Re: RAID 5 7+1 with oracle

I'm not convinced the two graphs on page 10 are the results of empirical data. Notice how the plot lines are *exactly* the same, only with different labeling on the X axis. This can be seen best on the "10krpm write" lines in pink, near the 200 and 400 IOPS points in the left and right graphs, respectively.

The two graphs on page 9, however, appear to be slightly different, which one would expect from two different trials.

Perhaps just a typeset error, but that makes the entire paper suspect to me, or am I not seeing that correctly?

And I'm not just being bitter about the RAID5 setup here on our SAN, either... ;)

Rich

> I'm not convinced your interpretation of the graphs is correct.
>

--
http://www.freelists.org/webpage/oracle-l