Whereas active replication servers are active, in that all of them receive, process, and reply
to requests, passive replication (Budhiraja, Marzullo, Schneider, & Toueg, 1993) is an asymmetric
scheme. With passive replication, only one server processes requests and replies to
the client. This special server is called the primary replica, and the other replicas are called
the secondaries or backup replicas. Client requests are sent to the primary, which processes
them. After the primary has processed each request, it checkpoints its state to the secondaries,
and after the checkpoints terminate, it replies to the client. This checkpointing ensures that
all replicas will have the same state after each client request. In case any secondary fails, no
special action needs to be taken, but if the primary replica fails, some extra recovery actions
need to be performed before resuming the service activity: At least one of the secondaries
has to be promoted to the primary role, and clients have to locate it and bind to it.
The main advantage of active replication over passive replication is that active replication
does not require any recovery action when failures occur. On the other hand, its main
disadvantage is the need of a group communication protocol that sometimes can be quite
expensive.
Pages:
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532