TagmaStore USP Problem

[timwah]

Hi all,
Don't know if we have any hardware gurus reading this...
I have a USP single frame system that I'm doing a full hardware reconfg on. The system keeps calling out failed cache memory dimms in locations that have none installed. Hardware and SVP are configured correctly (this is one of 4 systems I'm working on. The other three work fine). Have replaced the cache, cache board, shared memory, 16HS channel board, and DC power supplies. Anyone ever run across a failure like this?
I'm running out of things to try.

 

[timwah]

Update
Bad cache memory in the base positions of cluster 2 was causing the failures in cluster 1 ??? I configured the system down to base cache, and the failures moved from cluster 1 to cluster 2. Fixed the cluster 2 failure, and brought the system back to 20gb cache one feature at a time. Seems to be working fine now.

 

[thomas28]

Hi timwah,
I'm working for more than 20 years with Hitachi Storage systems, never heard of such a problem. It seems to me that some wrong configuration information was hold in shared memory which caused the error. I don't think it is possible that a bad cache dimm can cause a problem in the other cluster. Your problem seems to be solved, so it doesn't matter anymore!
kind regards
Thomas

 

[timwah]

Hi Thomas28,
Thanks for the reply.I've been working with Hitachi equipment for a bunch of years... back to the 7700 standard days. I agree that a bad dimm in one cluster should not be able to cause failures in the opposite cluster. Could be that I had some bad memory on both sides, and bringing it down to base removed the bad dimms from cluster one. I was working with an inventory box of about 100 dimms... but no matter now. System is on the way to South Africa? China? somewhere far from Chicago.
Best Regards,
Tim