Reduced performance usually occurs when the operating system or machines are overloaded as a result of increased demands on the system. Reduced performance can result from adding applications or Replication Server components, executing very large transactions, or even upgrading the operating system.
Performance problems can be critical or non-critical. You can eliminate non-critical performance problems by optimizing your replication system. For non-critical performance problems, see the Replication Server Administration Guide Volume 2.
If critical performance problems are left unresolved long enough, performance degradation can lead to fatal problems, such as full stable queues (see “Poor performance and stable queue size increasing”), in which replication stops. A replication, materialization, or dematerialization failure can be caused by a critical performance problem.
Critical performance problems can stem from:
Introducing new components, such as Adaptive Servers, databases, Replication Servers, RepAgents, or Replication Agents may cause resource contention and overload any component.
Changing the operating system—such as upgrading the operating system, applying patches, changing kernel parameters, or rebuilding the kernel—may adversely impact your replication system, memory allocation, and resources.
Adding applications to the replication system may impact memory requirements and use resources.
Replicating a very large database may produce a very high latency. Large transaction or an open transaction is also a possible cause.
