F5 loadbalancer probing caused unnecessary logs into our Gateway. How to get rid of these logs

Document ID : KB000122399
Last Modified Date : 05/12/2018
Show Technical Document Details
Question:
The customer in this case has F5 load balancer doing a health check access via  "/". 

They are getting a lot of access log entries of the form :  

2018-12-05T15:40:46.077+1100 INFO    7651 com.l7tech.server.MessageProcessor: Request URL not resolved: http://odoma04-gw93a.ca.com:8080/

And do not want to see these in the log. 
 
Environment:
API Gateway 9.2
API Gateway 9.3 
Answer:

General Approach.
Generally the logging is at INFO level for auditing, unless there is an entry in the policy to raise it to a higher level. 

And when so then there is no entry written for the "Message Processed Successful". 

It's possible you have for the "Catch All /* service an entry raising the audit status. 

General recommendation is to add specific service for the F5 URL, so you can set the audit settings for that one separately to your "Catch all" default policy.

The customer did not want to adopt this approach.

Alternative Approach
 
The error messages are written by the "MessageProcessor" class.

We found we can set the log level for the MessageProcessor to a level above INFO then it will not print the "URL not resolved"  log message.

Here is how I did it : 
    The class in MessageProcessor, and can be set in global cluster property log.levels : 
   If we set the log level to SEVERE ie above INFO messages are not displayed. 

    com.l7tech.server.MessageProcessor.level=SEVERE

As per: 
User-added image

Then here in the ssg_0_0.log  you can see after the cluster wide property setting we no longer get the "URL not resolved" message: 

User-added image

In that class rhere are only two "info" level messages: 
    "Request URL not resolved" 
and
   "Tracing enabled on service but no trace policy configured" 
So those are the two messages that you will not see anymore after setting that log level .