Michael Dean wrote:
> could you explain to me, simply, the advantages of lids over selinux?
> Thanks
For a substantial treatment on the relative merits of a broad selection
of intrusion prevention technologies, including several LSM packages
(SELinux, DTE, and SubDomain) consider this book chapter:
"Survivability: Synergizing Security and Reliability". Crispin
Cowan. Book chapter in "Advances in Computers", Marvin V. Zelkowitz
editing, Academic Press, 2004. Buy "Advances in Computers" 60 here
<http://www.elsevier.com/wps/find/bookdescription.cws_home/702750/description>.
Chapter here PDF <http://immunix.com/%7Ecrispin/survivability.pdf>.
Unfortunately, the chapter does not specifically address LIDS. LIDS
borrows some design elements from SubDomain (modeling access control on
relating programs to accesses instead of users) but inverts it:
SubDomain specifies the files a program can access, while (IIRC) LIDS
specifies the programs that may access a file.
I don't know where the notion of listing the files that a program may
access came from. It has antecedents in Janus (Goldberg, Wagner, Thomas,
and Brewer, USENIX Security 1996) and TRON (Bermin, Bourassa, and
Selberg, USENIX 1995).
The latter concept of listing the programs that may access a file was
originally introduced in the PACLs (Program Access Control Lists) paper
in 1990:
@inproceedings{
pacl90,
author = "D.R. Wichers and D.M. Cook and R.A. Olsson and J.
Crossley and P. Kerchen and K. Levitt and R. Lo",
title = "{PACL's: An Access Control List Approach to Anti-viral
Security}",
booktitle = "Proceedings of the 13th National Computer Security
Conference",
address = "Washington, DC",
pages = "340--349",
month = "October 1-4",
year = 1990
}
Which of these concepts is "better" depends on what you are trying to
assure.
Crispin
--
Crispin Cowan, Ph.D. http://immunix.com/~crispin/
CTO, Immunix http://immunix.com
This archive was generated by hypermail 2b30 : Sun Jun 13 2004 - 11:21:08 PDT