Hard Disk Monitoring

Hello,

the "health status" of a server running HD-Sentinel v4.40 just dropped a point from 100% to 99% these days. The reason is apparently the "wear indicator" of a Crucial M4 SSD.

Questions:

1.) Why the reduction in overall "health"? Isn't this a perfectly normal - and expected - occurence over the lifetime of a SSD? I would classify a SSD with 80% remaining life expectancy no less "healthy" than one with 100% remaining. Is there something else behind this?

2.) Why don't I see the attribute change in any log? It's not mentioned in "Log" (this is blank, stating "no problems logged") nor is it in "Alerts" (also blank).

I have set the "health calculation method" to "analyse data field (recommended for servers)", if that makes a difference.

Regards

Hello VGER,

The memory cells in solid state devices experience wearing during each write operations and each cells tolerate only a limited number of overwrite passes.
The "wear-leveling" feature of the SSD tries to hide/minimise this effect but generally the status of the memory cells of the SSD slowly but surely degrade.
Most SSD devices report an overall health of the memory cells by various attributes. Hard Disk Sentinel reads them and displays that (as no problems found), the health of the SSD is determined by this overall health of the memory cells.

> 1.) Why the reduction in overall "health"?

Because of the writes affecting the SSD.

> Isn't this a perfectly normal - and expected - occurence over the lifetime of a SSD?

Yes, this is completely expected and normal. This is why the text description shows that there are no problems - but the health of the SSD decreases.

> I would classify a SSD with 80% remaining life expectancy no less "healthy" than one with 100% remaining. Is there something else behind this?

I completely agree you. Exactly because this is normal, I'd not say an SSD with 80% health is "bad". It is still good, just experienced some wear.

> 2.) Why don't I see the attribute change in any log? It's not mentioned in "Log" (this is blank, stating "no problems logged") nor is it in "Alerts" (also blank).

Exactly because of the above. As the change of this value (and thus the health) is completely normal and expected for an SSD, it would be wrong idea to issue alert on this, especially because if the health is very high, it causes absolutely no problems.
Just an indicator that the SSD is used and experiencing wear.

Of course you can configure alert(s) when the health reach lower values (yellow and red alarm levels) - as then you'll be alerted when the SSD experienced really lots of wear and a replacement may be considered.