dump routine for mlx devices

[Andrew Kinney]

On 23 Apr 2003, at 8:43, Andrew Kinney wrote:

> That last one (ccd) could be a point of interest. Is it possible to
> create a ccd device from a spare mlxd device?  If so, is it also
> feasible to swap and dump to that ccd device or would it fail because
> the "real" device didn't support dumps?

Footnote: I decided to try it and here are the results, but you 
probably all knew this already.

dumping to dev #ccd/0, offset 9372672
dump failed, reason: device bad

Of course, this begs the question, why does the 
ccd driver code indicate it supports crash dumps?

Looks like I'm just going to have to break down 
and install an IDE disk to get those crash dumps 
I need, unless someone wants to address this 
part of my question:

> dumping to dev #mlxd/0x20001, offset 9372672
> dump failed, reason: device doesn't support a dump routine
>  
> Before I go try to write my own dump routine: Has anyone else 
> already invented that wheel or is the wheel too difficult a pursuit in
> this case due to peculiarities of this hardware?

I'd be willing to write the dump routine and pass it to the code 
maintainer for that driver, but I'd like to know if anyone has tried it 
before and run into a show-stopper.  

It would valuable for the FreeBSD community at large to be able to 
use a mlx RAID for a crash dump, though some of us value it more 
than others.  The obvious problem with using an IDE hard drive for 
crash dumps is that it represents a single point of failure that could 
bring the system down if the disk went bad while it was included in 
the swap space.

Sincerely,
Andrew Kinney
President and
Chief Technology Officer
Advantagecom Networks, Inc.
http://www.advantagecom.net