Keep your RAID safe and healthy ..

You are running your production server and has some sort of RAID .. I hope it is RAID and JBOD as I have seen before … The main reason for use of RAID configuration is to improve reliability and availability of your data. There is many different setups or configurations for RAID — each of them has pros/cons and price tag attached. Therefore you have to research and figure out for your self what is your data worth and how much you want spend to protect it. For example the typical RAID in use are RAID-1, RAID-10 or stripping with parity RAID-5, if you use some sort of them the server can recover from a single hard drive failure with no loss of data.

However, nothing is perfect so sometime happens that the bad luck comes around and one or two disk will fail — bad, bad situation. Hopefully your backups were working and you can trust them. Therefore is a good idea that you perform some sort of check or maintenance as you would do on anything else if you want to increase the reliability. One of the maintenance steps to keep you RAID array healthy and happy is running consistency check on RAID on regular basis. Once a month seems as a good idea. A consistency checks help spot blocks that have gone bad or are corrupted. This is excellent opportunity for you to remove the hard drive with bad blocks from RAID array and exchange. You do not want to leave the bad hard drive in the RAID and let it build bad blocks.Can you image you have RAID-5 array, bad blocks on one of the drive, you never check for consistency, bad blocks are being built and suddenly game over moment – drive is gone. You replaced the bad drive but you may be out of luck to rebuild your RAID-5 array — as bad blocks are everywhere … If you have RAID-6 – which is more expensive, use more disks and uses two parity stripes instead of one .. you would be protected even against this situation. Therefore, if you do not, you better start and accommodate into your monthly maintenance plan to run consistency checks on your RAID arrays in servers. If you never tun these check, you will never know what is happening – until it is too late.

If you are running DELL servers .. you should have the DELL OpenManage Administrator software installed, simply go to STORAGE item , click VIRTUAL DISK and from the drop down menu select task = CHECK FOR CONSISTENCY — it will take some time …

If you have a HP servers .. they have a software solution the NetRAID assistant and the software does the check on weekly basis for you. Also good point to consider .. if your RAID controller does not have a battery backup … the safest course is to set the write cache size to zero.

  • All about RAID and different configurations can be found here
  • More info on how to run consistency checks here
  • Some good tips for RAID and checks here
  • More info on RAID here

Leave a Reply