[Techtalk] /usr Partition

Akkana Peck akkana at shallowsky.com
Fri Jan 20 06:09:02 EST 2006


A typical partition scheme for me (obviously it depends on disk size):

* One /boot, small (maybe 64-128M).
* Several (three or four) / partitions, 4.5-10G each.
* A swap (I'm still confused as to how big to make swap).
* One or more (see question below) huge partitions for storing
  personal stuff like source code and photos.

The point of having a separate /boot and several / partitions
is that you can try out a new distro without affecting your
current one at all. Curious about gentoo or linspire? No problem,
just install to one of the spare partitions. And as long as /boot
is its own partition, grub can keep track of all your various
installs.

Currently I'm experimenting with making the "huge partition for storing
stuff" be /home, as Val suggested. In the distant past I've had bad
problems with that, because different distros would have
incompatible versions of software (say, gnome or kde) so every
time I rebooted into a different distro I'd get seven warning
dialogs telling me my settings had been migrated from an earlier
version, or my settings included things the current version didn't
understand. I got in the habit of keeping /home on the root
partition for each distro, and making symlinks from there to
the big partition where the images/music/source is stored.
But it looks like things have gotten better and sharing /home
works a bit better now.

I've never seen any point to making /usr a separate partition from /.
(In earlier Unix days there were reasons for it, but I'm not aware
of any good reasons that apply to a modern Linux system.)

If I have several physical disks to work with, then I'll sometimes
put / and /home on separate disks (and on different IDE channels)
so that if I'm compiling, seeks in the source tree (somewhere under
/home) and in system files (the compiler and /usr/include and
/usr/lib) are done at different times. I'm not convinced this
actually makes a measurable difference in compilation time, but
it makes me feel like I'm doing something to optimize the system.

Val Henson writes:
> Unfortunately, a lot of people are still partitioning up
> their modern 60GB super fast drives running ext3 like it was 1972, and
> all it does is make for headaches when one partition or another fills
> up.

Val, I've been wanting to ask you: I remember there used to be
arguments for not making filesystems over a certain maximum size,
so if you had a big disk, it was better to break it up into several
smaller filesystems.

In these days of 160G+ disks, is it kosher to have a single 100M
filesystem?  On ext2/3 it can be a major pain because fsck takes
forever, but what about reiser? With ext3 (which I know you like
better than reiser), how important is it to fsck often?

What about swap? What's the current guideline for how much swap you
should have?

	...Akkana


More information about the Techtalk mailing list