I've been running multiple servers on Wandboard Quads for a while now and it seems like lately I'm going through microSD cards like crazy. Usually everything runs fine for a while, then I'll get the occasional problem that makes me SSH in, at which point attempting a "sudo reboot now" usually gives one error or another. I can pull the plug and things will be fine again for a while, but eventually will get to the point where the server won't boot. In a recent case, Windows refused to even reformat the card once I pulled it. It recognized that I'd inserted one and said I needed to format it, but when I told it to go ahead and do so it then said it was unable to.
My assumption is that my ridiculously large library (fast approaching 100,000 tracks) is resulting in so much write activity that the cards just wear out.
Weirdest error so far: I had a card that would boot and let the system run just fine. Every time I rescanned the library or did something else that should involve a disk write it seemed to work, but after rebooting I'd find that everything was back to an older state. That card never did actually "die" in the sense that the server wouldn't boot, it just became useless because no changes at all would survive a reboot. Even deleted files would magically re-appear.
I'm exclusively using SanDisk and Samsung cards bought from Fry's, so I don't believe I'm working with counterfeits or knockoffs. FWIW, Samsung seems even more delicate, though they're often less expensive.
I've gotten in the habit of keeping a fully-configured backup 32GB image on my hard drive so I can get back up and running in the time it takes to write a new card and maybe scan the library again.
I have yet to experience this type of trouble running Max2Play on a Raspberry Pi, but that server doesn't get used quite as much.
Is this just the way it is when trying to work with a big library using a fragile medium or is the universe picking on me out of envy over my extreme handsomeness?
My assumption is that my ridiculously large library (fast approaching 100,000 tracks) is resulting in so much write activity that the cards just wear out.
Weirdest error so far: I had a card that would boot and let the system run just fine. Every time I rescanned the library or did something else that should involve a disk write it seemed to work, but after rebooting I'd find that everything was back to an older state. That card never did actually "die" in the sense that the server wouldn't boot, it just became useless because no changes at all would survive a reboot. Even deleted files would magically re-appear.
I'm exclusively using SanDisk and Samsung cards bought from Fry's, so I don't believe I'm working with counterfeits or knockoffs. FWIW, Samsung seems even more delicate, though they're often less expensive.
I've gotten in the habit of keeping a fully-configured backup 32GB image on my hard drive so I can get back up and running in the time it takes to write a new card and maybe scan the library again.
I have yet to experience this type of trouble running Max2Play on a Raspberry Pi, but that server doesn't get used quite as much.
Is this just the way it is when trying to work with a big library using a fragile medium or is the universe picking on me out of envy over my extreme handsomeness?