On Thu, Sep 20, 2001 at 11:10:02AM +0200, Herbert Liechti wrote:
> I tried it. See my actions below. The main performance boost is
> reached by creating an index and disabling the sequential scan:
> Without any index;
> real0m18.128s
> user0m0.010s
> sys 0m0.010s
>
> Same statement wi
Hi Andreas,
I'm running PostgreSQL 7.1.3 here on a PC with nearly a gig of ram, and
running Linux Mandrake 8.0
First thing I did was to increase the amount of shared memory and stuff
which Linux allows things to use :
echo "kernel.shmall = 134217728" >> /etc/sysctl.conf
echo "kernel.shmmax = 13
On Thu, 20 Sep 2001, Einar Karttunen asked me for query plans for
both M$ SQL and postgresql:
M$ SQL:
|--Compute Scalar(DEFINE:([Expr1002]=Convert([Expr1005])))
|--Stream Aggregate(GROUP BY:([Hauptdaten_Fall].[MeldeKategorie])
DEFINE:([Expr1005]=Count(*)))
|--Index
Scan(OBJE
Hi Andreas,
Sorry, I haven't seen the history of this thread. One question which
might be relevant is, have you adjusted the postgresql.conf file from
the default memory settings to be something better?
If these are the times you're getting from a default configuration, you
might be able to get
On Thu, 20 Sep 2001, Justin Clift wrote:
> Sorry, I haven't seen the history of this thread. One question which
> might be relevant is, have you adjusted the postgresql.conf file from
> the default memory settings to be something better?
I adjusted two parameters:
shared_buffers = 2048
(When I
On Thu, 20 Sep 2001, Herbert Liechti wrote:
> I tried it. See my actions below. The main performance boost is
> reached by creating an index and disabling the sequential scan:
Thanks. I tried this and it helps in dead (see below).
> ---
> crea
On Wed, 19 Sep 2001, Tom Lane wrote:
> No. In the first place, there's no extra sort: the planner is well
> aware that our current GROUP BY implementation produces ordered output.
> In the second place, there's no guarantee that GROUP BY will always
> produce ordered output in the future --- we
Tille, Andreas writes:
> SELECT Hauptdaten_Fall.MeldeKategorie, Count(Hauptdaten_Fall.ID) AS Anz
> FROM Hauptdaten_Fall WHERE (((Hauptdaten_Fall.IstAktuell)=20)) GROUP BY
> Hauptdaten_Fall.MeldeKategorie ORDER BY Hauptdaten_Fall.MeldeKategorie;
> (which should just measure the time needed for th
