Unsolved
This post is more than 5 years old
1 Message
0
2918
PowerEdge 1950 FBD Link Failed to Train
I recently brought my old PowerEdge 1950 out from storage, booted it up and found the following error: "Error: Memory failure detected. Memory size reduced. Replace the faulty DIMM as soon as possible.
Then lists my memory config (8x2GB)
The FBD link to the following DIMM failed to train: DIMM 6 & DIMM 2
FBD training error - The following memory branch has been disabled: Branch 0"
Can anyone help with this error? Do I need to re-configure the DIMM slots?
DELL-Daniel My
Moderator
Moderator
•
6.2K Posts
1
June 17th, 2016 17:00
Hello
It is unlikely that the error is caused by a population issue. Link training is the process of a controller trying to find a common communication method with a device. Link training failure typically indicates either device failure, slot failure, device not fully seated, or an incompatibility between the controller and device.
I would start by reseating memory, if the error persists then I would test single DIMMs in slot 1. If the error goes away when using a single DIMM then test each DIMM by themselves in slot 1 to see if you have a faulty DIMM.
Thanks
TechnicalEngineeringKLM
70 Posts
1
June 20th, 2016 00:00
Hi ACONLEY,
What Bios version are you on? You said, we brought an old server back to live. It's a known issue that Bios version 2.x is very buggy concenrning memory errors. i would advise you to first upgrade the Bios.
kgrds
Peter
CharlesIsWorking
167 Posts
0
January 24th, 2022 09:00
Any updates to this thread? I know it's old. I have a Poweredge 1900 I was looking at using again, and got the same error with slots 2 and 6.
DELL-Chris H
Moderator
Moderator
•
8.4K Posts
0
January 24th, 2022 11:00
CharlesIsWorking,
I would start with trying to swap dimms 2 and 6 with other matching dimms in the server, then restart and verify if the error follows the dimms. If the error remains, I would verify in the hardware log if there is any errors mentioning Logging being Disabled, if so then clear the hardware log and then restart. If still on those slots then there may be an issue with the slots themselves. Now if the error follows the dimms to the new slots I would look at replacing those dimms.
Also I would recommend updating the server as well, but let me know the revisions you're currently at if you are really far behind, as jumping too many updates or going out of order can cause issues as well.
Let me know what you see.