not necessarily directed towards nikolay, but just picking up this thread with the latest reply i found in my mail reader (ie his). On Wed, 11 Jun 2008 13:56:54 +0300 Nikolay Kichukov <hijacker@oldum.net> wrote: > Someone on IRC once advised that 2.6.25 would not be very well > integrated with vserver. So we are better of waiting until 2.6.26 ;-) just noticed from debian's new package list [1] (really the corresponding rss feed [2]), that debian has enabled vserver builds for kernel 2.6.25. from the build log [3] it appears they are using vs2.3.0.34.11, presumable from herbert's experimental patches [4]. [1] http://packages.debian.org/sid/main/newpkg [2] http://packages.debian.org/sid/main/newpkg?format=rss [3] http://buildd.debian.org/fetch.cgi?pkg=linux-2.6;ver=2.6.25-5;arch=amd64;stamp=1213288874 [4] http://vserver.13thfloor.at/Experimental/ i only refer to debian's choice of vserver patches because i use it as a vote of confidence and presume that somebody has done due-diligence and it's not totally borked. i'll test it on my workstation, but won't put it into "production" on my server (though i might need to on a new server for reasons of hardware support only found in kernels more recent than 2.6.22) > Me myself use debian testing and always compile my own kernel from > eu.kernel.org with the latest experimental patches. on my workstation, for occasional testing purposes, i use whatever debian testing has available (and for 2.6.25 that seems to be experimental, assuming it eventually migrates to testing/lenny). on my server i use the latest stable patch and matching vanilla kernel. > I enjoy the vserver project. it enables some extremely beneficial functionality (user-land virtualization and separation, besides security). corey -- undefined@pobox.com