2

Do we need to perform periodic disk checking,for bad blocks and bad sectors?.

Our servers havent been restarted for last 4 months,also no disk checking have been performed on any disks.

Can that can cause any problem in the future or will it increase probaility of Bad Disk.

Is it really practical to perform disk checking on production servers.

claasz
  • 510
  • 3
  • 11
Kevin Parker
  • 757
  • 1
  • 13
  • 32

2 Answers2

4

This seems practical only if you can afford the downtime for the disk check. Also note that you would need to automatize this somehow.

So for me it looks more reasonable to monitor the SMART data of your disks. Check out smartd, for example.

If you have a RAID (which is standard for production servers), then you should configure your RAID software to create an alert (e.g., an e-mail) in case of a problem.

claasz
  • 510
  • 3
  • 11
-1

It is recommended to perform a disk check every 180 days or every 25-30 mounts (reboots).

Hex
  • 1,949
  • 11
  • 17
  • 1
    comment for downvote please! I have been told in school long time ago for this but this does not mean that I have practiced it myself. If you think it is wrong tell me why. – Hex Oct 31 '12 at 09:18
  • 2
    Bit outdated... it was true for ext2 IIRC :-) all modern filesystems with journaling don't need checks and some even doesn't have it (like XFS). Watching system logs/ipmi/smart data is a MUST! – Kamil Šrot Oct 31 '12 at 13:31
  • 1
    You may think that modern FSes with journalling don't need checking, but I can assure you that ext3/4 FSes can still develop errors that require a fsck. Kernel bugs can do it, as can bad hardware (not just bad controllers, so smartctl (while very useful) can't always help), and even just cosmic ray events. The journal makes such errors harder to generate, but I have bitter experience that confirms they can still happen. – MadHatter Oct 31 '12 at 15:24