[MERGE] Updates to the "auto-buffer" logic of GraphIndex

[John Arbash Meinel]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

...

> 
> I think one other major factor is the size of the preferred read versus the
> average size of the index. And the fact that when padding the readv requests,
> we actually add bytes to the end, rather than "centering" the request. Because
> we are bisecting, this causes a bit of asymmetry in the reads. Anyway, I'd
> certainly like to be done with this portion, and focus back on btree indexes
> as the way forward.
> 
> I'll merge this refined with 50% auto-buffer, and we can tweak it in the future.
> John
> =:->

I would like to mention that having the recommended size be hidden under
Transport is a negative. Specifically, it would be nice if we could see ahead
of time "oh, we are going to read >50% in the first attempt, switch to a GET".
Rather than thinking we were only going to request 800 bytes, and have that
turned into a 64kB request that consumes most of the index.

However, I think we can approach it that way for BTree. Because BTree has nice
*aligned* pages, it is much easier to avoid re-sampling the same location
multiple times. And certainly I think btree is the one that should know which
pages it already has buffered, and not request them again.

John
=:->
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFIvYOvJdeBCYSNAAMRAkjcAKCA5BDywhwPdQSaQRQD6IPhACACfwCggj/R
cXWZ7wirhubQ31aNAnfpTAE=
=25b6
-----END PGP SIGNATURE-----