Final Word on the Samsung HD204UI

[BRiT]

Nope. The BUGGED and Fixed Firmwares report identical info.

The only sure fire way to know is to apply the fixed firmware yourself and then try to replicate the exact conditions which reproduce the bug. This is why for peace of mind the drive should be avoided.

[CDLehner]

Nope. The BUGGED and Fixed Firmwares report identical info.

The only sure fire way to know is to apply the fixed firmware yourself and then try to replicate the exact conditions which reproduce the bug. This is why for peace of mind the drive should be avoided.

Hmm  Just picked up an EARS for the same price (if I can get Egg to waive the re-stock...since I've never opened the drive). Now...what's wrong with the 20EARS? LOL

CD

[johnny121b]

Nope. The BUGGED and Fixed Firmwares report identical info.

Incorrect.  See my posts on 2/19 in this thread. 

[Superorb]

And it may even be that the update has already been done, right? Jan 2011 and after or something? Sorry...I'll go back and look through the thread.

CD

Yes, but since the version numbers are the same there's no way to verify if the drive has been patched or not. Better to patch any drive with the ...001 version number than to chance it. Also, make sure you're using the JP1 patch as that is the only one that changes the version.

There's not? Not even with a like a drive-software tool from Samsung?

CD

Well, it patches the drive with the SAME firmware, so ...001 looks just like the patched ...001 firmware

I finally was able to move all data from the F4 to the new Hitachi drive so I could patch it today, and it worked like clockwork. I could even verify that it successfully flashed since the firmware is now ...003 as reported by HD Tune and Speedfan. If/when you flash, be sure to use the JP1 firmware.

[BRiT]

Nope. The BUGGED and Fixed Firmwares report identical info.

See my posts on 2/19 in this thread.  

That is firmware for the JP1 market/drive and not originally intentioned for other markets/drives. That merely makes it even more of a fuster-cluck on Samsung's part.

Has anyone tried RMAing a drive flashed with the JP firmware to the US division? Has it been accepted? With how Samsung seems to run their drive firmware, I wouldn't be surprised if someone was denied warranty coverage because of the mismatches.

Granted, it might be akin to the Seagate forced firmware issue.

I still stand by my words of "for peace of mind the drive should be avoided."

[Superorb]

Nope. The BUGGED and Fixed Firmwares report identical info.

See my posts on 2/19 in this thread.  

That is firmware for the JP1 market/drive and not originally intentioned for other markets/drives. That merely makes it even more of a fuster-cluck on Samsung's part.

Has anyone tried RMAing a drive flashed with the JP firmware to the US division? Has it been accepted? With how Samsung seems to run their drive firmware, I wouldn't be surprised if someone was denied warranty coverage because of the mismatches.

Granted, it might be akin to the Seagate forced firmware issue.

I still stand by my words of "for peace of mind the drive should be avoided."

Samsung has even stated that we can use the JP1 flash on our non-JP1 drives, so I don't see why they wouldn't honor an RMA unless departments don't talk (and many times they don't). But it's probably best to avoid these unless you already have one.

[642]

I'm currently using one of my Samsung HD204UI (japanese patched, firmware revision : 1AQ10003) as the new parity drive of my unRaid server.
And, while the system is currently calculating the new parity using this drive, I use a loop that is reading the parameter of the drive every 1.0 second :

watch -n 1 hdparm -I /dev/sdn

with sdn currently the parity drive.
With around 600 minutes of calculating, the smart identify should be done around 36000 times.

Normally, if there is "the well known" write problem with this drive, it should be discovered on the next parity check, that I'll made just after sync.
In between, I don't use the server as anything else.

More infos soon. 

[642]

More infos soon. 

OK

Now after the parity sync, that write everywhere in the HD204UI, and a parity check that read if the value are correct, everything's fine.
Of course, to have an opportunity of a write failure, I've follow the process from HERE -> "How to reproduce"
After around 36000 smart "identify device" while writing all the disk (parity sync), and then reading all the disk back (parity check), there was no problem at all.
The Samsung was connected on one SuperMicro AOC-SASLP-MV8, and I've made sure that write-cache on this disk was enabled.
Now I will made a test directly on the MB, with chipset AMD 785G/SB710 but I want to be sure the NCQ is enabled.
How may I verify this?

[axeman]

so - i was looking at getting the F4 drive and was googling to find performance info when I found this. works great for me, since it would eventually go into an UnRaid array.

So even with the firmware fixes, and the impending release of v5, this drive is considered to be not a good buy? Strictly going by reviews on NE, the EARS drive gets lotsa terrible reviews.

At the moment, though it seems like that's the only viable option?

Seeking your guidance....

[Superorb]

^^ There's nothing wrong with the drive as long as you get the JP1 firmware onto it. Another member here ran hours of SMART queries on the drive while moving data and had zero errors. What is this v5 that you speak of?

[axeman]

thanks - i'm still nervous - i'm the guy that wants to run esxi, w. UR, then use those shares as the datastore. So i'm already staring at a hefty margin of error.

v5 - i was talking about UnRaid v5 - the one that's in 5, beta 6a... hopefully it will be released soon.

[Houseplant]

I too, like a few other people on this thread, bought an HD204UI without looking at the forums. I upgraded the firmware before putting it into the array so that I would hopefully avoid any problems. I actually have mine running as the parity drive and can see similar performance compared to what everyone normally reports from an un-Cached unRAID setup. I have seen no signs of corruption and the drive itself works just fine with every form of data I've thrown at it (small and large).

[Superorb]

I actually bought my F4 last summer before anyone knew of the problems which were discovered in December.

[axeman]

but - that's just the thing - if this is known since december, and now it's March - shouldn't Samsung have patched the drives that are now coming out?

how would i know if the drive I get has the right firmware, if the patched drive reports the same thing?

i have several samsung products (2 F3s, a BDRom, cell phone, tv, etc) and are relatively happy with the quality of each of those, but something like this is just scary. clearly we are all here and part of the unRaid experience because we are either paranoid about data loss, or have actually suffered data loss in the past, and are looking to avoid it in the future. I know people that don't back anything up, don't have a redundant copy of any important photos, etc and go on living just fine.

All that being said, I'm going to more than likely take a gamble with the F4s. I should keep a therapist's number handy incase something hoses my soon-to-be-built array.

[Superorb]

but - that's just the thing - if this is known since december, and now it's March - shouldn't Samsung have patched the drives that are now coming out?

how would i know if the drive I get has the right firmware, if the patched drive reports the same thing?

i have several samsung products (2 F3s, a BDRom, cell phone, tv, etc) and are relatively happy with the quality of each of those, but something like this is just scary. clearly we are all here and part of the unRaid experience because we are either paranoid about data loss, or have actually suffered data loss in the past, and are looking to avoid it in the future. I know people that don't back anything up, don't have a redundant copy of any important photos, etc and go on living just fine.

All that being said, I'm going to more than likely take a gamble with the F4s. I should keep a therapist's number handy incase something hoses my soon-to-be-built array.

According to Samsung, the drives have been patched from the factory for some time now. It's pretty painless to patch the drives with the JP1 firmware, so that shouldn't be a problem. I think there's as much gamble with this drive as there is with a EARS or any other "good" hard drive on the market now.