ivan marchesini wrote:
> Concerning the problem of the same "running time" I'm really surprised
> but I can confirm that the EXPLAIN ANALIZE time is different from the
> query execution time.
> probably it happen something like this:
> http://bytes.com/topic/postgresql/answers/422841-explain-analyze-total-runtime-walltime
> But we have tried to remove the fk constraints without any improvement
>
> Concerning the client problem, e have tried to execute the query
> directly on the server using psql.
> The difference is still in the order of seconds: around 10 ms VS 5000 ms
>
> many thanks...
>
> other suggestions?
As I said -- could the time be spent on the client?
Yours,
Laurenz Albe