freebsd stable - Oct 2010 - ATA error messages

Hi, all,

recently I set up a FreeNAS box:

FreeBSD freenas.intern.punkt.de 7.3-RELEASE-p3 FreeBSD 7.3-RELEASE-p3 #0: Sun
Oct 17 12:57:33 CEST 2010    
root@dev.freenas.org:/usr/obj/freenas/usr/src/sys/FREENAS-amd64  amd64

This is a HP/Compaq Microserver (Proliant N36L). After setting up
the system with 4 used 500 GB Seagate drives to find out if FreeNAS
would fit our needs, I replaced the drives with WDC RE4 2 TB
(WD2002FYPS) and put the system in production to store backups.

Yesterday the file system perfomance suddenly came to a crawl
and the system log fills with entries like this:

ad8: TIMEOUT - READ_DMA48 retrying (1 retry left) LBA=3907027968
ad8: WARNING - SETFEATURES SET TRANSFER MODE taskqueue timeout - completing
request directly
ad8: WARNING - SETFEATURES SET TRANSFER MODE taskqueue timeout - completing
request directly
ad8: WARNING - SETFEATURES ENABLE RCACHE taskqueue timeout - completing request
directly
ad8: WARNING - SETFEATURES ENABLE WCACHE taskqueue timeout - completing request
directly
ad8: WARNING - SET_MULTI taskqueue timeout - completing request directly
...

Only one of the for drives shows these timeouts.

My question: does this necessarily mean I have an isolated problem
with that particular drive, since the other three are running just fine,
or could there still be someting more fundamental going on?

atapci0: <ATI AHCI controller> port
0xd000-0xd007,0xc000-0xc003,0xb000-0xb007,0xa000-0xa003,0x9000-0x900f mem
0xfdfffc00-0xfdffffff irq 19 at device 17.0 on pci0
atapci0: [ITHREAD]
atapci0: AHCI Version 01.20 controller with 4 ports detected
ata2: <ATA channel 0> on atapci0
ata2: [ITHREAD]
ata3: <ATA channel 1> on atapci0
ata3: [ITHREAD]
ata4: <ATA channel 2> on atapci0
ata4: [ITHREAD]
ata5: <ATA channel 3> on atapci0
ata5: [ITHREAD]
...
ad4: 1907729MB <WDC WD2002FYPS-02W3B0 04.01G01> at ata2-master SATA300
ad6: 1907729MB <WDC WD2002FYPS-02W3B0 04.01G01> at ata3-master SATA300
ad8: 1907729MB <WDC WD2002FYPS-02W3B0 04.01G01> at ata4-master SATA300
ad10: 1907729MB <WDC WD2002FYPS-02W3B0 04.01G01> at ata5-master SATA300

Thanks,
Patrick
-- 
punkt.de GmbH * Kaiserallee 13a * 76133 Karlsruhe
Tel. 0721 9109 0 * Fax 0721 9109 100
info@punkt.de       http://www.punkt.de
Gf: J?rgen Egeling      AG Mannheim 108285

On Tue, Oct 26, 2010 at 11:33:25AM +0200, Patrick M. Hausen
wrote:
> recently I set up a FreeNAS box:
> 
> FreeBSD freenas.intern.punkt.de 7.3-RELEASE-p3 FreeBSD 7.3-RELEASE-p3 #0:
Sun Oct 17 12:57:33 CEST 2010    
root@dev.freenas.org:/usr/obj/freenas/usr/src/sys/FREENAS-amd64  amd64
> 
> This is a HP/Compaq Microserver (Proliant N36L). After setting up
> the system with 4 used 500 GB Seagate drives to find out if FreeNAS
> would fit our needs, I replaced the drives with WDC RE4 2 TB
> (WD2002FYPS) and put the system in production to store backups.
> 
> Yesterday the file system perfomance suddenly came to a crawl
> and the system log fills with entries like this:
> 
> ad8: TIMEOUT - READ_DMA48 retrying (1 retry left) LBA=3907027968
> ad8: WARNING - SETFEATURES SET TRANSFER MODE taskqueue timeout - completing
request directly
> ad8: WARNING - SETFEATURES SET TRANSFER MODE taskqueue timeout - completing
request directly
> ad8: WARNING - SETFEATURES ENABLE RCACHE taskqueue timeout - completing
request directly
> ad8: WARNING - SETFEATURES ENABLE WCACHE taskqueue timeout - completing
request directly
> ad8: WARNING - SET_MULTI taskqueue timeout - completing request directly
> ...
> 
> Only one of the for drives shows these timeouts.
>
> My question: does this necessarily mean I have an isolated problem
> with that particular drive, since the other three are running just fine,
> or could there still be someting more fundamental going on?
It's impossible to tell at this stage.

Please install ports/sysutils/smartmontools (version 5.40) and provide
the output from "smartctl -a /dev/ad8".  We'll start there.

Also, you might be interested in using ahci.ko (not ataachi.ko) since
your controller supports AHCI.  You'd gain NCQ capability and be given
some of the perks of using CAM.  Your disks would be renamed from things
like ad4 to ada0 and ad6 to ada1, however.  Just FYI.  This is unrelated
to your problem though.

-- 
| Jeremy Chadwick                                   jdc@parodius.com |
| Parodius Networking                       http://www.parodius.com/ |
| UNIX Systems Administrator                  Mountain View, CA, USA |
| Making life hard for others since 1977.              PGP: 4BD6C0CB |