[motles]

> a one-time payment for an external hard drive may be a more economical alternative over time.

If you care about your data, you are playing with fire.

Drives do not store data forever. Data must be read and rewritten occasionally to maintain it, from old media past its lifespan to new media.

Good storage software, with the ability to write your data with either mirroring or striping of some sort is able to routinely scan your entire data set and detect bit rot - and rewrite sectors that contain bitrotted data to new media.

You simply do not get that level of protection buying a single external hardrive alone.

Most enterprise storage systems do this. Most cloud storage does this. It’s worth paying for if you have data that is valuable.

[HelloNurse]

I don't have the impression that anybody here is discussing a single external hard drive.

Also, there's nothing particularly "enterprise" or technically exceptional in backup software that can read old versions of files (for testing) while writing new ones (for current backups) and can copy old backups to new storage locations.

[FooBarWidget]

I use the 'bitrot' app to store checksums of files and to check for bitrot. It works by checking for data changes that aren't accompanied with a file modification time change. With this you don't need something like RAID or ZFS or detect bitrot.

https://github.com/ambv/bitrot

But of course you still need backups. The way to use 'bitrot' in combination with backups is that you don't backup the bitrot DB file. Instead, you run 'bitrot' separately on the main disk and the backup disk.

As for fire-proofing my data: I store a disk at an off-site location (parents' home), and I regularly swap my main disk with the backup disk.

[janwillemb]

I have used local drives for my data for decades. I have data since 1995. Nowadays I also have a remote offline backup somewhere. It seems to work for me so far.

[justinclift]

With your setup, is that having your data on a single local drive at any time, or was it generally copied onto a few?

aka "making sure you're not hosed if a drive goes bad"

[janwillemb]

It started with floppy disks in a large box. They actually went bad quite often. From the moment we had a 20MB hard disk[1] for about €400 all went on that hard disk, I think we made copies on the floppy disks just in case (not sure though). It was not important: private stuff, school stuff, programming for fun.

After I got a windows computer, I started hoarding stuff on the harddrive. I don't think I bothered to make copies. When I started doing paid stuff, I bought an external drive for backups. Have been doing that since, in various configurations. My configuration is now: all computers in the house backup to a central server in the house. I rsync to an external HD from that server every now and then. The external HD is stored on a different location.

I have realized that the data is the most vulnerable when I am doing the backup. If the house goes up in flames at that moment, I loose everything. I should get a second HD to prevent that.

[1] https://www.computinghistory.org.uk/det/47582/Atari-SH205-Ex...

[jackjeff]

I have a cron job running snapraid scrub. And smart demon with email alert. Once I had to quickly buy a HDD (next day Amazon) because it was dying.

Big appeal of cloud storage is that the data won’t go up in smoke if there’s a fire. But everything else you can do if you’re patient enough.

[WesolyKubeczek]

> Big appeal of cloud storage is that the data won’t go up in smoke if there’s a fire.

Unless it’s OVH…