<div dir="ltr">mark - good point about the controller cache and the 1-bye/block, but remember my files in this case are only 1 block. <div style>dave - a long time ago I found/reported a bug in the way linux was doing their block stats. They were basically incrementing the byte counts as block were written to cache and the reported numbers were ridiculously high. I had wondered if something like this could be going on, but also remember the numbers being reported by xfs are much too high. I did 10K 1K writes, which I do understand are really 1-4k page each, but that's still only 40MB. If I add up all the 500MB/sec bytes xfs logging is reporting (even one of which is over 10 times larger), I see something on the order of of 10GB. But again this IS with the older kernel and so may not be worth worrying about.</div>
<div style>-mark</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Jun 17, 2013 at 8:15 PM, Dave Chinner <span dir="ltr"><<a href="mailto:david@fromorbit.com" target="_blank">david@fromorbit.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On Mon, Jun 17, 2013 at 10:28:19PM +0200, Stefan Ring wrote:<br>
> > I asked around and people believe the cache is on the order of a few GB and<br>
> > the test ran I was intentionally large enough to overshadow any cache<br>
> > effects, running for about a minute and doing 100K 1K file creates. The<br>
> > disk write data was close to a sustained 475MB/sec and would have easily<br>
> > filled the cache in the first handful of seconds which would have produced<br>
> > enough backpressure to slow down the write rate which it never did.<br>
><br>
> If a limited number of blocks gets written over and over, you won't<br>
> ever fill the cache.<br>
<br>
</div></div>Right, and the XFS log is circular and about 1GB in size on the<br>
filesystem under test, so should fit completely in cache. Hence<br>
speeds faster than a physical disk are acheivable if the log stays<br>
resident in the cache....<br>
<div class="HOEnZb"><div class="h5"><br>
Cheers,<br>
<br>
Dave.<br>
--<br>
Dave Chinner<br>
<a href="mailto:david@fromorbit.com">david@fromorbit.com</a><br>
</div></div></blockquote></div><br></div>