Final Word on the Samsung HD204UI


KYThrill

Recommended Posts

The Hitachi looks like it was 25% full which will skew the results away from it since the fastest part of the disk is full.

 

You are using Windows 7 (possibly vista, but I don't like to believe that OS still exists) which supports the advanced format, where unraid does not, so it wont properly show the advantages / disadvantages.

 

Also you seem to have only tested the read speeds, which is not the issue with these drives. Re run the bottom tests again but select "Write" and you should get different results.

 

The drives work fine, they write slower then the other 2 TB choices. They also use less power and tend to run cooler. If you have them, they will work fine.

Link to comment
  • Replies 121
  • Created
  • Last Reply

Top Posters In This Topic

They also tend to report impossible temps, so before you blindly trust the temps they report, make sure the temps are reasonable (ie: not below room temp).

 

Running the tests on an empty drive versus a drive that is filled even a little bit is not a fair test. Also, unRAID can not run the drives in the proper alignment so write performance will suffer.

Link to comment

Ok thanks.

 

The weird thing is that the Hitachi reached 40c in idle and it was nosier, whilst the F4 reached only 27c!

 

But all this is useless because they both need to tested whilst in the Unraid server.

Not that weird.  The Hitachi is not a "Green" drive.  It is trading off a more powerful motor's energy consumption for speed.  That generates more heat.

 

I have 4 of the Hitachi drives in one of my servers.  When first spun up after being idle they read 23, 24 24 and 24C in a room where ambient temperature is 22.5C.  The same server has three Seagate 7200RPM drives.. They measure 24,24, and 25C when first spun up after being idle.

 

(I'm assuming you installed your two disks in the same slot in the server, since in different slots different air-flow could easily affect the temperatures.)

 

Joe L.

Link to comment

ummm... what if the drive is formatted with ReiserFS under another distribution that allows proper alignment?

 

It will not be recognized by unRAID as one of its disks, therefore, it would offer to clear it if you attempted to add it to its array.

 

Currently, in all versions of unRAID, the partition MUST start on the second cylinder leaving the first cylinder unused except for the first sector that holds the MBR.  All modern drives (so far) report their geometry as having 63 sectors per cylinder, as in this example:

fdisk -l /dev/sdd

Disk /dev/sdd: 1000.2 GB, 1000204886016 bytes

1 heads, 63 sectors/track, 31008336 cylinders

Units = cylinders of 63 * 512 = 32256 bytes

 

They all fake the number of heads, and the number of cylinders, but basically unRAID leaves the first track unused.  It the drive reported 64 sectors per cylinder, I think it would start the first partition on sector 64 instead of sector 63.

 

 

 

 

 

Link to comment

After reading the forum posts about Samsung and their temperatures I decided I had to get to the bottom of it. Since I have 24 Samsung drives ranging from f2's all the way to f4's I need to know if the temp reading can't be trusted. I searched the internet and could not find anything stating that all Samsung Drives report an incorrect temperature anywhere.

 

- Do not use the temperature reading on your AC's thermostat that is 3 rooms away to gauge ambient temperature.

 

Since no one was reporting how they were getting their temperature readings I went shopping.

 

http://www.amazon.com/gp/product/B000MX5Y9C?ie=UTF8&tag=wwwgalttechco-20&linkCode=as2&camp=1789&creative=390957&creativeASIN=B000MX5Y9C

I also had a digital thermometer with a probe on hand just for another set of data.

 

Temperature at the thermostat was 25.5c (down the hall)

Temperature in the room was 27.2c (sitting on my desk)

Temperature outside the server was 25.1c (12 inches from the server)

Temperature inside the server was 24.7c (reading was from behind the backplanes on a SC933)

 

I use a mixture of Samsung drives HD502HJ, HD154UI, HD203WI and HD204UI.

When I first woke the array up, all drives reported temperatures of 19c-22c. This includes a Seagate 9BJ13G 7200rpm I had in for pre-clearing. I let the array "warm up" while I went and got a beer and heated up a snack downstairs, then refreshed the page. One of the HD154UI's was reporting a temperature of 24c, the rest ranged from 25c up to 30c. All drives of the same model were within 1 degree of each other.

 

Since nothing looked out of the ordinary, and I wanted to spend more time with my new $80 toy I decided to gather more info.

 

I am sure there is a better way to do this but this is what I did. I took a Screenshot of the array then stopped it. I then pulled one of each of the 4 types of drives and used the Infrared thermometer on the bottoms of them since the bottom temperature reading was higher. I took readings from each of the 4 drives in the same spot right next to the spindle.

 

Each reading was within 2c of the temperature listed in unRaid. The 154UI and 204UI were colder then unRaid reported, while the 502HJ and the 203WI were warmer.

 

I'm not really sure what controls were used by the people reporting this issue, but based on my results I do not see anything here that says Samsung's have false temperature readings. Either people didn't get an accurate ambient reading, or they had a faulty drive.

Link to comment

I guess I'll go against popular opinion also. I have a couple of Samsungs & they report their temperature accurately. They read around the same temp as my WD's and Seagate drives. We are not the Norm or Samsung did release a few bad batches of drives. All manufacturers do do that from time to time.

Link to comment

Ok, I returned the F4 2TB to Amazon.  Got to say brilliant service, they took the drive back no quibbles and even paid for the postage.

 

So completely confused now.  I need more storage soon, and am running 750GB drives at the mo.  So I'll be replacing the parity drive.  So it needs to be a drive I can trust 100%. 

 

If the F4 2TB is out of the question what should I be looking at?  The 154UI 1.5TB drive, the F3 2TB, or switching to WD - the 2TB WD EARS??  Bloody annoying TBH since the F4 offered best in class for performance, efficiency and cost.  Or I wait in anticipation that a patch will be released to make the newer drives work - but that could be a long time :(

 

These are the reviews that I've been reading from April of this year:

http://hothardware.com/Reviews/Definitive-2TB-Hard-Drive-Roundup/?page=1

 

Cheers, Matt.

Link to comment

Since no one was reporting how they were getting their temperature readings I went shopping.

 

http://www.amazon.com/gp/product/B000MX5Y9C?ie=UTF8&tag=wwwgalttechco-20&linkCode=as2&camp=1789&creative=390957&creativeASIN=B000MX5Y9C

I also had a digital thermometer with a probe on hand just for another set of data.

 

That's an expensive toy!  I recently bought this one on sale for $15.  Probably lower quality, but it seems accurate so far.

 

Thanks for the extensive testing.

 

johnieutah: The only 2 TB drives that I know of that will work out of the box are the Samsung F3s, the Hitachi drives, and WD EADS drives.  The WD is the only green drive in that bunch.

 

I also recommend the WD EARS drives with a jumper (as they are cheaper than the EADS).  After installing the 2 cent jumper, the WD EARS will perform exactly like an EADS.

 

The Seagate LPs are OK, but upgrading the firmware is a huge pain.  I tried and failed, so I'm currently running one with the old firmware and just hoping for the best.  I ordered another one during the Black Friday sales for just a bit over $60, which is the cheapest I've seen any 2 TB drive.

 

I agree that this is an annoying time to purchase a new drive.

Link to comment

I also think the f3's are one of the better drives. Seagate has scared me away since that 1.5 fiasco, as soon as Samsung burns me I'm sure I'll forgive Seagate.

 

I had 2 drives in my Raid 5 array fail within hours, then my external drive with the only backup of a few files failed as I was copying the data back. All of them were 1.5 Seagates. Needless to say I was unhappy, but they sent me replacements...

Link to comment

Meh. I don't get scared off from brands based on things like that. It was certainly a serious debacle, and I don't begrudge anyone the choice to shy away from Seagate, but it's ironic that Hitachi is listed as being a trouble free alternative, since just a few years before the Seagate debacle, the Deskstar line [then owned by IBM], went through a serious boondoggle of its own. We called them "Deathstars" at the time, if that gives you a hint. Pretty much every drive company has gone through its lumps to one degree or another.

 

For that matter, I had a Samsung 1TB that I RMAed back, and the drive I got from them failed to pre-clear properly, and I had to send that one back as well. That is an experience that could sour someone on a brand. But they handled it reasonably well, and I now I have a working drive, so it's all good.

 

The bottom line is that the Seagate LP has been outstanding as my unRAID parity drive. That's while running the CC34 firmware. I haven't seen from any source that the CC35 firmware is required for all drives. What I've mainly seen is stuff about trying to fix periodic clicking while idle, which is the drive recalibrating. And updating the firmware didn't always fix the issue. My drive hasn't done this at all, so I haven't bothered to try the firmware update.

Link to comment

WARNING: There is a really scary issue with all the Samsung F4 drives, as reported in this thread: http://lime-technology.com/forum/index.php?topic=9339.msg89029

 

The bug report indicates Samsung F4's can generate Bad Blocks when hdparm or smartmontools or any other disk tool is used on them. They will silently corrupt your data!

 

Stay away, far away from these drives if you value your data!

 

Yup evil bug makes these drives a shambles waiting to happen. 

 

More info :-

 

http://sourceforge.net/apps/trac/smartmontools/wiki/SamsungF4EGBadBlocks

http://hardforum.com:80/showthread.php?t=1566067

http://hardforum.com/showthread.php?t=1551753

 

Link to comment

That firmware release is quite sloppy and will cause another mess. They did not change the firmware revision at all. There is no way to know if new drives have the updated and claimed to be fixed firmware or not. For now one will have to flash the firmware regardless of what it claims to be, even months from now.

 

I'm guessing and implying malicious intent that they purposefully did not change the firmware revision in an attempt to sweep this issue under the rug without anyone noticing.

 

Shame on Samsung for this future headache.

Link to comment

That firmware release is quite sloppy and will cause another mess. They did not change the firmware revision at all. There is no way to know if new drives have the updated and claimed to be fixed firmware or not. For now one will have to flash the firmware regardless of what it claims to be, even months from now.

 

I'm guessing and implying malicious intent that they purposefully did not change the firmware revision in an attempt to sweep this issue under the rug without anyone noticing.

 

Shame on Samsung for this future headache.

With that in mind how do you even know if the update took place???

 

Without some way of confirming you have fixed firmware I'd stay WAY clear of these drives...    Even if you run the fix yourself... did it take? is it corrected?  do you have to test it by forcing an error?  Without a firmware revision number change you have no clue it worked at all.

 

As I said.. too many other disks in the market to use instead. 

 

Joe L.

 

Joe L.

Link to comment

A shame, really, esp. since the F3's were so good. Why is it that manufacturers seem to take willing turns in shooting themselves in the foot? Release a drive with major problems (IBM Deathstar, Seagate, and now Samsung), and then are sneaky about the issue and the fixing process? They try to cover it up, which only makes matters worse.

Link to comment

Here is the reason from the Samsung HDD department for the firmware number staying the same:

 

I just wanted to let you know the firmware version will not change after flashing. The version did not change because

 

it takes a long time to get customers approval for change notices. It would have taken over a month before we could

 

have released the fix.

 

Please make sure that for each drive the flashing process is successful…

 

We have already implemented the fix to our production. So drives available on the market in 2 months time will have the

 

fix applied. To be sure check the manufacturing date on the drives, January 2011 would be safer…

Link to comment

Here is the reason from the Samsung HDD department for the firmware number staying the same:

 

I just wanted to let you know the firmware version will not change after flashing. The version did not change because

 

it takes a long time to get customers approval for change notices. It would have taken over a month before we could

 

have released the fix.

 

Does this mean they've only released the current fix as a (dirty) stop gap? And there will be a proper release with a revised firmware # in a month or so once it's been through their internal procedure?

 

If this has happened because of their internal release policy - that needs fixed.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.