Re: Calculating tree nodes

On 9/4/07, Shawn O. Pearce [off-list ref] wrote:

Andreas Ericsson [off-list ref] wrote:

quoted

Jon Smirl wrote:

quoted

On 9/4/07, David Tweed [off-list ref] wrote:

quoted

On 9/4/07, Jon Smirl [off-list ref] wrote:

quoted

Git has picked up the hierarchical storage scheme since it was built
on a hierarchical file system.

...

quoted

quoted

quoted

One of the nice things about tree nodes is that for doing a diff
between versions you can, to overwhelming probability, decide
equality/inequality of two arbitrarily deep and complicated subtrees
by comparing 40 characters, regardless of how remote and convoluted
their common ancestry. With delta chains don't you end up having to
trace back to a common "entry" in the history? (Of course, I don't
know how packs affect this - presumably there's some delta chasing to
get to the bare objects as well.)

While it is a 40 character compare, how many disk accesses were needed
to get those two SHAs into memory?

One more than there would have been to read only the commit, and one more
per level of recursion, assuming you never ever pack your repository.

If you *do* pack it, the tree(s) needed to compare are likely already
inside the sliding packfile window. In that case, there are no extra
disk accesses.

Even better, lets do some back of the napkin math on the Linux
kernel tree.  My local (out of date but close enough) copy has
22,730 files in the tip revision.  Values shown are uncompressed
and compressed (gzip -9 | wc -c), but are excluding deltification.

                 Current Scheme       Jon's Flat Scheme
                 -----------------    -----------------
commit raw       932                  932 + 22,730*20 = 455,532
(compressed)     521                  456,338

root tree raw    876                  0
(compressed)     805                  0

This is not a fair comparison. The current scheme is effectively
diffed against the previous version. You aren't showing an equivalent
diff for the flat scheme. Both schemes are dealing with the same
22,000 SHAs.

The size win is from diffing, not compressing.

I'm not even bothering with the individual subtrees.  The numbers
will fall off quickly when you start to do subtree elimination and
only load the levels you need.

You are talking about doing disk IO for less than 4KiB with
the current scheme, and almost 456 KiB for the flat scheme.
That's before deltification.  So if you also assume deltification
its going to be higher as you need to read back to a base object
that is roughly the final size and then unpack the smaller deltas
to reach the real commit.

Remember, SHA-1s can be stored as 20 bytes of binary data but they
are also generally uncompressible.  That's why the root tree does
not compress very well, the SHA-1 data inside the tree cannot be
compressed and only the filenames have any shot at being compressed.

--
Shawn.

-- 
Jon Smirl
jonsmirl@gmail.com