we have several servers running openSUSE, and the ones running 12.1 all have the same problem. The number of Apache processes is growing constantly and after a few days there are around 100 processes and the Webserver cannot be reached anymore until I restart Apache. On the servers with a heavier load this is happening faster. I cannot find any hints in the error_log.
You’re not providing enough information for anyone to troubleshoot.
Memory Leaks (which is more or less what you seem to be describing) can be caused by many things… A “memory leak” typically is when the machine allocates resources to accomplish a task, but when that task is completed the resources aren’t returned for re-use in a timely manner.
This may also be partly intentional because modern OS and applications often cache objects which are likely to be called again by a similar request/task, and often this is done by design to the full capacity of available resources, but you should never see this affect performance.
In other words, just because available resources are shrinking and “free memory” and the like is disappearing, that by itself is usually not anything to get excited about… But, if you start to see latencies and performance degradation then it <is> a problem.