Large clone chewing CPU

[Bryan O'Sullivan]

On Wed, 2005-08-31 at 07:50 -0700, Eric M. Hopper wrote:

> I'm going to be doing some investigation.  This has been going for 16
> hours or so now.

Whoa!  That's way outside the normal range.  What's the load on the
server look like when this is going on?

>   There seem to be long periods of CPU chewing (lots of
> that spent in the kernel because of the mmaps) with short periods of
> time where actual data is transmitted.  I may be wrong about this, but
> my guess is that the cycles correspond to changesets.

If you could send some strace output for the contents of one cycle, that
would be interesting.

> Oh, the mmaps appear to be mapping MAP_ANONYMOUS pages...

Hmm, then it's probably allocating memory.

> And this one is all
> proprietary code and jars and gifs and stuff.

Can you at least characterise its contents?  Are the jars compressed?
How many revisions are in the repo?  What's the distribution of file
sizes?  How many of the binary files have multiple revisions?  What are
the sizes of the .d and .i files under .hg?

I've seen problems with performing hg operations over slow connections,
but nothing anywhere near this bad.

	<b