[SOLVED] Well sort of...Unraid unresponsive with constant disk activity

[Ryal7925]

Recently upgraded an old disk in my server, which worked perfectly for a few days and has shown no errors. But now it seems to be constantly reading from it and within 5-10 minutes of booting up I lose web/smb/nfs/etc access to the server, yet the HDD activity light is constantly on.

 

Any help would be appreciated.

*ssh/telnet still work, syslog attached

syslog.txt

[Ryal7925]

*update*

Cloned the disk and replaced it in the server but the issue keeps going. So I'm thinking it has to be something to do with the data itself.

[testdasi]

Did you install any plugin / docker / VM along the way?

[Ryal7925]

No plugins, I also did a fresh install of unraid just to make sure but didn't make any difference.

[Ryal7925]

Still don't really know what the problem was but after clearing the disk and putting it back in the array it all works again. Luckily that drive was just storing my game library so it can all be re-downloaded anyway (eventually).

 

As for the cause, my best guess is I hit some sort of file system limit as even trying to open the disk through a utility in windows took hours and I actually just gave up in the end. It was stuck on "reading file tree".

 

So in case anyone comes across this with an answer later it was a reiserfs disk containing 1 folder with approx 10,000 files within it. Can't be precise.

 

*Thanks to testdasi for trying to help.

[RobJ]

Your syslog was immediately after a boot, so not much to see, but no big issues.  You are connecting at 100mbps, don't know if that's what you expect, as you have a gigabit NIC.

 

Next time, it would help if you would include the complete diagnostics zip (Need help? Read me first!), and make sure it includes the period when the issue is occurring.  The diagnostics includes the SMART reports for all drives, which would reveal something about their health.

 

From your just posted comment, it's possible there's file system corruption on a drive.  You may want to check each data drive, see Check Disk File systems.

[Ryal7925]

Ah, I missed a section in that guide where you can save a syslog to the USB. Couldn't figure out how I was supposed to get a full syslog with the server freezing up all the time.

 

It's doing a parity sync for the moment but I will run a check on the drives after that, thanks for info.

Oh and yes I know it's 100mbps all too well my switch died last week and I'm still waiting for a replacement.