20,000 hard drives die every day... how will that change?

[Patrick]

Not sure if everyone saw the Seagate post: Tips for What to Do If Your Hard Drive Fails | Seagate which basically said that every day 20,000 hard drives fail.

My question is, how does this change with SSDs?

A few thoughts:
1. SSDs have no moving parts. There are no regular "oops it fell off the table" hard drive crashes.
2. SSDs have UBER rates that are becoming superior to disks. 10^-17 is quickly becoming standard
3. On big disk arrays, hard drives are having lower duty cycles because SSDs are now standard as a caching tier

Any thoughts on on the future?

 

[Patriot]

Well... Those fancy carriers that hold the HP drives actually record error at time of failure, which then gets passed back to the manufacturer to make changes for future disks.

SSD's are cost effective as a cache layer but with SMR being 20% cheaper than traditional hard drives the chasm between spinning and solid state has grown again.
Tiered storage needs to get more seamless as it is not going away... yet.

Things like phase change memory and memristor are the future though... It will bring the costs down and have the capacity required.