The Log Server is used to consolidate logging from all the other parts of Ceptor. Every component uses SLF4J to log all errors and debug statements, and SLF4J is implemented by either Logback or Log4h - both can be configured to forward log events to the Log Server.

The administration interface within the configuration / statistics server can be used to view the last 1000 items logged, the rest is written to a log file, or logged by other means, using a logback/log4j appender. See the section on logging, and the configuration section for details on how to set up logging to other destinations, like databases, NT event log or others.

Clustering

Log servers are (like Dispatchers) easy to cluster, but it has the drawback that it makes management a little harder, since instead of having a single consolidated logfile in which log from all components are stored, you now end up with multiple files. The log servers do not communicate with each other, and are not aware of each others existence.

As with Configuration Servers, there are no performance gains of clustering log servers – clustering them is only useful for failover purposes.