Excessive single bit memory errors in system event log - IBM eServer xSeries 240 and IBM Netfinity 5600

Resolving The Problem

RETAIN tip H136493

• Excessive single bit memory errors in system event log
• System may hang

The system is any of the following IBM eServer xSeries servers:

• an xSeries 240, type 8664, any model

The system is any of the following IBM Netfinity servers:

• a Netfinity 5600 server, type 8664, any model

The system is configured with more than two of the following option:

• IBM 512MB 133 MHz SDRAM ECC RDIMM, option part number 33L3062, replacement part number (FRU) 33L3063

All the RDIMMs have header code xxY6Y (where xx does not matter) as shown on the large barcode label, example:

11S38L2942ZY1Y6Y78400K

Note: The 5-digit header code field of xxY6Y begins at the 12th letter.

The system has the symptom described above.

Make sure the system has no more than two 512MB memory RDIMMs with header code xxY6Y, and that both RDIMMs are in memory slots DIMM 4 and DIMM 3, also labeled as J1 and J2, respectively.
 
The third and forth RDIMMs can be installed in the remaining slots.
 
Note: If the third and/or the forth RDIMM is a 1GB (1 GigaByte) memory module, the system may issue the following message during POST: "290 - Memory configuration is not optimized"
 
Note: This message does not affect the function of this configuration.

The behavior described above is caused by marginal loading on the memory bus.