Summary
What began as a simple Linux Mint installation turned into one of the strangest USB failures I have encountered.
I wanted to prepare my better HP laptop for my Spain Recon trip. The laptop currently runs Omarchy, but I wanted to test Linux Mint 22.3 Cinnamon as a potentially more familiar and dependable travel system. I selected the Mint ISO in BalenaEtcher, chose my 128 GB Samsung BAR Plus USB drive, and waited for Etcher to report “Flash Completed.”
When I connected the USB to the HP laptop, however, it continued booting Omarchy.
At first, this looked like a normal boot-order problem. I opened the HP Startup Menu with Esc, selected F9 for Boot Device Options, and checked the UEFI settings. The USB installer was not appearing as expected. Back on my Debian desktop, lsblk revealed something even stranger: the Samsung drive still contained the old OMARCHY_202608 and ARCHISO_EFI partitions.
The first suspicion was that Etcher might have written to the wrong device. Its internal log proved otherwise. Etcher had selected the correct Samsung USB, used the correct Linux Mint ISO, and reported writing all 3,091,660,800 bytes without errors.
To remove Etcher from the equation, I identified the USB by its permanent serial-numbered path, unmounted its Omarchy partitions, and wrote Mint directly with dd:
sudo dd if=/home/lvydvy/Downloads/linuxmint-22.3-cinnamon-64bit.iso \
of=/dev/disk/by-id/usb-Samsung_Flash_Drive_<USB_SERIAL>-0:0 \
bs=4M status=progress conv=fsync
The write completed successfully. A complete byte-for-byte comparison between the ISO and USB also returned no differences.
Then came the decisive test: I safely powered off the USB, physically unplugged it, waited, and reconnected it.
The same Samsung drive—with the same serial number—had reverted to Omarchy again.
The USB controller was acknowledging writes and temporarily returning the newly written Mint data, but the data was not being retained after power was removed. Earlier USB enumeration errors, including Linux kernel error -71, provided additional evidence that the drive’s controller or NAND storage had failed.
Omarchy did not cause the problem, and GParted could not repair it. This was a hardware failure below the filesystem and partition level.
The Samsung drive had been purchased from Amazon only about eleven weeks earlier. Although the normal return window had closed, the product advertised a five-year limited warranty. Amazon Customer Service created a warranty ticket, contacted the manufacturer, and said Samsung should contact me within 48–72 hours regarding a free replacement.
There was one more useful outcome. I successfully burned the Linux Mint ISO to a DVD+R using my internal ASUS DVD writer and verified it as Linux Mint 22.3 Cinnamon 64-bit. That DVD remains good, although my Dell external USB DVD drive appears to have a separate disc-reading problem.
For now, the better HP laptop will continue running Omarchy. I have until October 26 to test it thoroughly before Spain Recon. When the replacement USB arrives, it will become a dedicated Tails Persistent travel drive. A separate USB will be used for Mint or other recovery installers.
The biggest lesson was simple: a successful flash message is not the final test. Always disconnect and reconnect boot media before trusting it, verify the contents after a real power cycle, and keep separate USB drives for persistent systems and disposable installers.