Don??™t nag.
Monitoring server: All right, you two. Let??™s see some ID and a current status!
Primary server: Primary here. I??™m OK.
Secondary server: Secondary here. I??™m OK, too.
Monitoring server: OK, I??™ll check back in 30 seconds.
Secondary server: Boy is that guy annoying.
Primary server: Hey secondary, are you OK?
Ad infinitum.
If having some sort of automated failover is required, this type of communication is
a necessity. However, it adds a small resource drain on all participants, plus any disruption
in the conversation can result in a false indication of system failure.
Log shipping avoids both resource requirements and false-positive system failures.
Neither the primary server nor the secondary servers need to know if the other even
exists, much less what its status is. A monitoring server may be involved in a log-shipping
scenario, but it too doesn??™t care so much about the state of the servers. Instead, it cares
about the success or failure of the jobs that facilitate log shipping in the first place. Plus,
primary and secondary servers have no need to answer annoying requests for status from
a monitoring server.
Pages:
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290