[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Proposed changes to ESP (andf a little AH too)
Bill,
You are right that an error that occurs during decryption will not
be detected by this reversed ordering of the operations. However, the
purpose of the authentication/integrity service in IPSEC is to detect
malicious modification of the data, not to protect against errors that
might occur during protocol processing within IPSEC. Such errors copuld
also occur after both operations were completed, irrespective of the order
of processing, and thus would be undetected by the IPSEC module.
Yes, an encryption key mismatch would not be detected under the
proposed processing order, but I expect the encryption and authentication
keys for ESP will be derrived algorithmically from a common source, e.g.,
as described in Jim Hughes I-D. In such cases, it would be unlikely for
the sort of error you describe to occur. I admit there is a greater chance
of this for manually configured keys, but since this would trash all
incoming traffic over the Sa in question, this should be an easy error to
track down.
Steve