Re: [Gnumed-devel] get_encounter speedup

[Syan Tan]

I've updated my client code, hope it is fully upto date. the child_tables in gmPG makes sense, as this is a

general problem  . 

there seems to be some delays in

select * from v_pat_narrative, the same problem being sequential scan of child tables.

 

and a select with the condition "age(last_affirmed) < %s seconds" is a lot slower than "last_affirmed > now() - %s seconds "

because I think the later uses an existing index for last_affirmed, whereas age(last_affirmed) is  recalculated sequentially on

all rows for each call.

 

 

On Thu May 4 20:08 , Karsten Hilbert sent:

> Syan,
>
> I have eventually included your speedup code. It is,
> however, commented out by default. There's comments on how
> to activate it in get_encounters().
>
> The function does, however, also try another approach at
> speeding up the encounter access by approaching the problem
> somewhat differently.
>
> Karsten
> -- 
> GPG key ID E4071346 @ wwwkeys.pgp.net
> E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346
>
>
> _______________________________________________
> Gnumed-devel mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/gnumed-devel