Re: iSCSI : Login response with zero DataSegmentLength

["Julian Satran" <Julian_Satran@il.ibm.com>]

Priya,

The answer you suggest in step 4 is correct - as you are required to conclude the negotiation. If your answer would have had T=0 and a zero length
Data Segment that would have violated the rule you quote. 

Basically the rule is there to help avoid unnecessary traffic.

Julo

"Priya Viswambharan" <priyav@marantinetworks.com> Sent by: owner-ips@ece.cmu.edu 06/09/02 01:09

To:        <ips@ece.cmu.edu>         cc:                 Subject:        iSCSI : Login response with zero DataSegmentLength

Hi all,

 

Consider the following scenario.

 

1) I -> T    : login request: T=0, csg = 1, some text parameters that require negotiation

2) I <- T    : login response: T = 0, csg = 1, response for the initiator's text parameters + target's own text parameters that require negotiation.

3) I -> T    : login request: T = 1, csg = 1, nsg = 3, response for the parameters proposed by the target in step 2

 

Assume that at the end of the login request in step 3 all text parameters that need negotiation have been negotiated correctly at both ends. Is it valid for the target to send a login response as follows signifying the end of login phase? If not, what should be the correct response from the target ?

 

4) I <- T : login response : T=1, csg = 1, nsg = 3, DataSegmentLength = 0

 

I observed that the draft says that "A target or initiator SHOULD NOT use a Text or Login Response or Text or Login Request with no data segment (DataSegmentLength 0) unless explicitly required by a general or a key-specific negotiation rule."

 

To me the above scenario does not seem to be covered by this rule. Please clarify.

 

Thanks

Priya