Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-04 Thread Alexander Reichle-Schmehl
Hi!

john schrieb:

 I'd be interested to hear some recommendations for IDS to run on
 internet facing servers. Especially from the point of view of ease of
 installation, ease of maintenance, quality of the tool, and ability to
 have it deliver really useful information to the admin. I've used
 SNORT a bit in the past and my feeling was that it was so chatty that
 it was actually hard to tell if something bad was happening or not.

Don't think it really counts as IDS, but I like to use tiger and rkhunter.
  They perform some checks on the system on a regular basis. That is not a
really good protection against unauthorized access (well; it might catch
stupid cracker ;) but at least it helps to protect the systems from myself,
e.g. when I tweak some configuration option during a maintenance task in an
insecure manner (e.g. allow root login via ssh until I'm finished setting
up the system) tiger will remind me to reset the save values :)


Best regards,
  Alexander


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread john
On Tue, Jun 2, 2009 at 4:45 PM, Josh Lauricha j...@lauricha.com wrote:
 I'm surprised more people aren't running tripwire or other IDS.

I'd be interested to hear some recommendations for IDS to run on
internet facing servers. Especially from the point of view of ease of
installation, ease of maintenance, quality of the tool, and ability to
have it deliver really useful information to the admin. I've used
SNORT a bit in the past and my feeling was that it was so chatty that
it was actually hard to tell if something bad was happening or not.

John


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Boyd Stephen Smith Jr.
In 2be970b50906030853t29dfb90atd60089611f98e...@mail.gmail.com, john 
wrote:
On Tue, Jun 2, 2009 at 4:45 PM, Josh Lauricha j...@lauricha.com wrote:
 I'm surprised more people aren't running tripwire or other IDS.

I'd be interested to hear some recommendations for IDS to run on
internet facing servers.

I inherited a tripwire installation at some point.  It was one mail message 
per day (and if you didn't get that message you knew something was wrong).

It required a bit of tuning to not report errors regularly, but once I spent 
that time it was fairly hands-off.
-- 
Boyd Stephen Smith Jr.   ,= ,-_-. =.
b...@iguanasuicide.net  ((_/)o o(\_))
ICQ: 514984 YM/AIM: DaTwinkDaddy `-'(. .)`-'
http://iguanasuicide.net/\_/



signature.asc
Description: This is a digitally signed message part.


Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Steven Brunasso
Remember, that a HIDS (host IDS) is just a detective control on the  
host.  It shows that you have been hacked, you will probably want a  
good NIDS (network IDS) to see what attacks are being attempted over  
the wire.


HIDS is good to quickly detect a compromise...


http://sourceforge.net/projects/aide
http://packages.debian.org/search?keywords=aide



On Jun 3, 2009, at 9:55 AM, Boyd Stephen Smith Jr. wrote:


In 2be970b50906030853t29dfb90atd60089611f98e...@mail.gmail.com, john
wrote:
On Tue, Jun 2, 2009 at 4:45 PM, Josh Lauricha j...@lauricha.com  
wrote:

I'm surprised more people aren't running tripwire or other IDS.


I'd be interested to hear some recommendations for IDS to run on
internet facing servers.


I inherited a tripwire installation at some point.  It was one mail  
message
per day (and if you didn't get that message you knew something was  
wrong).


It required a bit of tuning to not report errors regularly, but once  
I spent

that time it was fairly hands-off.
--
Boyd Stephen Smith Jr.   ,= ,-_-. =.
b...@iguanasuicide.net  ((_/)o o(\_))
ICQ: 514984 YM/AIM: DaTwinkDaddy `-'(. .)`-'
http://iguanasuicide.net/\_/




--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Rick Moen
Quoting Boyd Stephen Smith Jr. (b...@iguanasuicide.net):

 I inherited a tripwire installation at some point.  It was one mail message 
 per day (and if you didn't get that message you knew something was wrong).
 
 It required a bit of tuning to not report errors regularly, but once I spent 
 that time it was fairly hands-off.

One way to use Tripwire in conjunction with a slightly more modern and
lightweight file-based IDS alongside it:
http://linuxgazette.net/issue98/moen.html

(That article is not, however, a comparative review, which is apparently
what the original poster is seeking.)

-- 
Cheers,  Notice:  The value of your Hofstadter's Constant 
Rick Moen(the average amount of time you spend each month 
r...@linuxmafia.com  thinking about Hofstadter's Constant) has just 
McQ!  (4x80) been adjusted upwards.


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Izak Burger
On Wed, Jun 3, 2009 at 5:53 PM, john lists.j...@gmail.com wrote:
 I'd be interested to hear some recommendations for IDS to run on
 internet facing servers. Especially from the point of view of ease of
 installation, ease of maintenance, quality of the tool, and ability to
 have it deliver really useful information to the admin. I've used
 SNORT a bit in the past and my feeling was that it was so chatty that
 it was actually hard to tell if something bad was happening or not.

We use aide.


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Nikolai Lusan
On Wed, 2009-06-03 at 08:53 -0700, john wrote:
 On Tue, Jun 2, 2009 at 4:45 PM, Josh Lauricha j...@lauricha.com wrote:
  I'm surprised more people aren't running tripwire or other IDS.
 I'd be interested to hear some recommendations for IDS to run on
 internet facing servers. Especially from the point of view of ease of
 installation, ease of maintenance, quality of the tool, and ability to
 have it deliver really useful information to the admin. 

It really depends on what you want. I'm using a combination of PADS
(Passive Attack Detection System) and fail2ban ... these can both be run
on either a host or a router, and integrate with netfilter. You can
customise what they are looking for to report and ban. Fail2ban is good,
it lets me blackhole people attempting nasty things in quick order ...
even better when combined with ipset and a decent firewall setup.
-- 
Nikolai Lusan niko...@lusan.id.au


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Jeremy Melanson
I really like OSSEC. It's licensed under GPL V3. The agent runs on
multiple platforms. It's easy to install, relatively easy to configure.
The agent is a self-contained HIDS, rootkit detector, log and file
monitor.
It can also decode Snort, Cisco PIX/ASA, IPTables, and a a whole lot of
other logs. This means that it can act as a centralized security
monitoring and alerting system.
There are tons of other features that I'm not going to mention here.

Oh yeah, and you can get commercial support for it if needed.

-
Jeremy Melanson



On Wed, 2009-06-03 at 10:14 -0700, Rick Moen wrote:

 Quoting Boyd Stephen Smith Jr. (b...@iguanasuicide.net):
 
  I inherited a tripwire installation at some point.  It was one mail message 
  per day (and if you didn't get that message you knew something was wrong).
  
  It required a bit of tuning to not report errors regularly, but once I 
  spent 
  that time it was fairly hands-off.
 
 One way to use Tripwire in conjunction with a slightly more modern and
 lightweight file-based IDS alongside it:
 http://linuxgazette.net/issue98/moen.html
 
 (That article is not, however, a comparative review, which is apparently
 what the original poster is seeking.)
 
 -- 
 Cheers,  Notice:  The value of your Hofstadter's Constant 
 Rick Moen(the average amount of time you spend each month 
 r...@linuxmafia.com  thinking about Hofstadter's Constant) has just 
 McQ!  (4x80) been adjusted upwards.
 
 


Re: Recommend good IDS? was Re: /dev/shm/r?

2009-06-03 Thread Nicolas GRENECHE
Hi,

If you run large nuber of hosts, i suggest samhain.
You have many features builtin (monitoring of files, system.map
altering, suid bits, appending only on log files etc.).
It works on client server model (a server who centralize hosts
integrity database).

Communications are secure (AES for ciphering and SRP for authentication).

The deployement procedure is very convenient : you can build samhain
instances on dedicated machines and deploying them on network with a
sing script. Everything done over SSH.

Give it a try ;-)

Ressource : http://la-samhna.de/samhain/


2009/6/3 Jeremy Melanson jmelan...@systemhalted.com:
 I really like OSSEC. It's licensed under GPL V3. The agent runs on multiple
 platforms. It's easy to install, relatively easy to configure.
 The agent is a self-contained HIDS, rootkit detector, log and file monitor.
 It can also decode Snort, Cisco PIX/ASA, IPTables, and a a whole lot of
 other logs. This means that it can act as a centralized security monitoring
 and alerting system.
 There are tons of other features that I'm not going to mention here.

 Oh yeah, and you can get commercial support for it if needed.

 -
 Jeremy Melanson


 On Wed, 2009-06-03 at 10:14 -0700, Rick Moen wrote:

 Quoting Boyd Stephen Smith Jr. (b...@iguanasuicide.net):

 I inherited a tripwire installation at some point.  It was one mail
 message
 per day (and if you didn't get that message you knew something was wrong).

 It required a bit of tuning to not report errors regularly, but once I
 spent
 that time it was fairly hands-off.

 One way to use Tripwire in conjunction with a slightly more modern and
 lightweight file-based IDS alongside it:
 http://linuxgazette.net/issue98/moen.html

 (That article is not, however, a comparative review, which is apparently
 what the original poster is seeking.)

 --
 Cheers,  Notice:  The value of your Hofstadter's
 Constant
 Rick Moen(the average amount of time you spend each
 month
 r...@linuxmafia.com  thinking about Hofstadter's Constant) has just
 McQ!  (4x80) been adjusted upwards.





-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Izak Burger
On Mon, Jun 1, 2009 at 12:26 PM, Vladislav Kurz
vladislav.k...@webstep.net wrote:
 Well, this really looks suspicious. Look for unexpected processes running,
 open ports, etc. Directory /dev/shm/ is world-writable like /tmp, so chances
 are that the attacker did not gain root yet. But he might have shell
 listening on some port and trying hard to get root using some local exploit.

I agree, chances are the box hasn't been exploited just yet, but I
would be worried about just how he got that file there in the first
place. We know that directory is world writable, so it could have been
written by anything, but what? Sometimes the ownership of the file
will give it away, for example, if the file is owned by www-data, you
know some exploit in apache (usually php!) was used to gain file
system access.


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Guntram Trebs

Izak Burger schrieb:

On Mon, Jun 1, 2009 at 12:26 PM, Vladislav Kurz
  
I agree, chances are the box hasn't been exploited just yet, but I

would be worried about just how he got that file there in the first
place. We know that directory is world writable, so it could have been
written by anything, but what? Sometimes the ownership of the file
will give it away, for example, if the file is owned by www-data, you
know some exploit in apache (usually php!) was used to gain file
system access.
  
Yes, chances are, that it's just some unsecure script in a webspace. Not 
good, but if you are a webservice provider, you always have some special 
customer.
I even know companies which buy a cms and don't think of who cares for 
it over the time as long as it's running ...


On the other hand, you should keep in mind, that it could be someone who 
has gained root provileges and hides some of his activities. If he is 
root, then there has to be some other traces left of him.


So you should collect other information:
- lsof and /proc, if you find suspicious processes
- intrusion detection software
- logfile scanning software and manual examining log files including 
firewall logs


Good point is, when you can trace times of activity. But always keep in 
mind, that the information could be wrong.


--
Guntram Trebs
freier Programmierer und Administrator

g...@trebs.net
+49 (30) 42 80 61 55
+49 (178) 686 77 55 




--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Johann Spies
On Mon, Jun 01, 2009 at 07:23:27AM -0400, Michael Stone wrote:

 Yes, that's a typical location for intruders to drop files. Easiest  
 thing to do is reinstall after thinking about how the compromise may  
 have occurred. (Did you update regularly, including kernel updates? Did  
 all accounts have strong passwords? Do you have web applications not  
 managed by the system that weren't being updated? etc.)

We had a serious situation on this computer and several others. Ssh
and sshd were replaced by the cracker's own version and in once case
nearly all the pam-related stuff were replaced also.  Through this
customised versions of ssh the cracker harvested every password that
was used during ssh logins and ssh sessions.

We are winning the battle and will in the next few weeks try do the
analysis of what went wrong.

Regards
Johann
-- 
Johann Spies  Telefoon: 021-808 4599
Informasietegnologie, Universiteit van Stellenbosch

 Thou wilt show me the path of life: in thy presence 
  is fulness of joy; at thy right hand there are  
  pleasures for evermore. Psalms 16:11 


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Wade Richards
Although it's worse if an attacker has root, don't think that just because
the attacker doesn't have root, it's no big deal.  If an attacker can run
(even as an ordinary user) unauthorized software on your machine, then
your machine may be part of a botnet.  And having unauthorized user access
to a machine leaves the door open for trojan horse type programs which may
give the user root access.  Finally, depending on the user account which
is compromised, the attacker may already have access to sensitive data.

Don't obsess on root access.  Any unauthorized use is a problem.

 --- Wade

On Tue, June 2, 2009 00:35, Guntram Trebs wrote:
 Izak Burger schrieb:
 On Mon, Jun 1, 2009 at 12:26 PM, Vladislav Kurz

 I agree, chances are the box hasn't been exploited just yet, but I
 would be worried about just how he got that file there in the first
 place. We know that directory is world writable, so it could have been
 written by anything, but what? Sometimes the ownership of the file
 will give it away, for example, if the file is owned by www-data, you
 know some exploit in apache (usually php!) was used to gain file
 system access.

 Yes, chances are, that it's just some unsecure script in a webspace. Not
 good, but if you are a webservice provider, you always have some special
 customer.
 I even know companies which buy a cms and don't think of who cares for
 it over the time as long as it's running ...

 On the other hand, you should keep in mind, that it could be someone who
 has gained root provileges and hides some of his activities. If he is
 root, then there has to be some other traces left of him.

 So you should collect other information:
  - lsof and /proc, if you find suspicious processes
  - intrusion detection software
  - logfile scanning software and manual examining log files including
 firewall logs

 Good point is, when you can trace times of activity. But always keep in
 mind, that the information could be wrong.

 --
 Guntram Trebs
 freier Programmierer und Administrator

 g...@trebs.net
 +49 (30) 42 80 61 55
 +49 (178) 686 77 55



 --
 To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
 with a subject of unsubscribe. Trouble? Contact
 listmas...@lists.debian.org





-- 
Wade Richards  (w...@wabyn.net)



-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Izak Burger
On Tue, Jun 2, 2009 at 6:42 PM, Wade Richards w...@wabyn.net wrote:
 Don't obsess on root access.  Any unauthorized use is a problem.

You are right of course. Right after I sent my message saying that
perhaps the machine hasn't been exploited yet I realised how wrong
such a view is. Someone gained access to an area they should not have
access to, it has been exploited already.

I have been fortunate enough to only be in this situation twice in the
last ten years. The first time was due to a weak password, and luckily
our attacker only installed an irc bouncer (renamed to bash so it
wouldn't stand out in a process listing). We could literally get away
with not reinstalling the entire machine, because the damage was
limited to one user account (yes, we did check for replaced binaries
:-) ).

The second time was caused by a php hosting control panel, which gave
the attackers (Turkish crackers unhappy with that Danish cartoon) the
ability to create ftp accounts and deface websites. Once again, the
damage was limited and we got away without a full reinstall. It was in
this sense that I hoped Johann (a former colleague of mine) might be
lucky enough to get away with limited damage.

Wait, there was a third time. On a CentOS box, I found a core file in
/etc/cron.d. I immediately realised what it was as I had an argument
about which kernel versions is affected with someone just the previous
week (thread here:
http://lists.clug.org.za/pipermail/clug-tech/2006-July/032952.html).
In this case, we eventually found that a former employee of the
organisation tried several exploits on the machine and left some
tell-tale signs behind. In this instance, though it seemed none of the
exploits succeeded, we decided to trash the CentOS install and move to
Debian :-)

In any case, enough about me. Good luck Johann, and I look forward to
more information on exactly what happened here.

regards,
Izak


--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Guntram Trebs

Hello,

there are few chances of replacing sshd without being root. In your 
place i would install every server new.


I think, he spied out passwords and maybe got root-Passwords in this 
way. Possibly he has even accessed servers where you didn't find him and 
left backdoors there. (manipulation of ~/.ssh/authorized_keys, another 
user with userid 0, replaced software, replaced suid-accesses, added 
software, ...)


When you reinstall your servers think of not using Login+Password but 
public-private key for user authentication.
You can even forward such a combination. But be aware, that if key 
forwarding is enabled it can be used by attackers too.
You should then protect the private key by password and use it only from 
trusted machines.
Avoid keyforwarding unless you need it, for instance when copying files 
from one server to the other it could be useful.
Lock your trusted working computer always when leaving it and think of 
encrypting the filesystem. (Be aware that a local attacker can install a 
password sniffer even if you encrypted your system partition)


good luck and think of not accessing the new installed computers from 
old ones ...

Regards,
Guntram

Johann Spies schrieb:

On Mon, Jun 01, 2009 at 07:23:27AM -0400, Michael Stone wrote:

  
Yes, that's a typical location for intruders to drop files. Easiest  
thing to do is reinstall after thinking about how the compromise may  
have occurred. (Did you update regularly, including kernel updates? Did  
all accounts have strong passwords? Do you have web applications not  
managed by the system that weren't being updated? etc.)



We had a serious situation on this computer and several others. Ssh
and sshd were replaced by the cracker's own version and in once case
nearly all the pam-related stuff were replaced also.  Through this
customised versions of ssh the cracker harvested every password that
was used during ssh logins and ssh sessions.

We are winning the battle and will in the next few weeks try do the
analysis of what went wrong.

Regards
Johann
  



--
Guntram Trebs
freier Programmierer und Administrator

g...@trebs.net
+49 (30) 42 80 61 55
+49 (178) 686 77 55 




--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-02 Thread Josh Lauricha
I'm surprised more people aren't running tripwire or other IDS.

On Tue, Jun 2, 2009 at 1:37 PM, Guntram Trebs g...@trebs.net wrote:
 Hello,

 there are few chances of replacing sshd without being root. In your place i
 would install every server new.

 I think, he spied out passwords and maybe got root-Passwords in this way.
 Possibly he has even accessed servers where you didn't find him and left
 backdoors there. (manipulation of ~/.ssh/authorized_keys, another user with
 userid 0, replaced software, replaced suid-accesses, added software, ...)

 When you reinstall your servers think of not using Login+Password but
 public-private key for user authentication.
 You can even forward such a combination. But be aware, that if key
 forwarding is enabled it can be used by attackers too.
 You should then protect the private key by password and use it only from
 trusted machines.
 Avoid keyforwarding unless you need it, for instance when copying files from
 one server to the other it could be useful.
 Lock your trusted working computer always when leaving it and think of
 encrypting the filesystem. (Be aware that a local attacker can install a
 password sniffer even if you encrypted your system partition)

 good luck and think of not accessing the new installed computers from old
 ones ...
 Regards,
 Guntram

 Johann Spies schrieb:

 On Mon, Jun 01, 2009 at 07:23:27AM -0400, Michael Stone wrote:



 Yes, that's a typical location for intruders to drop files. Easiest
  thing to do is reinstall after thinking about how the compromise may  have
 occurred. (Did you update regularly, including kernel updates? Did  all
 accounts have strong passwords? Do you have web applications not  managed by
 the system that weren't being updated? etc.)


 We had a serious situation on this computer and several others. Ssh
 and sshd were replaced by the cracker's own version and in once case
 nearly all the pam-related stuff were replaced also.  Through this
 customised versions of ssh the cracker harvested every password that
 was used during ssh logins and ssh sessions.

 We are winning the battle and will in the next few weeks try do the
 analysis of what went wrong.

 Regards
 Johann



 --
 Guntram Trebs
 freier Programmierer und Administrator

 g...@trebs.net
 +49 (30) 42 80 61 55
 +49 (178) 686 77 55


 --
 To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
 with a subject of unsubscribe. Trouble? Contact
 listmas...@lists.debian.org





-- 
Josh Lauricha


--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



/dev/shm/r?

2009-06-01 Thread Johann Spies
I am a bit worried that my computer have been compromised. 

Rkhunter reported:

[10:35:47] Warning: Suspicious file types found in /dev:
[10:35:47]  /dev/shm/r: ASCII text
[10:35:48]   Checking for hidden files and directories   [ Warning
]
[10:35:48] Warning: Hidden directory found: /etc/.java
[10:35:48] Warning: Hidden directory found: /dev/.udev
[10:35:48] Warning: Hidden directory found: /dev/.initramfs


I think the last three lines are not problematic but in /dev/shm/r I found:

spawn /bin/bash
interact

Do I have reason to be worried?

Regards
Johann
-- 
Johann Spies  Telefoon: 021-808 4599
Informasietegnologie, Universiteit van Stellenbosch

 Thou wilt show me the path of life: in thy presence 
  is fulness of joy; at thy right hand there are  
  pleasures for evermore. Psalms 16:11 


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-01 Thread Michael Stone

On Mon, Jun 01, 2009 at 10:46:54AM +0200, Johann Spies wrote:
I am a bit worried that my computer have been compromised. 

...

I think the last three lines are not problematic but in /dev/shm/r I found:

spawn /bin/bash
interact

Do I have reason to be worried?


Yes, that's a typical location for intruders to drop files. Easiest 
thing to do is reinstall after thinking about how the compromise may 
have occurred. (Did you update regularly, including kernel updates? Did 
all accounts have strong passwords? Do you have web applications not 
managed by the system that weren't being updated? etc.)


Mike Stone


--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-01 Thread Marcin Owsiany
On Mon, Jun 01, 2009 at 12:26:49PM +0200, Vladislav Kurz wrote:
 On Monday 01 of June 2009, Johann Spies wrote:
  spawn /bin/bash
  interact

Note that this seems to be a simple expect(1) script which runs a
shell. Not necessarily an indication of anything apart from a possible
attacker trying to exploit something using expect.

-- 
Marcin Owsiany porri...@debian.org http://marcin.owsiany.pl/
GnuPG: 1024D/60F41216  FE67 DA2D 0ACA FC5E 3F75  D6F6 3A0D 8AA0 60F4 1216


-- 
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org



Re: /dev/shm/r?

2009-06-01 Thread Michael Stone

On Mon, Jun 01, 2009 at 12:31:04PM +0100, Marcin Owsiany wrote:

Note that this seems to be a simple expect(1) script which runs a
shell. Not necessarily an indication of anything apart from a possible
attacker trying to exploit something using expect.


It's also an indication that the attacker could write to the 
filesystem...


Mike Stone


--
To UNSUBSCRIBE, email to debian-security-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org