Re: [GENERAL] Large table search question[Scanned]

2004-06-01 Thread Stijn Vanroye
Thanks for the reply. I was afraid it would come down to testing each individual situation.ยต The table I mentioned (6 million+ records) actually is a phonebook. And searching and filtering is possible on almost any combination of fields. So there's an index on each individual field now and that'

Re: [GENERAL] Large table search question

2004-06-01 Thread Richard Huxton
Stijn Vanroye wrote: I don't want to but in, I just find this an interesting discussion and would like to add my 2 cents: I have read this in the manual: (PostgreSQL 7.4beta4 documentation, Chapter 11.3 Multicolumn Indexes) Qoute: "Multicolumn indexes should be used sparingly. Most of the time, an

Re: [GENERAL] Large table search question

2004-06-01 Thread Stijn Vanroye
ards, Stijn Vanroye -Original Message- From: Richard Huxton [mailto:[EMAIL PROTECTED] Sent: dinsdag 1 juni 2004 10:44 To: John Wells Cc: [EMAIL PROTECTED] Subject: Re: [GENERAL] Large table search question[Scanned] John Wells wrote: > Guys, > > I have a general question about desig

Re: [GENERAL] Large table search question

2004-06-01 Thread Richard Huxton
John Wells wrote: Guys, I have a general question about designing databases for large data sets. I was speaking with a colleague about an application we're preparing to build. One of the application's tables will potentially contain 2 million or more names, containing (at least) the fields first_n

Re: [GENERAL] Large table search question

2004-05-30 Thread Tom Lane
"John Wells" <[EMAIL PROTECTED]> writes: > A common lookup the application will require is the full name, so prefix + > first_name + middle_name + last_name. > My friend's suggestion was to create a "lookup field" in the table itself, > which would contain a concatenation of these fields created d

[GENERAL] Large table search question

2004-05-30 Thread John Wells
Guys, I have a general question about designing databases for large data sets. I was speaking with a colleague about an application we're preparing to build. One of the application's tables will potentially contain 2 million or more names, containing (at least) the fields first_name, last_name,