Re: iSCSI: more on StatRN

[Michael Krause <krause@cup.hp.com>]

At 05:41 PM 10/30/00 -0600, Stephen Bailey wrote:

>Again, the spec really should STILL not leave anything to chance in this 
>regard.  It
>should spell out what should be detected and what should result from
>detection.

If interoperability is a strong desire, then for each error, there should 
be one recovery / notification process.  If this error is best handled by 
dropping the connection / session, then that is the only behavior that 
should be allowed.  Multiple execution paths leads to interoperability 
problems as well as hardware implementation problems which directly impact 
the technology adoption rate.  Hence, I agree with Stephen.

Mike