HI all!  I just got my license upgrade so I could add a new drive.  I added a Seagate Freeagent 750 pro that is a couple years old via esata cabling.

 

I'm getting an error a couple times per minute moving files to it. as per the attached syslog.

I tried jiggling the cables, no luck. 

I don' t have a spare esata cable anywhere to try.

 

So I thought I'd see if anyone had some ideas about these errors.  ALL I can see is that it is in UDMA/33 which sounds like a slow old IDE drive( of which there are none in my system)

 

Thanks for looking!

 

DonkeyDon

You've apparently used a 40 conductor cable intended for CD or DVD use, or a round IDE cable not able to handle the speeds of today's drives, or a defective cable.

 

The server has reset the drive many times and reduced the speed it is using, in an attempt to get good data from it.  The only solution is to replace the cable with a good quality flat 80-conductor cable. (do NOT use a round cable)

 

(Typically, an 80 conductor cable has a blue connector on one end and blank and gray on the other.  The blue connects to the MB.  If you have only one disk, it connects to the END of the cable, not the connector in the middle)

I added a Seagate Freeagent 750 pro that is a couple years old via esata cabling.

 

You've apparently used a 40 conductor cable intended for CD or DVD use, or a round IDE cable not able to handle the speeds of today's drives, or a defective cable.

 

The server has reset the drive many times and reduced the speed it is using, in an attempt to get good data from it.  The only solution is to replace the cable with a good quality flat 80-conductor cable. (do NOT use a round cable)

 

(Typically, an 80 conductor cable has a blue connector on one end and blank and gray on the other.  The blue connects to the MB.  If you have only one disk, it connects to the END of the cable, not the connector in the middle)

 

 

I believe this is a SATA drive, In that case A round or 40 conductor is not the issue. It may be the eSATA/USB chipset interface.

it may be also be a PATA internal in the Freeagent hardware and there is an interface programming incompatibility.

 

I would try a known quality eSATA cable, try it on another machine or open the box and review the current hard drive type.

If it's an SATA drive, you can try removing and mounting it internally in your machine to insure the drive itself is good.

 

 

Boy, I missed the fact that it is an SATA drive.  (got to not answer posts before second cup of morning coffee takes effect)

 

Sorry.

 

Joe L.

 

Weebo...I never thought for a second anyone made a 750 gig drive with PATA.  Had to look it up to believe it.  I guess I might see if I can crack open the case and plug it straight into the IDE port.

 

Joe, I was confused by the log myself   No IDE cables in sight, then UDMA/33 error?

 

hopefully the separate buses will give me more parallel paths to the drives for more bandwidth, in addition to fixing errors.

 

I guess it won't be the same drive ID , so i'll have to empty it of data first and preclear again on IDE bus? 

 

 

 

Chances are high it's an SATA drive. So there might be a chipset/interface incompatibility somewhere.

The cable could be too long and/or not of the right quality.

 

This is why I suggested trying it on another machine with eSATA, getting another cable or if you have the room in your case. Mount it internally.

 

 

 

Weebo...I never thought for a second anyone made a 750 gig drive with PATA.  Had to look it up to believe it.

I've got two 750Gig PATA drives in my array... I did not think twice about it.

  I guess I might see if I can crack open the case and plug it straight into the IDE port.

 

Joe, I was confused by the log myself   No IDE cables in sight, then UDMA/33 error?

 

hopefully the separate buses will give me more parallel paths to the drives for more bandwidth, in addition to fixing errors.

 

I guess it won't be the same drive ID , so i'll have to empty it of data first and preclear again on IDE bus? 

No need to empty/clear/reload.   If it gets assigned a new "device" just go to "devices" page and assign it to the old (and not empty) slot on the array.

 

Odds are you have in your BIOS an IDE emulation mode enabled.  Set it to AHCI mode and native SATA and you'll find performance much better.

 

The person in this post had a similar situation:  http://lime-technology.com/forum/index.php?topic=4150.msg36717#msg36717

I posted a screen-shot of a section of the manual for his MB showing the defaults were to emulate an IDE drive.  You might have the same for your MB.

 

Joe L.

ok.. took a few hours for me to give up on trying to save that Freeagent drive case, and just break off the impossible plastic snaps.

 

Bare drive is back in the server with regular SATA cable. Appears in unRAID with a slightly different name, which I assigned to the same slot, disk3.

 

It wants to do a data rebuild.  Parity, Disk1 and disk 2 show green dots, only disk 3 is blue dot, so as long as drives 1 and 2 are good, only disk 3 will get rebuilt, correct?

 

FWIW, i tried it once before I changed bios to AHCI, per Joe's suggestion above.  Both before that and after, i got 3 green balls, and 1 blue one for disk3.

 

Should i try anything else before starting the rebuild?

 

Thanks again,

Don

ok.. took a few hours for me to give up on trying to save that Freeagent drive case, and just break off the impossible plastic snaps.

 

Bare drive is back in the server with regular SATA cable. Appears in unRAID with a slightly different name, which I assigned to the same slot, disk3.

 

It wants to do a data rebuild.  Parity, Disk1 and disk 2 show green dots, only disk 3 is blue dot, so as long as drives 1 and 2 are good, only disk 3 will get rebuilt, correct?

Correct

FWIW, i tried it once before I changed bios to AHCI, per Joe's suggestion above.  Both before that and after, i got 3 green balls, and 1 blue one for disk3.

 

Should i try anything else before starting the rebuild?

 

Thanks again,

Don

With any luck, you'll see improved performance, and the errors will not occur any longer.