Understanding SMART

Topic

New Reply

I have an ancient laptop that’s still going. It’s started to slow up a bit lately, and I’m suspecting the HD. I’ve run HDTune, to examine the SMART data. But I don’t understand what it’s telling me. for 01-Raw Read Error Rate (which seems like a good place to start), it’s “current” is 100, “Worst” is 100 and “Threshold” is 46.
Seems bad, but HDTune marks it as ‘OK’.
What is “Threshold” – above which is bad? Why 46? Seems very unclear. The HDTune website doesn’t really explain it either.
Any insights most welcome.
Thanks
Peter

Reply | Quote

Replies

Hi Peter, your not alone. I suggest you look here.

Reply | Quote

Thanks. That website doesn’t give me any real information.
I keep finding conflicting information. For example:
Looking just at Raw Read Error Rate, my HD is Current 100, Worst 100, Threshold 46, Status Ok.
So wikipedia says that lower is better. In which case 100 sounds bad, it’s miles above the threshold, but the Status says Ok. That doesn’t make sense.
But then ariolic says that “Lower values indicate that there is a problem”. In which case I’m OK.
But then this page says ” If the Value has reached and OVER this threshold value, with very high probability the harddisk is in trouble”.
This page does make a little sense with “More errors (i.e. lower attribute value) means worse condition of disk surface”
I can’t find any authoritative information – does any one know where I could find that?
Peter

Reply | Quote

SMART isn’t really that smart, and to boot, you have to interpret it’s attributes in comparison to previously displayed results, that is, if you have actually saved and stored a copy somewhere.

SMART may only be effective in detecting predictable failures. (those failures that
occure over a period of time) SMART has no use in non predictable type failures that
occure suddenly.

What is S.M.A.R.T.?

The attributes available for the disks can change from disk to disk and not all attributes are available for a disk. The values for each attribute are, most of the time, normalized values, with a maximum of usually 100, 200 or 253. The table in HD HeartBeat shows five values for each attribute:

• Current – This is the value read for that attribute
• Worse – This is the worse value found for this attribute
• Threshold – This is the minimum value for the attribute. If the attribute reaches the threshold value, the disk is about to fail
• Data – This field stores extra data for the attribute. What’s in it depends on the attribute and the manufacturer. For example, attribute 194 stores the temperature in degrees Celsius. For some manufacturers, it also stores the minimum and maximum working values. Attribute 9 stores the number of hours, minutes or seconds (depending on the manufacturer) the drive was powered on, and others, like 5 store the number of reallocated sectors
• Flags – This field shows the flags associated with each attribute. This is set by the manufacturer and may vary from drive to drive. The flags can be a combination of these letters:
• F – this is a pre-failure check: values close to threshold indicates imminent failure. When the threshold is reached, the disk is about to fail
• C – result of online collection
• P – this measures disk performance (lower values indicate low performance)
• E – measures an error rate
• V – measures an event count – Ex. Attribute 5 indicates how many sectors had errors and were relocated, in the Data field
• S – attribute is preservable and is restored in each S.M.A.R.T. test Lower values indicate a decrease in the reliability of the disk.

Reply | Quote