Array has 1 disk with read errors

[seestray]

So woke up this morning to a pair of notices:

• Array has 1 disk with read errors
• Alert [TOWER] - Disk 1 in error state (disk dsbl)

 

I've got a few lines in the ST8000DM004 log with different sector numbers, but they are the red highlighted ones.

May 1 03:58:21 Tower kernel: print_req_error: I/O error, dev sde, sector 2256536312

Dashboard shows 144 errors for this disk - the others are all at 0.

 

 

Checked warranty, and it's still within a couple months of the end date (so if it's the disk I'd like it replaced before it runs out.)

Ran a short smart test, it passed.

 

Is there anything I should do before running the long smart test - the polling time is 952minutes, so if there's something I should do first - I'd like to do that before the 15+hours needed to run it.

 

tower-diagnostics-20190501-0951.zip

[itimpi]

Did not spot anything obvious in the logs other than the fact the drive was getting lots of read errors.   I would suggest that you run an extended SMART test to see if the drive passes that without errors.   If you want to consider getting it changed under warranty you probably need to have done that anyway.

[seestray]

The extended smart test came back clean - so in theory the drive is ok.

 

I'm thinking next steps;

1) swap the cables between a pair of drives,

2) rebuild the drive,

3) monitor for more errors - if they happen, then I should be able to narrow things down.

 

Does that make sense?

 

[JorgeB]

3 minutes ago, seestray said:

1) swap the cables between a pair of drives,

2) rebuild the drive,

3) monitor for more errors - if they happen, then I should be able to narrow things down.

Sounds like a good plan to me.