VSAM RLS assigns locked resources to an entry value in the lock table, and uses this entry value to quickly check whether a resource is already locked. If the lock structure (and thus the lock table) is too small, many locks can be represented by a single value, making 'false' lock contention possible. False lock contention occurs when two different locks on different resources attempt to use the same lock entry. The second lock requester is suspended until VSAM RLS determines that there is no real lock contention on the resource. False contention can be a problem for workloads with heavy R/W interests among systems.

For the best performance, you want to achieve the least possible amount of global lock contention, both real and false. The amount of real lock contention is application-dependent; it depends on record access patterns. False lock contention is almost entirely determined by the size of the lock table, with a larger lock table having less false lock contention than a smaller one. A good goal is to have total (real and false) global lock contention of less than one percent. The false contention component of the total global lock contention should be less than five percent, and ideally, should be substantially less than this. The documentation references one half of one percent as the false contention rate to be aware of. However, in some scenarios, a false contention rate of more than one half of one percent may be normal and will not cause performance degradation.