[nfsv4] Question On the seq_window field in a context creation response of the RPCSEC_GSS Protocol( RFC 2203)

Next message: Mike Eisler: "[nfsv4] Re: Question On the seq_window field in a context creation response of the RPCSEC_GSS Protocol( RFC 2203)"
Previous message: Talpey, Thomas: "Re: [nfsv4] more re: NFS4ERR_EXPIRED"

uddin@us.ibm.com

Message-ID: <OFBE9A737B.0C9A6EAC-ON88256E75.006A145C-88256E75.006A37C1@us.ibm.com>
From: Kaukab Uddin <uddin@us.ibm.com>
Subject: [nfsv4] Question On the seq_window field in a context creation response of the RPCSEC_GSS Protocol( RFC 2203)
Date: Tue, 13 Apr 2004 12:20:10 -0700

We would like to validate our interpretation and understanding of the
definition of a context sequence window
(as represented by the seq_window field in a context creation response).

We interpret this to mean the maximum number that could possibly be
supported on a security context and not
necessarily the exact number that needs to be supported on a security
context.

So for example:

1) If seq_window returned by the server  is 'n'.

2) If the client increments its sequence number by '1' for its requests,
then the effective number
    of simultaneous requests supported on this context would in fact be
'n'.

3) However if the client has a sequence number algorithm, that increments
the sequence number
    by more than '1', then the effective number of simultaneous requests
supported on this context
    would actually be less than 'n'

 Is our understanding here correct?

Regards;
Kaukab

_______________________________________________
nfsv4 mailing list
nfsv4@ietf.org
https://www1.ietf.org/mailman/listinfo/nfsv4