-**Fault tolerant** - Jobs must be replicated to N nodes before Enqueue() returns. Jobs must be ACKed or they'll be re-queued automatically within a specified Retry Timeout.
-**Distributed** - Disque pool. Multiple producers, multiple consumers.
-**Fault tolerant** - Jobs must be replicated to N nodes before `Add()` returns. Jobs must be `ACK()`ed after `Get()` or they'll be re-queued automatically within a specified `RetryAfter` timeout.