Read this:jussi wrote:My Samsung fight is getting dirty I have bit of troube RMAing faulty discs so I had to write an email to Samsung support and try to convince them that I got faulty drives.
I'll include the email here, maybe it contains something of value to others:
The serial number of the four discs was almost sequential and they had F/W revision 1AA01106I bought four HD103UJ drives about four months ago from Germany. I live in Finland.
I installed the drives to a RAID5 array (linux software raid) and for a while they seemed to work fine.
After a month or two I found out that there was something wrong with the drives when two of the discs dropped from the array (one had failed before the other, but I didn't notice that before the 2nd went down). I was able to recover the data thanks to linux raid tools.
This is when I started to diagnose what went wrong. I quickly found out that writing to discs would often increase SMART "Current Pending Sector Count" attribute. One disc was worse than the others, but all of them had problems with that.
During testing process I changed everything in my hardware configuration; PSU:s, cables, SATA controllers, etc. and the result was always the same; increased Pending Sector Count value and read errors.
It turned out that I was able to reset the SMART attributes by filling the discs with zero (dd if=/dev/zero of=/dev/xxx). Surprisingly this worked! The Pending Sector Count would reset to 0 and there were no Relocated Sectors or anything else that would indicate a bad drive.
But partitioning the drives and actually using them in a real filesystem would soon result in pending sectors and eventually disc would drop from the raid array.
Having changed everything else but the discs obviously led me to conclusion that there was something wrong with the drives themselves. I tested them with HUTIL 2.10 and got lots of "Surface scan ECC error"s and "M.C. ECC error"s on all of the four discs.
I contacted the seller and asked for replacement units. They wanted me to contact Samsung Finland and make sure that I had tested the discs correctly.
I made a call to Samsung Finland support and they confirmed that if HUTIL 2.10 shows a sector error that would indicate a failed drive which can be RMA'd.
The seller agreed and I sent the discs back to them.
Now the seller emailed me and say that the discs are not faulty. They had contacted Samsung Germany and were informed that HUTIL 2.10 does not work correctly on all Samsung discs. The new testing method would be running HDAT2.
Apparently the four discs that I sent would all pass the HDAT2 test, and now the seller refuses to replace the discs.
My question is can you please confirm the seller to RMA the discs? I tested the discs as was instructed by Samsung Finland and they proved to be faulty.
Also, I can provide you with Linux logs that indicate the drive failures.
I'll keep you posted in case I actually get a reply from Samsung. Until then I'm out of discs and 1000 euros.
Q: What is time-limited error recovery and why do I need it?
A: Desktop drives are designed to protect and recover data, at times pausing for as much as a few minutes to make sure that data is recovered. Inside a RAID system, where the RAID controller handles error recovery, the drive needn't pause for extended periods to recover data. In fact, heroic error recovery attempts can cause a RAID system to drop a drive out of the array. WD RE2 is engineered to prevent hard drive error recovery fallout by limiting the drive's error recovery time. With error recovery factory set to seven seconds, the drive has time to attempt a recovery, allow the RAID controller to log the error, and still stay online.
You need the Raid version of the drive.
Samsung HE103UJ
WD RE2 or RE2-GP
Seagate ES