If you receive I2C messages in the TTY debug logs of the series 8 RAID controllers that were introduced with 12G servers, this might not be caused by a hardware failure.
There are three I2C buses given:
If the H310 busy to do data transactions, specially when both JBOD and VD access simultaneously, the I2C interrupts every 5 seconds from CEM (no functional impact here). As it is being missed, this is causing those events.
During each boot up, the iDRAC has to communicate with the RAID controller to do inventory, status, and other functions. Communication between the RAID controller and the iDRAC is done via a bus called I2C.
Several messages were put in the RAID controller log during development to assist in resolving issues but many of the messages were not removed at the end.
Here is a sample of what will typically be seen:
09/04/12 18:55:37: I2C 1 reset!
09/04/12 18:55:42: I2C 1 reset!
09/04/12 18:55:42: I2C 1 reset!
09/04/12 18:55:43: I2C 1 reset!
09/07/12 20:42:46: One Wire Device detected fd000002aa470533
09/07/12 20:42:46: EepromInit: Family=33, SN=0547aa020000
09/07/12 20:42:46: One Wire Device detected fd000002aa470533
09/07/12 20:42:46: One Wire Device detected fd000002aa470533
09/07/12 16:00:34: ds1961s_ReadMemoryAuthenticated: EEprom CRC error in Reading AUTHENTICATED data
Given the nature of I2C this can potentially lead to thousands of TTY log entries around I2C device detection and negotiation.
These messages are innocuous and don't indicate that the RAID controller or the iDRAC is/are malfunctioning. No hardware has to be replaced based on the observance of these messages.
These TTY Log messages are filtered by newer firmwares of the PERC 8 Series controllers, starting with revision 20.12.1-0002 , which improves (reduces) I2C messaging in TTY log.