[CentOS] RE: Setting up RAID using mdadm on a proliant DL320 G4
William L. Maltby
CentOS4Bill at triad.rr.com
Mon Sep 3 14:42:13 UTC 2007
On Sun, 2007-09-02 at 17:27 -0400, Ross S. W. Walker wrote:
> From: centos-bounces at centos.org
> [mailto:centos-bounces at centos.org] On Behalf Of William L. Maltby
> > On Wed, 2007-08-29 at 16:38 -0400, Ross S. W. Walker wrote:
> > > ><snip CentOS hdrs and some lead-in>
> > > 5) create 4GB LV called swap, formatted swap
> > OK. This particular item has aggravated me over many
> > different posts and
> > I've withstood the urge to holler "WHY"?
> > LVM adds another layer of (unnecessary) overhead. For swap to
> > be usable,
> > it must be formatted first, so it can't be expanded
> > on-the-fly with LVM
> > in use. Further, expansion on-the-fly can be (effectively)
> > accomplished by adding another partition or swap file (ugh! more
> > unnecessary overhead) as needed when LVM is not in use.
> LVM is hardly any overhead at all.
I knew it wasn't much, from using it on 4.X and 5.0 and observing system
responsiveness. Also, top seemed to confirm that overhead was not
Regardless, I operate on the premise that none is better than some
unless there are substantial offsetting benefits (see below where you
talk about the benefits of LVM).
> You could always stop swap, and add more, or like you said create
> another swap LV and add it to the mix and later, and here is the
> important part, you can REMOVE the swap partition if it proves
> unneeded, OR you can stop swapping on it, reduce it, mkswap it
> and re-add it.
Yes. In my particular situation, I seldom have to reconfigure. Perhaps
because I've been doing this so long and no longer have to worry about
servicing others. So for me, these benefits a minimal and infrequently
needed for swap. I have found the benefits of LVM (1 and 2) *very* handy
for other file system needs as I'm always playing with CentOS (and LFS)
to learn more. Naturally, "work expands to fill the space (and time)
allotted" and I've frequently made use of LVM reconfiguration
capabilities in addressing the effects of that. I especially liked the
snapshot feature as I protected myself from the possible (likely?)
brain-farts that can occur when I'm involved in throwing things around
> Better storage management, better hardware abstraction, and in some
> cases when used with software raid, better performance, yes better
> performance as it provides for some more intelligent re-ordering of
> io requests pre-scheduler.
I hadn't thought of that last item. It reminds me too that learning
about some software raid (striping and mirroring) might be in order.
Although my swap usage is very low (almost non-existent), having swap
and other file systems striped might provide noticeable performance
gains, in certain situations I encounter, and provide some more
education for me.
> In the future once grub can handle booting right into LVs there will
> be no more need for physical partitioning at all, create one big VG
> out of the system disk and allocate LVs.
Other than the boot drive, which on my systems tend to be older, slower
and smaller, I often have unpartitioned drives. I just make the file
system on the raw device. I've not tried making an LVM on an
unpartitioned drive yet, but I guess it ought to work. Another thing
I'll be trying.
> Partitioning sucks.
LOL. It seems that when a new needed facility comes that eases the tasks
of someone, the previous facilities they had to use "sucked". Examples
abound, like LILO vs. Grub. I still like and use LILO in some cases. I
still like and use partitioning in some cases. If I were still working
professionally and administering rapidly changing systems with large
dynamic user bases and variable needs, I would certainly feel
> <snip sig stuff>
Thanks for taking the time. I'll be playing with some of this on my
CentOS 5.0 and LFS system(s). The faster one with SATA and IDE ports
seems a good candidate, as it has nothing of importance on it.
Right now, I'm trying to figure out how to get decent graphic
performance on that node. I recently installed the nvidia driver (tried
to use the Rpmforge rpm, but the graphic card demanded the .96xx
driver). Regardless, my 4.5 AMD 2200XP unit with a Radeon gets appx. 670
FPS, while the AMD 3200XP with the nvidia (and matching driver from
nvidia's site) gets about 300 FPS. It's the same with the nv driver. I
cured this on the 4.5 by using frame buffer, dri and some other things
in the Xorg.conf. I've got to resolve this for the faster system.
Oh! Also, mplayer won't go full-screen with the nvidia driver, but will
with the nv driver.
Anyway, that's all OT, so I'll close now.
Again, thanks for the interest.
More information about the CentOS