>> Well, if the problem is technically hard to solve, it doesn't become any
>> easier just because it is on everybody's wish list. Look at how much of
>> effort has been put by the database vendors like Sybase and Oracle into
>> replication, with not much practical success. And yet it is the
>> number-one feature wanted of any RDBMS today.
The database replication problem is a harder problem than what we are
asking you to solve with AFS.
Could you explain, in detail for those of us with extensive experience
with afs internals, why a vlserver which loses quorum should not
continue to allow lookup requests on its database?
The only counterargument I can think of is that if you have weird
asymmetric routing, then a client with access to the vldb sync site
can perform vldb operations which affect cache managers which cannot
reach a vlserver with quorum, causing the client to see old data or
have a volume "disappear". I think this is a worthwhile risk.
Marc
