Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

In one of our lab machines,I found a disk i/o error suddenly.It was running for a while doing lot of i/o on the disk due to our application processing huge data.My machine is hardware raided which is a RAID5.I heard from my friend an RHCE professional that, huge i/o would also cause disk crash.Is it true?

Also, i have another question about recovering my data in case single disk failure in my RAID5 machine.Can i recover data in my filesystem?

If so,can anyone guide me to recover the lost data due to disk i/o error?

share|improve this question

migrated from stackoverflow.com Nov 15 '09 at 20:20

This question came from our site for professional and enthusiast programmers.

    
This looks like an ServerFault question. –  monksy Nov 15 '09 at 17:38

1 Answer 1

recovering my data in case single disk failure in my RAID5 machine. Can i recover data in my filesystem?

This is exactly the sort of failure RAID5 was built to survive. If one disk in your RAID5 array fails, no data is lost and your system continues to work as if nothing happened (the array is said to be "degraded").

Once you have a replacement disk, just plug it in, format it and re-add it to the array. The operating system will take care of re-synchronizing the new disk and soon you have a healthy RAID array again.

I heard from my friend an RHCE professional that, huge i/o would also cause disk crash. Is it true?

Yes and no. Quite obviously putting more stress on hard drives increases the chance of failure. But in many cases, these failures are manufacturing defects, not a result of stress -- so it is a a matter of time when the defect manifests in an error. Stressing your hard drives simply triggers the defect earlier.

It turns out that disks which survive their first year in a highly utilized environment are also much more reliable in the long term. If you want to know more about hard drive failure rates, see the Google's paper Failure Trends in a Large Disk Drive Population (PDF).

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.