amtm AMTM not able to fix disk error?

[maxbraketorque]

After introducing a third disk to my AC86U, I started seeing disk errors for one of the pre-existing drives. All partitions on all three drives were previously formatted as ext4 using the AMTM disk formatting routine. I thought that AMTM would be able to repair the errors during the AC86U boot process, or during a disk check command run from from the command line, however disk check only ever reported "recovering journal", and the disk errors would persist during usage of the partitions. When I attached the problematic disk to my computer and ran an ext4 disk repair routine, it found multiple errors in all the partitions on the disk, and was able to repair them. So far, the partitions are all performing properly now.

Is the disk check routine in AMTM limited on the errors it can find and repair?

 

[ColinTaylor]

After introducing a third disk to my AC86U...

The router only has two USB ports. So you must be using a hub of some kind. That might be the source of the problem.

Is the disk check routine in AMTM limited on the errors it can find and repair?

It uses e2fsck's -p option which can only fix simple errors. Anything more complex requires human intervention. But it ought to indicate this in the log and then carry on without mounting the filesystem. If it doesn't do this then I suggest you run e2fsck -v manually from the router's command line and look for errors.

 

[maxbraketorque]

The router only has two USB ports. So you must be using a hub of some kind. That might be the source of the problem.


It uses e2fsck's -p option which can only fix simple errors. Anything more complex requires human intervention. But it ought to indicate this in the log and then carry on without mounting the filesystem. If it doesn't do this then I suggest you run e2fsck -v manually from the router's command line and look for errors.

Yes, all three drives are plugged into a (powered) USB hub, and the hub is connected to the USB 2 port on the AC86U. I just repaired the problematic partitions on the one drive yesterday, and so far, no disk errors reported in the router log. I'll report back if disk errors reappear in the router log.

ok on "-p". I read the description earlier, and I assumed that dc would report disk errors that could not be repaired, but it didn't. The worst I ever saw was "recovered journal" and then "clean" in the check of a partition. Maybe a limitation of the e2fsck implementation on the router? Or maybe -v needs to be used at the same time?

If I run dc from the amtm command line, does dc attempt to unmount the partitions before checking and repairing? I don't see a umount command in the disk-check script or the disk_check.mod scripts. I'm asking because I'm trying to figure out the limits of dc. Seems that if I want to attempt to repair a partition using the router, then I'll need to use the router command line as you suggested and go through the typical steps.

 

[ColinTaylor]

The only thing dc does is install the disk-check. It's not a repair menu.

If you're seeing the "clean" message in the log it means that e2fsck command has finished. So the problem is at some point after that.

 

[maxbraketorque]

ok. If I see issues again, I'll try running e2fsck from the command line.

 

[maxbraketorque]

The disk errors are back. Any thoughts as to whether the errors are due to 3 drives on USB or 7 partitions (2 drives with 3 partitions each and 1 drive with 1 partition)?

 

[ColinTaylor]

The disk errors are back. Any thoughts as to whether the errors are due to 3 drives on USB or 7 partitions (2 drives with 3 partitions each and 1 drive with 1 partition)?

Show us the contents of /jffs/addons/amtm/amtm-disk-check.log as well as a complete copy of the router's syslog. Without that information it's impossible to say.

 

[maxbraketorque]

ok, but I'm going to first try moving the recently added drive to the USB3 port to see whether that resolves the issue. I suspect that it won't, but I want to try.

 

[doczenith1]

What about using e2fsck -f -y? Years ago I had some issues on my flash drive and someone from this site suggested I use e2fsck -f -y. It fixed my issue if I remember correctly.

 

[ColinTaylor]

What about using e2fsck -f -y? Years ago I had some issues on my flash drive and someone from this site suggested I use e2fsck -f -y. It fixed my issue if I remember correctly.

The -y parameter should never be used in an automated script or as a matter or course. Only after running e2fsck manually and assessing the errors can you make an informed decision whether or not the use of -y is appropriate.

 

[doczenith1]

The -y parameter should never be used in an automated script or as a matter or course. Only after running e2fsck manually and assessing the errors can you make an informed decision whether or not the use of -y is appropriate.

Thank you for the additional information. I had just copied and pasted what someone mentioned to try. I will make note of not using the -y until the issue has been diagnosed.

 

[maxbraketorque]

With the third drive on the USB 3 port, the drives have gone a few days now without errors appearing, whereas when all three drives were on a USB hub connected to the USB 2 port, errors would appear within a day. So my tentative assessment is that this revised configuration prevents the error from occurring. I don't know if its a partition or device limit on one port, but as long as I have a working solution, I don't need to know.