arc-1680ix-24 timeout issues ongoing with seagate sata

[markm75]

I have the areca arc-1680ix-24 card in a supermicro 24 bay chasis..

I have tried all I can think of to keep this from occurring but timeouts occur almost daily on several ports.. even after replacing the harddrives with either the same model and new or a different model and new.

Originally this issue was caused by needing a different firmware on these SATA drives than they came with..

In particular the ST31000340NS.. I had run the update firmware for this drive via the web interface and it updated it to AN05 as originally recommended by areca.

Currently, slot-4 has this same drive but with the SN05 firmware.. I ran the update on the drives but this one wont update.. (I think its an issue with a certain vendor sending “special” versions of this drive that cant be updated)..
My slot-18 has a newer model drive.. ST1000NM0033-9ZM173 with firmware SN04.. For some reason its showing sata speed 150, Slot-3 has another one of these drives and it has “Failed” and shows sata150..

Earlier today even Slot-1 with firmware AN05 (st31000340NS) reported a timeout..

I show firmware is v1.45 and boot rom v1.46.. sas 4.4.3.0


At this point, I’m trying to figure out if this is a controller failure (if i were to replace with exact same model controller and would the arrays come back without any effort or even a newer model), or maybe its related to firmwares on the drives or a setting that needs set on the controller, or possibly the chasis itself?

Anyone have any suggestions,

Thanks in advance

 

[/dev/null]

have you set idle timeouts to 0 or disabled them?

 

[markm75]

have you set idle timeouts to 0 or disabled them?

I had not.. where is the configuration for this.. i dont see an option for this in the areca web interface?

 

[Blue Fox]

You probably need to update to a later firmware version to reveal these options. 1.51 is a few years newer than what you're running.

 

[markm75]

have you set idle timeouts to 0 or disabled them?

I've updated the firmware..

At this point, at least from here forward.. in HDD Power management.. stagger on power control is set to 0.7, time to hdd low power idle disabled (default), time to hdd low rpm mode disabled (default).. time to spin down idle hdd disabled (default).

I've reactivated as hot spares the formerly "timed" out drives.. its back to a rebuilding state.. time, pun intended will tell, if more timeouts occur.

 

[markm75]

I've updated the firmware..

At this point, at least from here forward.. in HDD Power management.. stagger on power control is set to 0.7, time to hdd low power idle disabled (default), time to hdd low rpm mode disabled (default).. time to spin down idle hdd disabled (default).

I've reactivated as hot spares the formerly "timed" out drives.. its back to a rebuilding state.. time, pun intended will tell, if more timeouts occur.

Well the timeouts left at default didnt fix the issues..

I'm still getting timeouts and failed arrays..

One drive in particular, the st100nm0033 has timed out and failed, showing sata 150 speeds for some reason (pretty sure there are no jumpers set on this)..I have another exact same drive of this type on this card and it reports sata 300 (Sn04 firmware) and is working.


I had an issue with one ST31000340NS with couldnt take the AN05 firmware, i pulled it out (must be some weird non upgradable version, which we recently got as well).. despite this the nm0033 failed/didnt go back to 300 ( i thought was maybe causing the issue)..

Unsure what to do next..

Option 1:
One thought was to replace the card with another (ebay used is all i can find).. if i simply swap out the card, i'm hoping/assuming that all the arrays will just magically be there with the new one since its identical?

Option 2:
Get another exact card.. put it in at the same time as the old one is still in there.. either buy 3-4 new drives to try to migrate and test to see if timeouts occur with new drives.. or maybe move a few arrays over to the new card and hope they are working after switching..
A sub option here is to try switching mini-sas cables..

The other thought i had was maybe the supermicro 24 bay chassis (SC846TQ-R900B) has some sort of issue..

Either way this is messy without a 2nd server to migrate things to, then test (backup server would take 2+ days i'm sure to restore 5TB over gigabit (dpm).