aac0 (TYPE 502) TIMEOUT after try coppy large file

[dev246]

Hi
I have FreeNas 9.1.5 stabile witch 8xSata Adaptec 5805 raid controler. (FreeNas is on usb flash drive)
I use only software raid (raidZ on all 8 stata discs), witch windows share on that one big volume.
When I try copy large single file (larger than 400GB) usually I get an error aac0: COMMAND 0xffffff8001ac0500 (TYPE 502) TIMEOUT AFTER 125 SECONDS and server stops responding.

Can You help me with this error?

 

[dlavigne]

Is the firmware version of the RAID controller at its latest from Adaptec?

 

[Nick2253]

It's probably too late for this now, but you should probably be using RAIDZ2, not RAIDZ. RAIDZ is sufficient to protect against UREs, but isn't sufficient to protect you from drive failure *because* of the UREs (unreadable read error).

My guess is the problem has to do with your firmware. Make sure you update your firmware. Are you using your RAID card to manage the individual disks at all? Or is it just JBOD passthrough?

Might want to take a look at this link. Has some information related to the error. Might be able to point you in the right direction: http://freebsd.1045724.n5.nabble.co...xxxxx-TIMEOUT-AFTER-xx-SECONDS-td3883828.html

 

[cyberjock]

Umm... is that controller working in JBOD mode or did you do a bunch of RAID-0's? I will warn you that Adaptec is the second worst choice of controller that will appear to work on the surface but is actually a cause of major problems, frustration, and lost data for forum users.

Can you query the SMART data for your disks? If you can, can you post one for us to see?

 

[dev246]

Hi
I'm sorry that so late write off.
The controller had the current version firmeware but the old build, I've updated it to the latest build.

After that the messages aac0: COMMAND 0xffffff8001ac7f80 (TYPE 502) TIMEOUT AFTER 121 SECONDS I get less, but still present. Now that message not suspend the server but only SMB share is not available for a long time, then everything returns to normal until the next error
All disc-s are connected directly into controller. I don't use any JBOD.
All discs connected to Adaptec are configured as software RaidZ because I care about large storage space even at the cost of performance.

Code:

Can you query the SMART data for your disks? If you can, can you post one for us to see? 

Can you give me the command to get SMART data form disc ?

 

[cyberjock]

It'll be something like "smartctl -a /dev/yourdisksdevice". But unless your Adaptec controller is something I've never seen before it won't work.... and that's one of the primary reasons why we tell people not to use RAID controllers. You can't monitor the health of the disks, so you wake up one morning and find your pool failed last night and all of your data is gone.

And unless you have religious backups and aren't using a RAID controller like yours, you should be doing RAIDZ2 at the minimum.

 

[dev246]

This is smart data from first disc connected to Adaptec

Code:

smartctl -d scsi -a /dev/pass0
smartctl 6.2 2013-07-26 r3841 [FreeBSD 9.2-RELEASE-p4 amd64] (local build)
Copyright (C) 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org
 
=== START OF INFORMATION SECTION ===
Vendor:              WDC
Product:              WD30EFRX-68A
Revision:            80.0
User Capacity:        3,000,592,982,016 bytes [3.00 TB]
Logical block size:  512 bytes
(pass0:aacp0:0:0:0): MODE SENSE(6). CDB: 1a 00 84 00 40 00
(pass0:aacp0:0:0:0): CAM status: CCB request completed with an error
Serial number:        WD-WCC1T0350923
Device type:          <31>
Transport protocol:  SAS
Local Time is:        Tue May 20 10:13:52 2014 CEST
SMART support is:    Available - device has SMART capability.
SMART support is:    Enabled
Temperature Warning:  Disabled or Not Supported
 
=== START OF READ SMART DATA SECTION ===
(pass0:aacp0:0:0:0): LOG SENSE. CDB: 4d 00 40 00 00 00 00 00 04 00
(pass0:aacp0:0:0:0): CAM status: CCB request completed with an error
SMART Health Status: OK
 
Error Counter logging not supported
 
Device does not support Self Test logging

 

[cyberjock]

Yeah.. see how you had to do -d scsi and then do pass0.. that won't fly with FreeNAS.

Also notice how have ZERO actual SMART data on the drive. That renders SMART useless.

So by using that RAID controller you just neutered the primary means of identifying a failing disk.

So go find some new hardware bro. If you don't know what to choose we have tons of stickies on the topic.

Edit: and just to come full circle with this, it's possible that this timeout is due to a failing disk. When I saw that you mentioned your problem and that you had an Adaptec I immediately focused on that because that's a common problem. You *need* to be able to identify if you have a failing disk or not. But, since you have no way to actually prove it's a failing disk, you can't do much for troubleshooting until you fix your hardware snafu.

 

[dev246]

Now I known that adaptec is not the best choice for freenas but Can I do something with what I have ?

 

[cyberjock]

Nope. Get rid of the controller first. Then we can see what's your status once we can actually see your status.

If you read the hardware requirements for FreeNAS it says not to use a RAID controller....

 

[dev246]

So, from your experience what would be the best to connect 8-16 discs ?

 

[cyberjock]

So, from your experience what would be the best to connect 8-16 discs ?

So go find some new hardware bro. If you don't know what to choose we have tons of stickies on the topic.