On Sat, Jan 9, 2010 at 15:20, tom <toab...@googlemail.com> wrote: > Hi Adrian, > > thanks for your answer! > The problem is, that i don't know how many projects will be created. > Every user > of the application can create and delete new projects. > Maybe it's possible to execute some raw sql commands whenever a > project is created or deleted (to create the new table and insert > trigger). > Do you think that's possible? > Do you have any good tutorial about postgres and partitioning? I'm > completly new to this topic. > > One possibility is to have one child table for each existing project. And one additional table for (all) the newly created projects (just as a temporary step until someone or a script creates the new tables). Later, a midnight script can check for new projects, create the needed new tables and rewrite the trigger(s) . I see no reason why automating everything wouldn't be possible .
By the way : how many projects do you expect to have in the database? Thousands of projects means thousands of tables , and a huge insert trigger that tries to determine where to insert each new record. PostgreSQL should be able to handle lots of tables, but that trigger will become a bottleneck at some point in future ... On the other hand, try to avoid being stuck into the partitioning idea: maybe improving the performance for your database can be done in other ways (maybe you are simply missing an index ! ) . Sorry, I have no links about partitioning to suggest. However, i suggest you to join the pgsql-performance mailing list and ask there for advice. That is the PostgreSQL list for discussions about specific performance issues. You'll find information about how to subscribe on the postgresql website . Just make sure that you post there enough information about your database: the columns of the table, the estimated number of rows , and a specific select query that is executed too slowly . And of course the execution plan of that query ( explain analyse select ..... from .... where .... ). So, make sure that you are describing a concrete problem : for example you didn't mention what is actual the problem that you hope to solve through partitioning. Obviously, you are seeking for faster database response time. But when exactly : when you insert data, or when you query the data ? There are many postgres experts on pgsql-performance , but you need to describe well what are you trying to do ;) -- Adrian Maier--
You received this message because you are subscribed to the Google Groups "Django users" group.
To post to this group, send email to django-us...@googlegroups.com.
To unsubscribe from this group, send email to django-users+unsubscr...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/django-users?hl=en.
