Filemaker Server Advanced 11 & OS X Server 10.6 locking up, causing data loss
I'm running FMSA 126.96.36.199 on a Mac Mini "server edition" with OS X Server 10.6.4. It was a fresh FMSA install on a brand-new server. It's never been used for anything else.
At irregular intervals, but pretty reliably twice a week, the server becomes unresponsive. My users get their Filemaker connections dropped (and we usually lose a half-day's worth of data). I can't SSH or VNC into the server, but it will ping, and I can connect to Bjango's iStat Server running on the Mini. The machine requires a hard reboot.
This is what it looks like when it's running normally: http://yoki.com/pics/istat_good.png
This is what it looks like when it blows up: http://yoki.com/pics/istat_bad.png
This is what it looked like about an hour ago: http://yoki.com/pics/istat_worst.png
I have serverdensity.com monitoring the server, and in the snapshot for one of these incidents that happened today, the two processes that are going haywire are:
/System/Library/Frameworks/JavaVM.framework/Versions/1.5/Home/bin/java -Djava.util.logging.config.file=../conf/logging.properties -server -d32 -Xmx512M -XX:+UseParallelGC -Xbootclasspath/p:../../admin-server/WEB-INF/lib/jacorb.jar:../../admin-server/WEB-INF/lib/logkit-1.2.jar:../../admin-server/WEB-INF/lib/avalon-framework-4.1.5.jar -Dorg.omg.CORBA.ORBClass=org.jacorb.orb.ORB -Dorg.omg.CORBA.ORBSingletonClass=org.jacorb.orb.ORBSingleton -Djacorb.home=../../admin-server -DFMS.COMPONENT=fmadminserver -Djava.util.logging.manager=org.apache.juli.ClassLoaderLogManager -Xdebug -Xrunjdwp:transport=dt_socket,address=3999,server=y,suspend=n -Djava.endorsed.dirs=../../../Common/Tomcat/common/endorsed -classpath :../../../Common/Tomcat/bin/bootstrap.jar:../../../Common/Tomcat/bin/commons-logging-api.jar -Dcatalina.base=.. -Dcatalina.home=../../../Common/Tomcat -Djava.io.tmpdir=../temp org.apache.catalina.startup.Bootstrap start
/Library/FileMaker Server/Database Server/bin/fmserverd -d
both using about 98% of the CPU (I assume they're running on different cores). Neither process routinely uses a fraction of that. No other processes are using more that 1% of CPU capacity.
The incident occurred around 2:12 PM. There's nothing in the FM logs between 6:01:34 AM and then. There's nothing in the Mac system logs between 5:35:09 AM and then. It's all pretty routine-looking stuff.
If anyone has any idea what's going on, or if additional info would be helpful, please let me know. I'm very inclined to go back to FMSA 10.