My TrueNAS build runs exclusively the types of drives you mention. I buy Sun, Oracle, HGST (all of which are HGST He Data Center drives), usually in the ~4-5 year old range. So, right now I’m in the process of updating the 6X 8TB disks it was built with to 10TB disks. After a certain point, it becomes more expensive to find the used 8TB size, so I just update to 10TB, or whatever the next size up would be.
Say what you will about the Backblaze HDD stats, but the HGST drives regularly lead the pack on the lowest AFR, so those are the ones I buy. I’ve bought a total of 10 I think, and I’ve had 2 failures in about 4 years. I use RAID Z2, with a hot spare, so that I can tolerate 2 failures, and have the hot spare immediately start to replace one of them, then I’ll buy two more, replace the hot spare, and replace as-needed.
Both the HGST drives failed at between 55k-60k hrs. I run them 24/7 as well. Normally I’ll receive them with somewhere between 35-50k POH.
3 Likes
Hard drives are a commodity but the HGST He series seems to be exceptionally reliable. The lower drag / power to run one is likely a factor.
The NAS uses a Z3 here and I don’t bother with a hot spare since it’s a SOHO system and it takes me all of 30 seconds to replace a bad drive with a “new” one. I will continue to replace drives as they start failing their SMART tests unless someone convinces me otherwise.
It’s like our dishwasher that I repaired 6 times before replacing it at 18 years of age or the CAC system that has been running relatively trouble-free for 19 now. Either may be outside the expected lifetime window but why replace unless there are obvious issues? In the case of the dishwasher, the plastics were giving way by turning brittle with age, submersion, and due to the heat. The CAC system has needed periodic maintenance, a new condenser fan motor, and new starting caps for the compressors. We can perform similar maintenance on our NAS’ as they age, i.e. replace failing drives, fans, and so on.
1 Like
