I'm starting the design of an n-n replication system as part of my
database design and am curious if IB/FB would have any features that
would make a char vs. a numeric ID more beneficial.

ID would identify a server and a sequence, by either:

ABC000000001 or 100000000001

Where either 'ABC' or 100 would be the server identifier and then up
to 9 digits (1B) of sequence ID.

Which would IB/FB have faster select for identifying a particular
server, (either ID starts with 'ABC' or ID mod 1000000000=100)?

Anything else to consider here?

Thanks all,

Jerry