In general, techniques are described for managing address spaces across network elements. A network device including a processor may be configured to perform the techniques. The processor may execute a pool manager that automatically distributes a first block of network addresses to a first network element acting, for a first network, as a first address allocation server to assign the first block of network addresses. The pool manager may further automatically distribute a second block of contiguous network addresses to a second network element acting, for a second network, as a second address allocation server. The pool manager may then dynamically manage a size of the first block of network addresses and a size of the second block of network addresses to address exhaustion of available network addresses within either or both of the first block of network addresses and the second block of network addresses.