[Gllug] partition table screwup

[Martin A. Brooks]

Henry Gilbert wrote:
> which led me going to the other extreme
> having just one mount point / and swap

That's actually quite normal in production environments where the loss 
of a single machine is irrelavent.

For example, when I worked for Fast Search & Transfer, each search 
cluster was setup in such away that more than 50% of the machines would 
have to fail before search performance would be affected.

When a machine did fail, it was a case of swapping out whichever 
component had failed and powering it back on.  The box would boot, 
automatically reinstall it's operating system and software, and just 
carry on.
-- 
Gllug mailing list  -  Gllug at gllug.org.uk
http://lists.gllug.org.uk/mailman/listinfo/gllug