Horton

Re: Declarative partitioning - another take

Maksim Milyutin <m.milyutin@postgrespro.ru>

View thread View on postgr.es
From: Maksim Milyutin <m.milyutin@postgrespro.ru>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Michael Paquier <michael.paquier@gmail.com>, Andres Freund <andres@anarazel.de>, Erik Rijkers <er@xs4all.nl>, Amit Langote <amitlangote09@gmail.com>, Amit Langote <Langote_Amit_f8@lab.ntt.co.jp>, Rajkumar Raghuwanshi <rajkumar.raghuwanshi@enterprisedb.com>, Ashutosh Bapat <ashutosh.bapat@enterprisedb.com>, Pg Hackers <pgsql-hackers@postgresql.org>, pgsql-hackers-owner@postgresql.org
Date: 2016-12-09T10:46:58Z
Lists: pgsql-hackers 
Hi, everyone!


08.12.16 18:25, Robert Haas wrote:
> Of course, this is the beginning, not the end.  I've been thinking
> about next steps -- here's an expanded list:
>
> - more efficient plan-time partition pruning (constraint exclusion is too slow)
> - run-time partition pruning
> - partition-wise join (Ashutosh Bapat is already working on this)
> - try to reduce lock levels
> - hash partitioning
> - the ability to create an index on the parent and have all of the
> children inherit it; this should work something like constraint
> inheritance.  you could argue that this doesn't add any real new
> capability but it's a huge usability feature.
> - teaching autovacuum enough about inheritance hierarchies for it to
> update the parent statistics when they get stale despite the lack of
> any actual inserts/updates/deletes to the parent.  this has been
> pending for a long time, but it's only going to get more important
> - row movement (aka avoiding the need for an ON UPDATE trigger on each
> partition)
> - insert (and eventually update) tuple routing for foreign partitions
> - not scanning the parent
> - fixing the insert routing so that we can skip tuple conversion where possible
> - fleshing out the documentation
>
I would like to work on two tasks:
  - insert (and eventually update) tuple routing for foreign partition.
  - the ability to create an index on the parent and have all of the 
children inherit it;

The first one has been implemented in pg_pathman somehow, but the code 
relies on dirty hacks, so the FDW API has to be improved. As for the 
extended index support, it doesn't look like a super-hard task.

-- 
Maksim Milyutin
Postgres Professional: http://www.postgrespro.com
Russian Postgres Company

Commits

  1. Fix typo.

  2. Document trigger-firing behavior for inheritance/partitioning.

  3. Fire per-statement triggers on partitioned tables.

  4. Set ecxt_scantuple correctly for tuple routing.

  5. Fix interaction of partitioned tables with BulkInsertState.

  6. Avoid core dump for empty prepared statement in an aborted transaction.

  7. Fix some problems in check_new_partition_bound().

  8. Remove unnecessary arguments from partitioning functions.

  9. Fix reporting of constraint violations for table partitioning.

  10. Fix tuple routing in cases where tuple descriptors don't match.

  11. Invalid parent's relcache after CREATE TABLE .. PARTITION OF.

  12. Doc: improve documentation about inheritance.