Re: [GENERAL] UNION within VIEW workarounds?

On Thu, Dec 07, 2000 at 07:45:02PM -0600, Ed Loehr wrote: > One advantage is that you can hide internal db schema from applications > that may need to access your db. That frees you to change the internal > schema somewhat as long as you honor the agreed "interface" of the > view... another adva

[GENERAL] My patches are now on a webpage.

Hello, I finally found time and put all my patches for 7.0.x to a webpage. I use these patches in one of heavy loaded system (it handles approx. 1M of queries each day). That's why I can say that they are a little bit tested. Also all of these patches are included in current 7.1 CVS. The link i

[GENERAL] Meaning of EXPLAIN information...?

Can anyone help with my understanding of the processing score reported by EXPLAIN and the way it relates (or doesn't) to processing time? I test-run queries against an old, slow linux box in order to help spot time or processor consuming queries, because I don't understand the figures that EXPLAI

Re: [GENERAL] PostgreSQL Book?

> What's going on with Bruce' PostgreSQL book? It says on the amazon.com > website that it is due on December 1, 2000. That was last Friday! I would love to know myself. I received my first copy on Wednesday, so it should be going out soon. No idea what is delaying it. -- Bruce Momjian

[GENERAL] Index question

Is there any reason an index would not be used in optimising a query? I have an SQL which does a join and correctly picks up an index on a numeric column but fails to use another index that is on a character column (char I think) Maybe its because its character column - as most joins are done on

RE: [GENERAL] Index question

Hi Martin, You could try doing this: SET ENABLE_SEQSCAN OFF ... and then run EXPLAIN with your SQL statement again and compare the two costs. That may well show you why the sequential scan is chosen over the index scan. Also, have you VACUUM ANALYZE'd your tables recently? You can find out wh

Re: [GENERAL] ilike and --enable-multibyte=KOI8

> Hello, Can anybody check and confirm(or not) this: > > if pgsql(DEVEL, from cvs-server) is built with locale and multibyte then > > "select * from table where row ilike '%a%'" crashes pgsql. > "select * from table where row ilike 'a%'" doesn't craches. > "select * from table where row like

Re: [GENERAL] Oracle-compatible lpad/rpad behavior

> Returns not empty string but NULL: The two are equivalent in Oracle. Try select 'a' || null || 'b' from dual and compare it to postgres. -Jonathan

[GENERAL] Memory Usage

I am having a small issue with PostgreSQL 7.0.3 on FreeBSD 4.2 Stable. When I perform an update on a table with roughly 2 million rows, the postgres process starts eating up memory until it eventually uses up all of the memory and exits without finishing. I have also seen the same thing happen i

Re: [GENERAL] Memory Usage

"Nathan Barnett" <[EMAIL PROTECTED]> writes: > I am having a small issue with PostgreSQL 7.0.3 on FreeBSD 4.2 Stable. When > I perform an update on a table with roughly 2 million rows, the postgres > process starts eating up memory until it eventually uses up all of the > memory and exits without