Mendix application stopped producing logs and is unreachable
Hello, I'm working on this Mendix 4.3.2. application which runs locally on our clients server. Last monday the application was unreachable and I found out that there where a lot of active processes. Today I was doing research to find out why the application was unreachable and how it is possible that there where so many active servers actions. I found out that the Java heap memory never was to high and that there where no error logs in the application logs. The only thing which was different was that the logs suddenly stopped at 14:40 and then they started at 07:00 the next day again. While normally there are many logs in between this times. 2015-03-22 14:40:25.154 INFO - Core: Login OK: user 'user' (Number of concurrent users: 21). 2015-03-23 07:00:06.051 INFO - Connector: New login request from [Source-IP: 127.0.0.1, X-Forwarded-For: 10.10.1.171, User-Agent: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 6.3; WOW64; Trident/7.0; .NET4.0E; .NET4.0C)] My questions are: How is it possible that a Mendix application is unreachable without producing any error logs? How is it possible that a Mendix application stopped producing logs without stopping the service?
What I have seen is that sometimes the log thread stops and then no logging output is generated anymore.
The reason why the logs are stopped is because the server is to busy. When resources get scarce the server has no longer time for the logging process. First thing you should check in a state like this is what is the oldest (longest) running microflow. This will give you an indication where to start looking. Try to examine the microflows on loops etc. Because if a microflow gets in a loop it will eat up one core. The user does not see any result and does press the button again resulting in a second core at 100 %. This way your application can go down fast.