Re: some linux tuning (fs benchmarks)

From: Doug Renner <dmr@dont-contact.us>
Date: Mon, 28 Jun 1999 13:29:52 -0600 (MDT)

> bonnie runs that fit entirely in memory are virtually meaningless as
> benchmarks.

I don't share that feeling. I'm sure they're useless when you are trying
to extrapolate specific data (I.E. like pure FS speed), but when all that
is important is how fast a given environment can cram stuff on the disk,
it's good for me. *I should really stop calling that I/O bandwidth when
it's obviously not. I don't know where I picked up that habit. :)*

I realize that there are caveats to that thought because it's almost
impossible to document every aspect of the environment. That's why I
qualified the trustfullness of the test in the last post.

In light of this particular thread, maybe one filesystem is slower than
another but one of the platforms handles the actual filesystem operations
better and they come out the same.

Overall, it's only cache speed that we care about, right? If I have
maximum_object_size set to 4K, that's going to be the majority of my I/O.
I will want to duplicate my environment as much as possible.

I just received Alex's reply and I think that he's on the money as far as
a better methodology. However, the real root of the thread was based on 1
vendors card being faster or slower than the "md" driver in Linux. From
what I understand, the bake-offs won't have those bounds. That's what I'm
really looking for: justification that I'm making the right platform
choices.
Received on Mon Jun 28 1999 - 13:28:33 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:47:02 MST