[postgis-devel] Database Architecture

[Chris Hodgson]

I'm not sure that separating your data into separate schemas will do 
anything more for you than just separate tables.

You should probably look into the table partitioning and inheritance 
features of Postgresql, it seems that it should be very applicable to 
your use case.

http://www.postgresql.org/docs/8.3/interactive/ddl-partitioning.html

Chris

OscarUK wrote:
> Many thanks
>
> Any pointers on question 1, i.e. Does splitting the database across schemas
> sound a sensible approach? Is it necessary etc.?
>
>
>
>
>
> strk-2 wrote:
>   
>> On Tue, May 04, 2010 at 01:39:51AM -0700, OscarUK wrote:
>>     
>>> Hi forum users
>>>
>>> Please note I am a novice user.
>>>
>>> I wish to set up a very large PG database. For example, one 500x500km
>>> square
>>> had 60 million lines and 25 million polygons. However, at any one time
>>> users
>>> will be viewing or extracting only small areas, say 100m by 100m.
>>>
>>> Because of the database size, I intend to break this down into various
>>> schemas within a single database. Where I need to grab map extracts
>>> across
>>> schemas, I hope to script views (or queries) on the fly that return the
>>> required union between schemas.
>>>
>>> My questions are therefore:
>>>
>>> 1.  Does this sound a sensible approach?
>>> 2.  How can I create a second schema that contains all the PG functions?
>>> (When I create a new schema it contains no PG functions and I cannot work
>>> out how to 'get them in').
>>>       
>> Postgis functions are tipically installed in the 'public' schema,
>> as long as you have 'public' in your search_path you should find them.
>>
>> --strk; 
>>
>>   ()   Free GIS & Flash consultant/developer
>>   /\   http://strk.keybit.net/services.html
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at postgis.refractions.net
>> http://postgis.refractions.net/mailman/listinfo/postgis-devel
>>
>>
>>     
>
>