In my Pentium 4 box 2GHz with 256 MB of RAM I run a pg server 8.0.3
under netbsd 2.0.2 in which among other tables there's a "huge" table
letture02 made of 657,000 records and 98 numerical columns.
Using psql
and the "heavy" view letture24btnondom (see below) I face the following
fatal error:
..........
# select * from letture24btnondom ;
out of
memory for query result
..........
Now, it happens that I pg_dumped
the database from an original ** freebsd ** installation on another
older pc (pentium 3 128 MB) where ** under psql I set up the same query
and ran it successfully many times.
Besides, via ODBC I can open as a
table the letture24btnondom view without any problems.
What should I
do for psql?
Ciao
Vittorio
letture24btnondom view:
SELECT letture02.
contatore, letture02.data, (letture02.let1 + letture02.let2 + letture02.
let3 + letture02.let4) / 4::numeric AS ore1, (letture02.let5 +
letture02.let6 + letture02.let7 + letture02.let8) / 4::numeric AS ore2,
(letture02.let9 + letture02.let10 + letture02.let11 + letture02.let12)
/ 4::numeric AS ore3, (letture02.let13 + letture02.let14 + letture02.
let15 + letture02.let16) / 4::numeric AS ore4, (letture02.let17 +
letture02.let18 + letture02.let19 + letture02.let20) / 4::numeric AS
ore5, (letture02.let21 + letture02.let22 + letture02.let23 + letture02.
let24) / 4::numeric AS ore6, (letture02.let25 + letture02.let26 +
letture02.let27 + letture02.let28) / 4::numeric AS ore7, (letture02.
let29 + letture02.let30 + letture02.let31 + letture02.let32) / 4::
numeric AS ore8, (letture02.let33 + letture02.let34 + letture02.let35 +
letture02.let36) / 4::numeric AS ore9, (letture02.let37 + letture02.
let38 + letture02.let39 + letture02.let40) / 4::numeric AS ore10,
(letture02.let41 + letture02.let42 + letture02.let43 + letture02.let44)
/ 4::numeric AS ore11, (letture02.let45 + letture02.let46 + letture02.
let47 + letture02.let48) / 4::numeric AS ore12, (letture02.let49 +
letture02.let50 + letture02.let51 + letture02.let52) / 4::numeric AS
ore13, (letture02.let53 + letture02.let54 + letture02.let55 + letture02.
let56) / 4::numeric AS ore14, (letture02.let57 + letture02.let58 +
letture02.let59 + letture02.let60) / 4::numeric AS ore15, (letture02.
let61 + letture02.let62 + letture02.let63 + letture02.let64) / 4::
numeric AS ore16, (letture02.let65 + letture02.let66 + letture02.let67
+ letture02.let68) / 4::numeric AS ore17, (letture02.let69 + letture02.
let70 + letture02.let71 + letture02.let72) / 4::numeric AS ore18,
(letture02.let73 + letture02.let74 + letture02.let75 + letture02.let76)
/ 4::numeric AS ore19, (letture02.let77 + letture02.let78 + letture02.
let79 + letture02.let80) / 4::numeric AS ore20, (letture02.let81 +
letture02.let82 + letture02.let83 + letture02.let84) / 4::numeric AS
ore21, (letture02.let85 + letture02.let86 + letture02.let87 + letture02.
let88) / 4::numeric AS ore22, (letture02.let89 + letture02.let90 +
letture02.let91 + letture02.let92) / 4::numeric AS ore23, (letture02.
let93 + letture02.let94 + letture02.let95 + letture02.let96) / 4::
numeric AS ore24
FROM letture02
WHERE letture02.contatore >= 2000
AND letture02.contatore <= 2999;
---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
choose an index scan if your joining column's datatypes do not
match
