In other words, the coordination
middleware is not becoming a bottleneck.
Summary
Online analytical processing must cope with huge volumes of data and at the same
time allow for short response times to facilitate interactive usage??”and the requirement
that the data analysed should be up-to-date is becoming more and more important.
This chapter has presented database clusters as a scalable infrastructure for
interactive decision support systems that are capable of analysing up-to-date data.
We discussed central architectural issues and performance aspects, such as several
physical design alternatives, possible query routing algorithms, and innovative update
propagation protocols.
Most data warehouses nowadays offer a compromise between freshness of data and
maintenance costs. But recent developments such as FAS (freshness-aware scheduling)
or relaxed currency constraints allow to explicitly trade freshness of data for
query performance, while at the same time not sacrificing correctness:
1. The requested freshness limit of queries is always met; and
2. Data accessed within a transaction is consistent, independent of its freshness.
The presented middleware-based cluster architecture proved to be very scalable
and it also shows that freshness-aware scheduling effectively allows users to trade
freshness of data for faster query response time.
Pages:
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470