It should be noted that RAM failures are extremely rare. The most common causes for RAM failure are humidity, electrostatic discharge, and dust; in a climate-controlled data center these conditions are nonexistent. The more common failure is the DIMM slot on the motherboard itself. However any of the symptoms listed on this page should be taken seriously and we are always available to provide hardware support for any suspected failures.
System locks during normal operation under different operations and loads: the most common symptom of a bad RAM module is the system hard freezing during normal operation. Due to the random nature of RAM access, it may take a few minutes or a few hours (or even days in some cases) for the bad portion of memory to be used by the server. As soon as the faulty area is accessed, the system halts completely.
Crashes during memory intensive operations.
Failure to boot. This can be accompanied by repeated long beeps, which is the accepted BIOS beep code for a memory problem. In this circumstance, you cannot test the memory with diagnostic software, so your only option is testing by replacement. Of course this can only be determined locally, but it one of the most obvious signs when a server halts since the beeps are very loud and easily decoded.
If you suspect at any time that your RAM is bad or going bad you should open a hardware diagnostic support ticket via MyCP so we can assist you immediately. We have special PCI diagnostic cards and CDROM bootable Linux distributions with MemTest86 which can help determine if the memory is going bad.