HDD dropping out of pool, too many errors...

[jagdtigger]

Hi all.

So as the title describes one of the hdds dropped out of pool a few weeks ago. Ran a long smart test but it completed without error. Just to be on the safe side after resilvering swapped it into a different slot(with the machine shut down). Yesterday it dropped out again, so its not the back plane or cables (each hdd has its own sata connection and only this hdd has issues). Both 12 and 5V is within tolerance. Ran a long smart test again but it again completed without errors. After resilvering i started a scrub and already there is a checksum error on the same hdd.

Can i consider that hdd faulty or i missed out something?

Thanks in advance!

 

[Samuel Tai]

Yes, that's the most likely scenario, if the errors follow the disk to a different port and cable.

 

[jagdtigger]

https://icybox.de/en/product.php?id=333

This is the backplane, each bay has its own sata connection, only thing common is power. But in that case all of them would have issues i think.

Never mind, its a zombie hdd i think, watch 195 Hardware_ECC_Recovered:

da1 ecc

youtu.be

 

[sretalla]

also the sata connections are separate but end up in the same sas connector on the hba.

SAS breakout cables can have issues with one single "broken-out" cable... do the movement of the disk to a different slot as suggested or swap the entire breakout cable.

 

[jagdtigger]

SAS breakout cables can have issues with one single "broken-out" cable... do the movement of the disk to a different slot as suggested or swap the entire breakout cable.

Seems like unintentionally did that, the position where it was is on a different sas breakout (5 hdd bay and one breakout has 4 sata connector on it). I found to problem anyway, seer the video in my previous post. That attribute counting up like that aint normal IMO.