MultiDatabase shard count limitations

[Eric Wong]

Olly Betts <olly at survex.com> wrote:
> It would be better to just look up each term once.  I think to achieve
> that cleanly will need a bit of refactoring.

Understood, and no rush.  I'll probably go with a combined index
which should allow space savings for cross-posted messages,
anyways.

I'll be around to provide any more info.  My current test case
requires ~10GB of git clones and ~100GB after indexing, though :<

> But we've identified something to address so it's probably most sensible
> to address that and then try to reprofile.

Alright, thanks for all your help!