[rhelv6-list] Align partition question on vmware

[Matthias Saou]

On Thu, 20 Mar 2014 13:59:18 -0300
francis picabia <fpicabia at gmail.com> wrote:

[...]
>    Device Boot      Start         End      Blocks   Id  System
> /dev/sda1   *        2048    24578047    12288000   83  Linux
[...]
> /dev/sdb1              63   427810949   213905443+  83  Linux
> 
> The guys running a system from there are claiming it is slower than
> the Sun 208R they were used to running, and it has no load on it.
> 
> Most articles I see about this are taking about the boot loading
> disk, while I'm looking at the issue for sdb.  Should I be concerned
> and looking at redoing the partitioning?

If you are seeing worse performance on sdb than on sda, and you
shouldn't (if they come from the same disk or SAN device), then you
should definitely consider re-doing that partition.

I think that if you run fdisk with the -c option (and the -u too), it
should then offer 2048 as the default start for a single partition,
just like what you've seen for the boot disk.

Another option would be to use parted, with an msdos partition table,
and create the partition starting at 1M and ending at 100%. If your
block device was larger than 2.4TB, you'd have to do it this way with a
guid partition table anyway.

This is one of those typical annoyances where anaconda does the right
thing by setting some options, while the tools' defaults don't. Another
typical example would be for ext filesystems and the acl and user_xattr
default mount options. Luckily that's trivial to change later on,
unlike partition boundaries!

Cheers,
Matthias

-- 
            Matthias Saou                  ██          ██
                                             ██      ██
Web: http://matthias.saou.eu/              ██████████████
Mail/XMPP:  matthias at saou.eu             ████  ██████  ████
                                       ██████████████████████
GPG: 4096R/E755CC63                    ██  ██████████████  ██
     8D91 7E2E F048 9C9C 46AF          ██  ██          ██  ██
     21A9 7A51 7B82 E755 CC63                ████  ████