1307928028 J * Democracy ~pukey@cpe12-98.sweetwaterhsa.com 1307928055 A * Democracy ***GENERAL STRIKE 6/15/11***SOLIDARITY*** 1307928099 P * Democracy 1307928119 J * Democracy ~pukey@cpe12-98.sweetwaterhsa.com 1307928177 A * Democracy ***GENERAL STRIKE 6/15/11***SOLIDARITY***\ 1307928190 Q * Democracy autokilled: This host violated network policy. Mail support@oftc.net if you think this in error. (2011-06-13 01:23:09) 1307930084 M * Bertl off to be now ... have a good one everyone! 1307930088 N * Bertl Bertl_zZ 1307931564 Q * quasisane Quit: leaving 1307931787 J * quasisane ~sanep@c-76-24-80-97.hsd1.nh.comcast.net 1307940904 J * ichavero_ ~ichavero@148.229.9.250 1307946196 J * ncopa ~ncopa@3.203.202.84.customer.cdi.no 1307947565 Q * ichavero_ Quit: This computer has gone to sleep 1307948366 Q * Hunger Quit: _._ 1307949418 J * petzsch ~markus@p57B63907.dip.t-dialin.net 1307951125 J * Hunger ~Hunger@Hunger.hu 1307951966 Q * petzsch Quit: Leaving. 1307952505 J * petzsch ~markus@p57B63907.dip.t-dialin.net 1307952545 J * derjohn_mob ~aj@p4FFD0F28.dip.t-dialin.net 1307953361 Q * petzsch Quit: Leaving. 1307954167 J * bonbons ~bonbons@2001:960:7ab:0:d43a:c43:e235:bcbc 1307954273 J * manana ~mayday090@nat049-252-205-109.tvoe.tv 1307956037 Q * manana Ping timeout: 480 seconds 1307956642 J * manana ~mayday090@nat049-252-205-109.tvoe.tv 1307957117 Q * derjohn_mob Ping timeout: 480 seconds 1307957227 J * LuckyLuke ~luca@host65-83-static.228-95-b.business.telecomitalia.it 1307958467 J * sannes ~ace@cm-84.209.81.224.getinternet.no 1307958972 N * Bertl_zZ Bertl 1307958978 M * Bertl morning folks! 1307959370 Q * manana Ping timeout: 480 seconds 1307959857 J * petzsch ~markus@p57B63907.dip.t-dialin.net 1307959909 J * manana ~mayday090@nat049-252-205-109.tvoe.tv 1307960566 Q * petzsch Quit: Leaving. 1307961402 J * derjohn_mob ~aj@ip-81-210-228-18.unitymediagroup.de 1307962022 M * Bertl daniel_hozac: you still alive? 1307962079 M * daniel_hozac Bertl: hey, sorry. i passed out pretty early Friday and had my bachelor party on Saturday... 1307962103 M * Bertl no problem, was just a little worried 1307962111 M * daniel_hozac i actually have some time right now, if you want to go over the changes. 1307962113 J * FireEgl ~FireEgl@173-16-9-3.client.mchsi.com 1307962139 M * Bertl 'your' bachelor party, so time to get married? 1307962144 M * daniel_hozac indeed 1307962157 M * Bertl well, congratulations then! 1307962164 M * daniel_hozac two weeks from yesterday 1307962193 M * daniel_hozac thank you :) 1307962204 M * Bertl yeah, I have time now as well, so .. how do we proceed? 1307962260 M * daniel_hozac so the reject file is what's left to apply? 1307962306 M * Bertl kind of, I applied the long overdue spinlock fix as well 1307962323 M * daniel_hozac right, i saw that. 1307962347 M * Bertl and I fixed the last two hunks of the reject file as well 1307962378 M * Bertl (and removed them, i.e. the current one is up-to-date) 1307962385 M * daniel_hozac ah 1307962465 M * daniel_hozac so how do you suggest we proceed? 1307962500 M * daniel_hozac do you want patches from me for e.g. COW? 1307962539 M * daniel_hozac the block/genhd.c seems like something we can just drop. 1307962668 M * Bertl yep, that should be handled upstream I guess 1307962688 M * Bertl (obviously missed that one, despite the fact I submitted it upstream :) 1307962735 M * Bertl hmm, but I see the same nonsense in the code ... 1307962795 M * Bertl so I guess it is still useful, will adapt it, no problem there 1307962862 M * Bertl yes, patches for CoW would be fine, of course 1307962930 M * Bertl for the networking part, it seems to me that now flow* is mutable and has separate entries for v4 and v6, so we probably need to adjust the checks and source IP selection to that 1307963132 M * arekm btw, is this true? "# - IPv4 source address selection for multihomed vservers is completely broken 1307963135 M * arekm # meaning routing table is ignored except for default 1307963138 M * arekm " 1307963174 M * Bertl might be true for certain kernels 1307963200 M * arekm 2.6.38.6 1307963357 M * Bertl should be fine at first glance, but I'll suggest 2.6.38.8 1307963391 M * Bertl s/ll/d/ 1307963851 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-ipv4src-fix02.diff 1307963862 M * daniel_hozac should fix it on 2.6.38.6 1307963873 M * daniel_hozac but an upgrade to -rc17 is best. 1307963892 M * Mr_Smoke daniel_hozac: congrats :) 1307963931 M * daniel_hozac Bertl: okay... 1307963997 M * daniel_hozac sounds like there is some work there at least 1307964111 M * arekm we have one more problem with vmlinuz-2.6.38.7-vs2.3.0.37-rc15-8 - cause the same thing as https://lkml.org/lkml/2011/5/23/398 1307964116 M * arekm vanilla is fine 1307964213 M * Bertl well, I blame grsec, do you have any debug info on that issue yet? 1307964597 M * arekm vmlinuz-2.6.38.7-vs2.3.0.37-rc15-8 had no grsec, https://lkml.org/lkml/2011/5/23/398 indeed had grsec but we ruled it out 1307964604 M * Bertl daniel_hozac: and there is also a hidden problem with create/unshare namespaces, as mainline now passes a task_struct for dubious reasons, which we do not have/want for our unshare purpose 1307964612 M * arekm not yet, reproducing requires some binary commercial xilinx software 8-) 1307964676 M * daniel_hozac Bertl: ah... 1307964683 M * daniel_hozac how did you solve that for now? 1307964785 M * daniel_hozac wow, am i reading it correctly? there is now support for entering a namespace in mainline? 1307964798 M * Bertl if it works, yes 1307964821 M * Bertl I have a patch here which doesn't pass the tsk, instead it only passes the required space info 1307964889 M * Bertl the tsk is currently only used to pass the user_ns for uts namespace related cloning ... a strange choice IMHO 1307965043 M * daniel_hozac the user_ns is needed for the uts namespace? 1307965050 M * daniel_hozac that sounds pretty backwards. 1307965052 M * Bertl correct 1307965084 M * Bertl IMHO they still do not know where to put the new namespaces :) 1307965148 M * daniel_hozac uts has been around for so long though... 1307965166 M * Bertl yeah, but user_ns is new 1307965177 M * daniel_hozac yeah... 1307965194 M * Bertl and instead of putting it into the proxy, which would require redesigning a few checks, they attach it to uts 1307965229 M * daniel_hozac really? 1307965269 M * daniel_hozac i thought the user_ns was in the cred. 1307965316 M * Bertl and it is in the uts :) 1307965381 M * daniel_hozac it seems to be to make sure you have the capability in the right user namespace for sethostname... 1307965577 J * petzsch ~markus@p57B66D6C.dip.t-dialin.net 1307966672 J * ichavero_ ~ichavero@148.229.9.250 1307967059 M * Bertl daniel_hozac: okay, uploaded three patches, fixing the genhd, the gettimeofday and probably the nsproxy issues 1307967160 M * Bertl hmm, no, forget the time fix, it is wrong 1307967181 J * DonAlex ~DonAlex@195.149.63.237 1307967190 M * daniel_hozac do we have any known uses of the virtualized time? :) 1307967194 M * DonAlex Afternoon Peeps. 1307967246 M * DonAlex Looked a lot of places for this but how do I suppress vxW warnings about hidden lookups.. They are flooding my syslog files.. 1307967287 A * arekm sets virt_time for each guest but hm, likely never changed it from guest 1307967309 M * DonAlex I understand they are just informational and happen because a user logs into the vserver and starts a shell but must be a way to trim the warnign a little it is reminding me every second anytime anyone is logged in 1307967351 M * Bertl you get hidden lookups from user logging into the guest and starting a shell? 1307967391 M * Bertl what do they look up? 1307967392 M * daniel_hozac Bertl: vx_ns_capable looks incorrect to me. i think we would need to do something similar to what ns_capable does... although i realize it isn't used yet :) 1307967442 M * Bertl well, the vx_capable() check adds an alternative to capable() no? 1307967456 M * Bertl so the vx_ns_capable should do about the same I guess 1307967460 M * daniel_hozac include/linux/utsname.h uses ns, but it should be old_ns 1307967483 M * Bertl ah, good, catch 1307967494 M * daniel_hozac we'd have problems with nested user namespaces inside the guest 1307967509 M * DonAlex Yes. 1307967516 M * daniel_hozac e.g. someone running their own user namespace would suddenly have capability to do things that it would otherwise not. 1307967516 M * DonAlex nothing just log in bash.. 1307967527 M * DonAlex run a ps start pine.. notmal stuff 1307967539 M * daniel_hozac you'd have to show us the warning. 1307967549 M * DonAlex which pastebin you using? 1307967557 M * Bertl daniel_hozac: okay 1307967572 J * ghislain ~AQUEOS@adsl2.aqueos.com 1307967573 M * Bertl DonAlex: paste.linux-vserver.org 1307967599 M * daniel_hozac i think we should make an ns_capable-clone that checks for ccaps if the namespace is the current_vx_info()->nsproxy->user_ns. 1307967617 M * daniel_hozac or... 1307967620 M * DonAlex ok http://paste.linux-vserver.org/19991 1307967650 M * daniel_hozac DonAlex: don't use vserver ... enter 1307967654 M * daniel_hozac your problems will go away. 1307967655 M * Bertl yup 1307967687 M * DonAlex I am running the vserver TO be logged into .. ssh access to mail etc.. 1307967705 M * Bertl ssh into the guest is no problem 1307967726 M * DonAlex You mean don't start Vserve form the command line? 1307967728 M * Bertl but if you really think you need/want to use vserver ... enter, then disable the warning in the kernel config 1307967743 M * DonAlex inside the vserver ? 1307967758 M * DonAlex I thought these messages were coming from the host kernel? 1307967768 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net 1307967798 M * Bertl there is only _one_ kernel 1307967824 M * Bertl you get this message because the kernel is configured to log access to hidden entries 1307967852 M * Bertl in this case it is a devpts entry, and you get that because you bring your pts into the guest via 'vserver ... enter' 1307967867 M * Bertl you can disable this specific warning with CONFIG_VSERVER_WARN_DEVPTS 1307967869 M * DonAlex Yes I understand that. I can see the /dev/pts devices are dynamically shown.. so you are saying I can enter the vserver and suppress the syslog entries? 1307967924 M * Bertl but as we said, if you do not enter the guest _directly_ i.e. use ssh or similar to reach the guest, then you won't get that warning at all 1307968007 M * Bertl (unless boinc scans the pts in a brute forward way that is) 1307968017 M * DonAlex Yes yes.. but that is not and option as the vserver is running like a separate machine.. That CONFIG_VSERVER_WARN_DEVPTS is that an option in the host kernel ? 1307968048 M * Bertl yup, and it is enabled right now (assuming you have a recent enough kernel of course) 1307968098 M * DonAlex 2.6.32 1307968117 M * DonAlex not shwing in the config file.. http://paste.linux-vserver.org/19991 1307968155 M * Bertl not recent enough :) 1307968159 M * DonAlex Crap.. 1307968173 M * daniel_hozac why is it not an option to run sshd in the guest? 1307968178 M * DonAlex *sighs* Ok.. so what's the min version then ? 1307968189 M * DonAlex it is in the guest? 1307968198 M * DonAlex perhaps I am not making myself clear.. 1307968204 M * Marillion start sshd 1307968215 M * Bertl daniel_hozac: I presume the boinc stuff is scanning pts 1307968222 M * DonAlex when you ssh into the vserver and start a shell then I get the warning.. no one logs in no warnings.. 1307968268 M * DonAlex boinc in running on the host not the vserver.. Hmmm 1307968288 M * daniel_hozac Bertl: yeah... 1307968327 M * DonAlex that's a little odd isn't it/ 1307968328 M * DonAlex ? 1307968339 M * Bertl no, not really odd 1307968364 M * DonAlex Well why would a host process flag vxW warnings? 1307968381 M * Bertl because it looks up hidden devpts entries :) 1307968398 M * Bertl but yeah, it is desputable if we should log those at all 1307968418 M * DonAlex Ahh.. hmm and there was I thinking it was coming from inside the vserver.. 1307968419 M * Bertl in any case, even if we fix that right now you would need a newer kernel :) 1307968445 M * DonAlex yeah.. looks like i have no choice.. oh well.. irritating but might as well. 1307968455 M * DonAlex Thanks for your help.. 1307968458 M * Bertl to answer the original question, you can separately disable this warning with 2.6.36.4 and newer 1307968480 M * Bertl but you can disable all the Linux-VServer specific warnings with your kernel as well 1307968482 M * DonAlex love vserver btw.. kudos peeps. keep up the good work. 1307968506 M * Bertl if you update, I'd suggest to go for 2.6.38.8 1307968508 M * DonAlex don't want to disable them all .. never know I might find a bug ;) 1307968532 M * DonAlex ok .. see what I can manage..Cheers.. 1307968537 M * Bertl you're welcome! 1307968544 M * Bertl feel free to hang around 1307968677 M * DonAlex Well gonna take me a bit to pull a new kernel apply patches.. compile etc ;) 1307968697 Q * derjohn_mob Ping timeout: 480 seconds 1307968699 M * DonAlex Debian does nto seem to have a 2.6.38.8-vserver image ;) 1307968717 M * Marillion sure 1307968807 M * Marillion DonAlex: you must it make themselves 1307968826 M * Marillion make-kpkg is your friend 1307968842 M * DonAlex Yeah.. I know.. S'ok not something I have not done before.. Just a hassle you know. ;) 1307968876 M * Marillion some images can you find on backports.debian.org 1307968945 M * DonAlex Yeah might as well make it myself though.. then next time it is easier ;) 1307968978 M * DonAlex Git is really cool like that don;t you think? pull exactly the kernel version you need for your patch.. 1307969004 M * DonAlex so if 2.6.36.4 is needed. it is what Ican get. :) 1307969056 M * daniel_hozac curl is also capable of doing that... 1307969073 M * Bertl and wget :) 1307969075 Q * Romster Quit: Geeks shall inherit properties and methods of object earth. 1307969443 M * Marillion DonAlex: you don't want 2.6.36.4, when i quote Bertl ;) 1307969586 M * Marillion take the .config from grml.org for 2.6.38.7 , wget 2.6.38.8 Kernel, patch it, and build then 1307969593 M * Bertl daniel_hozac: updated the nsproxy-fix01 and added a capinit fix 1307969606 M * daniel_hozac just ran into the capinit thing myself :) 1307969607 M * Marillion DonAlex: the easiest way 1307969722 Q * DonAlex Ping timeout: 480 seconds 1307970143 M * Bertl short break ... bbl 1307970148 N * Bertl Bertl_oO 1307970350 J * DonAlex ~DonAlex@195.149.63.237 1307971529 P * petzsch 1307973775 Q * ghislain Quit: Leaving. 1307974084 M * DonAlex ok ..so while I have you guys here.. should I use the vs2.3.0.37-donotuse or not? 1307974107 M * DonAlex what is the difference between that and vs2.3.0.37-rc17 ? 1307974218 M * daniel_hozac donotuse should be a hint :) 1307974250 M * daniel_hozac you want 2.3.0.37-rc17 1307974259 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com 1307974638 M * DonAlex fair enough.. just I am building against 2.6.39 and well had to ask.. 1307974650 M * daniel_hozac don't use 2.6.39. 1307974654 M * DonAlex oh/ 1307974659 M * DonAlex There are issues? 1307974661 M * mnemoc aim for 3.0 :) 1307974664 M * daniel_hozac there is no patch. 1307974670 M * daniel_hozac so if you want Linux-VServer... 1307974677 M * DonAlex hehe.. quite.. 1307974705 M * DonAlex well look I don't really mind how recent it is but I don't want to break anything else if possible.. 1307974719 M * mnemoc 16:16:19 < daniel_hozac> you want 2.3.0.37-rc17 1307974721 M * DonAlex so will vs2.3.0.37-rc17 work against 2.6.38 ? 1307974771 M * daniel_hozac it is listed as being for 2.6.38.8 for a reason. 1307974826 M * DonAlex so 2.6.38-rc8 then ? 1307974945 M * daniel_hozac no... 1307974947 M * daniel_hozac 2.6.38.8 1307974948 M * DonAlex Just a slight mismatch in the naming convention between git and the tarballs 1307974962 M * daniel_hozac you're probably just looking at the wrong tree entirely. 1307975159 M * DonAlex well looking at the tagged releases.. ? 1307975215 M * daniel_hozac are you looking at the 2.6.38.y.git tree? 1307975331 M * DonAlex Well pulled the linux-2.6 git tree from linus himself.. How I have always done it in the past 1307975348 M * daniel_hozac Linus' tree doesn't contain stable releases. 1307975365 M * daniel_hozac never has. 1307975370 M * DonAlex git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git linux-2.6 1307975398 M * DonAlex No? Hmm odd.. been doing it quite a few years now.. 1307975412 M * DonAlex usually against point releases like 2.6.35 etc.. 1307975435 M * daniel_hozac 2.6.35 isn't a point release. 1307975444 M * DonAlex You know what I mean ;) 1307975547 M * DonAlex So Ihave to pull the tarball then ? 1307975658 M * daniel_hozac or use the correct tree. 1307975799 J * dowdle ~dowdle@scott.coe.montana.edu 1307976032 Q * DonAlex Ping timeout: 480 seconds 1307976576 J * DonAlex ~DonAlex@195.149.63.237 1307977148 Q * ichavero_ Quit: This computer has gone to sleep 1307978488 N * Bertl_oO Bertl 1307978498 M * Bertl ha, I see I missed some entertainment :) 1307979262 M * arekm hehe 1307979862 J * derjohn_mob ~aj@p4FFD0F28.dip.t-dialin.net 1307981368 M * daniel_hozac Bertl: some fixes here... http://people.linux-vserver.org/~dhozac/p/k/ 1307981457 M * Bertl starting with btrfs, I presume? 1307981659 M * daniel_hozac yeah 1307981722 M * Bertl did I really miss a spinlock ... gee, I'm getting old (what else is new :) 1307981729 M * daniel_hozac hehe :) 1307981783 J * pmenier ~pmenier@ACaen-751-1-14-236.w86-220.abo.wanadoo.fr 1307982113 Q * derjohn_mob Remote host closed the connection 1307982419 Q * pmenier 1307984754 Q * nkukard_ Ping timeout: 480 seconds 1307984756 J * derjohn_mob ~aj@p4FFD0F28.dip.t-dialin.net 1307984951 J * nkukard_ ~nkukard@196-209-37-148.dynamic.isadsl.co.za 1307985105 J * _WildPIkachu_ ~nkukard@dsl-247-39-29.telkomadsl.co.za 1307985315 Q * derjohn_mob Ping timeout: 480 seconds 1307985474 Q * nkukard_ Ping timeout: 480 seconds 1307985900 J * petzsch ~markus@dslb-088-075-173-027.pools.arcor-ip.net 1307986544 Q * arekm Ping timeout: 480 seconds 1307986910 J * hparker ~hparker@2001:470:1f0f:32c:beae:c5ff:fe01:b647 1307986945 M * DonAlex Aww crap.. 1307986948 M * DonAlex what is going on now.. 1307986970 M * DonAlex vc_set_sched(): Function not implemented 1307986988 M * DonAlex I upgraded Vserver-utils and compiled them to latest libs.. 1307987016 M * daniel_hozac apparently not to a recent release. 1307987070 M * DonAlex From your website.. ;) how much more recent can it get :P 1307987088 M * DonAlex Should I not use stable then ? 1307987115 M * DonAlex 0.30.215 is not good enough? 1307987245 M * Bertl 0.30.215 is about 3 years? old by now 1307987267 M * DonAlex Triffic.. So stable is in facrt ancient ;) 1307987270 M * Bertl so no, it's not a good choice for any kernel past 2.6.5 or so :) 1307987296 M * DonAlex so 0.30.216-pre2981 good enough? 1307987398 M * Bertl should be fine I guess 1307987404 M * DonAlex We'll see then ;) 1307987561 M * DonAlex hmm seems a little slower starting up.. ? Is that normal ? 1307987649 M * Bertl slower compared to what? 1307987658 M * daniel_hozac and what's slower? 1307987662 M * DonAlex Tohow it used to start up 1307987679 M * DonAlex seems to pause a lot longer between daemons. 1307987686 J * nkukard_ ~nkukard@41-133-248-130.dsl.mweb.co.za 1307987716 M * DonAlex Might just be perception.. I will see how it performs as users log on 1307987724 Q * _WildPIkachu_ Ping timeout: 480 seconds 1307987780 J * Piet_ ~Piet__@659AACC4W.tor-irc.dnsbl.oftc.net 1307987850 M * DonAlex ok well at least the pty messages have stopped ;) now I won't have syslog files 300Mb in size ;) 1307988179 Q * Piet Ping timeout: 480 seconds 1307988305 Q * hparker Quit: Quit 1307988895 Q * DonAlex Ping timeout: 480 seconds 1307989189 Q * petzsch Read error: Connection reset by peer 1307989200 J * petzsch ~markus@dslb-088-075-173-027.pools.arcor-ip.net 1307989456 Q * petzsch Read error: Connection reset by peer 1307989470 J * petzsch ~markus@dslb-088-075-173-027.pools.arcor-ip.net 1307989559 N * Piet_ Piet 1307990035 J * arekm ~arekm@ixion.pld-linux.org 1307990318 Q * Piet Quit: Piet 1307990431 J * Piet ~Piet__@659AACC7O.tor-irc.dnsbl.oftc.net 1307991998 J * ichavero_ ~ichavero@148.229.1.11 1307992079 Q * sung Read error: Operation timed out 1307992193 Q * imcsk8 Ping timeout: 480 seconds 1307992628 J * derjohn_mob aj@88.128.37.124 1307992702 J * sung ~sung@doot.realfuckingnews.com 1307994278 Q * FireEgl Ping timeout: 480 seconds 1307996189 Q * sannes Remote host closed the connection 1307999021 Q * petzsch Quit: Leaving. 1308001121 J * wmp ~wmp@62.244.136.94 1308001126 M * wmp hello 1308001161 M * wmp i want to try Vserver, 2.3 is really experimental? 1308001178 M * wmp can i make on this product server? 1308001183 M * Bertl it is as experimental as the mainline kernel :) 1308001216 M * wmp vserver have own kernel verison? 1308001270 M * Bertl you need a patched kernel, yes 1308001296 M * wmp so, downlaod from kernel, patch and it shoud be stable? 1308001318 M * Bertl as stable as the kernel you patched, yes 1308001343 M * wmp so, lets go ;) 1308001371 M * wmp "vs2.3.0.37-donotuse" donotuse? 1308001391 M * Bertl yep, take 2.6.38.8 for now 1308001399 M * wmp ok 1308001410 M * Bertl the 2.6.39 will happen after 3.0 I guess 1308001429 M * wmp ok 1308001473 M * wmp Bertl: in Vserver is possible to clone my instalation but change username/password and IP? 1308001502 M * wmp i dont want to make from zero my VMs 1308001511 M * Bertl yes, although it's usually simpler/better to create a new guest 1308001535 M * wmp could give me link to howto or manual or else? 1308001542 M * Bertl you can build guests in a variety of ways, sec 1308001580 M * Bertl http://linux-vserver.org/Building_Guest_Systems 1308001608 M * wmp could give me link to howto or manual or else?;) 1308001610 M * Bertl also check 'vserver - build --help' 1308001612 M * wmp nice* :) 1308001621 M * Bertl it will show you even more 1308001795 M * wmp Bertl: in ubuntu on mother i shoud add repository? 1308001852 M * Bertl what you are asking is if we have an ubuntu repository? yes? 1308001894 M * wmp i compiling kernel with vserver patch,and i must add repository to ubutnu? 1308001938 M * Bertl no, the best choice is to build the kernel (as you plan to) and the tools (util-vserver) from scratch/source 1308001952 M * Bertl of course, you can make a deb from both 1308001981 M * wmp ok 1308002007 M * wmp ls 1308002008 M * wmp linux-2.6.38.8 linux-2.6.38.8.tar.bz2 patch-2.6.38.8-vs2.3.0.37-rc17.diff 1308002011 M * wmp patch < patch-2.6.38.8-vs2.3.0.37-rc17.diff 1308002018 M * wmp and i have FAILED 1308002056 Q * bonbons Quit: Leaving 1308002107 M * wmp http://pastebin.com/0kD6HP91 1308002207 M * wmp ok, my fail 1308002744 M * wmp Bertl: i dont know how to set timing frequrency in kernel on host. I shoud set this to 1000Hz or 250Hz? 1308002826 M * Bertl depends on your requirements 1308002836 M * Bertl higher rate means more interactivity 1308002843 M * Bertl lower means more batch performance 1308002850 M * wmp ohhh, in Vserver i want to make game servers 1308002944 M * wmp Bertl: i want to give to my VM public eth0 alias, i shoud use Automatically Assign Loopback IP ? 1308003041 M * Bertl that's unrelated, but you probably want to enable the auto loopback 1308003061 M * Bertl it allows you to use 127.0.0.1 inside a guest without too much config work 1308003084 M * wmp so, if i have public ip to my VM, i dont need this option? 1308003244 M * mnemoc without a loopback you'll have all your services exposed on that public ip 1308003283 M * wmp ok 1308003292 M * wmp and i can set this IP? 1308003368 M * wmp and i need this option: Automatic Single IP Special Casing ? 1308003440 M * Bertl better turn that off in your case 1308003453 M * wmp ok 1308004081 Q * manana Remote host closed the connection 1308005797 M * wmp what is vserver-Rootdir ? 1308005925 M * Bertl the place where you store your guest data 1308005937 M * wmp how i can change this? 1308005940 M * Bertl usually guests go into /vservers/ even 1308005986 M * Bertl you can specify a default at util-vserver buildtime and override that for each guest 1308005987 Q * derjohn_mob Ping timeout: 480 seconds 1308006009 M * wmp --with-vrootdir= this option? 1308006025 M * Bertl yep 1308006043 M * wmp why this isnt in installation directories: 1308006062 M * Bertl the default or the guest override? 1308006068 M * wmp default 1308006088 Q * dowdle Remote host closed the connection 1308006149 M * Bertl no idea 1308006163 M * wmp ;) 1308006199 M * wmp hmm 1308006204 M * wmp i havent: /etc/init.d/vprocunhide 1308006246 M * wmp ok, my fail 1308006540 M * wmp Bertl: after make install install-distribution i shoud do: make install? 1308006589 M * Bertl shouldn't be required ... i.e. install and install-distribution will have installed everything required to the places you configured 1308006601 M * Bertl note that this is /usr/local by default IIRC 1308006612 M * wmp yes, i change this to /usr 1308006619 M * wmp ./configure --sysconfdir=/etc --prefix=/usr --with-vrootdir=/home/vservers --localstatedir=/var 1308006641 M * Bertl it shows the configuration at the end IIRC 1308006668 M * Bertl you can also query it with 'vserver-info - SYSINFO' 1308006716 M * wmp hmmm, i have problem: http://wklej.org/id/546236 1308006722 M * wmp sorry 1308006756 M * wmp http://wklej.org/id/546237 1308006786 Q * ryker Quit: Leaving. 1308006846 M * Bertl upload 'vserver-info - SYSINFO' 1308006860 M * wmp http://pastebin.com/EL647gm6 1308006881 M * Bertl # 1308006883 M * Bertl util-vserver: 0.30.215; Jun 14 2011, 01:08:21 1308006888 M * Bertl too old for everything 1308006896 M * Bertl Use dietlibc: no (you have been warned) 1308006913 M * wmp so, what i shoud doing? 1308006931 M * Bertl get a more recent version, preferably pre2981 1308006938 M * Bertl and built with dietlibc 1308006975 M * wmp and make uninstall? 1308006987 M * Bertl yep, that's the first step 1308007005 J * ryker ~Adium@c-76-16-115-27.hsd1.in.comcast.net 1308007106 M * wmp succeeded ;) 1308007108 M * wmp nice ;) 1308007282 M * wmp [000]* (xfs format failed) [000]. [001]* [000]* (jfs format failed) 1308007294 M * wmp this is no bad? I want to use ext4 1308007314 M * Bertl it mainly means that you do not have the tools or filesystem compiled 1308007332 M * wmp yes, i havent 1308007340 M * Bertl if you are just interested in ext4, specify -Fext4 1308007347 M * wmp ok 1308007543 M * wmp Bertl: in first quest server context can be 1? 1308007563 M * Bertl no, xid/nid=1 is special 1308007571 M * Bertl you may start with 2 though 1308007574 M * wmp ok 1308007584 M * wmp context is machine id? 1308007598 M * Bertl yep, it is the unique identifier for that guest 1308007601 M * wmp ok 1308007632 M * wmp Bertl: and how to give my machine public IP address? 1308007657 N * ensc Guest4493 1308007662 M * Bertl e.g. with the --interface specification at build time, or vie the guest config 1308007667 J * ensc ~irc-ensc@p5DF2F4A1.dip.t-dialin.net 1308007678 M * Bertl http://www.nongnu.org/util-vserver/doc/conf/configuration.html 1308007690 M * wmp so, if i want to give this ip: eth0:5 : 188.165.18.196 1308007734 M * Bertl you either specify --interface 5=eth0:188.165.18.196/24 (or whatever netmask you have) 1308007743 M * wmp i must doing: --interface eth0:188.165.18.196/24 ? 1308007751 M * wmp ok 1308007770 M * Bertl or you set name/dev/ip/prefix to 5/eth0/188.165.18.196/24 1308007779 M * Bertl (in the interfaces directory) 1308007799 M * Bertl note that the eth0:5 is not required, i.e. eth0 is fine 1308007802 M * wmp in installed system? 1308007811 M * wmp guest system*? 1308007838 M * Bertl in the config, the guest itself does not configure the network interfaces 1308007850 M * wmp ok 1308007907 M * wmp Bertl: and how to give RAM limit on this machine? 1308007917 M * wmp i want to give it 4GB 1308007939 M * Bertl http://linux-vserver.org/util-vserver:Cgroups 1308007948 M * Bertl you do that with the cgroup memory controller 1308008005 M * wmp and i don't set this in vserver ? build 1308008012 M * wmp vserver build? * 1308008037 M * Bertl nope, but in the resulting config 1308008045 M * wmp ok 1308008074 Q * Guest4493 Ping timeout: 480 seconds 1308008188 M * wmp Bertl: vserver www build -m debootstrap --context 2 --hostname mysql.sored.pl --interface eth0:188.165.18.196/32 -- --arch amd64 -d lucid -m http://archive.ubuntu.com/ubuntu/ 1308008190 M * wmp yes? 1308008327 M * Bertl except for the /32 it looks good, but if you are sure that you don't want the IP to be on a network, then okay 1308008335 M * wmp /usr/lib/util-vserver/vserver-build: unrecognized option '--arch' 1308008360 M * wmp hmmm, i want to have access from network to this machine ;) 1308008376 M * Bertl the --arch goes last 1308008385 M * wmp and in /etc/network/interfaces i have: netmask 255.255.255.255 1308008388 M * Bertl i.e. it is an option to debootstrap, so 1308008448 M * Bertl vserver www build -m debootstrap --context 2 --hostname mysql.sored.pl --interface eth0:188.165.18.196/32 -- -d lucid -m http://archive.ubuntu.com/ubuntu/ -- --arch amd64 1308008472 M * wmp so, /32 or /24? 1308008488 M * Bertl what is your prefix for that IP? 1308008501 M * wmp http://wklej.org/id/546245 1308008531 M * Bertl that would indeed suggest /32 which is unusual 1308008564 M * Bertl note that if you plan to take that one up/down via the normal network scripts 1308008573 M * Bertl you want to specify nodev for util-vserver 1308008588 M * wmp so, /32? I dont understand you ;) My english is bad 1308008590 M * Bertl otherwise util-vserver will bring that IP up and take it down on guest start/stop 1308008631 M * Bertl yes, if that config you showed me is correct, then /32 1308008678 M * wmp ok 1308008846 M * wmp Bertl: and installed and configured guest i can clonning and run on other ip? 1308008882 M * Bertl sure, just change the config (in interfaces) 1308008886 M * wmp ok 1308008903 M * Bertl or specify a new --interface on build 1308008925 M * wmp O, this is simple ;) 1308008949 M * wmp Vserver is very simple ;) 1308008969 M * Bertl yep, it is simple and modular 1308008982 M * wmp yes, and i like this ;) 1308008998 M * wmp is possible to run VM with VNC? 1308009045 M * Bertl you can run a vnc server inside a guest, but as there is no hardware virtualization in Linux-VServer, there is no guest specific vga card :) 1308009085 M * wmp so, if in guest i crach kernel, i cant look on kernel panic in vnc? 1308009086 M * Bertl so, what you can do is put a vnc based X11 server in a guest, which will give you x11 over vnc 1308009106 M * Bertl the guest and the host share the same kernel, so no 1308009111 M * wmp ok 1308009130 M * Bertl i.e. if the guest causes a kernel panic (unlikely) then it happens on the host 1308009142 M * wmp ughhh... 1308009146 M * wmp no good 1308009219 M * Bertl well, the chances for that to happen are the same as for the normal kernel to panic because an unpriviledged user does something evil 1308009224 M * wmp Bertl: so, if i give to me user root in guest, this can kill my host? 1308009230 M * Bertl nope 1308009246 M * Bertl not unless the kernel is exploitable from normal userspace 1308009261 M * wmp yes 1308009269 M * wmp I: Base system installed successfully. ;) 1308009270 M * wmp nice 1308009277 M * wmp and this system in running? 1308009293 M * Bertl probably not yet, you need to start it first 1308009305 M * Bertl and for ubuntu, you probably want to configure upstart as well 1308009308 M * wmp vserver www start 1308009309 M * wmp RTNETLINK answers: File exists 1308009340 M * Bertl that is because you most likely specified eth0 in the interfaces line, but the IP is already up 1308009350 M * wmp so all is good? 1308009361 M * Bertl be careful when taking down the host as util-vserver will remove the IP from the interface in this case 1308009369 M * Bertl s/host/guest/ 1308009399 M * wmp from /etc/inetwork/interfaces? 1308009441 M * wmp vserver www enter 1308009443 M * wmp 'vserver ... suexec' is supported for running vservers only; aborting... 1308009450 J * derjohn_mob aj@88.128.38.70 1308009458 M * Bertl means the guest terminated shortly after it was started 1308009473 M * Bertl usually because no process (service) was left running inside 1308009477 M * wmp ... 1308009497 M * Bertl try with a debian guest first, this saves you from doing the upstart config 1308009521 M * wmp ok 1308009533 M * wmp can i install this on this same ip? 1308009537 M * Bertl sure 1308009590 M * wmp vserver www-debian build -m debootstrap --context 3 --hostname mysql.sored.pl --interface eth0:188.165.18.196/32 -- -d squeeze -m http://mirror.ovh.net/debian/ -- --arch amd64 1308009591 M * wmp yes?