Date: Thu, 23 Oct 1997 09:44:14 +1000 (EST) From: "Daniel O'Callaghan" <danny@panda.hilink.com.au> To: "George M. Ellenburg" <gme@inspace.net> Cc: Charles Mott <cmott@srv.net>, freebsd-hackers@FreeBSD.ORG, freebsd-isp@FreeBSD.ORG Subject: Re: Password files and virtual IP addresses Message-ID: <Pine.BSF.3.91.971023093155.524P-100000@panda.hilink.com.au> In-Reply-To: <01bcdf41$9f805fb0$f828cccf@caffeine>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 22 Oct 1997, George M. Ellenburg wrote: > | > |I do it by building virtual machines using a hacked inetd(8) which does a > |getsockname(2) followed by a chroot(2) to the virtual machine. The vm > |needs to have ld.so and lib/* etc, etc, etc. It is great for allowing > |telnet access to web sites while preventing customers from peeking at > |each other's stuff. > What about the problem with "username re-use" with the effective UIDs of > the users? Wouldn't 'webmaster@somedomain.com' and > 'webmaster@anotherdomain.com' effectively have the same UID (excluding > Sendmail tables/ tricks)? That is, if both users physically log in to the > server with the user of 'webmaster'. How would you bypass the UIDs > physically recorded in the UFS directory structure? No. You have separate /etc directories for each vm and you can use different uids. Even if the uid is the same from one vm to another, how much does it matter? It only matters in that you, the sysadmin, can't tell who owns a file specifically without doing a pwd to find out which vm you are in. Danny
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.91.971023093155.524P-100000>