Re: OpenSSH

2002-06-27 Thread Neil Ford

On 26/6/02 8:41 pm, Chris Ball [EMAIL PROTECTED] wrote:

 Chris == Chris Devers [EMAIL PROTECTED] writes:
 
   Chris Weren't there two recent SSH vulnerabilities? (Hence upgrades
   Chris from 3.2-3.3, and now on the heels of that 3.3-3.4?) Does
   Chris this setting inoculate users against both problems?
 
 No, there was one ssh _advisory_ from Theo, saying that we should all be
 running 3.3 with privsep because of some bugs in the network handling
 code.  He refused to give a patch for the bug.  Today's _exploit_ uses
 that bug to get remote root access.  It's the same vulnerability both
 times.  Turning off challengeauth protects against today's exploit of
 the network code vulnerablity.
 
 AIUI,
 
 - Chris.

It should possibly be pointed out that SSH protocol 1 Blowfish support may
have been broken around 3.3.

Now I know most same people will be using protocol 2 but there may be
occasions when there is only a windows client around and you have no choice.
To get round this you have to use 3des which obviously brings it's own
problems.

I haven't seen his reported anywhere yet, but am aware of it from
discussions elsewhere.

Neil.
-- 
Neil Ford
[EMAIL PROTECTED] | [EMAIL PROTECTED]





Re: OpenSSH

2002-06-27 Thread the hatter

On Thu, 27 Jun 2002, Neil Ford wrote:

 It should possibly be pointed out that SSH protocol 1 Blowfish support may
 have been broken around 3.3.

 Now I know most same people will be using protocol 2 but there may be
 occasions when there is only a windows client around and you have no choice.
 To get round this you have to use 3des which obviously brings it's own
 problems.

You mean there are still people using teraterm, rather than putty ?


the hatter





Re: OpenSSH

2002-06-27 Thread Natalie S. Ford

On Thu, Jun 27, 2002 at 08:47:01AM +, the hatter wrote:
 You mean there are still people using teraterm, rather than putty ?

Yes, on the rare occasions that I use windows to ssh.  Putty cannot print.

-- 
Natalie S. Ford   ø¤º°`°º¤ø,¸¸,ø¤º°`°º¤ø  [EMAIL PROTECTED]
http://www.natalie.ourshack.org/  http://natalief.livejournal.com/




Re: OpenSSH

2002-06-27 Thread Paul Makepeace

On Thu, Jun 27, 2002 at 02:10:46PM +0100, Natalie S. Ford wrote:
 Yes, on the rare occasions that I use windows to ssh.  Putty cannot print.

It does have a Copy All to Clipboard feature which would mean the times
you want to print it'd be all there. Or of course, copy/paste in the
main window if you just want that.

http://the.earth.li/~sgtatham/putty/latest/x86/putty.exe

Paul

-- 
Paul Makepeace ... http://paulm.com/

What is the sun's temperature? Spanish speaking dolphins.
   -- http://paulm.com/toys/surrealism/




OpenSSH

2002-06-26 Thread Alex McLintock



Sorry if this is old hat to everybody (I only get the digest of this list 
once or twice a day so you may already be discussing it) butthere is a 
vulnerability in recent versions of OpenSSH.

http://slashdot.org/articles/02/06/26/1547242.shtml?tid=172

Apparently you can temporarily fix the problem by making sure sshd_config says


ChallengeResponseAuthentication no

Dunno enough about security to say whether this is important or not.

Yeah yeah, I know nothing to do with perl but I bet this has the highest 
proportion of sys admins of any mailing list I am on.

And this comes less than a day after I upgraded Apache because of a 
different vulnerability.

Alex Mc





Re: OpenSSH

2002-06-26 Thread Chris Ball

 Alex == Alex McLintock [EMAIL PROTECTED] writes:

Alex Sorry if this is old hat to everybody (I only get the digest
Alex of this list once or twice a day so you may already be
Alex discussing it) butthere is a vulnerability in recent
Alex versions of OpenSSH.

We haven't discussed.  Yes, it is important for anyone running versions
of OpenSSH between 3.0-3.2 who _doesn't_ have:

Alex ChallengeResponseAuthentication no

in their sshd_config to upgrade now.  Most sane distributions (like
Debian) install sshd with this line as Alex sent it, which means that
you aren't vulnerable to today's exploit.  If you're running a standard
Red Hat sshd_config with OpenSSH 3.0-3.2, though, get upgrading.

OpenSSH 3.4 was released today, so it's worthwhile to upgrade to that
and enable privilege separation - at least, according to Theo.  :)

Comedy point: openbsd.org now advertises 'One remote hole in the default
install, in nearly six years!' rather than the ever-present 'No remote
holes in the default install in five years!'.

- Chris.
-- 
$a=printf.net;  Chris Ball | chris@void.$a | www.$a | finger: chris@$a
 Blessings to the chap who invented ice cream, ginger-pop and the rest!
 I'd rather invent things like that any day than rockets and bombs.
   -- Julian, Five on Finniston Farm





Re: OpenSSH

2002-06-26 Thread Chris Devers

Chris Ball wrote:
Alex == Alex McLintock [EMAIL PROTECTED] writes:

 
 Alex Sorry if this is old hat to everybody (I only get the digest
 Alex of this list once or twice a day so you may already be
 Alex discussing it) butthere is a vulnerability in recent
 Alex versions of OpenSSH.
 
 We haven't discussed.  Yes, it is important for anyone running versions
 of OpenSSH between 3.0-3.2 who _doesn't_ have:
 
 Alex ChallengeResponseAuthentication no
 
 in their sshd_config to upgrade now.

Weren't there two recent SSH vulnerabilities? (Hence upgrades from 3.2-3.3,
and now on the heels of that 3.3-3.4?) Does this setting inoculate users
against both problems?

 Comedy point: openbsd.org now advertises 'One remote hole in the default
 install, in nearly six years!' rather than the ever-present 'No remote
 holes in the default install in five years!'.

heh :)


-- 
Chris Devers   [EMAIL PROTECTED]
DO  NOT  LEAVE  IT  IS  NOT  REAL





Re: OpenSSH

2002-06-26 Thread Chris Ball

 Chris == Chris Devers [EMAIL PROTECTED] writes:

Chris Weren't there two recent SSH vulnerabilities? (Hence upgrades
Chris from 3.2-3.3, and now on the heels of that 3.3-3.4?) Does
Chris this setting inoculate users against both problems?

No, there was one ssh _advisory_ from Theo, saying that we should all be
running 3.3 with privsep because of some bugs in the network handling
code.  He refused to give a patch for the bug.  Today's _exploit_ uses
that bug to get remote root access.  It's the same vulnerability both
times.  Turning off challengeauth protects against today's exploit of
the network code vulnerablity.

AIUI,

- Chris.
-- 
$a=printf.net;  Chris Ball | chris@void.$a | www.$a | finger: chris@$a
 Blessings to the chap who invented ice cream, ginger-pop and the rest!
 I'd rather invent things like that any day than rockets and bombs.
   -- Julian, Five on Finniston Farm