1

I have a NAS which has been running 24/7 for a little over 3.5 years. So far I've had no drive failures, however I was wondering when I should swap them out with newer drives. Do WD Red drives tend to last a long time?

5
  • 5
    You swap them out the day before they fail; or, if your psychic abilities are not up to that task, you keep a good backup strategy & swap them the same day they fail ;)
    – Tetsujin
    Jul 15, 2019 at 16:44
  • 1
    Surely you have an appropriate RAID level for your setup...?
    – Daniel B
    Jul 15, 2019 at 17:05
  • Yes (sorry I didn't specify), I am using 4 drives in RAID5. Jul 15, 2019 at 17:28
  • You swap them out when they're broken or when you upgrade. My oldest drives have been spinning non-stop for more than 7 years. WD Red tend to be fine, but others are fine as well. I recently had a WD Red fail within a year. The most important is redundancy,\which will give you time to replace a broken disk without disruption. You have RAID-5 which is just about the cheapest form of redundancy (in terms of actual versus usable capacity).
    – StarCat
    Jul 15, 2019 at 18:01
  • 3
    Depending on drive size, RAID 5 is not appropriate. While rebuilding, more disks may fail due to the heavy load.
    – Daniel B
    Jul 15, 2019 at 18:48

3 Answers 3

1

As stated by DrMoishe Pippik, "There's no point in replacing a drive with good SMART statistics" (plus, un-needed replacement means un-needed heavy workload on the remaining disks to rebuild the array).

Then the choice is between:

  1. waiting for a disk failure to replace it
  2. periodically monitoring the SMART reports and replace a disk as soon as a report shows "pre-fail" signs

Drawbacks of 1):

  • if another disk is near-failure at that time, it has more chances to fail during the rebuild of the array (heavy workload...)
  • you have to replace the disk ASAP, in order to minimize the period during which the array runs without redundancy. During the rebuild, the NAS will have degraded performances

Advantages of 2):

  • you can more easily chose the moment for the replacement/rebuild, for instance at night or during the week-end, when the NAS is less solicited

Clearly, 2) is the way to go...

0

If you have setup email notifcations correctly you can configure your synology to email you if there are any issues with the disks. RAID 5 offers you one disk redundancy.

-1

At the very least you should expect to replace them by the end of the warranty period - that's how long the manufacturer believes a fair proportion of them will last.

For a (rather large) real-world data point, see Backblaze's hard drive stats.

2
  • 1
    That's no good if one fails & you didn't have a backup. MTTF is as good as a guarantee... ie no use at all if you get one that fails early... & a waste of money if you throw them all out after 5y1m just 'because the warranty ended'.
    – Tetsujin
    Jul 15, 2019 at 16:56
  • 3
    Device failure is driven more by the laws of probability than by date. Some drives fail early due to flaws, e.g. manufacturing defect or head crash, another portion accumulate minor problems (e.g. bad sectors) until failing, and a tail of the curve may last for decades. There's no point in replacing a drive with good SMART statistics. Jul 15, 2019 at 17:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.