RE: Re: Cisco IOS Denial of Service that affects most Cisco IOS routers- requires power cycle to recover

["Curt Purdy" <purdy () tecman com>]

If the packet expires in transit i.e. ttl 1 to router 2 hops away means it
never gets to that router.  Not possible to fill a queue with a packet that
is dropped by the previous router. Check out "Internet Core Protocols" at
Oreilly.

Curt Purdy CISSP, GSEC, MCSE+I, CNE, CCDA
Information Security Engineer
DP Solutions
cpurdy () dpsol com
936.637.7977 ext. 121

----------------------------------------

If you spend more on coffee than on IT security, you will be hacked.
What's more, you deserve to be hacked.
-- White House cybersecurity adviser Richard Clarke


-----Original Message-----
From: full-disclosure-admin () lists netsys com
[mailto:full-disclosure-admin () lists netsys com]On Behalf Of
lee.e.rian () census gov
Sent: Monday, July 21, 2003 4:55 PM
To: rnews () river com
Cc: incidents () securityfocus com; full-disclosure () lists netsys com
Subject: [Full-disclosure] Re: Cisco IOS Denial of Service that affects
most Cisco IOS routers- requires power cycle to recover



> The kickup to supervisor level happens when the packet is targeted
> directly at the router's IP address (per first Cisco advisory) or just
> has its TTL expire in transit past the router (per revised Cisco
> advisory).

Has anyone been able to verify that the problem occurs when the TTL expires
"in transit"?

I've been able to get packets stuck on the input queue by sending to the
router's interface address, sending to <network, 0> and <network, -1> but
sending to a router two hops away with a TTL of 1 just gives me an icmp ttl
exceeded & nothing new stuck on the input queue.

Lee





                      Richard Johnson
                      <rdump () river com>        To:
incidents () securityfocus com
                                               cc:
                      07/20/2003 03:20         Subject:  Re: Cisco IOS
Denial of Service  that affects most Cisco IOS routers-
                      AM                        requires power cycle to
recover
                      Please respond to
                      rnews






In article
<Pine.BSO.4.53.0307172223150.11409 () rhiannon precision-guesswork com>,
 Tina Bird <tbird () precision-guesswork com> wrote:

> information on the detailed structure of the evil packets in these
> protocols is not yet public AFAIK.


The router has problems if it receives a packet, content irrelevant,
that makes it to supervisor level claiming an IP protocol that it
doesn't have code to handle.

The kickup to supervisor level happens when the packet is targeted
directly at the router's IP address (per first Cisco advisory) or just
has its TTL expire in transit past the router (per revised Cisco
advisory).

Send enough packets (default 75), and the input queue is full.  hping is
enough of a launch platform for that--there's no need for
questionable-source exploit binaries when testing.


Richard

--
My mailbox. My property. My personal space. My rules. Deal with it.
                        http://www.river.com/users/share/cluetrain/






---------------------------------------------------------------------------
----------------------------------------------------------------------------

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.netsys.com/full-disclosure-charter.html

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.netsys.com/full-disclosure-charter.html