Figure 8-4. Shown is a less abstracted redo log buffer based on Oracle9i Release 2 and later. Similar to the log buffer shown in Figure 8-3, each strand of redo and the general log buffer have allocation and write markers. When DML occurs, space is allocated in a strand of redo, protected by one of the redo allocation latches. When a commit occurs, the related strand is flushed to the general log buffer, and then the general log buffer is flushed by the log writer background process.

By default, the number of redo strands is dynamic, but it can be made static by setting the hidden instance parameter _log_parallelism_dynamic to false. When Oracle is dynamically controlling the number of redo strands, the maximum number of strands is controlled by the hidden instance parameter _log_parallelism_max. The DBA can specifically set the number of redo strands via the hidden parameter _log_parallelism. The default number of redo strands is surprisingly low-perhaps two.

The number of redo allocation latches is controlled by the instance parameter processes. Around each eight processes defined, one redo allocation latch is created. Figure 8-5 shows both the query and output (with many output lines removed) of an Oracle Database 10g Release 2 system with 4,000 processes defined. Oracle created 488 redo allocation latches, which amounts to 8.2 processes for each redo allocation latch. This ratio of processes to allocation latches has been observed for many different processes parameter settings.