Re: [SLUG] Re: Virus Scanner

2011-04-04 Thread Jon and Hannah
On Sun, 3 Apr 2011 09:36:07 pm you wrote:
 I think it is going to come back and bite the Linux community if we go
 via the line that we are immune to viruses, like Apple users have done
 for many years.

Wasn't there a virus for unix systems a few years ago that slowed almost the 
entire internet to an almost halt?
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-04 Thread Erik de Castro Lopo
Jon and Hannah wrote:

 On Sun, 3 Apr 2011 09:36:07 pm you wrote:
  I think it is going to come back and bite the Linux community if we go
  via the line that we are immune to viruses, like Apple users have done
  for many years.
 
 Wasn't there a virus for unix systems a few years ago that slowed almost the 
 entire internet to an almost halt?

A reasonable telling of the history:

http://www.freerepublic.com/focus/f-chat/2634313/posts

That history lacks a telling of whether these were zero-day exploits
or exploits against old versions of software which could have easily
been prevented by keeping systems up-to-date.

Erik
-- 
--
Erik de Castro Lopo
http://www.mega-nerd.com/
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-04 Thread Troy Rollo
On Monday 04 April 2011 21:14:51 Erik de Castro Lopo wrote:
 Jon and Hannah wrote:
  On Sun, 3 Apr 2011 09:36:07 pm you wrote:
   I think it is going to come back and bite the Linux community if we go
   via the line that we are immune to viruses, like Apple users have done
   for many years.
 
  Wasn't there a virus for unix systems a few years ago that slowed almost
  the entire internet to an almost halt?

 A reasonable telling of the history:

 http://www.freerepublic.com/focus/f-chat/2634313/posts

 That history lacks a telling of whether these were zero-day exploits
 or exploits against old versions of software which could have easily
 been prevented by keeping systems up-to-date.

I suspect the original reference (with a generous meaning given to a few) 
was to this: http://en.wikipedia.org/wiki/Morris_worm
-- 

Regards, 
Troy Rollo 
Solicitor 
Parry Carroll 
Commercial Lawyers 
Direct: (02) 8257 3177 
Fax: (02) 9221 1375 
Switch: (02) 9221 3899 
E-mail: t...@parrycarroll.com.au 
Web: www.parrycarroll.com.au 

Liability limited by a scheme approved under Professional Standards 
Legislation 

This message and any attachments are confidential to Parry Carroll. If you 
have received it my mistake, please let us know by reply and then delete it 
from your system. You must not copy the message, alter it or disclose its 
contents to anyone. Thank you.

-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


[SLUG] Re: Virus Scanner

2011-04-03 Thread Richard Ibbotson
On Sunday 03 April 2011 11:41:34 Chris Allen wrote:
 What is the current consensus on using a virus scanner for Linux
 (specifically Ubuntu 10.10)?
 When I last asked this (about 2 years ago) the general opinion was,
 waste of time, Linux did' need it
 If scanners are recommended now, which is the favourite?

I always scan e-mail both ways.  I send e-mail with Exim.  Clamav and 
other anti-virus software is installed.  My firewall has a proxy which 
scans downloading web pages for viruses and worms.  This does not slow 
down my network connection.  I don't know about Australia but in 
England it's considered to be criminal not to scan for viruses.  The 
legal stuff doesn't say that it is criminal but it implies that 
something is wrong if virus scanning software is not installed on a 
workstation or somewhere else...

Firewalls

http://sleepypenguin.homelinux.org/blog/?page_id=174

-- 
Richard
http://www.sheflug.org.uk
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-03 Thread Morgan Storey
I think it is going to come back and bite the Linux community if we go
via the line that we are immune to viruses, like Apple users have done
for many years. Now there are Mac viruses appearing and a mac botnet.
Clamav and common sense can go a long way, don't install or run things
from an unknown source, and scan your system on occasion to be safe.
All the security in the world can not stop a misguided user from doing
something they shouldn't like giving a file execute permissions and
running it with sudo.

On Sun, Apr 3, 2011 at 8:57 PM, Richard Ibbotson
richard.ibbot...@googlemail.com wrote:
 On Sunday 03 April 2011 11:41:34 Chris Allen wrote:
 What is the current consensus on using a virus scanner for Linux
 (specifically Ubuntu 10.10)?
 When I last asked this (about 2 years ago) the general opinion was,
 waste of time, Linux did' need it
 If scanners are recommended now, which is the favourite?

 I always scan e-mail both ways.  I send e-mail with Exim.  Clamav and
 other anti-virus software is installed.  My firewall has a proxy which
 scans downloading web pages for viruses and worms.  This does not slow
 down my network connection.  I don't know about Australia but in
 England it's considered to be criminal not to scan for viruses.  The
 legal stuff doesn't say that it is criminal but it implies that
 something is wrong if virus scanning software is not installed on a
 workstation or somewhere else...

 Firewalls

 http://sleepypenguin.homelinux.org/blog/?page_id=174

 --
 Richard
 http://www.sheflug.org.uk
 --
 SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
 Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html

--
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-03 Thread Erik de Castro Lopo
Morgan Storey wrote:

 I think it is going to come back and bite the Linux community if we go
 via the line that we are immune to viruses,

Unfortunately, the alternative, virus scanners that look for
particular virus signatures is nothing more than security
theatre.

Firstly, inew viruses can be written so fast that the virus
detection engines have absolutely no way of keep up.

Secondly, self modifing polymorphic virses have been around for
at least a  decade. That means for an instruction set like x86,
for any set of 1000 instructions there are probably 10s of
thousands of ways to rewrite those 1000 instructions so they
behave the same but won't be detected by a scanner that detects
the original.

The *only* 100% safe way to guard against viruses to fix all the
security holes that viruses exploit. That means better coding 
practices.

Erik
-- 
--
Erik de Castro Lopo
http://www.mega-nerd.com/
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-03 Thread Morgan Storey
On Mon, Apr 4, 2011 at 6:46 AM, Erik de Castro Lopo
mle+s...@mega-nerd.com wrote:
 Morgan Storey wrote:

 I think it is going to come back and bite the Linux community if we go
 via the line that we are immune to viruses,

 Unfortunately, the alternative, virus scanners that look for
 particular virus signatures is nothing more than security
 theatre.

I agree that to an extent it is security theatre, it isn't 100%
foolproof, though little is. It does get the lowest common
denominator, the most common of viruses, that come along with screen
savers (gnome-look.org) and mouse cursor addon's (this is a windows
user thing usually).

 Firstly, inew viruses can be written so fast that the virus
 detection engines have absolutely no way of keep up.

Having worked with the clamav guys, this is true that new viruses are
coming out all the time, but they update their signatures so quickly
it is at least of some value. The record I have seen was two hours
from my submission of a sample to them publicly releasing a signature
update.

 Secondly, self modifing polymorphic virses have been around for
 at least a  decade. That means for an instruction set like x86,
 for any set of 1000 instructions there are probably 10s of
 thousands of ways to rewrite those 1000 instructions so they
 behave the same but won't be detected by a scanner that detects
 the original.

Yes polymorphic viruses have been around a long time, but look at say
the 100 biggest infectors at the moment, none of them I would say are
polymorphic, all of them can be picked up by signatures. Virus writers
are as lazy as the rest of us, so the majority of viruses will be
written to take over the majority of unsecured systems.

 The *only* 100% safe way to guard against viruses to fix all the
 security holes that viruses exploit. That means better coding
 practices.

Not 100% safe, you can still have users doing things they shouldn't
like giving a screensaver root privileges. You have to layer your
security on, so that your code is secure, and your users activities
are restricted and audited. One way to restrict is application white
or blacklisting, Av at the moment kind of fills the application
blacklisting, which isn't the best method due to malware now
outnumbering legitimate software, but whitelisting has its issues to
(apparmour/SeLinux), it can be hard for the average user to set this
all up.
100% secure code is also nigh-on impossible to write if you want it to
be flexible to the user. I know everyone likes to bash Microsoft, but
they spend billions on their secure development lifecycle, and they
still have holes that get exploited all the time, the only safe code
is code that doesn't run.

Most of the people on this list are not the average user, but my point
still stands if we continue along the line that Linux is immune to
viruses we will get bitten as Apple has, one day, and it will be
harder than the gnome-look screensaver of the Proftpd compromise.

 Erik
Most definitely not the average computer user.
--
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-03 Thread Nick Andrew
On Mon, Apr 04, 2011 at 06:46:36AM +1000, Erik de Castro Lopo wrote:
 Morgan Storey wrote:
  I think it is going to come back and bite the Linux community if we go
  via the line that we are immune to viruses,
 
 Unfortunately, the alternative, virus scanners that look for
 particular virus signatures is nothing more than security
 theatre.

I agree. What would a virus scanner look for anyway, if there
are no extant viruses on linux systems?

 Firstly, inew viruses can be written so fast that the virus
 detection engines have absolutely no way of keep up.

I don't think this is a strong argument. Windows viruses have lifetimes in
decades (if not forever). Statistically speaking, a given computer is very
unlikely to be infected by a young virus which cannot yet be detected by
a virus scanner. Much more likely that a computer will come into contact
with many well known viruses long after the viruses became prevalent.

 The *only* 100% safe way to guard against viruses to fix all the
 security holes that viruses exploit. That means better coding 
 practices.

Don't execute incoming data as code. That's rule #1, learned by hard knocks
as Windows systems happily executed auto-run files, email attachments,
word macros, PostScript documents, and so on.  Unfortunately we forgot
rule #1 with the invention of JavaScript and Flash. Your browser is now
happily executing untrusted third party code in your account.

That leads to rule #2 - defense-in-depth. The only hope we have to
survive this untrusted and potentially malicious code being executed by
our browsers is to implement sandboxes, language-level restrictions and
strict limits on authorization.

Nick.
-- 
PGP Key ID = 0x418487E7  http://www.nick-andrew.net/
PGP Key fingerprint = B3ED 6894 8E49 1770 C24A  67E3 6266 6EB9 4184 87E7
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-03 Thread Erik de Castro Lopo

I am subscribed to this list. Please don't CC me when replying.

Morgan Storey wrote:

 Yes polymorphic viruses have been around a long time, but look at say
 the 100 biggest infectors at the moment, none of them I would say are
 polymorphic, all of them can be picked up by signatures.

And all of those are hitting machines that are either unpatched or
have vulnerabitlites that should have been patched.

 Not 100% safe, you can still have users doing things they shouldn't
 like giving a screensaver root privileges.

A user dumb enough to do this shouldn't have root privileges.

 Virus writers are as lazy as the rest of us

But as soon as virus scanners catch all non-polymorphic viruses
virus writers will stop writing non-polymorphic viruses.

My thesis is that continuing the virus arms race with virus scanners
results in a situtaion where the virus scanners are unable to detect
99% of all viruses. I think effort should be invested in heading all
viruses that exploit code rather than users off at the pass by fixing
the software bugs viruses exploit.

The user problem needs to be dealt with separately.

 100% secure code is also nigh-on impossible to write if you want it to
 be flexible to the user.

I think the people here in Sydney who worked on the Sel4 project:

http://ertos.nicta.com.au/research/sel4/
http://www.sigops.org/sosp/sosp09/papers/klein-sosp09.pdf

would say that the level of difficulty is not nigh-on impossible
but is at the phd research in compsci level of difficulty.

Going by the experience of the Sel4 project, I would say that it
is currently not possible to economically write provably correct
code, but it possible. As more research goes on in the field of
provably correct code, the techniques will improve, become easier
to apply and become more widespread. Long term, that is the only
hope for secure computing.

 Most of the people on this list are not the average user, but my point
 still stands if we continue along the line that Linux is immune to
 viruses we will get bitten as Apple has, one day, and it will be
 harder than the gnome-look screensaver of the Proftpd compromise.

That was a user failure. Dumb users need to be locked down so they
can't compromise the systems they work on.

Erik
-- 
--
Erik de Castro Lopo
http://www.mega-nerd.com/
-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html


Re: [SLUG] Re: Virus Scanner

2011-04-03 Thread Peter Hardy
To answer OP's question, my Linux mail server uses spamassassin and ESET
for filtering. My Linux file server also periodically performs full
scans with ESET. I do not yet run any virus scanning on my desktop
though.

On Mon, 2011-04-04 at 11:30 +1000, Nick Andrew wrote:
 On Mon, Apr 04, 2011 at 06:46:36AM +1000, Erik de Castro Lopo wrote:
  Morgan Storey wrote:
   I think it is going to come back and bite the Linux community if we go
   via the line that we are immune to viruses,
  
  Unfortunately, the alternative, virus scanners that look for
  particular virus signatures is nothing more than security
  theatre.
 
 I agree. What would a virus scanner look for anyway, if there
 are no extant viruses on linux systems?

The biggest market for antivirus scanners on Linux is when the system in
question is acting as a server for other systems. This is evidenced by
the fact that there are quite a few commercial A/V systems available for
Linux, almost none of which are suitable for desktop use.

And, for better or for worse, you need to demonstrate that you're
performing regular virus scans on *all* of your systems on a regular
basis for quite a few security-related standards. Regardless of whether
you think you need it or not. Yes, I'm looking at you, Payment Card
Industry.

 That leads to rule #2 - defense-in-depth. The only hope we have to
 survive this untrusted and potentially malicious code being executed by
 our browsers is to implement sandboxes, language-level restrictions and
 strict limits on authorization.

I think that eliminating a reasonably large swathe of of your attack
vector through a regularly-maintained virus scanner is a good
contribution to solid defence-in-depth...

-- 
Pete

-- 
SLUG - Sydney Linux User's Group Mailing List - http://slug.org.au/
Subscription info and FAQs: http://slug.org.au/faq/mailinglists.html