:
:Matthew Dillon wrote:
:>     on the same directory over and over again.  Blogbench does its best
:>     to really mess up B-Trees :-).  I have narrowed the issue down to
:
:That's probably why they don't use B-trees for ReiserFS.
:
:What about crit-bit? Other trees?

    I am not too familiar with Reiser so I can't really come to that
    conclusion, but it doesn't seem likely that the reasons are similar.

    HAMMER's issue insofar as the B-Tree goes is mainly due to its history
    retention practices.  If I mount with -o nohistory then the issue
    becomes one of locality of reference due to HAMMER not immediately
    reusing space freed by the rename-over that blogbench does.

    After running the test over the weekend the culprit seems to pointing
    more towards HAMMER's low level storage allocation model and away
    from the B-Tree per-say.  I was already planning on making some major
    changes there so we'll see what pans out.

					-Matt
					Matthew Dillon 
					<dillon@backplane.com>