smartmontools with aac

[Don Bowman]

From: Scott Long [mailto:scottl at freebsd.org]
> Don Bowman wrote:
> > is there a way to use smartmontools with
> > aac?
> > 
> > e.g. smartctl wants to see 'ad0' or 'da0'
> > as the disk to connect to. For aac, it is
> > aacd0. Specifying:
> > 
> > smartctl -i -d scsi -a aacd0
> > gives
> > Smartctl open device: aacd0 failed: Inappropriate ioctl for device
> > 
> > I'm not really sure what ioctl are being used,
> > i guess i should look in the smartmontool source.
> > 
> > --don
> 
> AAC containers do not export SMART events.  You can monitor individual
> disks by enabling the aacp driver and watching the /dev/passN devices,
> or you can script the aaccli to report events for you.

The problem i am trying to work around is with camcontrol inq
on the pass through device (on releng_4). I would like to
get the serial number (also would like to monitor the drives,
but that's next, i have aaccli scripted into nagios for
drive failure).

Smartmontools doesn't seem to understand pass devices,
or at least, i don't know how to make it. I do have
the aacp configured, and it works somewhat.

The bug i am seeing with the serial number fetch via
camcontrol is:

(wheel)# camcontrol inq 0:0:0 -v
pass0: <IBM IC35L146UCDY10-0 S25S> Fixed unknown SCSI-3 device 
CAM status is 0x12

camcontrol: error getting transfer settings
CAM status is 0x6
(wheel)# 
(wheel)# 

shows the error, on a system with aac.

(wheel)# camcontrol devlist
<IBM IC35L146UCDY10-0 S25S> at scbus0 target 0 lun 0 (pass0)
<IBM IC35L146UCDY10-0 S27E> at scbus0 target 2 lun 0 (pass1)
<IBM IC35L146UCDY10-0 S27E> at scbus0 target 3 lun 0 (pass2)
<IBM IC35L146UCDY10-0 S27E> at scbus0 target 4 lun 0 (pass3)
<SUPER GEM359 REV001 1.09>  at scbus0 target 8 lun 0 (ses0,pass4)
<SUPER GEM359 REV001 1.09>  at scbus1 target 8 lun 0 (ses1,pass5)


>From the hitachi manual for the disk
(http://www.hitachigst.com/tech/techlib.nsf/techdocs/6BB69573F9D1537687256BD
600697AAA/$file/U146Z10_sp22H.pdf) it does support this
message, and it works with ahd (i think, i haven't 
recently checked).

--don