Re: RDB Problem?

[Ignatios Souvatzis <is%beverly.kleinbus.org@localhost>]

On Fri, Jun 16, 2000 at 04:38:19PM +0100, Lars Hecking wrote:
> David Brownlee writes:
> >     Is the bogus warning always with nheads=1 and nsectors off by 1?
>  
>  1) Nope. 2) Yes. This accounts for one reserved block per cylinder,
>  or so I undertand.

Well, for the nsparesperwhatever. Which is not always correct. And as I
understand, there used to be some very strange RDB preparation software that
created bogus geometry in the RDB.

>  For my SUN1.05 1G drives, I'm getting
> 
>     warning: found rdb->secpercyl(756) != rdb->nsectors(85) * rdb->nheads(9)

(85-1)*9 == 756

> > > > warning found rdb->secpercyl(1339) != rdb->nsectors(1340) * 
> > > > rdb->nheads(1)

(1340 - 1) * 1 = 1339

> > > > warning found rdb->secpercyl(4127) != rdb->nsectors(4128) * 
> > > > rdb->nheads(1)

(4128 - 1) * 1 = 4128

Btw: It's a warning only. A _warning_.

It's only activated because of "options DIAGNOSTIC" in the GENERIC kernel.

If the installation (onto the same disk) worked, you can savely ignore it.
In case something strange is happening with respect to disk operations, and
we see this warning, we can investigate whether a bogus RDB was the cause of
it.

        -is