[Tilesathome] t at h server performance

[Martijn van Oosterhout]

On Dec 22, 2007 3:43 PM, Christopher Schmidt <crschmidt at metacarta.com> wrote:
> In the long term, we need to come up with some more efficient way of
> storing blank tiles than the current database.

It should be said that we arn't using the current format as
efficiently as possible. For example, we could make the clients
produce less blank tiles, and we could get the server to store them
more efficiently, we're not taking full advantage of the fallback
system. Some random ideas:

- Clients create a blank sea/land tile whenever they find something
blank. The client could immediately check to see if the tile above it
is also blank and if so store a 0-byte file instead. That's one less
entry in the DB to clean up later.
- When the server has the list of blank tiles to insert, it could run
one pass over the list to see if it's mostly sea of mostly land and
whatever it is, insert a blank tile with that type at zoom-12 and then
only insert the other types.

The latter idea is I think where we could make some real gains bt it's
currently impossible to do on the client side because client only have
the option of uploading a blank tile *or* a real tile at zoom-12, not
both.

> For the record, I ran the cleanblanktiles script yesterday, so the blank
> tile db is as clean as that code makes it. I don't know if there are
> further optimizations that can be made there -- perhaps someone else can
> comment -- but I've done everything I know how to do.

I'd never looked at the cleanblanktiles before but it certainly looks
like its missing some possible optimisations. Especially my second
suggestion above would remove many z17 blank tiles, which are half the
current DB.

Have a nice day,

-- 
Martijn van Oosterhout <kleptog at gmail.com> http://svana.org/kleptog/