I'm wondering if anyone has a solution to this problem.  I've seen it on and off with various Access Manager environments and versions probably for over 10 years.  I've opened support cases now and then and they usually just linger with no solution until they are closed months later.  Basically what happens is that when rotatelogs runs it hangs on one of the log files and causes 100% utilization.  When this happens, all the proxies stop responding.  I think this happens because rotatelogs is a piped logger and apache just waits to service requests until the logging function returns (or maybe it is caching connections and runs out of memory to do that, I don't know, just guessing).   I typically get a response from support of either "you have too many log files" (I've seen it happen on servers with only several dozen) or "it's part of Apache, not our code, so we can't do anything."  So I'm trying to find other options.  Usually killing the stuck process frees things up. Sometimes I have to manually move the log files out of the way because the thread just restarts on the same log file and goes to 100%.  I can keep killing it but it just keeps coming back.  

I thought about using something else to rotate the logs, but if you move the active log I believe apache will cease to log.  You have to suspend apache to rotate the logs.

Short of stopping/restarting novell-apache2 nightly, has anyone come up with a solution to this?

Matt