disk remounts as readonly and then has errors

adrianmariano · 02-07-2016, 01:10 PM

I discovered that my disk was mounted read-only, which caused a lot of things to fail. Upon rebooting (perhaps that was a mistake) I had a ton of disk errors, that I had to correct by manually running fsck. After this was done I rebooted and everything appears normal.

I'm wondering why this happened? Should I be concerned that my disk is failing? Should I be concerned about the data on the disk being corrupted? I ran the smartctl long test and the drive passed. Is there something else to do? I can't even find anything in syslog about the disks being remounted, let alone something that would explain *why* they were remounted.

Keith Hedger · 02-07-2016, 05:20 PM

Usually after a certain amount of mounts without a disk check the system will force a check, usually not somthing to worry about as its just maintainence, but if as you say you had a load of errors, there could be an underlying problem, have you had any sort of system crash?

If smarctrl reports everything is ok I wouldn't worry too much but I would make sure I did a regular backup, just in case.

syg00 · 02-07-2016, 06:33 PM

It was re-mounted ro because you probably told it to - check fstab. Safety mechanism.

adrianmariano · 02-07-2016, 06:42 PM

Maybe I wasn't clear about what happened. While the system was up the file-system was somehow remounted as read-only. I was in my windowing system with a bunch of programs running that could suddenly no longer write to the disk. This made my running system fail because I couldn't write to /tmp, for example, so things started failing left and right. It did not happen during a boot process. I'm not sure how long the system had been up...but not particularly long, maybe a couple weeks.

After rebooting the system came up in "emergency mode"---text only and some very limited selection of programs available---and said that fsck couldn't be run automatically and that I had to run it manually. I ran fsck manually and told it to correct a lot of errors.

I've been running linux for 16 years and have never observed this before. It's not normal.

adrianmariano · 02-07-2016, 06:57 PM

Quote:

Originally Posted by syg00

It was re-mounted ro because you probably told it to - check fstab. Safety mechanism.

Ah, I see the errors=remount-ro option in fstab. That explains why it remounted, but not why there were errors.

astrogeek · 02-07-2016, 07:07 PM

You can find more about that in man mount:

Code:

 errors={continue|remount-ro|panic}
        Define the behaviour when an error is encountered.  (Either  ignore  errors  and  just  mark  the
        filesystem  erroneous  and  continue,  or remount the filesystem read-only, or panic and halt the
        system.)  The default is set in the filesystem superblock, and can be changed using tune2fs(8).

As to why it happened... possibly a power event, a strategically placed cosmic-ray (it happens), memory error (might be worth running memtest86), other hardware or PS hickup.

Other than running a good long memtest, I would update all my backups, reseat cables and cards, and monitor it but try to not worry about it if the hard disk looks OK - unless and until it happens again.

One event in 16 years is a tribute to just how good the hardware and software really are. Two events in a smaller time window is a problem worth investigating.