- cross-posted to:
- homevideo
- movies@hexbear.net
- cross-posted to:
- homevideo
- movies@hexbear.net
Oppenheimer and the resurgence of Blu-ray and DVDs: How to stop your films and music from disappearing::In an era where many films and albums are stored in the cloud, “streaming anxiety” is making people buy more DVDs, records – and even cassette tapes.
This is a pretty big overstatement.
This has not been true for years. SSDs are generally more reliable than HDDs except in write-intensive applications (and even then… It really depends on what exact models you are comparing). SSDs have a life-span mostly talked about in terms of TBW (terabytes written) rather than years for a reason, if they’re powered on and not written too they’ll last as long as or longer than a hard drive. (Note: Powered on regularly, SSDs can lose data if stored unpowered for a long time (months)). If you just have an archival drive you’re not constantly erasing and rewriting data to, an SSD is a great choice. Reads also barely affect the lifespan of at all, so you can still access the data you want to protect (hell, write-lock the drive even and it’ll last decades if powered on).
This is just plain silly. Yes, the mechanical wear of the drives spinning up and down means they’ll die faster. But we’re still talking MTBF measured in years. And replacing a hard drive that’s barely used every single year? That’s not just bad advice it’s creating e-waste for no reason. Also note drives fail on a bathtub curve… If you have two good drives that lasted a year, you are increasing your chances of a failure by swapping them for two brand new drives… The best thing you can do for your hard drives is to not power cycle them constantly, any typical usage is fine. Also mechanical parts can actually wear out from disuse as well. Even archival services don’t go to these extremes you’re recommending.
If you really care about saving your data follow 3-2-1. 3 copies of your data (live, archival (external HDD or similar), off-site), two-different forms of media (HDD, SSD, cloud (yes cloud is an HDD or SSD but they have their own redundancy)), one off-site (in the event of a fire etc.)
Honestly 99.9% of consumers would be fine with a 2-2-1 scheme, 2 copies (live and off-site/cloud), 2 forms of media, 1 off-site. If you don’t trust Google or don’t want to pay for cloud storage, set up a server with redundant disks at a friend’s house. Just keeping a second copy on a server with redundancy is plenty of fail over for most use cases. 3-2-1 is for data centers and businesses (and any cloud service you rent from will follow 3-2-1…) Let’s not overcomplicate how difficult it is to keep data intact, if I tell someone to buy a new 12tb HDD each year they’re just gonna give up on keeping it safe.
No, I’m not saying replace the drive annually. That would, indeed, be dumb. I’m saying copy the data back and forth between hard drives that are kept offline.
You’d be better off keeping the drives spun up. I’ve had more drive failures from drives not in use than powered up.
Mechanical stuff is just like people, it doesn’t like sitting around doing nothing. Just like it’s difficult for us to get moving after sitting for a long time, mechanical things struggle too. There’s things like stiction in high-precision moving surfaces.
I don’t trust drives that have sat around, I assume they’re dead or at least the data corrupted.
You don’t keep it completely powered off; you power it on occasionally, but don’t keep it constantly running.
Parts that are constantly moving wear, and will eventually fail. Things that are never used can seize. You want to have a happy medium. But that’s also why you want to have multiple mechanical drives that you’re cycling through; if any single device fails, you still have your data backed up.