database - Identical messages committed during a network partition -

i'm working on distributed database. i'm in situation where, during healing of partition (nodes beginning recognize ones split from) 2 different clients try , commit compare-and-set of 3 4, , both successful. logically, should not possible, i'm curious if there functional problem both returning successful. both clients correctly believe final state is, , command sent out successful. can't think of serious problems. there any?

the "standard" definition of cas (to extent there such thing?) guarantees @ 1 writer see successful response particular transition. couple examples depend on guarantee:

// generating unique id while (true) {   unique_id = read(key)   if (compare_and_set(key, unique_id, unique_id + 1)) {     return unique_id   } } 

if 2 clients both read 3 , execute compare_and_set(key, 3, 4), they'll both think they've "claimed" 3 unique id , may end colliding down road.

// distributed leases/leader election while (true) {   locked_until = read(key)   if (locked_until < now()) {     if (compare_and_set(key, locked_until, now() + ten_minutes)) {       // i'm leader ~10 minutes.       return;     }   }   sleep(ten_minutes) } 

similar problem here: if 2 clients see lock available , both cas acquire it, they'll both believe leader @ same time.