[OSM-dev] mysql partitioning

[Simon Hewison]

Nick Hill wrote:
> However, most of the database time seems to be spent pulling segments 
> from the database. For every node in the selected area, an index look-up 
> is performed for segments associated with that node. Then for segments 
> which have only one node associated with them, the other node is 
> downloaded. This is very, very db intensive.

Why do we even have such segments with only one node associated with them in 
the database? The only time I've ever seen them is where somehow there's two 
nodes bang on top of each other, and an editor has got a little bit confused 
as to what the user was pointing at.

Another rather nasty thing I've spotted is that when the new renderer is 
working out what tags are on ways attached a list of segments. It seems to 
be doing lots of individual calls for each row in get_multis_from_segments()
, because it's calling out to get_multi() for each of those.

I can see that those multiple SQL transactions could be reduced, with all 
the overhead of parsing SQL queries, by writing it as one SQL query.

Admittably, it looks like get_multi() was written with area support in mind, 
  which might complicate it a bit.

-- 
Simon Hewison