Allow CLUSTER on partitioned tables
commitcfdd03f45e6afc632fbe70519250ec19167d6765
authorAlvaro Herrera <alvherre@alvh.no-ip.org>
Sat, 2 Apr 2022 17:08:34 +0000 (2 19:08 +0200)
committerAlvaro Herrera <alvherre@alvh.no-ip.org>
Sat, 2 Apr 2022 17:08:34 +0000 (2 19:08 +0200)
treeaaa33106f858b9170192561addd9359417e9ac71
parentb7c485fb93726cb04b858442d73043b56e603711
Allow CLUSTER on partitioned tables

This is essentially the same as applying VACUUM FULL to a partitioned
table, which has been supported since commit 3c3bb99330aa (March 2017).
While there's no great use case in applying CLUSTER to partitioned
tables, we don't have any strong reason not to allow it either.

For now, partitioned indexes cannot be marked clustered, so an index
must always be specified.

While at it, rename some variables that were RangeVars during the
development that led to 8bc717cb8878 but never made it that way to the
source tree; there's no need to perpetuate names that have always been
more confusing than helpful.

Author: Justin Pryzby <pryzby@telsasoft.com>
Reviewed-by: Matthias van de Meent <boekewurm+postgres@gmail.com>
Discussion: https://postgr.es/m/20201028003312.GU9241@telsasoft.com
Discussion: https://postgr.es/m/20200611153502.GT14879@telsasoft.com
doc/src/sgml/ref/cluster.sgml
src/backend/commands/cluster.c
src/backend/commands/tablecmds.c
src/bin/psql/tab-complete.c
src/include/commands/cluster.h
src/test/regress/expected/cluster.out
src/test/regress/sql/cluster.sql