Sign in to follow this  
fpbear

Confused about TLER for onboard RAID

Recommended Posts

I have the Intel ICH9R onboard RAID controller. I'm planning to swap out the drives with larger ones, I'm confused whether I need to consider TLER drives.

My existing RAID 1 setup (two WD 750 AAKS drives) ends up having to do CHKDSK error scans at Windows startup more often as the drives get older. Not sure why, and I'm wondering if that's related to TLER. Write back cache is disabled.

I want to upgrade to a four drive RAID 10 array and looking at the WD Caviar Green 2TB drives, there are two models at the same price - I think WD20EADS supports channging the TLER setting and it has a 32MB cache. While WD20EARS does not have the option to change TLER it has a better 64MB cache.

I was reading on the Adaptec support site that the WD20EADS drive is compatible with their controller cards as long as the jumper is set to 1.5GB/s. Not sure if such advice applies to the Intel ICH9R onboard software RAID chip. For ultimate reliability in RAID 10 on the ICH9R controller, should I consider TLER and/or using the 1.5GB/s jumper when picking the new drives?

Edited by fpbear

Share this post


Link to post
Share on other sites

Many people recommend getting drives with TLER though personally I think it's pointless with good controllers. That said, the ICH9R is not a good controller, so it might be worth your while.

You won't need the 1.5gb/s jumper as the ICH9R has no real problems I know of with 3gb/sec. If the only options you have are a TLER drive with 32MB cache or no TLER with 64MB cache, I'd take the TLER one. The cache won't make a huge difference, and you *might* be a little safer with TLER.

Share this post


Link to post
Share on other sites

I have the Intel ICH9R onboard RAID controller. I'm planning to swap out the drives with larger ones, I'm confused whether I need to consider TLER drives.

My existing RAID 1 setup (two WD 750 AAKS drives) ends up having to do CHKDSK error scans at Windows startup more often as the drives get older. Not sure why, and I'm wondering if that's related to TLER. Write back cache is disabled.

I want to upgrade to a four drive RAID 10 array and looking at the WD Caviar Green 2TB drives, there are two models at the same price - I think WD20EADS supports channging the TLER setting and it has a 32MB cache. While WD20EARS does not have the option to change TLER it has a better 64MB cache.

I was reading on the Adaptec support site that the WD20EADS drive is compatible with their controller cards as long as the jumper is set to 1.5GB/s. Not sure if such advice applies to the Intel ICH9R onboard software RAID chip. For ultimate reliability in RAID 10 on the ICH9R controller, should I consider TLER and/or using the 1.5GB/s jumper when picking the new drives?

smartmontools 5.41

command --scan -d csmi (appear list of drives)

Some like /dev/csmi0,2; /dev/csmi0,3 (according to ports used for raid).

http://sourceforge.net/apps/trac/sma...ID-Controllers

http://smartmontools-win32.dyndns.org/smartmontools/

I have ASUS p5e X38 ICH9R and scterc work fine.

If your drives support ERC set you can make simple script and run it on win boot.

Share this post


Link to post
Share on other sites

TLER is just a way for the HDD not being disconnected by the raid HBA when trying to recover a damaged sector.

The most important thing is to buy drives that have the highest possible Unrecoverable Bit Error (UBE) per bits read ratio.

  • SandForce controller based SSD are in the 10^17 range (means they have a 0.009% probability of 1 unreadable sector per 1TB read)
  • Most Enterprise SAS are in the 10^16 range (0.086% probability of 1 unreadable sector per 1TB read)
  • Most Enterprise SATA are in the 10^15 range (0.86% probability of 1 unreadable sector per 1TB read)
  • Most Desktop SATA are in the 10^14 range (8.6% probability of 1 unreadable sector per 1TB read)
  • Some SATA are in the 10^13 range (don't use!)
  • Note : Drive's manufacturers do not always publish those figures, and a physical drive with a raid oriented firmware (and its price bump) may "looks" better than the same physical drive marketed for a desktop usage...

This is why you have a scrubbing feature usually in the RAID HBA. It checks the parity regularly and allows to fix bad sectors transparently to lower the risk of an unreadable sector at rebuild time(=after a HDD failure).

SATA II (3Gbs) offers some other differences than the throughput versus its SATA I previous edition, that may explains why the raid hba manufacturers don't want to support it at all...

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this