SOLVED Enabling S.M.A.R.T. and Viewing Test Results

[tgrosinger]

I have recently been setting up a ASRock C2750D4I and have connected 4 hard drives to the Marvell SE9230 controller. Since I only had 4 drives and I believe RAIDZ2 would only get me the same efficiency as RAID10, I decided to just setup the RAID controller with RAID10.

I enabled S.M.A.R.T. in the BIOS, however when I am logged into the machine it seems to show otherwise:

Code:

[root@freenas ~]# smartctl -a /dev/ada0                                                                                            
smartctl 6.3 2014-07-26 r3976 [FreeBSD 9.3-RELEASE-p5 amd64] (local build)                                                         
Copyright (C) 2002-14, Bruce Allen, Christian Franke, www.smartmontools.org                                                        
                                                                                                                                   
=== START OF INFORMATION SECTION ===                                                                                               
Device Model:     MARVELL Raid VD 0                                                                                                
Serial Number:    1e107f7399020001                                                                                                 
Firmware Version: MV.R00-0                                                                                                         
User Capacity:    8,001,406,238,720 bytes [8.00 TB]                                                                                
Sector Size:      512 bytes logical/physical                                                                                       
Device is:        Not in smartctl database [for details use: -P showall]                                                           
ATA Version is:   ATA/ATAPI-7 T13/1532D revision 4a                                                                                
Local Time is:    Wed Jan 14 20:13:23 2015 PST                                                                                     
SMART support is: Unavailable - device lacks SMART capability.                                                                     
                                                                                                                                   
A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.  

I setup email notifications and have been receiving one each morning saying:

The volume Storage1 (ZFS) state is ONLINE: One or more devices has experienced an error resulting in data corruption. Applications may be affected.

Though looking at Storage/Volumes shows that Storage1 is HEALTHY. I am sure it is obvious that this is my first time to FreeNAS, but please let me know if there is more information I can provide and any help is greatly appreciated.

 

[Fraoch]

SMART can't be read through a RAID controller. Plus ZFS and a RAID controller won't cooperate well together. There are also reports of some problems with the Marvell controller - use the Intel ports first.

 

[tgrosinger]

Using the Intel ports limits me to just using the SATA2 ports though. I am not sure if that would have a noticeable impact though.

So assuming I do that, would the best method be to not use RAID at all and use the OS to configure a ZFS pool? If I want to make this migration, is there a way I can do so without losing the jails I have configured? (I have the data on a temporary location still)

Thank you.

 

[Fraoch]

Mechanical hard drives are barely able to saturate the SATA I bus, let alone SATA II. There will be no impact.

However there doesn't seem to be any way around breaking the RAID array. Once you deactivate the RAID controller the data on the drives will be corrupt. Until you do, ZFS won't work at all, as you see.

You're lucky to have spotted this early before you place TBs worth of data on the pool.

 

[tgrosinger]

I don't mind breaking the RAID, just curious if I can migrate the jails I have created or if I have to start from scratch.

 

[Fraoch]

You'll have to start from scratch, without the RAID controller that data is unreadable to anything else. If there was a way to back up configurations of the jails (clone them?) that would work - haven't played with jails myself so someone else would have to say.

 

[tgrosinger]

Sounds good, I will plan on that unless I hear from someone else first. Thank you for your help.

 

[Ericloewe]

It should be possible to replicate the Jails dataset to a spare disk (ZFS, naturally) and then replicte it back.

 

[Eric Batte]

I've got the same mobo and have 6 drives connected. I keep getting email alerts on the nightly short test stating "failed to read SMART Attribute data" or "Read SMART Self-Test Log Failed". If I run smartctl -a /dev/ada? on any/all of the drives no errors are reported. It also notes that "No errors logged" in the SMART error log.

Since I'm getting these strange reports on nearly all of the drives, which are brand new, I suspect something else might be the problem. Any ideas what to check next?

Thanks

 

[Ericloewe]

Can you post the output of one of the drives on pastebin and link to it here?

 

[Eric Batte]

http://pastebin.com/5MFcbYJa



Error reported in email for the same device:

The following warning/error was logged by the smartd daemon:

Device: /dev/ada2, failed to read SMART Attribute Data

Device info:
WDC WD40EFRX-68WT0N0, S/N:WD-WCC4E6PVSZ1D, WWN:5-0014ee-20b4a1538, FW:82.00A82, 4.00 TB

For details see host's SYSLOG.

You can also use the smartctl utility for further investigation.
No additional messages about this problem will be sent.

 

[Eric Batte]

Here is the other error I've gotten:

The following warning/error was logged by the smartd daemon:

Device: /dev/ada2, Read SMART Error Log Failed

Device info:
WDC WD40EFRX-68WT0N0, S/N:WD-WCC4E6PVSZ1D, WWN:5-0014ee-20b4a1538, FW:82.00A82, 4.00 TB

For details see host's SYSLOG.

You can also use the smartctl utility for further investigation.
No additional messages about this problem will be sent.



They seem to be inconsistent. One day I'll get one error for a drive, the next day I won't, then a different error, etc.

 

[Ericloewe]

What controller are you using? Intel and/or one of the Marvells?

 

[Eric Batte]

Looks like I'm using the Marvell controller. Needed to check the setup to be certain. When I did I noticed that SMART was disabled in UEFI setup, so I enabled it. Not sure if using the Marvell or disabled SMART might be contributing to this issue.

 

[Ericloewe]

Try the Intel controller.

 

[Eric Batte]

Done. I will report back if errors persist.

Is it documented anywhere that the Marvell controllers do or don't work?

 

[Ericloewe]

Let's say they're flaky. Some people don't have problems, some do.

ASRock seems to have a new firmware for one of the controllers, which disables RAID and should be more stable. That might also help, though it may not have been officially released yet.
Let's say

 

[Zpid]

I have the same motherboard and have sucessfully disabled the raid using the offficial "disable marvell raid bios" here: http://www.asrockrack.com/support/ipmi.asp#Marvell9230
After alot of ahcich timeout in the logs before updating the firmware it is now working fine since a week, my original post is here:
https://forums.freenas.org/index.ph...t-getting-ahcichx-timeout-on-xx-port-0.26468/

 

[Eric Batte]

Good find. FYI, I did not enable raid in the bios on my setup. Did you?

In the meantime I moved all the cables to the intel SATA ports and for the first time I didn't get any error emails this morning.

 

[Zpid]

Good find. FYI, I did not enable raid in the bios on my setup. Did you?

In the meantime I moved all the cables to the intel SATA ports and for the first time I didn't get any error emails this morning.

No i did not enable raid, and now after the bios upgrade it's not possible to enable in bios.
Going to use the remaining intel SATA ports to another set of 4x4TB hdd's to run in raidz1 i think.