From: RJ Atkinson (rja@inet.org)
Date: 08/06/01-09:25:16 AM Z
Message-Id: <5.1.0.14.2.20010806101818.00a32b00@10.30.15.2> Date: Mon, 06 Aug 2001 10:25:16 -0400 From: RJ Atkinson <rja@inet.org> Subject: Re: Final comments on recharter of work group At 09:26 06/08/01, Brian Pawlowski wrote: > o Management > > Work is needed to provide better management and administration > capabilities for NFS via a management MIB. The WG shall consider undertaking development of MIBs for NFS and for related protocols (e.g. ONC RPC). Any MIBs developed will be modular rather than monolithic (e.g. any NFS MIB will be separate from any ONC RPC MIB, so that someone using ONC RPC but not NFS can still gain value from the work). I'd also like to see any re-charter stress the importance of not breaking interoperability with the installed base of pre-NFSv4 implementations and also stress the WG's need to more seriously consider the operational implications of the protocol changes and additions. This WG has had a real tendency of late for "feature creep" with a result that the current protocol is extremely difficult to deploy and maintain operationally. Continuing the current approach creates real incentives for users to NOT deploy NFS (or to un-deploy NFS) and instead deploy SMB/CIFS or AFS (on grounds of operational simplicity of either of the latter protocols as compared with current NFSv4). This is a bit of a heretical view among the protocol designers, but I'm speaking as one of the (very) few actual users/operators that participate in the NFS WG. Frankly put, I do NOT want any more features. I want a protocol spec that is stable and I want more focus on operational considerations. I know I'm not alone in this in the user/operator community. Yours, Ran rja@inet.org
This archive was generated by hypermail 2.1.2 : 03/04/05-01:49:00 AM Z CST