Using archive log only mode on your source system

By default, the CDC Replication Engine for Oracle databases is configured to read online redo logs while Oracle is writing transactions to them. This is required for low latency replication when CDC Replication has completely processed the log records stored in the archive logs.

Oracle online redo logs can be a source of physical disk contention in that both Oracle and CDC Replication are accessing these files simultaneously. Therefore, the operating system must share the underlying hard disk device between reader (CDC Replication) and writer (Oracle). In this case, if the physical disks do not have sufficient I/O operations per second, both products will be constrained and performance will degrade. The Oracle database log writer may find that writes will "queue" and take longer to complete, and CDC Replication may be seen to stall on disk reads.

When latency of replication is not a primary consideration, the CDC Replication Engine for Oracle databases can be configured to read from archive logs only. Oracle can be configured to write archive logs to a separate file system which is not shared with the online logs. This way, disk access by CDC Replication log reader does not affect Oracle log writing throughput. CDC Replication will only start reading archive logs after Oracle has completely finished writing the file. If your source database is overloaded, this type of configuration can remove log and disk contention to improve performance.

This type of configuration may be used to resolve a Log Reader bottleneck.