As vfp is a file based database there is no such
thing as clustering but the clustering a RAID
or SAN would offer.

You very probably know more than me about SAN.
I know it's a storage area network and that it can
somehow be compared to RAID in that it can
handle not only an array of disks but also several
servers as to make it one logical drive.

Well, if a file will be clustered by that on several
disks on several servers and everything is managed
transparantly by the SAN, so that foxpro works
on a table as if it was stored on one harddisk on
one server, that's no problem and could surely
accelerate access on data, reading perhaps more than
writing.

But you wouldn't and couldn't cluster a foxpro
dbc by having several copies of the dbc on
several servers. There is no replication built in,
so that those database copies would surely go
out of sync, if half of the users use one dbc
and half of the users the other...

Normally clustering of databases is done by
the database server, which handles where what
data will go and stores several parts of data
on several servers. This is not possible in foxpro
without writing some server component, that
would do that.

So you can do RAID or SAN, if it is transparent
for foxpro, that means a table remains one file, but
that is not what you normally understand by clustering
a database, although it would surely accelerate it.
But you'd still be bound to the 2GB limit for a table.
If this is the reason for clustering, SAN wouldn't help.

Bye, Olaf.