On Mon, 2005-10-24 at 02:06, Ryan Cresawn wrote:
> Hello,
> 
> I'm somewhat concerned with the current state of the RAIDframe
> documentation and I would like to get some answers from members of
> this mailing list so I can help to improve it.  Here is the URL:
> 
> http://www.netbsd.org/guide/en/chap-rf.html
> 
> I am particularly interested in this documentation because I use
> NetBSD/sparc64.  I use RAID-1 on the boot disks of three Suns for
> obvious reasons.  My concern is that some of the sparc64 specific
> parts of the RAIDframe documentation seem to be incorrect. 

Those examples are from the 2.x erra.  They were all thoroughly vetted
when I submitted them; but some changes have been committed since then. 
Updates rre welcome.

>  I'm hoping
> to get some firm answers to these questions so that I can recommend
> changes to the documentation with confidence.  Here are my specific
> concerns:
> 
> 1.  In my testing I have discovered that after I have created both
>     components of the mirror I have been unable to boot from the
>     original boot disk.  My solution was to create a small partition,
>     'e', which holds the boot block and ofwboot.  Has anyone
>     successfully booted an Ultra 1, an Ultra 5, or any other sparc64

Yea you want to avoid this.  If you get the offsets right, you can have
the boot blocks load right off of the RAID partition.

2.  The reason for choosing a particular dumpdev partition size is not
>     stated in the RAIDframe documentation and I believe it should be.
>     It is stated in swapctl(8); however, I believe that there are some
>     users who will not know how to properly select a size for this
>     partition and may select one which is too small.  I believe the


I left out the dumpdev stuff on my document because I thought it was
voodoo >:}  Someone went back and re-added it later.

My thoughts are that Software RAID 1 and kernel core dump devs are a bad
combination.

~BAS