Log4J Logging vs wM built-in Logging feature

Hi ,

Can some one please help me understand when do we go for custom logging and when to use built-in logging feature provided by webMethods. Are there any specific advantages / disadvantages? wht is the parameter based on which this decision can be taken

In all of my various projects, I’ve always used the wM-supplied logging facilities. The projects used both the usual server and error logs, as well as custom-named files that used the same underlying logging facilities of the pub.flow:debugLog service.

Others have indicated they’ve used log4j for various purposes. Perhaps one or more of them will share their thoughts.

I have used both built-in logs as well as custom logs.

When you wanted to log the entry/exit of each service along with the request and response xml, custom logs is more helpful. Also we can easily identify warning, info, fatal, error logs. Serverlogs can be mostly used for other IS logging.

One thing i would suggest is that, if you are using JDBC appender, and if you have run time processes (publish and wait) where you call the database to perform certain biz requirments, It might affect the performance.

I couldnt understand… what is that JDBC appender?? is it any inbuilt wM service :confused:.
And how will it affect the performance if we are using a custom logging when we call the database. Please could you explain in detail.

Thanks
Sue

JDBC appebder is not a wM service. These are lists of log4j appenders (http://www.allapplabs.com/log4j/log4j_appenders.htm).
For instance file appender logs all the events to a file. JDBC appender is used to log events directly to a database.

If you are appending all log events to a database, it will use some of the connection pool and it will decrease the availability of a connection pool for the integration server whenever a jdbc adapter service calls the db.

BTW
where are you planning to log the events using log4j?
Do you have time critical processes (such as search functionality, validation) where you call the database constantly?