Dealing with hard disk failures

I am having a lot of problems with hard disk failures in the clusters I am maintaining, especially Maxtor 6Y120L0 hard drives. So I have decided to install the smartd utility that allows to monitor the S.M.A.R.T. information given by the disks.

I don't know how reliable is the SMART information and how good it is anticipating hard disk failures but I will give it a try.


This is how I set it up:

• In RedHat 9:
  smartd is included in kernel-utils in RedHat and in the smartd
  

  
  rpm -i /usr/programs/redhat/RPMS/kernel-utils-2.4-8.29.i386.rpm
  chkconfig --level 3 smartd on
  service smartd start
  

  
  


• In Debian Woody:
  

  
  apt-get install smartsuite
  /etc/init.d/smartsuite start
  

  
  


• In Debian Sarge:
  

  
  apt-get install smartmontools
  

  
  Edit /etc/default/smartmontools and uncomment the line:
  

  
  start_smartd=yes
  

  
  

In RedHat logwatch checks the logs for smartd alarms.