Is this a normal HDD curve in speed test

[mrbilky]

Is this a normal HDD curve in speed test for a drive in the storage array, it's a WD RED 4TB the average was around 140mb

 

[JorgeB]

Yes

[mrbilky]

24 minutes ago, mrbilky said:

Is this a normal HDD curve in speed test for a drive in the storage array, it's a WD RED 4TB the average was around 140mb

 

Thank you!

[JorgeB]

And it's one of the newer faster models, which I suspect use short stroked 1.2TB platters.

[mrbilky]

Just now, johnnie.black said:

And it's one of the newer faster models, which I suspect use short stroked 1.2TB platters.

Well they were bought only a few moths ago and placed in the system together I was just shocked at the steep curve I've never seen a curve before always just a number associated with the final speed like in crystal disk mark

[SSD]

The beginning of your test was when the "outer tracks" were being read. More data can be written to the outer tracks because they are longer (the circumference of the outer track is longer than an inner track). As it reads from outer to inner, less and less data can be read in a single rotation. And since the rotation speed is a constant, less and less data is read per unit time.

 

All spinner hard drives are like this. SSD don't spin and you should get a very consistent speed from start to finish.

[mrbilky]

45 minutes ago, SSD said:

The beginning of your test was when the "outer tracks" were being read. More data can be written to the outer tracks because they are longer (the circumference of the outer track is longer than an inner track). As it reads from outer to inner, less and less data can be read in a single rotation. And since the rotation speed is a constant, less and less data is read per unit time.

 

All spinner hard drives are like this. SSD don't spin and you should get a very consistent speed from start to finish.

Ahh yes I actually remember now reading up on short stroking a hard drive and does improve read/write speeds anybody doing that in unRAID? Probably not effective in this use case?

[SSD]

Not sure. Short stroking makes the slowest tracks inaccessible. If this drive is indeed short stroked, the slowest speed is not as slow as it would have been had the drive not been short stroked.

 

The reason for short stroking is normally to increase the performance of a drive, But in this case, if indeed true, I think it would be to gain efficiencies in manufacturing by standardizing the platters and other components. The speed improvement would just be a nice side effect.

 

The WD REDS are slow as compared to 7200 RPM drives. They are even slower than other slower RPM drives like the Seagate Archive drives. But high speed is not altogether important for most use cases of a media server. The advantages of lower price, reduced energy consumption, and perhaps longer life are typically more important.

[Fireball3]

bjp999 you changed your nick?