"Simon Riggs" <simon(at)2ndquadrant(dot)com> writes:
> Long text fields that might use LIKE should be set to 100. CHAR(1) and
> general fields should be set to 10.
I could see arguing that either way. Longer fields are capable of more
precision and so may need more buckets to predict. On the other hand longer
fields take more space and take longer to compare so to make consistent use of
resources you would want to avoid storing and comparing large numbers of them
whereas you could afford much larger targets for small quick columns.
Ask me about EnterpriseDB's Slony Replication support!
In response to
pgsql-hackers by date
|Next:||From: Dave Page||Date: 2007-11-19 08:31:57|
|Subject: Server maintenance|
|Previous:||From: Matt Magoffin||Date: 2007-11-19 07:52:22|
|Subject: possible to create multivalued index from xpath() results in 8.3?|
pgsql-patches by date
|Next:||From: Andrew Dunstan||Date: 2007-11-19 13:31:43|
|Subject: Re: [HACKERS] fulltext parser strange behave|
|Previous:||From: Simon Riggs||Date: 2007-11-18 14:42:54|
|Subject: Re: Better default_statistics_target|