ACE Lab has implemented a new function in the WD utility of last upgrade, called TLER (Time Limited Error Recovery).

As many professionals in the DR field know, WD drives can respond very slowly when cloning surface, even if you do the slow fix patch and stop reallocation. This is because in SA there are many modules (more than 400) that get written with unimportant error logs by already weak heads... most of this modules are not important for data recovery process but they can be an obstacle to cloning speed and surface security (heads could also scratch surface while trying to record this error logs on those unimportant modules).
The idea of this TLER function is to disable writing of error logs by excluding those unimportant modules from mod dir, this will permit to clone surface of many WD drives with bad sectors and weak heads with a decent speed if we can't afford to swap heads.

See the video conference where AceLab engineer shows how to do this:
https://youtu.be/47eXzxZk-0E?t=1455

My question is... how to implement this function on MRT?
I tried to disable unnecessary modules on DIR module (01) on a Tresxlb drive without noting any improvements... speed is still very slow.
Does the TLER function change something else apart excluding unnecessary modules from dir?

_________________
My firmware database:
https://mega.nz/folder/O01DkBRI

I thought that TLER was WD's own terminology for the standard ATA feature called Error Recovery Control (ERC). Samsung and Hitachi call it Command Completion Time Limit (CCTL). This command "is a feature of hard disks which allow a system administrator to configure the amount of time a drive's firmware is allowed to spend recovering from a read or write error".

https://en.wikipedia.org/wiki/Error_recovery_control

I believe that HDAT2 can issue ERC commands. The standard setting is typically 7 seconds for both read and write.

WD also at one time made available a free tool called WDTLER which enabled users to specify their own time limits. As I understand it, WDTLER's edits are nonvolatile, but any changes resulting from the execution of ATA ERC commands do not survive a power cycle. WDTLER would only run against specific WD models.

I would get a copy of WDTLER, obtain a supported WD model, and then compare the contents of module 02 before and after running WDTLER. One other thing you could do is to use a hex editor to search module 02 for a value corresponding to 7000 milliseconds, ie 0x1B58 (or 0x58 0x1B). In fact I have found such settings in the ROM in WD's Arch 7 models.

Western Digital Time Limit Error Recovery Utility :
http://www.hddoracle.com/viewtopic.php?f=22&t=218

_________________
A backup a day keeps DR away.

I'm sure it will be added soon enough.... AFAICS, From a DE stance you can get pretty close to the same results by reducing timeouts - it's a useful feature to have in a single button press but I'm not overwhelmed by it's actual functionality, others may have had different experiences.

I'd have preferred ACE to have been working on solutions for the locked PCBs.

_________________
Data Recovery Services in the UK.
https://www.usbrecovery.co.uk/

Hello everyone ,
200ms was the time i used in one case for a weak head and i was damm impressed .

_________________
Regards
Amarbir S Dhillon , Chandigarh Data Recovery Labs [India]
Logical,Semi Physical And Physical Data Recovery
Website-> http://www.chandigarhdatarecovery.com

_________________
A backup a day keeps DR away.

only a few gets written but it takes time indeed. However, i doubt TLER prevents writing these, you probably need to handle this in another way
I haven't used TLER, coz i use other ways but it seems impractical if the time is defined in seconds...
As Lardman said, timeouts and resets are about just as useful.

pepe

_________________
Adatmentés - Data recovery

I'm cloning a drive since weeks (client is not in hurry, he can wait even months after i explained the situation).
You're right guys... reducing timings on DE side it helps a lot to improve reading performances !

I already knew that reducing timings in DE would help a lot, i'm doing that since years, but in this particular case i noted that it was not helping me much.
After playing more with settings, i've found that if i use the "fix slow responding in RAM" in MRT before starting the clone, it helps a lot.
The slow fix via ID and re-lo patch in this case is not enough... for some reason in this case i need to play with this option also in ram.

The cloning speed before was 10 - 20 Kb/s with a lot of errors (red squared), after the ram fix it speed up between 100 Kb/s - 1 Mb/s (in some zone it reaches 3 Mb/s).
However on many zone it still reads at few Kb/s...probably because those zone are really damaged.

_________________
My firmware database:
https://mega.nz/folder/O01DkBRI

I still have to try with HDAT2 and WDTLER tools.
In the meanwhile i've searched for value 1B58 / 581B in the ROM and in mod 02 before and after patch.
I've found pattern 1B58 in the rom at offsets 0x66733 and 0x115370, nothing in mod 02.
Pattern 581B was found in the rom at offset 0x8737 and 0x43052, nothing in mod 02.

_________________
My firmware database:
https://mega.nz/folder/O01DkBRI

TLER is only enabled in enterprise models. The default setting is 7 seconds. TLER would be disabled, or not present, in desktop models.

_________________
A backup a day keeps DR away.

After few hours of cloning it gets slow again... the slow fix in ram is not enought.
I have to do the slow fix on modules and in ram every time i see reading stuck at 10 Kb/s

_________________
My firmware database:
https://mega.nz/folder/O01DkBRI