rasdaemon
rasdaemon copied to clipboard
rasdaemon: support memory corrected error statistics
A high volume of Correctable Errors (CEs) indicates that the memory controller is frequently performing Error-Correcting Code (ECC) operations, which will increase memory controller latency. The CE statistics feature can report the number of CEs occurring per second. When the count exceeds a certain threshold, it signifies intensive ECC activity and triggers warnings.
New enviroment MC_CE_STAT_THRESHOLD to setup threshold.
LGTM. Reviewed-by: Shuai Xue [email protected]
Merged, thanks!