> I was more thinking of a tool, which test the whole disc surface > and reports every bad sector.
badblocks -wvs (which takes forever, but in my experience is quite good at make failing disks actually fail ;) During the test you can monitor the smart attributes (smartctl -A, esp. the reallocated sector count) to see what is actually going on. Warning: backup your data as badblocks in destructive mode will wipe the disk clean. Also, you can run a SMART offline long test (smartctl -t long), which will scan the whole surface. But on most disks the test stops at the first error, so you don't get a complete picture. > Smarts is more of statistical kind: It counts events and tries > to calculate dooms day from that ;) That's just a part of it. Attribute values and the error log actually provide useful information about the status of the disk. andrea
