Link to home
Start Free TrialLog in
Avatar of Michael McGovern
Michael McGovernFlag for United States of America

asked on

MD3000i RAID controller module Memory Parity Error!

We have a DELL MD3000i array with dual-controllers (A/A). Last week, one of the controllers went bad due to memory parity error (or aka. Memory Consistency Error) and we took a huge performance hit. We ordered a replacement and swapped it out, the error was still there.  We thought we may have received a defective part, so we got another controller, but the error doesn't go away. We took out the battery and memory from the controller, waited 2 mins, then popped it back in, but still no luck. Does any one know how to correct this issue? Does something else needs to be done beside putting the controller in offline mode before swapping?

Summary : RAID Controller Module Memory Consistency Error
Storage array:  MDPLSHELP01
Component          reporting problem:  RAID Controller Module in slot 0
 Status:           Online
 Location:  RAID Controller          Module/Expansion enclosure, RAID Controller Module in slot 0
Component          requiring service:  RAID Controller Module in slot 0
 Service          action (removal) allowed:  No 
 Service action LED on component:  No
 Replacement          part number:  
 Board ID:  1532
 Submodel          ID:  63
 Serial          number:  63V10XX
ASKER CERTIFIED SOLUTION
Avatar of Coolie Sheppard
Coolie Sheppard
Flag of United States of America image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of Member_2_231077
Member_2_231077

AFAIK the cache modules have ECC memory so can;t have parity errors. Memory consistency us more likely to be the cache on one controller not matching that on the other after mirroring one to the other. I would shutdown and power off, then clear the cache on the other controller. i.e. do the same battery removal on both.

You can set mirrorcacheenabled=false with SMCLI to get rid of the performance hit when only using one controller but obviously that is a risk as write cache is only stored on one controller in that mode.
Avatar of Michael McGovern

ASKER

We're going to try the method suggested by andyalder during our maintenance window tonight.
Powering down the array and pulling out both controller with battery reset didn't work. Updating the firmware cleared the error! Thanks everyone.