RE: IPS-All: Reminder - Security draft last call ends Monday, Jul y 1 at 8am EST

[Black_David@emc.com]

> When a target ends up in the negative authentication cache, does
> that imply that the initiator will not contact the target while
> it's in the cache, or does  it imply that the initiator can contact
> the target but should not use IPsec (if IPsec failed)? Or shouldn't
> use app level security if app level security failed?

The first - initiator will not contact the target while it's in the
cache under the assumption that authentication will just fail again.
As Bernard wrote, this protects against a denial of service attack
where a rogue iSNS server in league with a rogue iSCSI target causes
an initiator to repeatedly waste its time trying to talk to the target.

Would it help to require an administrative interface to clear entries
from the negative cache or the entire cache?

Thanks,
--David
---------------------------------------------------
David L. Black, Senior Technologist
EMC Corporation, 42 South St., Hopkinton, MA  01748
+1 (508) 249-6449 *NEW*      FAX: +1 (508) 497-8500
black_david@emc.com         Cell: +1 (978) 394-7754
---------------------------------------------------

-----Original Message-----
From: Jonathan Trostle [mailto:jtrostle@corp.iready.com]
Sent: Monday, July 01, 2002 6:17 PM
To: 'Bernard Aboba'
Cc: 'ips@ece.cmu.edu'
Subject: RE: IPS-All: Reminder - Security draft last call ends Monday, Jul y
1 at 8am EST




When a target ends up in the negative authentication cache, does that imply
that the initiator will not contact the target while it's in the cache, or
does  it imply that the initiator can contact the target but should not use
IPsec (if IPsec failed)? Or shouldn't use app level security if app level
security failed?
Jonathan 
-----Original Message----- 
From: Bernard Aboba [mailto:bernard_aboba@hotmail.com] 
Sent: Saturday, June 29, 2002 10:42 PM 
To: jtrostle@corp.iready.com; ips@ece.cmu.edu 
Subject: RE: IPS-All: Reminder - Security draft last call ends Monday, 
July 1 at 8am EST 


>One comment/question on the security draft below: 
> 
>p. 23: "Where iSNS is used without security, IP block storage protocol 
>implementations MUST support a negative cache for authentication 
>failures." 
> 
>Is it worth pointing out that when iSNS is used with security, then a 
>negative cache MUST NOT be used? An attacker can cause authentication >to 
>fail through a DoS attack which could result in an entry being >added to 
>the negative cache. 
There are two orthogonal issues here -- one is iSNS security, the other is 
IPS protocol security. If iSNS is not secured, then a peer might receive and

accept an iSNS packet from a rogue iSNS server. However, if the IPS session 
is subsequently secured, and mutually authenticated, the endpoint specified 
in the bogus discovery message will fail to authenticate. The argument is 
that this should result in a negative cache entry within the iSNS 
implementation, so as to prevent continual attempts to authenticate to bogus

peers. 
If iSNS is secured, then presumably this would preclude a rogue iSNS server,

and the negative cache is unnecessary. 
Do you have an issue with the negative cache in general, or just its use 
where iSNS is secured? 





_________________________________________________________________ 
Send and receive Hotmail on your mobile device: http://mobile.msn.com