New light on old thread 3

This was my old thread:

I may have figured why one drives kept dropping out of the array AND this applies to any of you doing RAID setups with desktop hard drives.

Apparently RAID controllers deal with hard drives in a different way than desktop controllers on motherboards. It's related to something called TLER, which I was not aware of until today.

Desktop vs. Enterprise drive and RAID

Thanks for that heads-up there Goom...


Ben
"If it works don't fix it! If it doesn't use a sledgehammer..."
How to ask a question, when posting them to a professional forum.
Only ask questions with yes/no answers if you want "yes" or "no"

Maybe everybody else knew this!? Don't know if TLER applies to PATA drives also, which is what I have in my old clunker.

The whole thing was making me question my motherboard (raid controller built-in) after the hard drive in question passed all tests.

There is also some mention that it's "more okay" to use desktop hard drives with a RAID controller if you're doing RAID 0 or 1, but they didn't exactly explain why. But where is the "master list" of RAID that is considered Consumer grade RAID vs. Enterprise RAID.

That's how I ran into this. Going to upgrade a small company server from single hard drive to a RAID1, which is basically consumer grade but used in a small business environment.

Well that explains a lot - potentially as I had a similar issue with a desktop mirror (RAID 1). Testing showed no issue and it re-built the array fine. I still replaced the "failed" drive though. No repeats (yet!) maybe I didn't need to?

It begs the question though, in my mind, do RAID drives with TLER not recover data from sectors going bad? I'm a bit lost as to what happens when sectors go bad in this situation. Can anyone throw any light on the matter?

Many thanks for that goombawaho - brilliant find.

A note along the same line - IDE drives used in RAID - years ago when I first joined my current employer, Netware had been installed on desktop grade towers, complete with RAID using IDE drives. One of these "servers" stopped responding during office hours, and things looked normal from the console, except it was at 100% cpu utilization. Turned out that one of the IDE drives had failed, and the system was using all of its resources trying to rebuild it. I disconnected the drives one at a time, and discovered that the drive that had failed was the one with the boot files. I was able to load DOS onto the 'good' drive from diskettes, boot DOS then Netware, and get the office back to work. Then I ordered a new server with SCSI controllers and drives, and built the replacement server properly.

Fred Wagner

I remember those Netware days very well. I can't for sure answer stduc's question about what happens in RAID, but my guess is this:

The hard drive marks the sectors as bad, writes the data to a new/remapped sector and moves on. I think that's the importance of TLER - it moves on more quickly than a non-TLER drive because it's likely not the only disk and data won't be lost if the sector is bad.

This was NEVER an issue, as far as I know, back in the parallel, SCSI drives I was using in my server RAID.

stduc said:

It begs the question though, in my mind, do RAID drives with TLER not recover data from sectors going bad?

Click to expand...

I know it's Wikipedia but

Wikipedia Overview section said:

If the drive itself is inherently reliable but has some bad sectors, then TLER and similar features prevent a disk from being unnecessarily marked as 'failed' by limiting the time spent on correcting detected errors before advising the array controller of a failed operation. The array controller can then handle the data recovery for the limited amount involved, rather than marking the entire drive as faulty.

Click to expand...