Re: Table Partitioning Feature - Mailing list pgsql-hackers

Hi Amit,

I overlooked the fact that you dropped composite partitions and subpartitions template from the proposal presented in http://archives.postgresql.org/pgsql-hackers/2008-01/msg00413.php.
Is it because this is too hard to support? or you don't see any immediate need for it?

Thanks,
Emmanuel

Hi Emmanuel,

Please find my comments in-lined:

On 1/23/09, *Emmanuel Cecchet* <manu@frogthinker.org <mailto:manu@frogthinker.org>> wrote:

   Amit,

   You might want to put this on the
   http://wiki.postgresql.org/wiki/Table_partitioning wiki page.


Sure.

   How does your timeline look like for this implementation?


 The implementation is planned as follows:
- Partition table commands
++ An intermediate patch in Feb end
++ Final patch in mid March
- Global Index: Mid March
- Optimizer changes for partitioned table: May

   I would be happy to contribute C triggers to your implementation.
   From what I understood in
   http://archives.postgresql.org/pgsql-hackers/2008-01/msg00269.php,
   you already have an implementation that parses the grammar and
   generates rules as if someone had written them. Is this code
   available?


We have just started with the implementation, i will post the grammar rules next week.


   Regarding the use of triggers to push/move data to partitions,
   what if someone declares triggers on partitions? Especially if you
   have subpartitions, let's consider the case where there is a
   trigger on the parent, child and grandchild. If I do an insert in
   the parent, the user trigger on the parent will be executed, then
   the partition trigger that decides to move to the grandchild. Are
   we going to bypass the child trigger?


We are not supporting sub-partitioning - There is just one level of partitioning.

   If we also want fast COPY operations on partitioned table, we
   could have an optimized implementation that could bypass triggers
   and move the tuple directly to the appropriate child table.


We will definitely consider to implement fast COPY after we are done with the planned tasks.

Thanks,
Amit


   Thanks for this big contribution,
   Emmanuel

       Hi,

       We are implementing table partitioning feature to support
       - the attached commands. The syntax conforms to most of the
       suggestion mentioned in
       http://archives.postgresql.org/pgsql-hackers/2008-01/msg00413.php,
       barring the following:
       -- Specification of partition names is optional. System will
       be able to generate partition names in such cases.
       -- sub partitioning
        We are using pgsql triggers to push/move data to appropriate
       partitions, but we will definitely consider moving to C
       language triggers as suggested by manu.
       - Global non-partitioned indexes (that will extend all the
       partitions).
       - Foreign key support for tables referring to partitioned tables.

       Please feel free to post your comments and suggestions.

       Thanks,
       Amit
       Persistent Systems



       ------------------------------------------------------------------------




   --    Emmanuel Cecchet
   FTO @ Frog Thinker Open Source Development & Consulting
   --
   Web: http://www.frogthinker.org

   email: manu@frogthinker.org <mailto:manu@frogthinker.org>
   Skype: emmanuel_cecchet

--
Emmanuel Cecchet
FTO @ Frog Thinker Open Source Development & Consulting
--
Web: http://www.frogthinker.org
email: manu@frogthinker.org
Skype: emmanuel_cecchet