It's possible you have one of those A2 cards where cache flush is being ignored on the write test. It would be interesting to know how such things were determined when those the cards were tested....and even the SD card that Raspberry essentially bears no responsibility in it's productionhttps://www.raspberrypi.com/news/sd-cards-and-bumper/
That was quick!
Neural quantum entanglement?![:lol:]( "Laughing")
Seriously now: What's the expected MTBF of a decent SD card? Usual endurance data are limited to industrial-grade cards
Moreover, a bit of an oddity: Late August I ran diagnostics on a fairly (two weak old installation) fresh SD card, a 256GB Kingston Canvas Go Plus:Code:
prepare-file;0;0;62178;121seq-write;0;0;62178;15seq-read;84290;20;0;0rand-4k-write;0;0;21198;5299rand-4k-read;17898;4474;0;0Sequential write speed 62178 KB/sec (target 10000) - PASSRandom write speed 5299 IOPS (target 500) - PASSRandom read speed 4474 IOPS (target 1500) - PASS
Is the random write test being skewed upwards by write caching?
Any clue what the numbers are on the official ones?
Statistics: Posted by ejolson — Tue Oct 08, 2024 5:40 am