[ILUG] dying hard disc - upgrade questions?

[Brian Foster]

      profuse apologies if this is a repeat post, but I've
      been battling the postfix(1) installation and know a
      bunch of e-mails bounced for assorted/silly reasons
      (the latest being that `/etc/resolv.conf' was not
      readable by the `postfix' user --- not that `postfix'
      ever actually said anything like that (or gave a hint),
      I had to use strace(1) to figure it out ....  ;-\  ).

------- Forwarded Message

Date: Wed, 26 May 2004 02:34:25 +0200
Subject: Re: [ILUG] dying hard disc - upgrade questions?
To: Irish Linux Users Group <ilug at linux.ie>
From: Brian Foster <ilug at blf.utvinternet.ie>
In-Reply-To: Message from Paul Jakma <paul at clubi.ie>
   of "Tue, 25 May 2004 21:30:12 BST." <Pine.LNX.4.58.0405252121430.2606 at fogarty.jakma.org>

  | Date: Tue, 25 May 2004 21:30:12 +0100 (IST)
  | From: Paul Jakma <paul at clubi.ie>
  | 
  | On Tue, 25 May 2004, Brian Foster wrote:
  | >   + extended partitions should only be be the last in each table
  | >     (and indeed, in the last (4th) slot of the MBRs table);
  | 
  | Hmm, really? I've often created extend partitions in between primary
  | partitions.

 YES.  remember, I was asking about “best practice‟
 for a collection of s/w (not all Linux); _not_ what
 was possible.  and above I was talking about order
 in the MBR's partition table, _not_ the physical
 arrangement of the partitions on the media.  e.g.,
 in the setup I now have, it is physically arranged
 as:  hdb1 (primary)  ≡ 0   ..N₁   (cylinders),
      hdb2 (primary)  ≡ N₁+1..N₂   (cylinders),
      hdb3 (primary)  ≡ N₃+1..End  (cylinders),
 and  hdb4 (extended) ≡ N₂+1..N₃   (cylinders);
 or graphically:

    ┌──────┬────────┬──────────────┬────────┐
    │ hdb1 │  hdb2  │     hdb4     │  hdb3  │
    ├──────┼────────┼──────────────┼────────┤
    0     N₁       N₂             N₃       End

 the extended hdb4 physically occurs before the
 primary hdb3, opposite of the order in the table.

 also, whilst I didn't try too hard, the SUSE 9.1
 installer would not seem want to let me even create
 a primary after the extended in the MBR's table.

  | >   + there is an indeterminate number of logical partitions per
  | >     extended;
  | 
  | There may be limits in how many an OS supports though, cant remember
  | if Linux has such a limit. It definitely had a limit in how many
  | partitions can be on a disk (least for scsi disks - was limited to 15
  | - probably still is).

 *sigh*  yep.  problem is, there seems to be no
 agreement on what these limits are (even allowing
 for changes over time and different s/w (O/Ses),
 this is amazing!); and quite a number of docs don't
 even discuss the matter at all.

 I was maybe unclear here:  when I said “indeterminate‟
 I meant I could not determine what a “safe‟ maximum
 was; where again, “safe‟ means “works with‟ arbitrary
 s/w (not just Linux).

cheers!
	-blf-

------- End of Forwarded Message

-- 
«How many surrealists does it take to    |  Brian Foster      Montpellier,
 change a lightbulb?  Three.  One calms  |  blf at utvinternet.ie      FRANCE
 the warthog, and two fill the bathtub   |    Stop E$$o (ExxonMobile)!
 with brightly-colored machine tools.»   |        http://www.stopesso.com