Two new drives do not bench the same?

E

ep0x73

2[H]4U
Joined
Sep 5, 2013
Messages
2,574
Bought two HGST NAS 4TB a week ago.

The one drive is consistent in atto peaking at 178MB read/write.

The second drive can only hit 168MB read/write.

Crystal the 2nd drive is benching 30MB slower.

A full scan shows nothing wrong.

Same drive, same firmware and made the same time.

Seems unusual, they should both be close to the same.

10MB is not much but they should be identical.
 
The fact that they have the same model number doesn't mean they are identical. There have been cases in the past where a company has reduced the number of platters on a series of drives without changing the model number, for example.

Another thing is to maybe try swapping the ports around to see if it's the ports that have different speeds for whatever reason, rather than the drives.
 
These are the same batch from the same time/plant so they are identical in all aspects.
I changed ports, I even tossed it on the Marvell controller and the same, it's running 10-15mb slower then the other drive.

HDtune on drive 1 is smooth, drive 2 it's erratic with huge swings up and down.

That is not normal.

I am still within my 30 days for an exchange.

Something just seems goofy.

I have a pair of 3TB and they are smooth in HDtune as well, I mean a nice curve, not erratic up and down.

drive 1
2lc1ft5.png


drive2
20h56om.png


both drives, same model, firmware and even controller.

Something is off with drive 2
 
Last edited:
Did you swap SATA cables? Were both drives full formatted? Also make sure there was no data on either drive during the benchmark so that no indexing or AV scanning would be running in the background.
 
My guess is that drive 2 is actually faulty but is not showing up in tests yet.

If I had 2 identical drives, and one of them was doing that, I would be getting it replaced.
 
drescherjm said:
Did you swap SATA cables? Were both drives full formatted? Also make sure there was no data on either drive during the benchmark so that no indexing or AV scanning would be running in the background.
Click to expand...

Both drives were put into the same hotswap bay so the cable is not the issue or the first drive would be acting erratic too.
I put 60 hours on drive one to let it "burn", pulled and put in 2 when I noticed this.

Both drives benched raw with HDtune, then formatted and tested again.

I even deleted the drive to raw again and ran HDtune and it was still erratic.

Suppose I'll have to contact NE and see how well they like to do swaps.

Both my 3TB NAS drives bench the same as the drive 1 in the graph.
 
I think that the QAs just missed to let that 2nd drive pass the quality check. :)
 
windft test came back fine for surface scan.

In 9 hours up time raw read has already dropped to 98 from 100.

196609 hardware errors already

"Stores data related to the rate of hardware read errors that occurred when reading data from a disk surface"

The disk is 100% blank, I never put a single file on it.

One of the heads must not be 100% which is why the graph is so erratic.

juxnqb.png


NE is going to exchange it for free.
 
That temp is murdering the drive, lol, 187...

In all seriousness, glad they are replacing it because it is defective for sure.
 
Here is the odd thing.

A minute after I took that screen the C1 errors were all fixed.

HGST does not show C3, hardware ECC recovered, it appeared to work though.

I never see any C1 errors on any of my wd drives.

Seagates quite a few but if the ECC number matches they are all "fixed"

2hrdg1w.png


Ran another slew of benches, still the same slow read/write and erratic behavior in tune.
 
It physically worked and with the retail packaging and being packed well it did not show any damage.
Appears to be media related, one of the platters or more do not have as strong a magnetic coating.
With WD moving to 1.25TB platters I wonder how long before HGST does the same since they will get WD tech?

I'd bet the NAS line gets updated before the end of the year and gets even faster.
 
Try to get it RMAed though and see if the replacement would give you better results like the first drive.
 
Already done, it's going back for a exchange from NE who is sending me a new one for free.

I liked the Samsung F3 stairs graph.

That is how a drive should behave

2u6zazb.png
 
You must log in or register to reply here.
Back
Top