Yes. There are multiple things we can do to determine the cause of this. On servers that mysteriously crash, we always install a monitoring tool we wrote that logs the output of specific debugging commands a system administrator would otherwise run when logged onto a server having an issue. The logging includes the process table, memory usage, disk and CPU statistics, active network, as well as other things such as the MySQL process list. If the server crashes, our administrator can restart it then review these logs to see what occurred to cause the crash.