On Thu, 31 May 2001, Tom Lane wrote:
> Kovacs Zoltan <kovacsz(at)pc10(dot)radnoti-szeged(dot)sulinet(dot)hu> writes:
> > It means that the 21st line of the result is problematic, because writing
> > 20 instead of 21 I got no problem.
>
> I think not. The current implementation of LIMIT fetches one more row
> than is really needed, IIRC.
Tom, the real problem is that I get _different_ output for
tir=# SELECT c.relname AS viewname, pg_get_userbyid(c.relowner) AS
viewowner, pg_get_viewdef(c.relname) AS definition FROM pg_class c WHERE
(c.relkind = 'v'::"char") limit 21 offset 1;
ERROR: cache lookup for proc 4303134 failed
and
tir=# SELECT c.relname AS viewname, pg_get_userbyid(c.relowner) AS
viewowner, pg_get_viewdef(c.relname) AS definition FROM pg_class c WHERE
(c.relkind = 'v'::"char") and c.relname = 'felhasznalo';
viewname | viewowner |
definition
-------------+-----------+------------------------------------------------------------------------------------------------------------------------------------
felhasznalo | postgres | SELECT szemely.az, szemely.nev,
szemely.teljes_nev FROM szemely WHERE ((1 <= szemely.felhasznalo) AND
(szemely.felhasznalo <= 2));
(1 row)
The second one also _should_ result an ERROR. (As you can see, this view
doesn't contain any function. I put an index on the table `szemely' but I
dropped it. There may be some relation between this error and the dropped
index...?)
In a consequence, I cannot pg_dump my database (which is under
production... :-( Please help! Unfortunately I cannot duplicate this
problem from scratch, but I may try to do it.
TIA, Zoltan
In response to
Responses
pgsql-hackers by date
| Next: | From: Zeugswetter Andreas SB | Date: 2001-06-01 07:43:03 |
| Subject: AW: Access statistics |
| Previous: | From: Joe Conway | Date: 2001-06-01 06:31:51 |
| Subject: Fw: Isn't pg_statistic a security hole - Solution Proposal |