OT: redhat

[Bill Vermillion]

On Sat, Nov 13 15:01 , Bill Campbell, showing utter disregard for 
spell-checkers gave us this: 

[much deleted - wjv]

> On Sat, Nov 13, 2004, Bill Vermillion wrote:
> >Shakespeare wrote plays and sonnets that will last an eternity, 
> >but on Sat, Nov 13 14:32 , Bill Campbell wrote:" 

> >> On Sat, Nov 13, 2004, Jay R. Ashworth wrote:

> >> On Linux systems, you have programs that will resize existing
> >> partitions, /sbin/resize2fs, and /sbin/resize_reiserfs on a
> >> SuSE 9.0 Professional box here which handle ext2/3 and reiser
> >> file systems. FreeBSD has a program, growfs, to handle the job.

> >I'm not familiar with the SuSE offerings but growfs will only
> >let you grow a file system.

> >The big confusion in the BSD world vs the rest of the world
> >is the definition of partitions.  In BSD those are slices.
> >And if you want to use growfs beyond the boundary of a slice
> >[called a partition in the rest of the world] you have to
> >resize the slice with fdisk, and then once slice is larger
> >you can enlarge the system with growfs.

> In that respect FreeBSD is similar to SCO Xenix and OpenServer
> which uses divvy to break a ``partition'' into logical file
> systems.

It's just that the term partition is used differently.  
But even Linux is similar to FreeBSD.  hdaN - with N being filesystems
The FreeBSD differentiated the HDs so that the typical IDE
will be adMsNX with M being the the letter following hd in Linux,
N being the slice number, and X a letter designating up to
8 filesystem designation.  If the drives were SCSI those
would start with sa and not ad.  FreeBSD just tends to be
more detailed in indentity.

> >If you look back at the attributuions I mentioned AIX was the only
> >one I knew of that could expand dynamically. growfs requires an
> >unmounted fs, and while not an AIX expert [or as that AIXpert?]
> >I understood that it's facility would expand on a running system.

> I think that AIX is based on the Veritas file system which has
> many neat features including ``freezing'' a partion for backup
> while journalling changes which can then be used to update the
> frozen system. Veritas has been ported to systems other than
> AIX.

Was Veritas around when AIX came out?  I remember when IBM
introduced the RS6000 and AIX.  They were proud of SMIT [for
administration] and the ability to grow filesystem on the fly.
Things that are much more imortant now than they were 10 years ago.

> >Then Jay mentioned SuSE 9 XFS.

> XFS was originally written by SGI and released to the Linux
> world as open source.

I've used it on the SGIs - on the MIPS platforms - and the machines
I used were basically a 40x64 architecture.  40 bits for HW memory
addressing but 64 bits on the SW side.

> >Then Mark [aka Fairlite] mentioned joining two physically disparate
> >partitions to be able to make a larger FS.   I can see that will
> >be problematic at least in the Intel based world as everything 
> >expects 4 physical partitions, one of which will be active.

> There can only be four ``primary'' partitions using PC standard
> partitioning or three primary and an ``extended'' partition which can
> contain multiple logical partitions.

And it depends upon the OS running on top of the drive if you can
use logical partitions.  Given the BSD heritage going back to
the late 1970s - it does't like logical partitions.  But the
logical partitions were the MS invention to handle drives
that were bigger than fileystmes that the default OS could handle.

Remember the FAT12.  And remember when you could only have
about 512 files total.  [I think it was 512 - it may have been
slightly different].  THen came 16 bit fat and of course the 2GB
limits were there.  So you made three 2GB partitions, and then
took whatever was left and made as many logical partitions to use
up the drive space.

So many of those limits are gone now.  Remember when the OS for
booting had to like within the first 1024 cylinders.  When you look
at the current FreeBSD booting limits you wonder how we survived so
long with the old limits.  In the 5.x and 6.x FreeBSD the booting
code has to lie withing the first 1.5 Terabytes of the first
filesystem.  The first filesystem does not have to be limited to
that amount, just the the boot code must like in the first 1.5TB of
it.

> The more advanced systems such as vinum and LVM can span
> partitions and/or hard drives so there's no physical
> limitation.

But there is the possible problem of losing a drive in a spanned
system.  But with the 400GB drives now available that should only
be a problem on really large systems, and then you will most likely
use a RAID system as a spanned FS with no RAID recovery could be
a long time to restore - even with the fast systems.

> >Each partition must occupy contiguous disk space.  So to
> >use space on more than one partition [BSD slice] that are not
> >physically adjacent the fileysystem creation programs would handle
> >that at the logical level.  I don't see the need for having to
> >physically join partitions to make a file system if you can
> >logically create a filesystem spread over more than one
> >physical partition.

> The only time I've used LVN on Linux to join disk partitions
> was to join two 2gb partitions into a 4gb logical partition
> where I didn't want to change the partitions with fdisk for
> fear of breaking things that knew about the partition numbers.

That makes sense.  Does Linux still change the numbers if you
remove a drive in the middle - say you had 4 drives and removed one
- do the drives above the middle one get renumbered.  That used to
be a problem but I'm not that deep into Linux at the moment.

Bill
-- 
Bill Vermillion - bv @ wjv . com