;-)
1148477086 M * Wonka vnamespace: vc_enter_namespace(): No such process
1148477126 M * Wonka with both "wohnt" and "124"
1148477136 M * Wonka which are $name and $xid
1148477137 M * doener_ yeah, sorry, missed the first part of you discussion
1148477141 P * Roey Leaving
1148477153 M * derjohn Wonka, as a payback you should list the information on the wiki ! :)
1148477180 M * Wonka derjohn: a) where?, b) sure, as soon as it's solved...
1148477205 M * derjohn Wonka, FAQ ?
1148477223 M * tokkee twice == frequently? ;-P
1148477235 M * derjohn or put it under the infmous link /some_hints_from_john :)
1148477242 M * tokkee No... I think that's a good idea.
1148477247 M * Wonka ;)
1148477249 M * tokkee FAQ
1148477250 M * Wonka 'k
1148477266 Q * keith80403 Quit: Leaving
1148477275 Q * Methos Quit: Methos
1148477286 M * derjohn Wonka, at the time the strategy ist "put it in", With the new wiki we may re-structure it anyways ...
1148477348 M * Wonka while we are at it... i also want a way to clone newconfig vservers.
1148477376 M * derjohn Wonka, thats easy ...
1148477420 M * Wonka i want to build a template, and just copy that...
1148477427 M * derjohn Wonka, cp -a /var/lib/vservers/foo /var/lib/vservers/bar . then enter /var/lib/vservers/bar and change all symlinks (I think 3) to the right place.
1148477443 M * Wonka symlinks? in /vserver?
1148477452 M * derjohn s/var\/lib/etc/
1148477458 M * derjohn sry :)
1148477459 M * Wonka ah, ok
1148477478 M * Wonka how do i make foo not start?
1148477485 M * Wonka call it .foo?
1148477493 M * derjohn and then copy /vserver/foo /vserver/bar and do a "chxid" on the /vserver/bar
1148477499 M * Wonka mhmh
1148477528 M * derjohn ahm and change /etc/vserver/bar/{context|interface...} to what you like
1148477533 M * Wonka yeah
1148477540 M * Wonka would be debian stable...
1148477548 M * Wonka debootstrap sucks a little
1148477563 M * Wonka copies files from the host to the new root
1148477564 M * tokkee Wonka: debootstrap is great! :-)
1148477572 M * Wonka especially /etc/hostname
1148477576 M * Wonka which is just wrong there
1148477588 M * Wonka tokkee: in principle, you're right
1148477593 M * derjohn Wonka, usually the is /etc/vserver/foo/apps/init/mark, while mark contains the word 'default', meaning it starts the vserver at boot time. check out some_hints_from_john ;)
1148477608 M * Wonka 'k
1148477620 M * derjohn tokkee, well, I you have a cleanly build template it easier to copy
1148477640 M * derjohn Wonka, and you could also do a "-m skeleton" on vserver build.
1148477651 M * Loki|muh derjohn: there will be a new wiki? sounds great \o/
1148477657 M * tokkee derjohn: Well of course - that saves you from doing a basic configuration twice...
1148477660 M * Loki|muh any details?
1148477671 M * derjohn Loki|muh, yes, you will have to install it for us :)
1148477676 M * derjohn Loki|muh, </kidding>
1148477700 M * Wonka derjohn: that conflicts with -m debootstrap, nor?
1148477707 M * derjohn Loki|muh, there are plans out. BenBen is the driving force behind it. It will probably be mediawiki.
1148477716 M * Loki|muh well, I think I could, if it is appreciated...
1148477718 M * Loki|muh ;)
1148477730 M * tokkee Wonka: yes... but you don't need debootstrap if you already have your basic stuff copied from another guest.
1148477780 M * Loki|muh I do always a -m skeleton and then rsync a debootstrapped sys over the skeleton
1148477780 M * derjohn Wonka, sure. -m skeleton creates /etc/vserver/whatever/... and only makes very basic stuff in /vserver/whatever, giving you the oppotunity to copy /vserver/bla into /vserver/whatever
1148477792 M * Wonka ah, ok
1148477800 M * derjohn Loki|muh, I apprentice it! Others too! Contact BenBen !!
1148477821 M * derjohn Loki|muh, then you can directly use -m debootstrap :)
1148477836 M * Wonka so, i would -m debootstrap a template, -m skeleton the real vserver, and rsync / cp -a the dir...
1148477846 M * Loki|muh derjohn: yes but my skeleton is customized ;)
1148477868 M * Wonka mine would be too
1148477871 M * derjohn Loki|muh, then copy it to the .default ....
1148477887 M * derjohn Wonka, dont forget , that a skeleton != template.
1148477897 M * Wonka -v
1148477919 M * Wonka derjohn: did you mean .defaults?
1148477943 M * Loki|muh derjohn: oh there is a possibility like that... good to know ;)
1148477968 M * tokkee ;-)
1148478025 M * derjohn well .. lart me , but I always assumed to can put the skeleton files to /etc/vserver/.defaults/... ?
1148478038 M * derjohn but ATM i kind find a skeleton subdir there.
1148478050 M * derjohn someone more enlightend maybe able to tell .... 
1148478076 M * derjohn (there is a "files" subdir ...  hmmm ... dunno :))
1148478098 M * derjohn Bertl ? More enlightend ?
1148478119 M * tokkee "sending signals across contexts [done]" Who would want something like that?
1148478146 M * derjohn IPC between vservers?
1148478178 M * tokkee That's how I interpret it...
1148478197 M * Bertl derjohn: didn't follow the discussion, what was the question?
1148478228 M * derjohn Bertl, where does -m skeleton get's it base files from? How to modify that set?
1148478243 M * derjohn *lol* "gets it's"
1148478253 M * Wonka "get it's" even ;)
1148478280 M * derjohn Wonka gets it's new vserver ....
1148478304 M * Wonka "where does -m skeleton get"...
1148478308 M * Wonka however
1148478317 M * derjohn Wonka, ok you won .
1148478337 Q * orzel Remote host closed the connection
1148478342 M * tokkee "get its" would be correct ;-)
1148478384 M * Wonka mmh, yes, there was something like that in 5th grade...
1148478414 M * tokkee ;-D
1148478484 M * Bertl derjohn: IIRC -m skeleton is hardcoded
1148478499 M * Bertl after all it's just the few dev files and /proc and /tmp
1148478514 M * derjohn Bertl, OMG! Is there a -m template or such?
1148478531 M * Bertl not that I know of, why would you need/want that?
1148478598 M * tokkee What does '-m copy' do?
1148478612 M * Bertl copy a template
1148478629 M * Wonka which?
1148478646 M * Bertl ah, sorry, copy is the copy from host
1148478652 M * Bertl it always confuses me
1148478666 M * Bertl 'clone' is the copy from template
1148478674 M * derjohn Bertl, then we have all :)
1148478712 M * tokkee So, copy creates a copy of the host system?
1148478724 M * Bertl yes
1148478730 M * tokkee Cool :-)
1148478760 M * derjohn Cool!
1148478783 M * tokkee clone is not present in 0.30.204... why do I have .204 installed btw?
1148478808 M * tokkee aptitude install -t sarge-backports util-vserver
1148478880 M * Wonka 209 seems not to have clone, too
1148478947 M * tokkee ... either :-P
1148478977 A * Wonka was quite good in englisch, but not best
1148478991 M * Wonka usually, i can make myself understood
1148478993 A * tokkee spent one year in the US ;-)
1148479003 M * tokkee But still my English is not the best ;-)
1148479004 M * Wonka this is also known as "broken english" ;)
1148479051 M * Wonka and every year at chaos communication congress, i see again that there's much more broken english than mine
1148479068 M * tokkee Definitly ;-)
1148479069 M * phedny I speak too "broken english" because difficult is it speak right the language :)
1148479222 M * tokkee My english is very sucking a lot, too, most of ze time ;-)
1148479770 Q * the_hydra Quit: 
1148481051 M * Bertl okay, off for now ... back later ...
1148481061 N * Bertl Bertl_oO
1148481668 Q * shedi Quit: Leaving
1148482853 Q * matti Ping timeout: 480 seconds
1148482875 Q * DarthVader Quit: Leaving
1148483027 J * matti matti@linux.gentoo.pl
1148483287 Q * dna Quit: Verlassend
1148483373 M * Wonka yipee! finally, there's util-vserver 0.30.210 in debian!
1148483378 M * Wonka (testing, that is)
1148483484 Q * ||Cobra|| Remote host closed the connection
1148483528 J * stefani ~stefani@tsipoor.banerian.org
1148483570 Q * Milf Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org
1148483613 J * dna ~naucki@p54BCD20B.dip.t-dialin.net
1148483902 M * Wonka util-vserver 0.30.210-10 also has no buld method "clone"
1148484438 J * lilalinux ~plasma@dslb-084-059-005-197.pools.arcor-ip.net
1148484958 M * cemil foo!
1148485118 M * doener_ bar!
1148485131 M * trash 42!
1148485145 J * kir ~kir@swsoft-mipt-nat.sw.ru
1148485183 N * otaku42 otaku42_away
1148485194 P * matti 8-X
1148486880 M * daniel_hozac -m clone is a q'n'd patch.
1148486919 M * daniel_hozac i haven't gotten around to making the real thing yet.
1148487372 M * daniel_hozac (http://daniel.hozac.com/vserver/util-vserver-0.30.210-clone.patch if you want it)
1148488283 M * kir daniel_hozac, don't you need to stop a source vserver to do a proper clone?
1148488322 M * kir daniel_hozac, i.e. do copy, then stop source vserver, then do rsync and start source vserver -- just to be sure everything is in solid state?
1148488377 M * kir daniel_hozac, or at least require source vserver to be stopped
1148488394 N * sars sarnold
1148488521 M * derjohn cemil, doener_ trash , 23 ! 
1148488567 M * derjohn kir, I think so, too (open files/active files), so a vserver bal stop would make sense :)
1148488745 Q * doener_ Quit: leaving
1148489410 J * Viper0482 ~Viper0482@p54976C59.dip.t-dialin.net
1148491069 Q * lilalinux Ping timeout: 480 seconds
1148491229 J * shedi ~siggi@inferno.lhi.is
1148491579 J * lilalinux ~plasma@dslb-084-059-005-185.pools.arcor-ip.net
1148491622 J * ben_ ~ben@88.134.54.56
1148491643 M * ben_ hello 
1148493058 J * jpacheco ~justin@CPE00146c1608af-CM00159a655a08.cpe.net.cable.rogers.com
1148493062 M * jpacheco hey everyone!
1148493134 M * jpacheco any of you guys know how to mount directories onto vservers while the vserver is running?
1148493208 P * michal` >/dev/yachill
1148493673 M * daniel_hozac jpacheco: vnamespace -e <xid or name> mount...
1148493695 M * daniel_hozac kir: as i said, q'n'd.
1148493862 J * schimmi ~sts@port-212-202-73-176.dynamic.qsc.de
1148493937 M * kir daniel_hozac, so you can just add a one-line check that vserver is stopped and bail out crying loudly if it is not :) probably even better than stopping a running instance because to some ppl that might be seen as a bad surprise
1148493965 M * daniel_hozac AFAIK this patch is not used anywhere.
1148493991 M * kir well I understand. just giving my $0.02
1148494067 M * jpacheco daniel_hozac: thx
1148494266 M * jpacheco daniel_hozac: that command will allow me to bind a directory into a vserver like entries in a vserver's fstab will?
1148494276 M * daniel_hozac yes.
1148494301 M * daniel_hozac http://linux-vserver.org/Namespaces
1148494683 M * schimmi hi! what's the right to mount directories into the vserver readonly which are writable in the host? (something like "-o bind" I guess?)
1148494750 M * daniel_hozac yep, -o bind,ro should work fine.
1148494812 M * schimmi sure? last time I checked the ro didn't change anything
1148494829 M * schimmi but trying again :)
1148494873 M * daniel_hozac well, it does require BME.
1148494885 M * schimmi BME?
1148494889 M * daniel_hozac (which is in devel and 2.0.2-rc6+ IIRC)
1148494899 M * daniel_hozac Bind Mount Extensions.
1148494924 M * schimmi ic
1148494926 M * schimmi interesting
1148494958 M * schimmi are there any more up to date sarge packages?
1148494989 M * daniel_hozac IIRC backports.org should have vserver binary kernels.
1148495077 M * jpacheco daniel_hozac: i tried a few variations of that command, some exit with no errors, however, the directory still won't unmount
1148495079 M * schimmi nice, even prebuilt vserver kernels. no need for another 4 hours build session :))
1148495335 J * Cocoademon ~cocoademo@alcopop.org
1148495346 J * Magicsofa ~magicsofa@c-66-30-95-227.hsd1.ma.comcast.net
1148495372 M * Magicsofa cookie
1148495372 A * Cocoademon munches a milk chocolate cookie
1148495377 P * Magicsofa 
1148496253 M * schimmi this BME patch is not part of recent 2.6 kernels, isn't it?
1148496589 M * hardwire meh
1148498119 Q * pusling Ping timeout: 480 seconds
1148498846 N * Bertl_oO Bertl
1148498851 M * Bertl evening folks!
1148498855 M * Bertl schimmi: not yet :)
1148498895 M * s0undt3ch can't we use nmap from within a guest?
1148498909 M * Skram s0undt3ch: kernel stuff
1148498909 M * Skram heh
1148498914 M * Skram edit bcappabilities or something
1148498922 M * s0undt3ch Skram: with what?
1148498931 M * Skram CAP_NET or CAP_NET_RAW
1148498932 M * s0undt3ch and where exactly?
1148498932 M * Skram i forget
1148499135 M * s0undt3ch and from outside the guest I can ping it but I cant nmap it, is this also normal?
1148499178 M * Bertl s0undt3ch: why would you want to use nmap from a guest instead of the host?
1148499321 M * s0undt3ch Bertl: because of my above comment
1148499323 M * s0undt3ch and from outside the guest I can ping it but I cant nmap it, is this also normal?
1148499390 M * s0undt3ch nmap 10.1.0.50
1148499396 M * s0undt3ch Nmap finished: 1 IP address (0 hosts up) scanned in 0.269 seconds
1148499405 M * s0undt3ch nmap 10.1.0.50 -P0
1148499410 M * s0undt3ch Nmap finished: 1 IP address (0 hosts up) scanned in 0.267 seconds
1148499421 M * s0undt3ch ping 10.1.0.50
1148499421 M * s0undt3ch PING 10.1.0.50 (10.1.0.50) 56(84) bytes of data.
1148499421 M * s0undt3ch 64 bytes from 10.1.0.50: icmp_seq=1 ttl=126 time=0.174 ms
1148499422 M * Bertl s0undt3ch: what are you trying to nmap`
1148499439 M * s0undt3ch to find out opened ports
1148499461 M * Bertl okay, ports of a guest, yes?
1148499509 M * s0undt3ch :) yes
1148499542 M * Bertl okay, and you want to do that from the same host or what?
1148499576 M * s0undt3ch I'm fine to do it from the root guest
1148499596 M * s0undt3ch *from the physical server
1148499597 M * Bertl there is no root guest, and the host context is of course on the same host too
1148499612 Q * Viper0482 Remote host closed the connection
1148499613 M * Bertl have you ever tried to nmap your own server?
1148499622 M * s0undt3ch yep and it works
1148499626 M * Bertl or your own machine that is?
1148499638 M * s0undt3ch oops :|
1148499677 M * s0undt3ch nmap 127.0.0.1 works, 10.1.0.1 doesn't
1148499702 M * s0undt3ch iptables issue? (forgoten conf)
1148499726 M * Bertl probably
1148499745 M * s0undt3ch yeah probably, well, I'm ok with it for now
1148499803 M * Bertl basically you won't get more information if you run nmap inside a guest (with NET_RAW) than when you do it from the host
1148499829 M * Bertl and you might get wrong (or at least different) results than from outside
1148499829 M * s0undt3ch since I'm new to vserver I'll leave bcapabilities alone
1148499862 M * Bertl so the best way to check a new guest (with nmap) is to do it from your workstation
1148499900 M * s0undt3ch k, thanks
1148499913 M * Bertl you're welcome!
1148499971 M * mire hello
1148499978 M * mire has anyone got centos vserver images
1148499992 M * mire or an instructions on how to build one that works on debian?
1148500014 M * Bertl there is a howto on the wiki (don't know if it works for debian)
1148500014 Q * hardwire Quit: Ex-Chat
1148500034 M * Bertl (probably not)
1148500050 M * Bertl IIRC, marlow had some older centos images
1148500072 M * mire no centos on marlow, only fedora 2
1148500111 M * Bertl but basically you can (if you really want debian as host system) install centos to a separate partition and/or in a virtual machine like QEMU
1148500152 Q * Dr4g Ping timeout: 480 seconds
1148500153 M * Bertl I've heard that ubuntu allows for a working rpm/yum and all rpm based distros have no problem with installing rpm and deb based stuff
1148500167 J * Dr4g ~Dr4g@ua-85-227-10-132.cust.bredbandsbolaget.se
1148500192 M * Bertl mire:
1148500214 M * Bertl you could do the follwing (smart move :)
1148500233 M * Bertl get an arbitrary rpm based guest image (e.g. the fc 2 one)
1148500257 M * Bertl chroot into that image, and install util-vserver there
1148500279 M * Bertl you will then be able to create other guests from within that image
1148500301 M * Bertl which in turn, can be moved out of the image and used as guest
1148500411 Q * shedi Quit: Leaving
1148500498 M * kir mire, you can try your luck with centos4 images for OpenVZ: http://openvz.org/download/template/cache/
1148500528 M * kir mire, should more or less work as is
1148500540 M * kir mire, or so I think (never tried it myself)
1148500747 M * Bertl well, if the images contain unmodified distro binaries (and therefor would work on a normal linux system) then they will work quite fine for Linux-VServer too
1148501233 Q * lilalinux Remote host closed the connection
1148501239 M * jpacheco hey Bertl!
1148501332 M * Bertl hey jpacheco!
1148501389 M * jpacheco what's happening?
1148501744 M * mire ok, lets try openvz images...
1148502308 J * DoomStar DoomStar@ppp83-237-50-168.pppoe.mtu-net.ru
1148502369 M * Bertl welcome DoomStar!
1148502460 M * DoomStar Hello )
1148502465 M * DoomStar so quiet
1148502515 M * DoomStar I came to ask. Is there a tool that waorks like vtop, but also show context id or name
1148502647 M * Wonka question about dlimits: in the config dir, only the max is saved
1148502657 M * daniel_hozac DoomStar: no.
1148502667 M * Wonka where is the current level saved?
1148502682 M * Wonka or is useage measured on start?
1148502698 M * daniel_hozac Wonka: /etc/vservers/<name>/cache/<xid><path with _ instead of />, IIRC.
1148502718 M * Bertl DoomStar: vps is as close as you can get
1148502719 M * daniel_hozac ah, cache/dlimits.
1148502736 M * Wonka ah, thx
1148502737 M * Wonka ok
1148503175 M * mugwump ooo, akpm accepting my patches... happy day
1148505015 Q * DoomStar Quit: 
1148505869 J * NetAsh ~NetAsh@88.222.136.221
1148505874 M * NetAsh hello
1148505912 M * Bertl welcome NetAsh!
1148505935 M * NetAsh bertl, is there any progres on sparcy shiping?
1148505961 M * Bertl nope, not yet
1148505981 M * NetAsh in case there is - inform me via email, ok?
1148505993 M * Bertl will do so, TIA!
1148506009 M * NetAsh I am infrequent guest in irc :)
1148506017 M * Bertl np
1148506068 M * Bertl do I ahve your email address?
1148506094 M * NetAsh you sent me once - I gues you have
1148506112 M * Bertl ah, yes, I remember ...
1148506132 M * Bertl sorry always a little confused lately :)
1148506302 M * NetAsh np ;)
1148506343 Q * ben_ Quit: Konversation terminated!
1148507082 J * pusling pusling@195.215.29.124
1148507388 J * zeng ~guest@p54A9678D.dip.t-dialin.net
1148507401 M * zeng hi all... 
1148507426 M * sarnold zeng: are you the fellow working on multi-seat X?
1148507438 M * zeng hi sarnold, yep... :)
1148507446 M * sarnold zeng: cool; got a quick pointer?
1148507488 M * zeng sarnold: yes, indeed, we are preparing an article for it.
1148507521 M * zeng sarnold: do you have time to "be beta tester" of the article ;) ?
1148507540 M * sarnold zeng: of an article, yes; of the actual software, configuration, etc, doubtful :(
1148507602 J * Aiken ~james@tooax6-180.dialup.optusnet.com.au
1148507653 M * zeng sarnold: i see... the point is, we need someone (beside me) who has time, resource and is willing to test to seat-up a multiseat according to the article.
1148507670 M * Bertl evening zeng!
1148507671 M * zeng I mean: to setup ;)
1148507676 M * sarnold zeng: heh, yeah, i completely understand ;)
1148507677 M * zeng hi Bertl...
1148507711 M * zeng Bertl: I'm looking for some beta tester ;)
1148507733 M * zeng Bertl: someone, who likes to feed the beast ;)
1148507774 M * Bertl ah, what will she have to test?
1148507823 M * zeng Bertl: someone (beside me) who has time, resource and is willing to test to setup a multiseat according to the article, I & milf prepared.
1148507893 M * zeng btw. it doesn't have to be a 3 head multiseat. 2 heads would be actually enough.
1148507920 M * Bertl ah, i.c. well, judging from the overwhelming interest on the booth, you should not have to search very long ...
1148508235 M * Skram what is this you all are talking about?
1148508238 M * Skram multiseat?
1148508250 M * zeng sarnold: please be "patient" with the multiseat OK :) It will be published soon on our wiki.
1148508273 M * sarnold zeng: cool :) thanks
1148508329 M * zeng Skram: http://linux-vserver.org/MoreUbuntu
1148508359 M * Skram o ok
1148508370 M * Skram so its specifically for a server and thin clients?
1148508372 M * Skram kind of
1148508381 M * Skram +?
1148508427 M * zeng Skram: nope, it just simple multiseat Xorg, but powered by Linux-Vserver ;)
1148508443 M * Skram i dont get it
1148508458 M * Skram so, the server runs X, what does the computer connected to the monitor do?
1148508545 J * weasel_ weasel@asteria.debian.or.at
1148508569 N * weasel_ Guest43
1148509092 P * stefani I'm Parting (the water)
1148509139 M * daniel_hozac Skram: multiple monitors, one computer.
1148509179 Q * weasel Read error: Connection timed out
1148509179 N * Guest43 weasel
1148509204 M * Skram right
1148509205 M * Skram okay
1148509223 M * Skram all of our servers for work are remote and most people work from home
1148509236 M * Skram but sounds very cool, as I spoke further with zeng 
1148509325 M * zeng Bertl: since we still have to use /dev/mem for Xorg, we might need some "isolation" here.
1148509382 M * Bertl yep, we talked about that on LT, we should do some tests to check what X is doing on /dev/mem
1148509435 M * NetAsh good night all
1148509442 Q * NetAsh Quit: 
1148509446 M * Bertl btw, a method to make it instantly secure (but maybe with less features) would be to use the framebuffer devices
1148509468 M * Bertl (which provide the required abstraction)
1148509469 M * zeng Bertl: with framebuffer you dont have acceleration 
1148509478 M * zeng no openGL :(
1148509480 M * Bertl that's correct
1148509500 J * doener ~doener@i5387C3EC.versanet.de
1148509510 M * Bertl evening doener!
1148509526 M * Bertl zeng: okay, the first 'test' we should do is on the userspace side
1148509549 M * Bertl let's run the Xorg with an strace -fF -o xorg.trace
1148509554 M * zeng hi doener...
1148509561 M * doener evening!
1148509574 Q * harti Quit: Client exiting
1148509583 M * Bertl and filter out the access to /dev/mem so that we get a first idea what X is doing there
1148509696 M * zeng i see... give me some minutes...
1148509708 M * Bertl I also assume that X will require some iopl (port range mappings)
1148509804 M * zeng BTW: here you can see, why Xorg need /dev/mem (DEV_MEM) http://webcvs.freedesktop.org/xorg/xserver/xorg/hw/xfree86/os-support/linux/lnx_video.c?rev=1.11&view=markup
1148509993 M * zeng Bertl: trace output follows
1148509997 M * zeng 9950  open("/dev/mem", O_RDONLY)        = 4
1148509998 M * zeng 9950  mmap2(NULL, 4096, PROT_READ, MAP_SHARED, 4, 0xf8) = 0x97f60000
1148509998 M * zeng 9950  munmap(0x97f60000, 4096)          = 0
1148509998 M * zeng 9950  close(4)                          = 0
1148510061 M * Bertl hmm, read shared 4k from 0xf8 ?
1148510089 M * zeng yep, that's all. I think the rest, is done by nvidia module ...
1148510095 M * Bertl interesting .. but that should be allowable
1148510115 M * Bertl but I doubt the nvidia interface is  available/secure
1148510123 M * Bertl and what about non-nvidia cards?
1148510166 M * zeng I have Sis one... need to switch seat ... brb...
1148510206 A * Bertl .o( zeng running around his multi-seat-desktop ... nice :(
1148510215 M * Bertl s/:(/:)
1148510257 J * Methos ~mjoconr@hacker.pineview.net
1148510265 Q * zeng Quit: ... und tschüß
1148510487 M * Loki|muh lol
1148510687 J * zeng ~sef@p54A9678D.dip.t-dialin.net
1148510715 M * Bertl wb zeng!
1148510902 Q * dna Quit: Verlassend
1148510962 M * tokkee Anybody in here going to LUG camp?
1148510977 M * Loki|muh hehe
1148510980 M * tokkee ;-)
1148511005 M * zeng yep... with sis driver, i see many access to /dev/mem
1148511154 M * Bertl write access too I assume
1148511175 M * Bertl okay, got a good idea what X tries with /dev/mem
1148511204 M * zeng i see... mind to share ?
1148511207 M * Bertl question is, if you do not give bcaps to the guest, is /dev/mem sufficient?
1148511228 M * zeng no, I need bcaps
1148511233 M * Bertl I think we could virtualize /dev/mem for X in a reasonable way
1148511250 M * Bertl but aside from that, what else is required?
1148511270 M * zeng to be exactly, I need SYS_ADMIN, SYS_RAWIO
1148511285 M * Bertl well, you can save the time to secure /dev/mem with that :)
1148511299 M * zeng these caps are reduced after X is started.
1148511309 M * Bertl what is SYS_ADMIN required for?
1148511338 M * Bertl rawio is probably used for the beforementioned iopl
1148511349 M * Bertl once it figures the pci regions from /dev/mem
1148511385 M * zeng well I think you SYS_ADMIN to be able to map /dev/mem 
1148511410 M * Bertl hmm, could be, haven't checked that yet
1148511437 M * Bertl you have a kernel tree at your hand (for some quick tests)?
1148511477 M * zeng hm... not really. Since I use the Ubuntu one, it takes very long to build... :(
1148511508 M * zeng it seems to me, the Ubuntu .deb builds all modules too...
1148511528 M * zeng but what do you want me to test with the kernel ?
1148511550 M * Bertl yes, okay, let's delay this till tomorrow, please prepare a kernel with a minimum of modules/drivers/etc which still works for you
1148511554 M * zeng I can prepare it and let it runs overnight.
1148511563 M * zeng i see...
1148511570 M * Bertl we probably need to recompile it a few times
1148511586 M * Bertl so also do not use the ubuntu build process, instead build it manually
1148511607 M * Bertl the kernel build system will then only rebuild the affected parts
1148511617 M * Bertl (usually takes a few seconds on simple changes)
1148511638 M * zeng mind to share what you have in mind regarding this kernel test?
1148511666 M * Bertl I will prepare a few hacks to circumvent security in the kernel to allow access to things I consider virtualizeable without too much efford
1148511688 M * Bertl if that succeeds to make X start and work as expected, we can refince that
1148511691 M * Bertl *refine
1148511696 M * zeng reading ...
1148511745 M * Bertl i.e. what I saw until now is that X opens /dev/mem to get the pci config space
1148511769 M * Bertl it will then detect the devices from that, and map/access the iospace/memory for those
1148511776 M * zeng yes... and based on that, it tries to map them right?
1148511796 M * Bertl yes, I assume so, so we will allow that without further security checks
1148511809 M * Bertl just making a few log entries
1148511823 M * Bertl if that works for X (i.e. you do not need the bcaps)
1148511830 M * jpacheco gezz, bertl's hacks are wicked, lol
1148511851 M * zeng but then the /dev/mem is still there in guest.
1148511854 M * Bertl then we have a good chance to virtualize that properly by assigning certain pci devices to certain guests
1148511917 M * zeng I see... it will be "virtualized" /dev/mem. Got it ...
1148511930 M * Bertl yes, that and the iopl permissions too
1148512041 M * zeng ok. any other information you might need?
1148512089 M * tokkee .oO( sugar and coffee... excellent ;-)
1148512140 M * Bertl hmm, I just saw that the xorg cvs entry you pointed me to only uses /dev/mem on powerpc as it seems
1148512175 M * zeng I can't follow you
1148512206 M * Bertl maybe I missed something, but the lnx_video.c just accesses /dev/mem on powerpc
1148512217 M * Bertl #if defined(__powerpc__) ...
1148512230 M * Bertl on x86 it would do:
1148512236 M * Bertl if (ioperm(0, 1024, 1) || iopl(3)) {
1148512277 M * Bertl so that one does not explain why xorg on x86 accesses /dev/mem
1148512330 M * daniel_hozac hmm?
1148512360 M * daniel_hozac mapVidMem is only !defined(__sparc__), AFAICT.
1148512377 M * zeng well, there are some other "Xorg modules", e.g. int10, see http://webcvs.freedesktop.org/xorg/xserver/xorg/hw/xfree86/os-support/linux/int10/linux.c?rev=1.6&view=markup
1148512412 M * zeng int10 is used to "soft reset" graphic cards
1148512415 M * Bertl ah, some code uses DEV_MEM other code has it literally :)
1148512431 M * zeng yep... open source ;)
1148512508 M * Bertl anyway, it looks like we can do this if xorg plays nice ... so we will test tomorrow, okay?
1148512550 M * zeng ok. time?
1148512580 M * Bertl let's aim at 2000 cet
1148512629 M * zeng cet? do I have cet? ;) sorry, still not use to it. only know GMT +/- offset ;)
1148512650 M * Bertl I guess so, I'm in CET :)
1148512687 M * Bertl I assume you are still in germany, so CET for you too :)
1148512697 M * Wonka CEST
1148512703 M * zeng ok... see you then... need some sleep know...
1148512704 M * Wonka since several weeks
1148512706 M * Bertl yes, correct
1148512710 M * sarnold TZ=Europe/Berlin über alle ;)
1148512724 M * sarnold tschui zeng
1148512733 M * zeng bye all...
1148512737 M * zeng bye Bertl
1148512738 M * Bertl good night!
1148512744 M * zeng bye sarnold
1148512764 P * zeng Leaving
1148513202 J * shedi ~siggi@dsl-og-108-50.du.vortex.is
1148513264 J * FCOJ ~mordur@dsl-og-108-50.du.vortex.is
1148513375 M * FCOJ I am setting up a vsever host on a Dell Poweredge 850 dual core machine. I have 2 Gb RAM. Would anyone please advice me on the size of a swap partition for this setup.
1148513623 M * Methos 2 or 4
1148513626 M * Wonka depends on your useage
1148513682 M * Methos normaly the same as the ram, thats what i use, it really does not matter if you make it larger. because its not going to affect your total disk space much
1148513738 M * Methos I build all my machines with 2xraid mirrors and I place half the swap on one and half on the other. As I understand it its not a good idea to place the swap in a raid
1148513816 M * FCOJ This machine has two hd which I have mirrored.
1148513860 M * FCOJ The debian installer can only find 1 gb of RAM because of no highmem support in install, but it suggests 2.8 gb for swap. 
1148513868 M * doener hm, swap in raid should survive a disk crash, with two swap partitions (one on each drive) when one disk crashes you might loose swap space that is in use... 
1148514224 M * Bertl FCOJ: the swap size depends on the mode you run the kernel in
1148514249 M * Bertl FCOJ: if you disable overcommitment, you will need a lot more swap space, although it will not be really used
1148514471 Q * bonbons Quit: Leaving
1148514703 M * Methos doener: Thats what I always thought which is why in the early years I did put my swap on the raid. But I read a while back that the designer of the software raid did not recommend it
1148514754 Q * shedi Quit: Leaving
1148514766 M * Aiken Methos do you remember the reason?
1148514784 M * Methos no
1148514919 M * Aiken one of my machines uses a file for swap sitting on raid5 instead of swap spread across the 3 disks
1148514941 M * Aiken so far am happy especially when on of the disks that previously had swap on it went bad
1148515000 M * Aiken I am wondering if it was a techical reason for a performance reason not to have swap sitting on raid like that
1148515038 M * Aiken I thought separate swap partition with the same priority was surposed to better than swap on raid0 due to performance
1148515063 M * Aiken as the kernel will effectively stripe the swap anyway in that situation