From: Brian Pawlowski (beepy@netapp.com)
Date: 08/17/01-12:12:57 PM Z
From: Brian Pawlowski <beepy@netapp.com> Message-Id: <200108171712.KAA18585@tooting-fe.eng.netapp.com> Subject: Re: Replication and migration requirements Date: Fri, 17 Aug 2001 10:12:57 -0700 (PDT) NFS V4 as it stands includes some info to support client to server interactions for client failover in a migration or replication situation. The work being proposed is a separate protocol to support the feature defined for NFS V4. Not adding anything new - only a server-to-server protocol to support what is there. > At 12:15 17/08/01, Spencer Shepler wrote: > > Are you saying that you don't need any replication or migration > >backend protocol to support data management or that you don't want > >those types of hooks/features in the NFSv4 protocol itself ? > > Good question. Myself, I don't need a replication or > migration backend protocol. I accept that some other users > probably do want the ability to replicate/migrate stuff > between/among servers. > > So my main concern is that we don't do this inside NFSv4, > which already appears to be more complex than necessary and > already is an operational concern. If replication (etc) were > done in a totally decoupled way so that no additional gorp > gets added to NFSv4, that would be OK with me. > > It has not been clear to me (I couldn't be in London due > to serious illness so am a bit out of sync) that this work was > being undertaken in such a totally decoupled manner. Are folks > intending to undertake this as a separate protocol, decoupled > from NFSv4 ? If yes, kindly pardon my confusion. > > Regards, > > Ran > rja@inet.org >
This archive was generated by hypermail 2.1.2 : 03/04/05-01:49:01 AM Z CST