Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.sys.ibm.pc.hardware.storage > #6424

Re: Cannot access new Seagate external HD

Cross-posted to 2 groups.

Show all headers | View raw

In article <d5icghF150uU1@mid.individual.net>, VanguardLH wrote:
> Wouldn't changing the sector size give you a larger partition size?

Yes, absolutely it would. 

The trouble is that you then need to use an OS and a BIOS that 
understands sectors of the new size. Windows [up to Windows 7 -- see 
below] categorically does not do this, and nor do most PC BIOSes (EFI 
firmware may, but if you have EFI you have less need to be able to do 
so).

> As I recall, advanced format HDDs were first supported in Windows
> Vista. Alas, the OP has Windows XP.
> 
> https://en.wikipedia.org/wiki/Advanced_Format

Advanced Format uses 4k physical sectors, but then (by default) presents 
these over the ATA interface as 512-byte blocks. The disk controller 
performs a mapping internally between the actual 4k-bye sectors and the 
pretend 512-byte sectors.

It's possible to disable the mapping, but as most software expects 
512-byte sectors it's on by default.

The Wikipedia article you cite calls the mapping scheme "512e" and the 
native unmapped addressing of actual 4k sectors "4Kn" (4kB native).

So, when it says:
> For example, Windows Vista, Windows 7, Windows Server 2008, and
> Windows Server 2008 R2 (with certain hotfixes installed) support
> 512e format drives (but not 4Kn) ...

It's saying that Vista /et al/ support AF disks but only in the mode in 
which they pretend to have 512-byte sectors. The extend to which they 
are "compatible" with AF is that their partitioning tools align the 
filesystem efficiently on the disk. 

However, it goes on to say:
> Windows 8 and Windows Server 2012 also support 4Kn Advanced Format.

Which I didn't know. Thanks for (indirectly) drawing that to my 
attention.

> Would Windows XP work with a 512e advanced format HDD?  The external
> interface would still be addressing via 512-byte sectors but the
> drive itself would be using 4096-byte sectors.  What would be the
> point of using 4096-byte physical sectors on the platter if only 512
> bytes of each larger sector was all the drive would deliver?  The
> physical sector size would be larger but you can only get 512 bytes
> out of each sector?
> I suspect this requires usurping the MBR bootstrap code with a
> translation utility.

The translation is actually performed by the HDD controller, so it 
should all just work. Note, though, that the translation is only 
efficient if the filesystem is arranged so that the 4k sectors on the 
disk correspond to the 4k allocation units of the filesystem, so you 
have to partition the disk carefully (i.e. not use XP's FDISK) to make 
it work well.

This article (linked from the Wikipedia article you cited)
https://lwn.net/Articles/377895/

is about Linux, but explicitly discusses some problems with XP. It's an 
interesting and informative read.

The POINT, though, is that using 4k physical sectors allows the disk 
manufacturers to achieve higher data densities on a track (because there 
is wasted space and ECC data stored between the sectors on a track, and 
you waste less space with a few large sectors than you do with a lot of 
small ones) so AF makes higher capacities per platter possible, and so 
makes disks both bigger and cheaper.

Your theorizing about dynamic volumes is interesting ... in theory I 
guess you could write some software layer that used a single very big 
HDD as storage (with either 512-byte or 4k sectors -- it doesn't matter 
because you're not going to show it to the OS) and presented it to the 
OS as though it were a RAID5 volume (or something). I don't know whether 
you could do that on Windows and make it work ... but I expect you could 
with enough effort (there remains a question over whether that much 
effort is worth it, when there are other ways to use big disks).

> Luckily the OP bought a drive with a capacity that exceeds the 2 TB
> partition size using 512-byte sectors in both the file system cluster
> size and cluster addressing and the 2^32 addressing limit in the MBR
> partition record.

ITYM: ...capacity that *does not exceed* the 2TB partition size using 
512-byte sectors ...

It's a 2TB drive, they're typically around 2,000,000,000 bytes (2 
"marketing" TB) whereas the partition size limit is 2,147,483,648, so 
they work fine with MBR partitioning (the one I'm using now -- a Hitachi 
HDS5C3020ALA632 -- is 2000398934016 bytes; Linux's fdisk describes it as 
1.8 TiB).

-- 
Cheers,
 Daniel.
 

Back to comp.sys.ibm.pc.hardware.storage | Previous | Next — Previous in thread | Next in thread | Find similar

Thread

Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-10 19:20 +0100
  Re: Cannot access new Seagate external HD Chronos <usenet@chronos.org.uk> - 2015-09-10 19:50 +0100
    Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-10 20:11 +0100
      Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-10 20:46 +0100
        Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-10 21:07 +0100
          Re: Cannot access new Seagate external HD Chronos <usenet@chronos.org.uk> - 2015-09-10 22:10 +0100
          Re: Cannot access new Seagate external HD Mike Tomlinson <mike@jasper.org.uk> - 2015-09-11 00:22 +0100
      Re: Cannot access new Seagate external HD Mike Tomlinson <mike@jasper.org.uk> - 2015-09-11 00:16 +0100
        Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-11 07:16 +0100
          Re: Cannot access new Seagate external HD Daniel James <daniel@me.invalid> - 2015-09-11 15:17 +0100
          Re: Cannot access new Seagate external HD Shadow <Sh@dow.br> - 2015-09-11 11:30 -0300
          Re: Cannot access new Seagate external HD Mark Perkins <mark@none.invalid> - 2015-09-11 09:35 -0500
            Re: Cannot access new Seagate external HD Mike Tomlinson <mike@jasper.org.uk> - 2015-09-11 21:22 +0100
  Re: Cannot access new Seagate external HD Shadow <Sh@dow.br> - 2015-09-10 16:18 -0300
    Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-10 21:08 +0100
  Re: Cannot access new Seagate external HD Ed Light <nobody@nobody.there> - 2015-09-10 14:30 -0700
  Re: Cannot access new Seagate external HD micky <NONONOmisc07@bigfoot.com> - 2015-09-10 21:17 -0400
    Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-11 07:23 +0100
  Re: Cannot access new Seagate external HD VanguardLH <V@nguard.LH> - 2015-09-10 21:11 -0500
    Re: Cannot access new Seagate external HD Daniel James <daniel@me.invalid> - 2015-09-11 15:17 +0100
      Re: Cannot access new Seagate external HD VanguardLH <V@nguard.LH> - 2015-09-12 05:18 -0500
        Re: Cannot access new Seagate external HD Daniel James <daniel@me.invalid> - 2015-09-12 20:46 +0100
          Re: Cannot access new Seagate external HD VanguardLH <V@nguard.LH> - 2015-09-14 04:14 -0500
            Re: Cannot access new Seagate external HD Daniel James <daniel@me.invalid> - 2015-09-14 13:03 +0100
            Re: Cannot access new Seagate external HD Mark Perkins <mark@none.invalid> - 2015-09-14 21:10 -0500
  Re: Cannot access new Seagate external HD Daniel James <daniel@me.invalid> - 2015-09-11 15:17 +0100
    Re: Cannot access new Seagate external HD Terry Pinnell <me@somewhere.invalid> - 2015-09-12 07:05 +0100

csiph-web