Our 6026T-6RF+ server crashed suddenly and caused some data loss. When boot, it was warning that it had encountered “Multiple ECC memory errors” on the LSI RAID controller. Since the LSI is an onboard RAID controller there is no physical DIMM stick to swap out for it since the memory is built into the motherboard. I then shut it off and pulled the power to fully reinitialize the server on POST but it repeated the ECC warning. I then proceeded to boot the server with the warning that data loss might occur and while the server came back online we did experience data loss on 2 databases that are hosted on it.
Is one of the DIMMs bad?
Supermicro software RAID controllers with LSI 2008-based controller uses the motherboard memory. If you saw ECC errors like that, one of the motherboard memory may be bad.
The hardware controllers with LSI 2108-based controller have the cache memory built in the card. The RAID controller may be bad.
In your server, it has the LSI 2108 controller. The error is from the cache memory of the SAS controller. Please try to reflash its firmware first.
Print
Share