Seagate ST3000DM001 Slow Write Speed

[Chymerix]

Hi,

I just purchased a Seagate ST3000DM001 and noticed that my write speeds are consistently slower than what I'm seeing in various reviews. Does anyone know what might be contributing to this difference? Thanks.

 

[Schmide]

AHCI mode?

 

[Chymerix]

It's already in AHCI mode.

 

[BFG10K]

Slow write speeds can be caused by reallocated sectors, AKA faulty drive. Run a health check on it.

 

[bradley]

I would also make certain your SATA cable is properly seated and not defective, also that the drive is properly aligned. Or try disabling APM using CrystalDiskMark or HDParm, and try updating the firmware. If all else fails, create an RMA ticket.

 

[Chymerix]

I have two of the same drive and testing both yields the same results, so I'd be surprised if they were both faulty. I updated the firmware on both drives and drivers on the Marvell controller before I ran the test in the original screen shot.

What's strange is that I get better sequential results on USB 3.0 so I'm starting to wonder if it's an issue with the SATA port on the external enclosure I have: Mediasonic ProBox 4
http://www.amazon.com/ProBox-Drive-E...pr_product_top

It might also be an issue with the SATA PCI-Express card I have: StarTech PEXESAT32
http://www.amazon.com/StarTech-PEXES...pr_product_top

The SATA port on the enclosure is SATA II, but I'm not reaching anywhere near the speed limitations of SATA II. Any other ideas? Thanks.

 

[Chymerix]

After some debugging I've determined that the issue is with the enclosure and possibly the esata ports. I'm not sure why, but whatever circuitry is needed in the enclosure is causing a slowdown. I was able to verify this on two enclosures that I own: The Mediasonic ProBox and a single drive Rosewill enclosure.

When I plugged the drive directly to an internal sata port the speeds were significantly faster and more consistent. I might return the ProBox in favor of a solution like this: http://kingwin.com/products/cate/mobile/racks/kf_1000_bk.asp

I haven't entirely ruled out the pci-express sata card, but I know for sure that the card is faster than the external esata ports provided by the motherboard (Asus P8P67 Deluxe - JMicron). The controller must be pretty bad, as the drive barely broke 90MB/s when plugged into this port.

The image below shows the different configurations in order of increasing performance. I hope this helps anyone in a similar situation. The bottom line is that it doesn't seem possible to achieve SATA speeds with eSATA.

 

[BFG10K]

It would have been really useful to tell us you’re using an enclosure and also add-in SATA cards. None of this was mentioned in the OP.

It is possible to get good eSATA speeds, you just need a back-plate that’s basically a pass-through and connects directly to internal Intel ports. Then it should be identical to internal speeds.

 

[Smoblikat]

I have two of the same drive and testing both yields the same results, so I'd be surprised if they were both faulty. I updated the firmware on both drives and drivers on the Marvell controller before I ran the test in the original screen shot.

What's strange is that I get better sequential results on USB 3.0 so I'm starting to wonder if it's an issue with the SATA port on the external enclosure I have: Mediasonic ProBox 4
http://www.amazon.com/ProBox-Drive-E...pr_product_top

It might also be an issue with the SATA PCI-Express card I have: StarTech PEXESAT32
http://www.amazon.com/StarTech-PEXES...pr_product_top

The SATA port on the enclosure is SATA II, but I'm not reaching anywhere near the speed limitations of SATA II. Any other ideas? Thanks.

:facepalm:

I dont even need to say anything......

 

[TakeNoPrisoners]

Yea that would have been great information to put in the OP as usually external anything is garbage compared to internal inputs. Drives benchmark differently when they are hooked up to internal ports compared to being hooked up to some random external controller.

 

[Chymerix]

I made the assumption that using an eSATA port in an enclosure was irrelevant given that everything was marketed as 3 Gbps SATA II. I was wrong.