Shell script for burn-in and testing of new or re-purposed drives:
https://github.com/Spearfoot/disk-burnin-and-testing
🧵 3/n this is bloody embarrassing! 😳
The -p parameter DOES NOT specify the test pattern!
I need to stop #VibeCoding! 🤣
…blame #Mistral small 3 #LLM #AI via #DuckDuckGo duck.ai 😅
Howerver, i've decided to make it extra stressful, by inverting the test pattern.
1 pass = testing inverting 0x55, then another inverting 0xAA - the corrected script in the screenshot.
edit: correct screenshot-final2-corrected2.png 🤣 🍻
🧵 2/n i'm still badblocksing my old HGST Ultrastar 7K6000 4TB (HUS726040ALA610/0F23090).
For some reason, it only produces read errors with the 0xAA test pattern. Reallocated_Sector_Ct increased to 19, after the 2nd pass (13/0/0).
I've started a 5-pack of 0xAA write-read passes - i intent to "burn out" (forcefully reallocate) every last remaining "weak" sector.
This is gonna take the whole week; imma stress this poor puppy until it's not degrading much further.
Oof! Turns out, my offsite backups have never been particularly reliable. I keep 'em on the 10TB HDD from the old mirror now.
I guess, i don't get to resell that old 4TB HDD, then. Meh, i'm not gonna use it myself, and eBay has this "defective / spare parts" setting - 20 bucks incl. shipping.
How to fix:
badblocks: Value too large for defined data type invalid end block (4394573824): must be 32-bit value
Bought a new large external disk as data sink for my server at home. The disk is 18TB and after the previous purchase had a hardware problem I wanted to make sure that this one is working. Created a partition, formatted it with ext4 and let's run a badblocks check on it. That failed, badblocks can't handle Big Data.
You know one thing I hate about #SSD #storage?
Testing it with #badblocks will most likely not tell you anything about its health and will degrade it significantly more than #HDD storage.
Are you willing to use refurb/used storage you can't test for proper function though?
I'd be happy to find out I'm wrong and someone has made a program to address this issue... but afaik no such thing has happened yet.
The resilvering finished yesterday and I moved everything from the 8TB drives onto the new array. Now wiping all of the remaining old 8TB drives with random data before selling them off. Also running #badblocks on an older 4TB drive (ALWAYS treat old drives as sus) and one of the 10TB ones since it encountered an error during the normal random data wipe.
So far Seagate Exos is proving to be very fast. They have a good rep, so hoping they all hold up well over time. Since most of the load is #Plex, it shouldn't be too punishing. #storage #homelab #zfs #TrueNAS #TrueNASCore
tokai (Christian Rosentreter)
joschi
OSTechNix
Multi Purr Puss 
Andreas Scherbaum
WWWCDORG
LisPi
Jesse Harris