View Single Post
Old 14th January 2009
jggimi's Avatar
jggimi jggimi is offline
More noise than signal
 
Join Date: May 2008
Location: USA
Posts: 7,977
Default

I don't see much gain -- security or otherwise -- by having multiple MBR partitions for a BSD, as opposed to some other systems, such as Linux or Windows. That's because BSD manages disk partitioning through a single disklabel anyway. (As for performance, an I/O is an I/O: it points to a sector or a set of sectors on a drive by LBA number.)

Partitioning (either MBR or disklabel) is a logical way of separating contiguous chunks of sectors. The only benefit I can see to multiple MBR partitions for a BSD would be to help you to manage two discontiguous chunks of disk space that are used by the same BSD. However, you could still address discontigous chunks via the disklabel anyway, and ignore MBR tables.

Not all BSDs allow multiple MBR partitions -- OpenBSD, for example, allows only one A6 (OBSD) MBR partition at a time. If you need to manage discontigous areas, you either reconfigure the drive so the data resides in a single MBR partition, or you manage them via disklabel(8). When I first began to build test OpenBSD systems on a single workstation, I didn't have much experience with OpenBSD, so I managed multiple systems with fdisk. Within a very short period of time, I switched to managing things with disklabel, where partitioning was considerably more flexible. I still use disklabel today, when I need to multiboot different OBSDs.

MBR partitioning is still important: for the boot sequence, or for mounting foreign file systems.
Reply With Quote