Re: Ada and PL/pgSQL ?

["björn lundin" <b.f.lundin@gmail.com>]

Den fredagen den 9:e maj 2014 kl. 21:59:30 UTC+2 skrev sdale...@gmail.com:
> 
>> I think you are bitten by mysql performance.
> 
> No there are similar results when exporting the schema from MySQL to >Postgres.  The fact is there is only so much you can logically do in a single >query.  You  can index and join and filter and sort etc.  But when push comes >to shove you need to implement business logic somewhere.

Yes. And for me, the correct place is in a separate layer - in Ada.
That also makes things a lot easier if you want to change db-engine again.
We are into the 5th db since the late 80's
(Mimer, Oracle RDB, Oracle, Sql-server, Postgresql)
 
> The tables are simple and not at all complex.  There are two tables involved > a transactions table and an order table.

Given that and 
> The minimum execution time for an order was 30s.
> However it was climbing as high as 5mins. 

makes me wonder about the number of records you have in the db.
With these timings, it must be in the tens or hundreds of million records?

explain analyze of the queries after a vacuumdb --analyze would be interesting to see.
But that is for Postgres performance mail list, not for this one.

Executing logic will take place in a postgres process.
Having your logic i c or ada executing locally on that server -
does that make a difference in the figures?
 
> The transactions table reflect credits/debits against accounts, whereas the orders table represents a literal "filled out order form".
> 
> It can't really be simplified much more than that.

No. And that is why I am curios about the amount of data.
The timings just seems very strange.

/Björn