Complex, multitier applications that consolidate the various parts of an enterprise system increase the load on high-volume database servers. When you address this by upgrading database hardware, from 4 to 8 CPU systems to 16, 32, or 64 CPU systems, you can drastically increase contention on such key resources as the log, the spinlocks that guard the log, and the address locks. Asynchronous log service (ALS) and optimistic index locking help resolve contention, increasing Adaptive Server scalability in systems using four or more online engines.
Optimistic index locking can resolve increased contention on some important resources, such as the spinlocks that guard address locks on the root page of an index.
Applications where this amount of contention might occur are typically those in which:
Access to a specified index constitutes a significant portion of the transaction profile, and many users are concurrently executing the same workload.
Different transactions, such as ad hoc and standardized queries, use the same index concurrently.
