What Does SMART "Retired Block Count" really mean?

[andrew.levine]

Hi;

I am using a couple of TS869L RAID arrays - a primary and a secondary.
Had a drive failure a few days ago - and hot swapped the drive out.
Took about 12 hours to rebuild: considering the data size and usage logs, I think that was pretty good.

However;

I now see a couple of other drives with Status of "WARNING" under the SMART monitoring capability.
They have "Retired_Block_Count" figures beyond the recommended totals. Both drives have 100, and the threshold is 36.

I don't understand what the issue is exactly.

I tried to understand what retried blocks are used for, and there is scant info on the SMART / Wikipedia and Qnap websites.

Before I go around providing recommendations on what to do next (yes, I hear you saying "replace those drives, they are failing"), I'd like to understand WHAT is failing on those drives? Are retired blocks physical or virtual? Are there 1000's of them or are they dynamically generated as needed? Where are they in the priority of "issues to solve now!". Is it a minor issue, major issue? Is this REALLY a predictor of a failure, or a good maintenance indicator? In other words, is this about failure or preferred operating parameters.

All input is welcome
AL.

[doktornotor]

Google does not work?

https://www.google.com/?gws_rd=ssl#q=re ... ount+smart