A failing HDD will indicate problems with it using either what you hear during operation or by what it records in an internal log of issues. To run a complete diagnostic costs nothing; it takes about five minutes to listen to the disk spin up and also to have a free utility read out from the built-in “black box” recording system. Here are both tools along with the responses that do not add insult to injury.
If your hard disk drive (HDD) makes clicking sounds or scraping noises or continues to spin up then come right back down then it failed your testing process; there’s no need to run an additional testing program since it only adds additional wear to the bad components.
First locate a quiet area, flip the computer on and devote the first minute to listening as the disk spins up; this is when the drive is most honest about its mechanical condition. A good disk starts out with a smooth humming noise and ends with almost no noise while it continues to make soft “tick” noises sporadically. There are only a few distinct sounds associated with a bad disk. If you hear a steady click (disk heads looking for a place they cannot find), a scrape or grind (an object is rubbing against something it should not be; stop the computer immediately), a beep or high whine (motor strained too much to move), or a repeated spin-up, sigh, silence (drive gave up, tried again) you have found your culprit. No matter what a utility says after that, you now know the drive has talked, so each additional second wasted is damage and the best possible repair location is a clicking-drive bench, not a download site.
If it’s quiet, read the SMART report. To view SMART on Windows use CrystalDiskInfo or on Mac you can use Disk Utility or smartctl, or use the makers’ version of it (SeaTools or WD Dashboard); whichever one you choose they’ll be getting their data from the same source, the HDD. Don’t be fooled by the one word statement and go by three different numbers: Reallocated Sectors (bad areas replaced with spares), Pending Sectors (areas which can’t be read at the moment, possibly containing your files), and Uncorrectable Sectors (sectors not read correctly no matter how many times it tried). A score of zero indicates health, and a low score which doesn’t change indicates a previous issue, however the important thing is to track change; any increase is a potentially dead drive. It’s also worthwhile glancing at Spin Retry; any value over zero indicates having difficulties with the spindle, which is more serious than having unreadable sectors. Review your age and temperature counters too. While these two pieces of information are informative regarding overall drive history, do not take them as final judgments. Lastly, many USB enclosures block SMART reporting entirely; therefore if your enclosure reports “no data”, this does not mean everything is okay.
When a warning appears you immediately want to test more: perform a full surface scan, run CHKDSK or use your manufacturer’s “fix” tool. Understand these actions as an exercise in repeatedly reading (and in CHKDSK’s case writing) on a damaged device which has indicated its damage. The mistake people make with CHKDSK is they mistakenly believe it is a health assessment, but it is truly a file-system repair utility; it can show you that the operation was successful while your damaged device is dying because of all the repeated loading. Do this instead: get the data out first. Save your important files, most valuable first, by backing them up to other storage devices while there are still readable files available. Once your important files are backed-up, then go back and diagnose the problem. Don’t perform any diagnostic tests before you have saved your important files. If the process of transferring files becomes unresponsive (the transfer stalls, your computer freezes, or a new noise starts), this will be the final warning you receive: you should now shut down your computer and send it to the professional hard drive recovery bench where specialised imaging kit performs a non-destructive read from failed disks much softer than any operating system.
Check twice a year for every critical asset. Also, check after any incident (a fall, moving house, or an unexplained freeze). Taking two minutes per check means the effort is not the frequency, it is the remembering. Associate this activity with one of your routines and take a screenshot of the status table each time you do it to allow for comparison instead of memory recall.
No. View any claims to be anything else with extreme scepticism. Reallocated or damaged sectors are forever; “repaired” sectors are merely hidden from your operating system’s eyes, and once a drive begins reallocation, there is no going back. Your best option here is recovery then retirement: take all the data off this failing drive and put it on new hardware and let this old drive go in peace.
Nobody can give you a true number. Some drives keep trucking along for months after receiving a warning; some are done by Sunday night; and ultimately, most often, the amount of time left before failure is determined by the use; each hour is another spin of the wheel. Don’t read “caution” as something to just keep an eye on; read “caution” as copy the irreplaceables off immediately, while you can still get at them easily.
We would like to hear from you about the noise(s) your HDD is making, or you can provide the SMART data screenshot for one of our engineers to review objectively. This will be a complimentary 48-hour diagnostic with a written price quote prior to performing any services.