cluster computing - SQL SERVER replication or mirroring configuration to achieve the following scenario? -

suppose have master server called master

a "slave" server called a

a "slave" server called b

a "slave" server called c.. , on.

every slave server have several clients.

the environment should work described below, taking server example:

when change made in a, replicates it's changes master server. then, master server replicates changes across slaves. (b,c,...).

if slave server lose connection master server, server clients can still operate , connecting .

when connection back, slave server should upload changes master server, , download every change made other slaves, through master server well.

any guidance on configuration should use appreciated.

what you're describing sounds merge replication.