probe dmesg status of RAIDs -- twe and amr

Danny Howard dannyman at
Mon Mar 14 11:16:11 PST 2005


I have a bunch of FreeBSD systems, many with hardware RAIDs.  The 
problem is, I have no easy way of checking to see that they are okay.  I 
tried once to get the 3ware binary to work with the 3ware kernel module 
to get to work with the 3ware firmware update that I had to ... ARGHH! 
It was awful.

On the other hand, in /var/run/dmesg :

twed0: <TwinStor, Normal> on twe0
twed0: 114472MB (234439600 sectors)
twe0: command interrupt

Or, on another system:

amrd0: <LSILogic MegaRAID logical drive> on amr0
amrd0: 140012MB (286744576 sectors) RAID 5 (optimal)

I know from experience that the "Normal" on the twed0 means "nothing 
broken" and I like to think the "optimal" on the amrd0 means "RAID5 is 
doing fine."  This is enough information for me . . . if a RAID is in a 
failure state, I can live with gracefully failing a server, if need be, 
and booting in to the BIOS to get the real information.

The problem is, I only get these messages when I boot . . . and I don't 
boot very often, of course.  I have tried poking around sysctl, 
atacontrol, camcontrol . . . all to see how I might probe out these 
messages on a regular basis, but I can not figure it out.  Can someone 
give me a tip on how to pull up bits of dmesg output while the system is 
running?  Or if someone has better advice on checking up on the status 
of various flavors of RAID . . .

Thanks in Advance,

More information about the freebsd-questions mailing list