I have several servers which are experiencing seemingly random events where the /boot filesystem is switching to read only mode. My disk layout is:
/dev/sda - ext4 - /boot (63Gb)
/dev/nvme4n1p2 / (890Gb)
These are massively over-specced, I know...
I am using the UUID to mount rather than the device names above.
I also have 24 x NVME drives mounted under /data (These are fine)
I have tried mount -o remount,rw /boot and get the following error:
# mount -o remount,rw /boot
mount: /boot: cannot remount /dev/sda2 read-write, is write-protected.
I have also tried querying the drive using hdparm and smartctl:
# smartctl --all /dev/sda
smartctl 7.1 2019-12-30 r5022 [x86_64-linux-5.4.0-131-generic] (local build)
Copyright (C) 2002-19, Bruce Allen, Christian Franke, www.smartmontools.org
Short INQUIRY response, skip product id
A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.
# hdparm -I /dev/sda
/dev/sda:
HDIO_DRIVE_CMD(identify) failed: Input/output error
I cannot find a way to trigger this event however the drive has not failed as I can reboot the server and it will mount read/write and continue to work for several weeks.
I am able to run an "ls" command on /boot and traverse the directory tree so the data is readable.
Device details:
Model Number: SATADOM-SL 3IE3 V2
Firmware Revision: S17411i