Subject: Re: type-length-value chains for FFS (was large Inodes)
To: John F. Woods <>
From: Bill Studenmund <>
List: tech-kern
Date: 03/24/1999 09:40:04
On Wed, 24 Mar 1999, John F. Woods wrote:

> > Yes, I mis-remembered. There is a flag u_int32_t, and 28 BYTES of data, or
> > 7 u_int32_t's. Added to the 2 spare u_int32_t's we have now, I think we
> > have enough space. We'd need 3 for Y2038, and one for acl's.
> OK, so whoever decides to lay claim to the 5th remaining u_int32_t gets to
> design the NetBSD Resource Manager.  Sounds fair to me. :-)

They should be used one at a time. :-)

> > Growing the fs to support more than 2^32 blocks (files > 2 Peta bytes or
> Isn't it 2 terabytes?  That's actually an achievable storage size (for a
> RAID system, anyway), if you've got the money.

Doh! I mis-read mc's output. :-( You're right that that size is achievable
now. ;-)

> If you want to be really picky, you only need 3 extra bytes per block pointer,
> since the file sizes are still constrained to a mere 64 bits. ;-)

I don't really want to go to that level of packing. :-(

Take care,