ATA_IDENTIFY - CAM status: Command timeout

[senior_hombre]

Hi,
as i tried to switch on my NAS today i noticed that was not able the freeNAS interface using my browser, so i plugged a montor on and noticed that it got stuck in e loop of showing strange (error) message containing:
ATA_IDENTIFY and CAM status: Command timeout.
I do not know what this means or what this is related to (i guess it is hard disc related) but i made a picture will my cell phone (i do not now how to access the log any other way) i hope someone can give me a hand here:

my hardware specs are:
Mainboard: ASRock C2750D4I
6 hard discs (encrypted)
freeNAS 9.10 stable installed

edit: it also seems that only 4 of the 6 hard discs are listed in the BIOS menu

 

[senior_hombre]

Ok i have an update on this.
I just restarted the system and this time the boot went trough! The system is acessable via the browser interface, i also can access all of the hard discs! The bios still only lists 4 of them thought, but i think that was always the case because the bios only lists one of the SATA controllers.

The question remaining is: are the error messages something i have to worry about? it is possible that this just was an exception (this did not happen before)?

I still would like to do some kind of hard disc or system check, is there something i can do, maybe run some tests or something like that?

 

[Robert Trevellyan]

Start with @joeschmuck's hard drive troubleshooting guide.

 

[senior_hombre]

Ok i am currently doing the long smartctl tests

 

[senior_hombre]

Ok i did the smartctl long tests on all drives, all finished without errors!
But are not the error messages itself and the fact that a restart made them go away, hints that the problems are not the hard drives itself but the mainboard (SATA controllers) or the connection/cables?
How can i test that?

I do not reboot my NAS very often and if the error is non critical (which means it only occurs during boot and when it does i can circumvent it by rebooting until it is gone, and there is not way that the error occurs on a running system ) then it is ok. But i do no want my running system crashing!
I just want to know if i have to take any measures or not. I still have warranty on the mainboard, so i could send it in and have it replaced.

 

[Robert Trevellyan]

How can i test that?

You can do some testing by swapping drive connections. If an error stays with a drive, the finger points at the drive. If the error stays with the SATA port, or with the cable, point the finger at that. The only way to rule out other parts, e.g. motherboard or PSU, is to replace them with known good parts.

If you post your full hardware specs, you'll have a better chance of getting help from other members.
https://forums.freenas.org/index.php?threads/updated-forum-rules-8-3-16.45124/

 

[senior_hombre]

Ok, but i have to wait to the error to occur again. My hardware spec are basically just what i posted in the intitial post.
Mainboard: ASRock C2750D4I (CPU is integrated)
32GB ram
6 hard discs (encrypted)
freeNAS 9.10 stable installed

 

[Robert Trevellyan]

Full hardware specs - please review the article I linked.

 

[senior_hombre]

Ok sorry, i will be more specific:
the mainboard is a: ASRock C2750D4I board with an integrated 8 core CPU (Intel Avoton C2750, Socket: FCBGA1283 SOC)
RAM: i have four 8GB (ECC) modules (type: Kingston ValueRAM DIMM 8 GB ECC DDR3-1600) installed, with a total of 32GB
the board has 3 SATA controllers: 1. an Intel C2750 (2x SATA3, 4xSATA2), 2. a Marvell SE9172 (2x SATA3) and 3. a Marvell SE9230 (4x SATA3)
I have six 4TB hard disc drives plugged into the two Marvel controllers (i do not use the Intel ports), all HDs are of the type: WesternDigital Red SATA 6GB/s 64MB cache WD40EFRX
the drives are in a RAIDZ2 pool (two redundant drives) with a total of 16TB RAM usable (of the full 24TB)
the drives are also encrypted
the board has two Intel i210 LAN controller (Gigabit LAN 10/100/1000 Mb/s)

This is the second ASRock C2750D4I board i am using. The first one (bought 2014) broke down last year, right after the warranty was over, so i had to buy a new one. But luckily all other components where ok so i just connected them to the new board and did not even had anything to configure, all worked right from the beginning.

ok i hope that i did not forgot anything important

 

[nojohnny101]

The marvel ports on the C2250 and C2750 from ASRock are subpar. The manual of your board states to avoid using them until all Intel sata ports are filled. So even ASRock is essentially acknowledging they are not as reliable as the intel ones. People have been running hard drives off of the marvell controllers no problem, but since you are having problems, I would eliminate that as being a problem and plug all your drives into the intel ports.

the process for that is to shut the machine down, plug all drives into the intel controller, then power on.

 

[Robert Trevellyan]

My guess is you might be able to get a refund for your failed board:
https://forums.freenas.org/index.ph...s-system-fault-may-lead-to-dead-system.50314/

 

[senior_hombre]

Ok i will try the intel ports then, thanks.

And iam also a bit worried about the problem with the Avalon CPUs, it seems that the board iam using will die sooner or later. I guess it would be the best to prepare for this now. I think i will buy a different SOC (maybe with an Intel C3000 family CPU) as soon as my current board dies.
But my main concern is the migration of my current raid72 pool to a new system. Do i have to consider special measures for encrypted hard drives (like extratcting the header from the HDs or exporting/importing the keys from the old FreeNAS system to the new one)?

 

[Robert Trevellyan]

If you take the drives and the boot device to a new system, it should be pure plug-and-play. However taking all possible precautions with an encrypted pool (per the documentation) makes sense.