1166054617 P * stefani I'm Parting (the water) 1166054757 Q * yarihm Quit: Leaving 1166054896 M * Radiance not vserver related but perhaps some one has a quick fix for now; unknown user is sending out spam from a shell server, using 127.0.0.1:25 ... is there a way to prevent this or rather discover him ? 1166054988 M * Hurga do you have shell access there? root even? 1166055120 M * bronson_ It works!! I have a fully up-to-date Edgy+vserver kernel. 1166055127 M * bronson_ I'll post the git tree later on today. 1166055154 M * bronson_ And I guess I'll set up a package archive. 1166055188 M * Hurga bronson_: cool 1166055206 A * Hurga would prefer dapper + ipv6 patch, though :) 1166055235 M * bronson_ Well, it probably wouldn't be too hard to throw ipv6 over the Edgy patch. 1166055298 M * bronson_ I'm using the Edgy kernel because 2.0.2.2 applies 2.6.17. But, I'm running the Edgy kernel on my Dapper server with no problems. 1166055325 M * Radiance Hurga, yes, you have a good idea ? right now i'm not in a position to spend hours or days of monitoring realtime myself 1166055337 M * Hurga lsof 1166055362 M * bronson_ Radiance: jnettop won't tell you where the connections are coming from? 1166055365 M * Hurga bronson_: Now that you mention it... 1166055368 M * Radiance already did but he's not active at this moment or for the last hours 1166055413 M * bronson_ Ah, what you want is a traffic pounce that will page you when he starts up again? :) 1166055480 M * Radiance well, whatever would help for now until i have more time to evaluate the situation 1166055497 M * Hurga bronson_: I'm currently running 2.6.17-11.30vs2.0.2.1 on dapper, uniklu package. 1166055515 M * Radiance last night it was 18k mails, the night before it was 8k mails 1166055534 M * Radiance it doesn't really generate spikes in traffic among the usual 1166055540 M * bronson_ I tried the uniklu package too but it bugged me that it was compiled without dynamic context IDs. 1166055661 M * daniel_hozac you shouldn't use dynamic contexts anyway. 1166055661 M * bronson_ daniel_hozac: no? How come? 1166055661 M * daniel_hozac they're deprecated and are going away. 1166055661 M * daniel_hozac and they cause problems. 1166055661 M * bronson_ So you'll have to manually assign a context ID for each guest? 1166055661 M * daniel_hozac yes. 1166055661 M * bronson_ gah. That's unfortunate. 1166055661 M * bronson_ How did they cause problems? 1166055703 Q * bronson_ Remote host closed the connection 1166055718 J * bronson ~bronson@adsl-75-36-147-248.dsl.pltn13.sbcglobal.net 1166055724 M * daniel_hozac it breaks xid tagging, and with networking checking against the nid (2.1+ only, IIRC), vserver ... enter won't work anymore. 1166055727 M * bronson First time I've seen xchat crash. 1166055823 Q * borgfish Ping timeout: 480 seconds 1166055866 J * chaotika_ chaotika@ccc2.rbg.informatik.tu-darmstadt.de 1166055927 M * Hurga Radiance: activate identd - I hope your MTA supports ident 1166055963 M * Hurga identd is made exactly to track that kind of abuse. 1166055965 M * bronson Trying to understand... xid tagging is to put limits on diskspace usage and nid is for network usage? 1166055986 Q * chaotika Ping timeout: 480 seconds 1166055996 M * bronson And if you have dynamic context IDs, then on-disk xids won't work. 1166056033 M * bronson Am I close? First time I've heard of xids and nids. 1166056050 M * bronson Er, did I miss anything before "it breaks xid tagging..."? 1166056127 M * raw hum, i have problems with syslog-ng: it can't open /proc/kmsg (operation not permitted (1)) .. any idea? 1166056418 M * bronson raw: http://www.paul.sladen.org/vserver/archives/200211/0078.html 1166056454 M * bronson Or, better: http://www.paul.sladen.org/vserver/archives/200211/0083.html 1166056581 M * raw ? 1166056593 M * raw comment what out? 1166056608 M * Hurga the oops decoding I guess. 1166056640 M * Radiance Hurga, but even if ident would work then it would only show the postfix user doing the delivery ? 1166056682 M * raw ah got it, i write it to the wiki 1166056691 A * sladen looks up 1166056776 M * Hurga Radiance: I'm not sure if I get the situation. I thought you had a local user spamming a local postfix on port 25. I don't really know if ostfix tracks ident on incoming port 25 connections (my ancient sendmail does), but in that case it should work. 1166056921 M * Radiance well, a rogue user is using most likely a ssh tunnel to localhost 25 sending out the spam...so what i meant is that even if ident would work with postfix then it would only show the uid of the postfix user delivering the mail to its targets 1166056986 M * Hurga in any case, ident is unlikely to work through a tunnel. hm. 1166057112 M * daniel_hozac bronson: no you didn't miss anything, and that's pretty close, yes. 1166057130 M * daniel_hozac the xid is the context id, much like a user id, but for the entire guest. 1166057147 M * daniel_hozac and the nid is the network context id, which identifies the network parts of the guest. 1166057162 M * bronson OK, that makes sense. 1166057184 M * bronson I'll get used to maintaining context IDs... I've been doing it until now. 1166057212 M * daniel_hozac you already give guests unique IP addresses, no? 1166057227 M * daniel_hozac personally i use the last octet as the xid for all my production guests. 1166057229 M * bronson Right now, if a guest specifies a xid it's used, otherwise a dynamic one is assigned? And soon, instead of a dynamic xid, it will just bomb out? 1166057260 M * daniel_hozac yep. 1166057275 M * daniel_hozac (just like it bombs out when you disable dynamic contexts right now) 1166057299 M * Radiance Hurga, i have to get some sleep now, but i'll try to find him or a solution which will find him too :) 1166057303 M * Radiance g'night all 1166057357 M * bronson Sounds good. My only hesitation is that a few months ago, when I first tried linux-vserver, I couldn't figure out why my guests weren't starting up... conflicting xids, of course. But the error message was useless. 1166057398 M * bronson I thought it was networking problems... tearing hair out... :) Bertl figured it out. 1166057409 M * daniel_hozac should've said something about "already running". 1166057569 M * bronson I agree. :) It was a very old linux-vserver patch. 1166057585 M * daniel_hozac 2.4? 1166057596 M * bronson Here's a small present for the wiki: http://wiki.u32.net/index.php?title=Linux-VServer/InitStyles 1166057612 J * dmax ~semaj@81.193.58.182 1166057619 M * bronson Update of http://oldwiki.linux-vserver.org/InitStyles Where should it go in the new one? 1166057641 M * daniel_hozac hmm, there are more initstyles though. 1166057644 M * bronson daniel_hozac: no, early 2.6. 1166057650 M * bronson Not sure exactly. 1166057657 M * bronson I'll add them. 1166057662 M * daniel_hozac gentoo and minit. 1166057672 M * bronson The old page said gentoo was deprecated. 1166057702 M * daniel_hozac it was :) 1166057715 M * daniel_hozac Hollow reimplemented it for 0.30.212+. 1166057733 M * bronson Heh. 1166057912 M * daniel_hozac i guess it should go in http://linux-vserver.org/util-vserver:InitStyles though. 1166057943 M * bronson OK, I'll put it there. 1166057959 M * bronson gentoo is like sysv... Do you know offhand what the difference is? 1166057968 J * s0undt3ch ~s0undt3ch@81.193.58.182 1166057996 M * daniel_hozac from sysv? it works on Gentoo :) 1166058023 M * bronson should have guessed. 1166058138 M * bronson ...and do you know the difference between minit and init? All I can find is that minit somehow helps if you're seeing X keyboard issues... 1166058180 M * bronson Oh, http://www.fefe.de/minit/ 1166058237 M * bronson So I would need to install minit on the host before using a minit-based guest? 1166058306 M * daniel_hozac no. 1166058326 M * bronson included with util-vserver. 1166058334 M * daniel_hozac no, your guest would need it installed. 1166058341 M * bronson Ah, install on guest. 1166058352 M * bronson So, build guest with sysv, install minit, then specify minit. Cool. 1166058386 M * daniel_hozac or you know, build a guest with minit. :) 1166058395 M * bronson er, right. 1166058405 M * bronson I'm spoiled by debootstrap. 1166058439 M * daniel_hozac what? 1166058444 M * daniel_hozac surely Debian has minit packages. 1166058461 M * bronson Yep, but dboostrap makes it so I don't ever have to think about defaults. 1166058468 M * bronson Somebody else picks everything for me. :) 1166058487 J * gerrit ~gerrit@c-67-160-146-170.hsd1.or.comcast.net 1166058509 M * daniel_hozac heh 1166058683 M * bronson OK, I'm done playing with it: http://linux-vserver.org/util-vserver:InitStyles 1166058807 M * daniel_hozac actually, the part about "the word isn't recognized" isn't true. 1166058826 M * daniel_hozac and i guess "withint" is a typo :) 1166058868 M * bronson doh. 1166059029 M * bronson OK, fixed. I meant to verify that and forgot. :) 1166059657 M * johnny now what's this about? * Was deprecated but has been reinstated around util-vserver 0.30.212. 1166059661 M * johnny for gentoo init 1166059670 M * daniel_hozac yep. 1166059760 M * daniel_hozac i think it requires baselayout 1.13 though. 1166059814 M * johnny hmm 1166059825 M * johnny i just installed baselayout 1.13 on my desktop machine 1166059832 M * johnny i'll ask hollow 1166059944 J * MrX ~urk@218.208.209.30 1166060354 J * Aiken_ ~james@tooax6-043.dialup.optusnet.com.au 1166060429 Q * johnny Quit: Ex-Chat 1166060681 Q * Aiken Ping timeout: 480 seconds 1166062795 J * marcfi1 ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1166063172 Q * RichyF Quit: ( www.nnscript.de :: NoNameScript 3.72 :: www.XLhost.de ) 1166063671 J * _dmax ~semaj@81.193.56.46 1166063679 J * s0undt3ch_ ~s0undt3ch@81.193.56.46 1166063756 Q * dmax Ping timeout: 480 seconds 1166063764 N * _dmax dmax 1166063787 Q * s0undt3ch Ping timeout: 480 seconds 1166063787 N * s0undt3ch_ s0undt3ch 1166064268 J * _dmax ~semaj@81.193.56.46 1166064276 Q * dmax Ping timeout: 480 seconds 1166064284 N * _dmax dmax 1166064337 Q * s0undt3ch Ping timeout: 480 seconds 1166064405 J * s0undt3ch ~s0undt3ch@81.193.56.46 1166064880 Q * Hurga Remote host closed the connection 1166066371 J * _dmax ~semaj@81.193.56.216 1166066431 Q * dmax Ping timeout: 480 seconds 1166066432 N * _dmax dmax 1166066542 Q * s0undt3ch Ping timeout: 480 seconds 1166066734 M * marcfi1 daniel_hozac: you awake? 1166067743 Q * Rich_Estill charon.oftc.net cation.oftc.net 1166067743 Q * renihs charon.oftc.net cation.oftc.net 1166067743 Q * Vudumen charon.oftc.net cation.oftc.net 1166067743 Q * yang charon.oftc.net cation.oftc.net 1166067868 J * yang ~yang@yang.sponsor.oftc.net 1166067868 J * Vudumen c28212698f@perverz.hu 1166067868 J * renihs ~penguin@83-65-34-34.arsenal.xdsl-line.inode.at 1166067868 J * Rich_Estill ~restill@c-24-11-195-139.hsd1.mi.comcast.net 1166070707 J * SNy_ 68cbdd900c@bmx-chemnitz.de 1166070788 Q * SNy Read error: Connection reset by peer 1166072205 Q * lilalinux Remote host closed the connection 1166073445 P * marcfi1 1166073687 N * Bertl_oO Bertl 1166075540 Q * FireEgl Remote host closed the connection 1166077126 J * marcfi1 ~mef@c-68-39-177-97.hsd1.nj.comcast.net 1166077142 M * marcfi1 hey bertl... 1166077149 M * marcfi1 are you up early or late. :) 1166077163 M * Bertl late .. what's up? 1166077179 M * marcfi1 nothing... everything is fine. 1166077194 M * marcfi1 just waiting for a machine to come up with rc9 1166077203 M * Bertl ah, good :) 1166077249 M * marcfi1 What is your opinion of the KVM support in 2.6.20 (http://www.heise.de/english/newsticker/news/82344/from/rss09) 1166077267 M * marcfi1 KVM = kernel virtual machine support for Intel VT and AMD Pacifica? 1166077268 M * Bertl well, hopefully a common base for VMware and Xen 1166077273 J * wind_lee ~wind@58.37.76.74 1166077284 M * Bertl welcome wind_lee! 1166077307 M * wind_lee :) 1166077307 M * marcfi1 that wasn't my take away... 1166077421 M * marcfi1 What I took away from this was that KVM made Linux itself the hypervisor, thereby eliminating the need for a Xen/VMware-ESX on VT/Pacifica processors. 1166077566 M * Bertl well, yes and no 1166077607 M * Bertl with PC and VT most parts of xen are obsolete 1166077754 M * marcfi1 true... except for I/O virtualization. 1166077763 M * marcfi1 or lack thereof. 1166077794 M * marcfi1 You still need to go through the equivalent of dom0 to safely access I/O devices, such as the disk drive, ether, etc. 1166077852 M * marcfi1 And presumably Xen's CPU scheduler is used to multiplex the processor. 1166077872 M * Bertl yep, that's the part which will stay xen specific 1166077910 M * marcfi1 I think of Xen as a microkernel, there will be a dom0 that does the heavy lifting wrt I/O, and then there are guest domains. 1166077953 M * marcfi1 However, with the KVM support then the linux that normally would be dom0 under Xen will itself be the hypervisor. 1166078019 M * marcfi1 In some twisted sense, this is like the switch that Windows NT made from 3.51 -> 4.0. I.e., from a microkernel based solution whose equivalent of dom0 was win32 to a monolithic solution with integrated win32kernel. 1166078070 M * Bertl don't know much about windows, except that it isn't deterministic and you cannot fix it :) 1166078140 M * marcfi1 well, going back to Linux, I think this means that with KVM the Linux scheduler itself will be used to schedule the various guest VMs. 1166078221 M * marcfi1 I believe these VMs will appear just like processes to the KVM enabled Linux. 1166078255 M * Bertl yes, that might be the ultimate goal 1166078260 M * marcfi1 If that is the case, then I suppose Linux will need a nice scheduler that can do fair share, guarantees, etc. .... hmm, sounds familiar. 1166078294 M * marcfi1 he he he :) 1166078935 M * Hollow morning 1166078947 M * Bertl good morning Hollow! 1166078977 M * Hollow morning Bertl, i upgraded to 2.3.0.3 yesterday, seems to work fine, but didn脛t run tests yet... though i noticed legacy is still in 2.3 ..? 1166079000 M * Bertl only small parts left 1166079058 M * Hollow but didn't you remove it in 2.2 already? it seems to be completely available in 2.3 still, even the config options etc 1166079092 M * Bertl that's because 2.2.x was cleaned up, 2.3 is still more in sync with 2.1 1166079123 M * Hollow ah i see 1166079139 M * Hollow but the new ata drivers in 2.6.19 rock :) 1166079227 M * Hollow could you give me brief update for the dmap feature? how is it supposed to work? 1166079244 J * FireEgl ~FireEgl@adsl-61-147-76.bhm.bellsouth.net 1166079257 M * Bertl it is a mapping function for device special nodes 1166079275 M * Bertl basically you can map /dev/zero to provide random data 1166079326 M * Bertl at the same time, you can control access and creation via this table 1166079391 M * Hollow so you can configure allowed devices per guest? 1166079430 M * Bertl yes, creation, opening, and remapping 1166079479 M * Hollow sounds nice.. some people already asked for "guest specific vprocunhide" in the past .. 1166079602 M * Bertl yes, that is something we might do in the future too 1166079663 M * Hollow well, if the device map feature is finished, that's up to util-vserver/vcd, no? 1166079714 M * Bertl no, as it only addresses special devices, not proc entries 1166079716 M * Hollow or does dmap not completely replace the vprocunhide mechainsm? 1166079727 M * Hollow ah right 1166079744 M * Hollow i meant the dev files, not proc 1166079757 M * Hollow still tired, just woke up :) 1166080937 Q * wind_lee Quit: 没有陌生，没有距离，我们相聚在网络，用彼此真诚的心默默交流----见证你我的网络情缘，Chatkey中文聊天软件 http://www.chatkey.com 1166081611 J * meandtheshell ~markus@85-124-233-240.work.xdsl-line.inode.at 1166081984 M * marcfi1 well, time for bed here... cheers and goodnight. 1166082017 M * Bertl yeah, guess I'll head off too ... 1166082028 M * Bertl have a nice one everyone ... cya later! 1166082039 P * marcfi1 1166082063 N * Bertl Bertl_zZ 1166082644 J * dna ~naucki@64-245-dsl.kielnet.net 1166082679 M * Borg- erm.... 1166082686 M * Borg- where is the op? 1166083004 M * bronson Borg-: no need! 1166083237 Q * shedi Quit: Leaving 1166083444 Q * dna Quit: Verlassend 1166083748 Q * Wonka charon.oftc.net arion.oftc.net 1166083748 Q * tso charon.oftc.net arion.oftc.net 1166083748 Q * michal` charon.oftc.net arion.oftc.net 1166083748 Q * Roey charon.oftc.net arion.oftc.net 1166083748 Q * meebey charon.oftc.net arion.oftc.net 1166083748 Q * h01ger charon.oftc.net arion.oftc.net 1166083748 Q * waldi charon.oftc.net arion.oftc.net 1166083781 J * Wonka produziert@chaos.in-kiel.de 1166083807 J * meebey meebey@booster.qnetp.net 1166083956 J * michal` ~michal@www.rsbac.org 1166084459 M * Borg- bronson: nah... watch wind_lee 1166084468 M * Borg- I would ban that asshole already 1166084481 J * waldi ~waldi@bblank.thinkmo.de 1166084513 M * bronson Huh, haven't seen him. Guess I've been lucky. 1166084528 J * Roey ~katz@h-69-3-4-130.mclnva23.covad.net 1166084709 J * h01ger ~holger@socket.layer-acht.org 1166085328 J * s0undt3ch ~s0undt3ch@81.193.56.216 1166085548 J * cdrx ~legoater@cimai.net4.nerim.net 1166085727 Q * Adrinael Remote host closed the connection 1166086033 J * Adrinael adrinael@st12-127.tky.hut.fi 1166087086 Q * raw Read error: Connection reset by peer 1166088127 J * witchdoc ~witchdoc@d463c2b3.datahighways.de 1166088134 M * witchdoc good moring 1166088165 M * witchdoc hmpf i use the quit simple methode to copy a vserver from one host to another host 1166088174 M * witchdoc err 1166088180 M * witchdoc i search for it 1166088191 M * Borg- howdy waldi ;) 1166088202 M * Borg- waldi: world is small I see.. 1166088222 M * witchdoc a simple "tar cvf vs.tar /var/lib/vservers/vs" brings strange errors (mysqld doesnt run, postfix doesnt chroot) 1166088276 M * Borg- witchdoc: first use --atime-preserve .. just in case 1166088287 M * Borg- next.. use -p to preserve modes and UID:GID 1166088295 M * Borg- and it should work.. no reason why it shouldnt 1166088318 M * witchdoc hmmm 1166088399 M * Borg- and I think its better to do cd /var/lib/vservers/vs && tar -czf vs.tar --atime-preserve * 1166088420 M * Borg- so you can extrart root vserver in another location.. just like that 1166088507 M * witchdoc nice idea 1166088558 J * Johnsie ~jdlewis@jdlewis.org 1166088613 Q * Johnnie Ping timeout: 480 seconds 1166088936 M * witchdoc i try this (tar -czfp --atime-preserve) , think that some flags with my "tar cvf" are lost 1166089073 M * Borg- -p is used with -x 1166089079 M * Borg- --atime-preserve is used with -c 1166089130 M * witchdoc ah, okay, -p is for extract ... 1166089150 M * Borg- yeah :) just read man tar to confirm things.. maybe you find some additional switches to make things even better 1166089194 M * witchdoc yeap, just to know where to read ;-) 1166090890 Q * cdrx Ping timeout: 480 seconds 1166091221 M * witchdoc Borg-: error: 'Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' 1166091234 M * witchdoc hmpff 1166091258 M * Borg- first.. did you stopped vserver.. tared it.. untared at location 1166091262 M * Borg- then 1166091274 M * witchdoc yes i did it so 1166091289 M * Borg- find /var/tmp -type s -exec rm -f {} \; 1166091294 M * Borg- find /var/run -type s -exec rm -f {} \; 1166091300 M * Borg- and maybe some other dirs? 1166091316 M * Borg- you need to remove temporary crap from /var /tmp dirs I think 1166091324 J * tso ~tso@241-059.dyn-fa.pool.ew.hu 1166091340 M * Borg- also ls -l /var on started vserver.. maybe some permission issues? 1166091389 M * Borg- I really do not see a reason why it doesnt work.. you moving it to same vserver kernel? 1166091529 M * witchdoc Borg-: no, its a littlebit newer 1166091542 M * witchdoc moment i take a look at the versions 1166091559 M * witchdoc old one 2.6.17.11-vs2.0.2 1166091580 M * witchdoc new one 2.6.18.5-vs2.0.2.2-rc9 1166091622 M * witchdoc Borg-: the mysql problem, i solved with simpel reinstall mysql-server 1166091645 M * witchdoc postsuper: fatal: scan_dir_push: open directory defer: Permission denied 1166091651 M * witchdoc strange ... 1166091735 M * Borg- hmm 1166091742 M * Borg- did you tar -czf /etc/vserver/vs too? 1166091757 M * Borg- or created new one from scrath? 1166091767 M * witchdoc i create an skeleton 1166091783 M * witchdoc an tar it into the skeleton v-root 1166091807 M * Borg- ahh.. 1166091818 M * Borg- hmm weird. 1166091829 M * Borg- no clue for now 1166092120 J * shedi ~siggi@dsl-149-109-85.hive.is 1166092523 M * witchdoc hmmm 1166092527 M * witchdoc postfix/postfix-script: warning: not owned by postfix: /var/spool/postfix/incoming/1 1166093077 M * witchdoc and thats true, it was own by root ... i must make an failture with the tarball 1166093492 M * witchdoc damn ... uid/gid trouble 1166093766 Q * Aiken_ Quit: Leaving 1166093806 Q * Arcticfox Ping timeout: 480 seconds 1166094211 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net 1166095894 J * dna ~naucki@44-234-dsl.kielnet.net 1166097653 Q * ensc Killed (NickServ (GHOST command used by ensc_)) 1166097662 J * ensc ~irc-ensc@p54B4ED7C.dip.t-dialin.net 1166099504 Q * shedi Quit: Leaving 1166099777 J * borgfish ~bla@141.12.9.118 1166100416 J * Arcticfox ~arcticfox@static243-170-124.adsl.no 1166100439 A * Arcticfox curses the weather 1166100444 J * lilalinux ~plasma@80.69.41.2 1166102013 J * chand ~chand@m167.net81-64-156.noos.fr 1166102103 Q * chand 1166102118 J * chand ~chand@m167.net81-64-156.noos.fr 1166102455 P * marcfiu 1166104703 Q * kugg Read error: Connection reset by peer 1166106092 Q * FireEgl Remote host closed the connection 1166107772 Q * cdrx Quit: Leaving 1166108216 J * dreamind ~dreamind@C2107.campino.wh.tu-darmstadt.de 1166108237 M * dreamind Hi folks :) 1166108256 M * daniel_hozac hello 1166108267 M * dreamind hi daniel_hozac :) 1166109689 J * cdrx ~legoater@cap31-3-82-227-199-249.fbx.proxad.net 1166109741 M * witchdoc bye 1166109746 Q * witchdoc Quit: cya 1166110451 M * daniel_hozac Hollow: could we just block every wiki change that was