Usually, anything above 8th port is routed through the expander.
I have been selling Areca PCIe 3.0 cards for several months already. Could get you one unless you always shop Newegg. :)
@Ducs
Areca controllers, at least the 12/16/24-port cards use a built-in expander. That external port you have mentioned is used exactly what you would need it for - for daisy chaining a bunch of external enclosures. I still do not get it as to why you need an In/Out port on a controller.
@mikeblas
Adding drives into a config one by one, the drives increase vibration being produced by the setup.
When vibration exceeds drive's tolerance threshold drives will start producing random timeouts and if they are part of a RAID set, then controller will start dropping drives out as soon...
There is a good chance that vibration limits your ability to add more drives in the setup. What kind of drives do you have? Are they all identical?
Try what VR suggested - remove the drives from your case and hook them up to the controller while they are all placed on a hard and solid surface.
Pull just the drives out of the 847J, leave everything else connected. If you get the stability back, then start plugging them back in one by one and checking for stability between each drive.
Out of curiosity, are the drives of the same brand? Samsung? Seagate? what model? If you get the...
Since you have Disk Write Cache enabled, it explains why the writes are so much faster. Yet, I'd certainly expect faster reads from 8x SSDs in RAID0.
Have you tested the drives individually to see if one (or possibly more than one) of them cripples the speed?
First of all, YES, you need to get rid of that drive! But not so fast.
Update the F/W and all other filesets along with it.
Since you do not know your RAID set initial parameters, it may take a few tries before you find the right ones.
In order to be able to make another attempt to recover...
@Zerosum
After all is said and done, try swapping the cable which you use to connect Areca card with an expander.
It may be at fault. Also, direct connection between the drives and an Areca controller might shed some light on the problem as well.
When you had the drives connected and created a new volume set, what SCSI/ID/LUN did you assigned it to? If it is not 0/0/0, see if changing it makes any difference.
Some time ago there was a bug in F/W where controller would not boot if out of all volume sets there was no one with 0/0/0.
Second that... Hitachi, when connected to an Areca RAID controller, will automatically give full control to the card and you do not have to move a finger for that.
I know there are people out there that might say it is possible to expand a RAID set by adding more than one drive at a time. A word of advice, I would strongly encourage you to expand your RAID set by adding one drive at a time.
As an Oracle DBA I get to work with enterprise grade storage...
Just remember - RAID by no means is a replacement for a backup. There are a lot of things that could go wrong with it and losing 12TB worth of data would be very painful.
I hope you have Hitachi Ultrastar drives for a setup this big as with regular Deskstar drives it is asking for troubles...
RAID30/50/60 are hard to maintain and once created they are not expandable. Better to build 2 independent RAID6 arrays and stripe them on OS level. Personally, I'd choose 2 individual RAID6 arrays over RAID60.
In case if you are only testing the drives and to get the RAID6 option available in the drop down list you need to have at least 4 drives in a RAID set. Delete the existing RAID set and re-create it with 4 or more drives as RAID6 or simply expand the existing one.
To make the transition to run...
Since RAID6 requires more drives than RAID5 for the same capacity you need to add one drive, expand the RAID set and only then you will get an extra option in the drop down list allowing to change the RAID level. Often times you may hear or see these acronyms OCE/OLCE (Online Capacity Expansion)...
The only way to change your RAID5 into RAID6 is by adding an extra drive. Then you should receive an option either to expand capacity of your RAID set or migrate the RAID level from 5 to 6.
When you create a volume set you need to choose support for volumes greater than 2TB (64LBA). You probably missed that part and now your controller present the LUN as 2TB only. Refer to User's manual.
The drive #11 might be at fault here and/or its cable. Remember, they are SATA drives and communicate in simplex mode with the expander, whereas SAS do it in full-duplex mode. Expander has to wait for a SATA drive to respond back once a command (I/O request) was issued and may as well go belly...
Try disabling Spin Down Idle option and see if it helps. With 11 drives configured for staggered spin up at 0.7 the controller may still think the drives (the RAID set that is) are not read for longer than 7 sec and starts dropping them off.
What I had recommended to you was once recommended to me by Kevin Wang.
However, if you checked the history, you would find that after you got an answer from Benjamin my next reply was to update the F/W and then use Benjamin's suggestion. :) It does not say a thing about pulling the drives...
One thing that jumped out at me right away - the drive's temp! It has exceeded 55C (out of its operational specs)!
Get better cooling for your system!
Secondly, never keep any setting at Auto. Choose either Enabled or Disabled option. This way it would be you making a decision, not the internal...
Since the controller has already found some errors, I'd stop the check and re-run it with scrubbing and parity re-calculation. See if it makes any difference. With all drives being in place there is a high probability that scrubbing and parity re-calculation could take care of the errors. Do...
OK, a good news for a change! :)
Now, first run the volume set check, but for the first run UNcheck the scrubbing and parity recalculation.
Once completed, do the same, but with both options checked. If it completes with no errors then consider the data as completely recovered. Still...