[OSM-dev] Proposed database improvement

[Nick Hill]

Since the query update, It appears the database server is now receiving 
and processing around 3-4x the rate of queries (at least according to 
IOwait and Ethernet statistics), and is nowhere near capacity constrained.

I wonder how much of that extra load is productive?

In other words, perhaps the amount of load which can be generated can 
increase much faster than the useful work. Should we consider how to 
improve the ratio? For this, we'd need to define useful work, which 
could be politically and practically difficult.



Raphael Jacquot wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> SteveC wrote:
>> seems a bit faster to me! Thanks guys.
> 
> that's an understatement dude. this is like the difference between being
> ashamed of showing it, and happily being able to say the thing is under
> active development ;D