1285632583 Q * orzel Quit: Konversation terminated! 1285635896 Q * user__ Remote host closed the connection 1285638666 Q * hello 1285641743 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fee0:9872 1285643085 Q * balbir_ Ping timeout: 480 seconds 1285643736 J * balbir_ ~balbir@122.172.37.122 1285643931 Q * hparker Ping timeout: 480 seconds 1285645322 J * hparker ~hparker@2001:470:1f0f:32c:215:f2ff:fee0:9872 1285647514 Q * hparker Quit: Quit 1285647797 J * ichavero_ ~ichavero@201.144.130.18 1285648937 Q * niki Quit: Ex-Chat 1285651920 N * Bertl_zZ Bertl 1285651924 M * Bertl morning folks! 1285653181 J * ncopa ~ncopa@180.40.189.109.customer.cdi.no 1285654225 Q * balbir_ Ping timeout: 480 seconds 1285654375 Q * ichavero_ Quit: This computer has gone to sleep 1285655400 Q * derjohn_mob Ping timeout: 480 seconds 1285656134 J * balbir_ ~balbir@122.172.37.122 1285656626 J * derjohn_mob ~aj@213.238.45.2 1285657381 J * ghislain ~AQUEOS@adsl2.aqueos.com 1285659721 M * morfoh moin moin 1285660656 M * hijacker morning 1285661893 Q * michal Quit: reboot :) 1285663477 Q * balbir_ Ping timeout: 480 seconds 1285663968 Q * AndrewLee Ping timeout: 480 seconds 1285664133 M * Bertl off for now ... bbl 1285664137 N * Bertl Bertl_oO 1285664434 J * michal ~michal@www.rsbac.org 1285664746 J * balbir_ ~balbir@122.172.37.122 1285667551 M * Chlorek is there any managing tool for cgroup, except manual files editing? 1285667609 J * petzsch ~markus@p4FF45325.dip.t-dialin.net 1285671368 J * manana ~mayday090@84.17.25.149 1285672123 Q * manana Remote host closed the connection 1285673185 J * ntrs ~ntrs@77.28.160.85 1285674243 Q * ntrs Ping timeout: 480 seconds 1285677024 Q * petzsch Quit: Leaving. 1285677801 Q * jrdnyquist Quit: Leaving 1285678129 J * balbir__ ~balbir@122.172.37.122 1285678170 Q * balbir__ Remote host closed the connection 1285678353 J * barismetin ~barismeti@zanzibar.inria.fr 1285678429 Q * barismetin Remote host closed the connection 1285678537 J * jrdnyquist ~jrdnyquis@slayer.caro.net 1285679715 J * petzsch ~markus@p4FF45325.dip.t-dialin.net 1285680434 Q * petzsch Quit: Leaving. 1285680947 Q * `kbad Ping timeout: 480 seconds 1285682157 J * petzsch ~markus@p4FF45325.dip.t-dialin.net 1285683046 Q * michal Ping timeout: 480 seconds 1285683163 J * michal ~michal@www.rsbac.org 1285683188 M * Alteisen hiho everyone 1285683195 M * Alteisen anyone using KSM? 1285683208 M * ard somebody did as test... 1285683217 M * ard running a vserver host in a kvm 1285683231 M * ard because that's the only way to get ksm to work with vserver 1285683239 M * Alteisen ah okay 1285683258 M * Alteisen then that is the reason, why i don't have shared pages 1285683305 M * Alteisen i echoed '1' > run, and waited for some other values to change 1285683321 M * Alteisen so the application has to support KSM 1285683350 M * Alteisen and i thought that it was a kernel feature for everything 1285683591 M * ard well, you have to do a syscall to indicate that certain memory is shareable 1285683608 M * ard it would be neat for java virtual machines or even dalvik on an android 1285683632 M * ard most of the stuff in a vserver setup is already shared if you hashified the setup 1285683667 M * fLoo anyone knows a solution howto manage several vps 1285683672 M * fLoo with just 1 dedicated ip ? 1285683682 M * fLoo so that i can access them from outside each one by one ? 1285683686 M * Alteisen i was just blue eyed, thinking that for example to lighttpd instances would be merged together in RAM automatically by KSM 1285683693 M * fLoo using the same port for every machine 1285683711 M * ard fLoo : for web access I use a proxy like pound 1285683720 M * ard for ssh it's impossible 1285683749 M * ard for rsync access it's something you can configure on the host :-) 1285683755 M * fLoo i thought there might be a trick to tell ssh when a connection comes for "1.tld" its for another machine then for "2.tld" 1285683766 M * fLoo like a dns-based-filtering 1285683773 M * ard you can always do a dnat base on source ip 1285683780 M * ard but that's not what you really want 1285683800 M * fLoo where is the difference if i may ask 1285683808 M * ard I've looked at account based ssh proxying, but it's not that handy 1285683864 M * ard the difference is that you cannot do a DNAT based on a reversed lookup, only on preconfigured ip' 1285683865 M * ard s 1285683881 M * ard for connections to your system you might as well look at openvpn 1285683916 M * fLoo good idea 1285683922 M * fLoo i mean the one with ovpn 1285683935 M * fLoo i forgot about that possibility .) 1285683954 M * ard You can even run openvpn within a vserver (for fun :-) ) 1285683973 M * ard or run the openvpn in a vserver in a seperate network namespace (extra fun and bonuspoints) 1285684023 M * fLoo haha 1285684029 M * fLoo already did that :P 1285684048 M * fLoo you can also run ipv6 inside a guest ;) 1285684096 M * ard well ipv6 works less handy than ipv4 1285684107 M * ard with ipv4 I put all /32 on lo 1285684117 M * daniel_hozac IPv6 is far easier. 1285684125 M * fLoo word 1285684131 M * daniel_hozac you just start a daemon on the host, and your guests get automatically assigned public IPs. 1285684131 M * ard with ipv6 you must put them all on the interface because else the icmp neighbour discovery fails 1285684139 M * fLoo daniel_hozac : not really 1285684153 M * fLoo i am using sixxs and have to assign ipv6-addresses of my sixxs-subnet manually 1285684192 M * daniel_hozac that could work too. 1285684195 M * fLoo like create a new network interface /etc/vservers/[...]/interfaces/(1..9)/ 1285684203 M * fLoo s/could/does :P 1285684210 M * daniel_hozac you don't need to do that, but if that's what works for you. 1285684326 M * fLoo btw Bertl_oO 1285684330 M * fLoo go for 2.6.35.6 :P 1285684449 M * ard daniel_hozac : what daemon actually? 1285684459 A * ard was probably sleeping again :-) 1285684472 M * daniel_hozac vip6-autod 1285686345 J * dowdle ~dowdle@scott.coe.montana.edu 1285687385 M * ghislain i was wondering, how dlimit work with quota ? Can we have the standard quota system for /var and use dlimit for /var/tmp and /var/log without issues ? 1285687406 M * ghislain do it cost in term of cpu/io to have dlimit on for a directory ? 1285687408 M * ghislain :) 1285687454 N * Bertl_oO Bertl 1285687568 M * Bertl first, you cannot use dlimit per directory, only per filesystem mount 1285687597 M * Bertl second, the overhead is similar to quota, as only relative changes are accounted 1285687635 M * Bertl (i.e. no measureable I/O overhead, minimal, probably undetectable CPU overhead) 1285687665 M * Bertl and yes, you can mix and match with quota and filesystem specific accounting 1285689725 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1285690811 J * `kbad ~kyle@apache2-cid.adonis.dreamhost.com 1285690848 M * `kbad has anyone noticed hpet going nuts in 2.6.35.4? 1285690899 M * `kbad it's one of the top things i see in oprofile and it's not even close to the top on 2.6.33.7 1285690938 M * Bertl do you only see that on Linux-VServer patched kernels? 1285691058 M * `kbad i haven't profiled 2.6.35.4 vanilla yet, just curious is anyone else has experienced it 1285691075 M * `kbad i'll check it out and let you know :) 1285691207 M * Bertl okay, thanks! 1285691225 M * Bertl nap attack ... bbl 1285691230 N * Bertl Bertl_zZ 1285691874 Q * derjohn_mob Ping timeout: 480 seconds 1285693107 J * derjohn_mob ~aj@d046038.adsl.hansenet.de 1285693759 M * _are_ daniel_hozac: i repeatet the vclone tests from 2 days ago on an ext4. 1st try 21m, 2nd try 20m. it seems ext4 performs way worse than ext3 in this case. ext3 1st try 2m36s, 2nd try 24s. Quite a surprise to me. 1285693974 Q * petzsch Quit: Leaving. 1285694736 J * ntrs ~ntrs@77.28.161.40 1285697714 J * ntrs_ ~ntrs@77.28.167.17 1285698147 Q * ntrs Ping timeout: 480 seconds 1285699331 J * petzsch ~markus@p4FF45325.dip.t-dialin.net 1285699815 M * harry bzed: not really 1285699815 M * harry ;) 1285699975 Q * Chlorek Quit: - 1285699982 J * Chlorek ~cokolwiek@c.sed.pl 1285700156 M * harry hmm... i lied 1285700159 M * harry i'm making one now :) 1285700349 M * harry grmbl... Bertl_zZ !!1 can you fix the patch! :) 1285700349 M * urbee yo 1285700367 M * harry i could do it, but i'm so dead tired. i can't do it now 1285700377 A * harry will makes mistakes and that would... well... suck :) 1285700394 M * harry so... tomorrow, new patch! for 2.6.32.23 1285700442 A * harry just needs the bertl-patch and it's done :) 1285702188 Q * imcsk8 Quit: Leaving 1285702509 J * niki ~niki@94.145.207.11 1285702520 Q * michal Ping timeout: 480 seconds 1285702628 J * michal ~michal@www.rsbac.org 1285703414 Q * derjohn_mob Ping timeout: 480 seconds 1285703600 M * Alteisen debian squeeze, current vserver-patch, util-vserver 0.30.216-pre2864-2+b1, error: root@alix:/usr/local/sbin# vserver-stat 1285703604 M * Alteisen vc_rlimit_stat(RLIMIT_AS): Invalid argument 1285703618 M * daniel_hozac get recent utils. 1285703649 M * Alteisen is there a backport for squeeze? 1285703678 M * daniel_hozac i have no idea. 1285703718 J * manana ~mayday090@84.17.25.149 1285703821 M * Alteisen will check sid 1285703913 J * petzsch1 ~markus@p4FF45EBF.dip.t-dialin.net 1285703945 M * _are_ lenny/squeeze: 0.30.216~r2772-6, sid 0.30.216-pre2864-2+b1 1285704103 M * _are_ and http://repo.psand.net/pool/main/u/util-vserver-basic/ has more recent ones, Ben wrote a mail abot that on mailingle 1285704109 M * _are_ *mailinglist recently 1285704221 M * Alteisen ah 1285704243 M * Alteisen btw, 0.30.216-pre2864-2+b1 is in squeeze 1285704250 M * Alteisen sid offers nothing else 1285704253 M * _are_ check whether you need the -debian-version of psand stuff, though 1285704279 M * _are_ guess i misunderstood m yaptitude then 1285704285 Q * petzsch Ping timeout: 480 seconds 1285704341 M * Alteisen can i use http://repo.psand.net/pool/main/u/util-vserver-basic/ in /etc/apt/sources.list? 1285704383 M * _are_ slightly different, go a few directories up till you find 'dists' 1285704421 M * _are_ deb http://repo.psand.net/ squeeze main contrib non-free 1285704424 M * _are_ should work 1285704455 M * _are_ if you feel uncomfortable with contrib and nonfree, you should get util-vserver even while leaving it out 1285704705 M * _are_ and if you want to prevent everything to be updated to psand versions, you might try pinning by adding to /etc/apt/preferences: 1285704708 M * _are_ Package: * 1285704708 M * _are_ Pin: origin repo.psand.net 1285704708 M * _are_ Pin-Priority: 1 1285704743 M * _are_ just did so myself as I used the opportunity to as an excuse just to do so ;) 1285705092 M * Alteisen i was just wondering, how to manage the preferences stuff... 1285705175 M * Alteisen wget http://repo.psand.net/pubkey.txt; apt-key add pubkey.txt; rm pubkey.txt 1285705185 M * Alteisen needs also to be done 1285705227 M * _are_ and I have to say, squeeze/sid Packages are 0 byte -> use lenny 1285705670 M * Alteisen hm 1285705696 M * Alteisen deb http://repo.psand.net/ lenny main 1285705708 M * Alteisen that is in sources.list, and the pubkey is in 1285705716 M * Alteisen apt-get update works 1285705748 M * Alteisen root@alix:/etc/apt# cat preferences.d/util-vserver 1285705748 M * Alteisen Package: util-vserver 1285705748 M * Alteisen Pin: origin repo.psand.net 1285705748 M * Alteisen Pin-Priority: 999 1285705768 M * Alteisen but package does not upgrade 1285705915 M * daniel_hozac of course, on squeeze, the util-vserver tarball itself works... 1285706035 M * Alteisen yes, but i had problems with original squeeze util-vserver deb 1285706075 M * Alteisen ah, you are talking about getting the source and compile it? 1285706154 M * Alteisen i would prefer a working package via debian package management 1285706491 M * Alteisen aha 1285706493 M * Alteisen util-vserver - user-space tools for Linux-VServer virtual private servers 1285706493 M * Alteisen util-vserver-basic - most simple util-vserver tools with 1285706493 M * Alteisen util-vserver-basic-debian - most simple util-vserver tools with 1285706505 M * Alteisen package has a different name?! 1285706573 M * daniel_hozac yes 1285706611 M * Alteisen which one is the one i want? 1285706656 M * Alteisen -debian sounds like the debian-fancy-stuff, that one should avoid 1285706721 M * daniel_hozac no 1285706723 M * daniel_hozac just debian paths 1285706809 M * Alteisen the /var/lib/vservers thing? 1285706816 M * daniel_hozac yes 1285706827 Q * petzsch1 Quit: Leaving. 1285706843 M * Alteisen and if i want /vservers instead, i use the package without -debian? 1285706852 M * daniel_hozac yes 1285706861 M * Alteisen :) 1285706997 M * Alteisen now libbeecrypt6 is missing 1285707367 J * petzsch ~markus@p4FF45EBF.dip.t-dialin.net 1285707645 Q * bonbons Quit: Leaving 1285707800 M * Alteisen libbeecrypt7 exists and installs, but the util-vserver-basic still wants the older packages. 1285708095 J * ViRUS ~mp@p579B5233.dip.t-dialin.net 1285708188 A * Alteisen will go ahead tomorrow 1285708210 M * Alteisen good night everyone, and thanks daniel_hozac and _are_ 1285708213 J * derjohn_mob ~aj@d046038.adsl.hansenet.de 1285708302 M * _are_ nn Alteisen 1285708948 J * imcsk8 ~ichavero@201.174.32.227 1285710203 J * cuba33ci_ ~cuba33ci@111-240-219-157.dynamic.hinet.net 1285710308 Q * cuba33ci Ping timeout: 480 seconds 1285711511 M * fLoo is it possible to check the inode size of a vserver inside a guest ? 1285711742 M * fLoo stupid question 1285711744 M * fLoo df -i does ;) 1285711860 Q * derjohn_mob Ping timeout: 480 seconds 1285711926 Q * Chlorek Ping timeout: 480 seconds 1285711976 Q * petzsch Quit: Leaving. 1285712214 N * Bertl_zZ Bertl 1285712217 M * Bertl back now ... 1285712473 J * Chlorek ~cokolwiek@c.sed.pl 1285713525 M * `kbad http://paste.linux-vserver.org/18010 1285713533 M * `kbad interesting? 1285713560 M * `kbad that's 2.6.33.7-vs2.3.0.36.30.4 1285713606 M * Bertl you'll probably get the same with just cgroups 1285713616 M * Bertl (i.e. without the Linux-VServer patch) 1285713627 M * `kbad yeah 1285713640 M * Bertl I never said that cgroups would be efficient :) 1285713652 M * `kbad ;) 1285713691 M * Bertl but it looks like your guest is out of memory 1285713697 Q * manana Remote host closed the connection 1285713726 M * Bertl i.e. it selects processes to kill/remove pages, which of course will hurt performance 1285714350 Q * ntrs_ Ping timeout: 480 seconds 1285714500 M * fLoo re 1285714550 M * fLoo uuh 1285714562 M * fLoo Bertl still no .6 patch ? :P 1285714921 M * Bertl doesn't the .5 apply just fine? 1285714940 M * Bertl (besides the obvious extraversion hunk) 1285715081 M * fLoo yep 1285715094 M * fLoo i just wanna push u to your limits :P 1285715111 M * Bertl so no worries, there will be a new patch soon, but I have to iron out some things first 1285715127 M * fLoo i am still waiting for hard cpu limits ;) 1285715129 Q * ViRUS Quit: If there is Artificial Intelligence, then there's bound to be some artificial stupidity. (Thomas Edison) 1285715138 M * Bertl we too, we too 1285715141 M * fLoo hehe 1285715151 M * fLoo kick the kernel team in their a**es 1285715170 M * Bertl but you might actually tell the folks working on that that they should get it done 1285715189 M * fLoo i did that a thousand times in my deepest dreams -.- 1285715199 M * fLoo didnt help at all 1285715202 M * fLoo obviously;) 1285715224 M * Bertl raising the topic on lkml might help :) 1285715301 M * fLoo haha 1285715311 M * fLoo i am not even signed up ;) 1285715334 M * fLoo webreader is much more comfortable to me 1285715409 M * Bertl fine for me, you are the one asking for hard cpu limits :) 1285716645 Q * ghislain Quit: Leaving. 1285717159 Q * imcsk8 Quit: Leaving 1285717794 Q * dowdle Remote host closed the connection