BL660 Gen9 servers with degraded memory modules

[CRD]

Good day all.

Hoping someone migth have some advice for me.

I have a client that is using BL660 Gen9 servers in a C7000 enclosure. 

Their servers are giving degrade memory errors but the thing is when you go and replace the 2 degraded memory modules for example on the one server and you power it back on the problem of the degraded memory module is moved to another slot or slot and CPU location.

Even if you swap a module that shows as Good with Degraded one then the problem wont nessasarily move you will find that both shows Good now after the move and there will be other 3 memory modules now showing degraded that was previously showing as Good in a different location.

C7000 components and server components are on the latest firmware.

Thanks

[CRD · Accepted solution]

Hi

I went to the DC and managed to solve the problem. Did the same process on all 4 servers and it help. 

1. Remove all memory and only populate 2 memory dimms per CPU and then start up server fully. Check status is good then shutdown servers.

2. Populate another 2 memory dimms per CPU and then start up server fully. Check status is good then shutdown servers.

3. Carry on like the above untill all dimm slots are fully populated 

This was a tedious process doing it on 4 servers with each server having 4 CPUs and a total of 32 dimms but at least it helped and all servers are healthy now.

I did try with the 3rd to skip and just populate the rest of the dimms without doing it in pairs per CPU, the server then gave degraded dimms agains so it seems the process mentioned above of only adding 2 dimms per CPU each time is important.

[Sunitha_Mod]

Hello @CRD,

Perfect! 

We are glad to know you were able to find the solution and we appreciate you for keeping us updated.