Setting the degree to a lowest common denominator value (for example, 2) provided
effective parallelism for higher user counts, but did not leverage resources fully when
fewer users are active.
Self-tuning adaptive parallelism
Oracle8i introduced the notion of self-tuning adaptive parallelism. This feature automatically
scales down parallelism as the system load increases and scales it back up
as the load decreases. When an operation requests a degree of parallelism, Oracle
will check the system load and lower the actual degree the operation uses to avoid
overloading the system. As more users request parallel operations, the degree they
receive will become lower and lower until operations are executing serially. If activity
decreases, subsequent operations will be granted increasing degrees of parallelism.
This adaptability frees the DBA from the difficult task of trying to determine the
optimal degree of parallelism in the face of constant changes in workload.
Adaptive parallelism takes two factors into account in determining the degree of parallelism
granted to an operation:
??? System load.
Pages:
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388