Barracuda XT, Barracuda, Barracuda Green, DiamondMax and other desktop SATA and IDE drives

[meowsauce]

Not sure if this is the best place to ask, but does anyone know the command or how to set/enable ERC (similar to TLER/CCTL) timeout on 7200.11 drives through the rs232 or sata connection? This may or may not help out with long pauses some 7200.11 users are experiencing, and would definitely help out RAID 5 users like myself (a little wary of setting up a 8x1.5tb R5 with no erc timeout). I've looked through the rs232 command list (here) and wasn't able to find anything (its noted that this command list is not 100% complete). I haven't asked seagate tech support about this, but I can't imagine they would be willing to help...maybe I will later.. Perhaps someone saw something I missed, knows how to enable this, or just to point me in the right direction?

Link to the BSY/0 LBA fix for those who are unfamiliar with the fix.

...if they edit/don't allow links, most of this is googleable.

...and yes, I know the es.2 drives have erc enabled and are better suited for my application, unfortunately that doesn't help me, or the 8x1.5tb 7200.11 drives I have.

hyram (im sure you'll come across this thread) and others: it seems ns has better non-recoverable error rate, and higher rotational vibration tolerances (due to the presence of a vibration sensor it seems), some also say firmware has a different cache algorithm

meow

[hyyram]

meow,

Fist let me say the the forums maintenance here is questionable... I changed my email address and had to enter a new screen name. What's with that???

Yep... I'm looking into this as well. Unfortunately, I've been side tracked for a while. I hope to be getting back into the thick of it as soon as I can. I'm kinda in the same boat as you... 8 x 1.5TB AS drives... if I hadn't bought them before knowing what I know now I would have gone with either the NS's or WD RE3's.

I can't remember what parameter I found that I thought I could use... maybe read/write timeout rather than ERC. I don't have the doc here that I can check. I'll see if I can find it again in a day or so when I have time to spend on this.

Best o' luck, 

hyram 

[meowsauce]

hyram,

lol...yeah the forum's config here does seem a little strange while registering for an account.

Yeah, I wish I looked into this issue before getting the drives, I just figured that everyone has quirks with any manfacturer, but usually they're able to be worked out. I also have two WD1001FALS, but I'm able to set tler on that through a wd tool that was released. I just wish I could do the same on the 7200.11 AS drives. Is the doc with the commands you are referring to the same doc that I linked to above, or is it a different one? Whenever you do get a chance, can you tell me the commands you saw/were looking into? I think after having these drives loaded down, there's a pretty good chance they'll drop out of the array semi often if erc is not set to a lower value.

I've decided to implement a raid 50 array with the drives, maybe if im lucky enough and only two drives decide to drop out at once, they'll be on the seperate r5 arrays. I wanted to go with raid 10, but my data is already approaching 4tb...that and my controller doesn't support raid 6. I might pick up a perc 6i just for this issue.

good luck,

meow

[hyyram]

meow,

I found what I'd intend to investigate... in the commands doc, page 19, show/set retry commands. I've no idea if this will work, and I'm somewhat unclear on what all the parameters are, but I figure it's worth a try. I'm thinking the ECC retry setting might do the trick.

First though... and I'm working on this now... I'm running all my drives thru seatools long test to see if the drives themselves think their good or bad. Gotta start with something I've got at least a little confidence in. Being a mac guy, I had issues with getting seatools to run on my mac pro. Resorted to borrowing a win box to do the tests on. FYI... for the 1.5TB drives, seatools long test takes ~4 hrs, a full erase takes >10 hrs. Once I've got 8 "good" drives out of seatools, I'll configure the RAID and use "disktester" to see what that'll tell me. If everytihng works out to this point, I'll problably stop and shoot the horse. If things still aren't working the way I want, only then will I proceed with messing around the drive inards.

And if you haven't read it yet, get the google white paper on drive failures. It's a good read.

hyram

[nick.lowe]

You need to issue the appropriate SMART commands to the drive to check ERC function availablity and to enable it. 

See http://www.t13.org/Documents/UploadedDocuments/docs2008/D1699r6a-ATA8-ACS.pdf

8.3.4 SCT Error Recovery Control command

Message Edited by nick.lowe on 10-01-2009 09:18 AM

[fzabkar]

It may be an idea to first check whether ERC is supported by the drive.

Section 4.1 indicates that bit 3 (Error Recovery Control (AC3) supported) and bit 0 (SCT Feature Set supported) of word 206 (SCT Command set support.) in the Identify Device data block should be set to 1.

[nick.lowe]

Absolutely!

hdparm -I device

... will give you this information if you want to check quickly.

[nick.lowe]

I've checked and can confirm that this feature is available on the Barracuda 7200.11, 7200.12 and LP series drives.

Also see http://forums.seagate.com/stx/board/message?board.id=ata_drives&thread.id=1069

Message Edited by nick.lowe on 10-01-2009 09:20 AM

[fzabkar]

I have two suggestions that someone may like to follow up. Unfortunately I don't have the hardware or the technical knowledge to do so myself.

The first involves obtaining a copy of WDC's wdtler.exe utility and "convincing" it to accept Seagate drives.

The second method would be to use MHDD version 4.5. It includes a "terminal" facility which enables you to send ATA commands to the drive.

I'm assuming that the ERC setting persists after a power cycle, otherwise there would be no point in trying the above.

Here is wdtler:

http://zacuke.com/files/wdtler.ZIP

The archive contains BAT files which consist of two lines:

hdaccess
WDtler -arg1 -arg2

I suspect that hdaccess is a TSR that enables DOS to communicate with SATA controllers, or RAID.

WDtler -r0 -w0 ------------- turns off TLER/ERC
WDtler -r7 -w7 ------------- turns on TLER/ERC for 7 seconds

This is the online help:

===================================================
Syntax:
WDTLER [-R] [-W] [-?]
Parameters:
-R Set read TLER seconds. Use 0 to disable.
-W Set write TLER seconds. Use 0 to disable.

/? This help screen.
===================================================

The wdtler.exe executable is compressed and needs to be unpacked with UPX:

http://upx.sourceforge.net/#download

I would think that one way for wdtler to identify a WDC drive would be to look for a "WD" prefix in the model number reported by an ATA Identify Device command. There is one location in the binary part of the .exe file that contains "WD". Maybe changing this to "ST" would work. I'd be very wary of trying this, though, as wdtler may do some low level writing that is very specific to WDC hardware. However, if you type "WDTLER" without any arguments (as in TLERscan.bat), then I suspect that wdtler will only read the current state of the read and write timeout counters.

[fzabkar]

Some good news.

I recently contacted the author of HDAT2 and he has graciously added an Error Recovery Control Menu to the latest version (4.7) of his software.

He has asked me to test it. Unfortunately I do not have the means to do so.

See http://www.hdat2.com/