well, you get the idea.
If mirroring is run with asynchronous communication, as it is in High Performance mode, there is
always the chance that the principal and the mirror will be out of sync due to network holdups, deliberately
pausing mirroring operations, or large amounts of data changes in a short period of time. If the
principal fails at any point while the two nodes are in a synchronizing state, there is a chance that the
mirror node will not represent all of the data that was present on the principal. Bosses and executives
generally don??™t like it when data is lost. Granted, High Performance mode does its best to stay in sync,
and in most normal operations it is in sync, but remember, what can go wrong probably will go wrong.
Always assume the worst when designing for disaster recovery.
Mirroring Mode:High Availability
Finally, the pi??ce de r?©sistance of mirroring modes: High Availability. This mode includes
a principal, a mirror, and a witness. When the server is in High Availability mode, communication
is synchronous. Not only is failover automatic, but both failover and failback
can be done literally at will.
Pages:
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401