Caching of Oracle Sequences

In the high concurrency database system, the cache of sequences should be adjusted correspondingly. Now look at the definition of a high-concurrency sequence of the database itself. When we send a request to the database, listen to the reception, and then Oracle initiates a background process (the process is usually referred to as the number of database concurrency, that is, how many processes the database can start at the same time to provide services externally, determined by the processes parameter) for the service, The corresponding session Sesson is then allocated, and the session's Audsid field (not 0 is an external request, internally 0) is obtained from the sys.audses$ sequence, which is defined as follows

CREATEwith210016 2000000000 1    10000Noorder;

With a cache of 10000, session allocation can be categorized as high concurrency in a database, because many sessions are generated at the same time during busy times, while development-defined sessions are typically between 20 and 50, in which case the sequence-waiting event occurs when a high concurrency situation arises , if there is a sequence waiting event, it is recommended that you adjust the number from 50 to one in 500 to 1000 for the first time.

Caching of Oracle Sequences