Understanding smartctl -a output

I find the self-test results very reliable, and they are self-explanatory (either the last one run failed, or passed).

The various vendor-specific attributes are just that. There is actually no standardized way to interpret them (which is why smartmon tools maintains a drive database with interpretations for those values). You can find a description of the meanings of many if the flags here: http://en.wikipedia.org/wiki/S.M.A.R.T.#Known_ATA_S.M.A.R.T._attributes

The line SMART overall-health self-assessment test result: PASSED is derived from the values printed below it, translated, normalized, and given thresholds by the drive database.

For the normalized values, lower is usually better, but not all the flags indicate things that will themselves portend mechanical failure (the ones with a threshold are more likely to). Things like uncorrectable read errors, spin-up failures, etc. are likely indicators.

Your drive looks to be in fine shape, from those results.

Seagate uses SER (Seek_Error_Rate) to code two distinct counters: 16 high weight bits are used for Seek Error Count, and 32 low weight bits used for a Seek count. You would prefer hexadecimal display to ease your reading of these two counters (6 nibbles = 2 for error count + 4 for seek count).

RRER (Raw_Read_Error_Rate) does not show an incremental counter, but the result of something like -10 log (number of errored sectors / total bits on disk), which explains there is a min and a max. Stay near the max is better.

More explanations on this here : http://www.users.on.net/~fzabkar/HDD/Seagate_SER_RRER_HEC.html