Hi andrew,

This is actually normal behaviour. When only one instance is using the tables, the file server can use read and write buffering on the tables.

When a second instance is opening the same tables, the server is going to disable at least the write cache. As soon you are going to write something to the table it is also going to disable the read chache. This file server mechanism is called opportunistic locking and is part of the file server redirector implementation.

Question. Is there any referential integrity in effect (cascading deletes perhaps)?

You can download filemon from www.sysinternals.com to see what messages go accross the network and you can find out what is taking so long. Also you can see what information comes from the cache and what is read from/written to the network.

Walter,



>Actually this is what I see:
>
>Open 1st instance of VFP
>Execute DELETE - runs fast
>
>Open 2nd instance of VFP
>USE the 2 tables
>
>Back into the 1st instance
>Execute DELETE - slow
>
>Close the 2nd
>In the 1st run DELETE - slow !!!
>Reopen tables
>Run Delete again - fast
>
>What's going on?