Subject: Re: dinode: di_spare[2]
To: Darren Reed <darrenr@cyber.com.au>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
List: tech-kern
Date: 12/01/1997 16:36:22
On Tue, 2 Dec 1997 09:29:09 +1100 (EST),
   Darren Reed <darrenr@cyber.com.au> writes:

>> >>>>> "JS" == Jonathan Stone <jonathan@DSG.Stanford.EDU> writes:
>> 
>> JS> Matthew Jacob <mjacob@ns.feral.com> writes:
>> 
>> >> This is a pretty interesting idea. I'd like to point out that there's
>> >> a performance penalty in this (if I understand what you're proposing)
>> >> while the implied 'spare' usage allows NAS (amongst others) to directly
>> >> use (locally) the spare fields for HSM.
>
>What is NAS and HSM ?

NAS=Numerical Aerospace Simulation (or something close)

A NASA group at NASA Ames, (at Moffet Field, just down the peninsula)
doing aerodynamic simulation, virtual wind-tunnels, and other neat and
data/CPU-intensive stuff.  Several NetBSD developers and a core group
member are employed there. I'm sure they'll correct the acronym.

HSM==Hierarchical Storage Management

Mainframe-style on-demand automigration of data between disks and
less-accesible media, like tape robots. I think NAS still use this;
they used to have a custom-built tape robot in the 80s, when their
online storage demands exceeded the capacit of COTS solutions.


>Do you (or anyone else ?) have any more detailed information on what AFS
>does here ?  I do know that SunOS 4 uses the last 16 bytes in their
>"dinode" structure for storing values used in the kernel only:

[snip]

It's not just in-kernel; AFS uses some of these fields these show up
on-disk as well, at least in AFS volumes on the server. (newer
versions may do it differently).  jtk points out there was no AFS
server-side release for NetBSD; i don't know if anyone ever built one
`unofficially'.

Maybe breaking AFS server compatibility is the right thing, but we
should be aware we're doing that, and/or make it an option.

I also don't know if these bits are ever used on-disk for the client
cache; I'm not sure, and jtk suggests not.


>but that shouldn't be a problem - it's an earlier verison of `UFS' (0),
>which already has certain differences to UFS versions 1 and 2.

Don't take this as gospel, but I believe UFS `version 0' was the
4.2BSD version which allowed more than 8 cgrotor positions; SunOS uses
UFS version 1, with a fixed-size rotor and more `spare' space; version
2 adds 32-bit uid/gids, and NetBSD prefers version 3, with in-inode
symlinks.

Anyway, NetBSD can mount SunOS partitions read-write; and some people
do use that for multi-boot installations on Suns.  It's ufs 0 we don't
support (or rather, only allow readonly mounts).