postgres table and index bloat - ghdrako/doc_snipets GitHub Wiki

• https://github.com/ioguix/pgsql-bloat-estimation/blob/master/table/table_bloat.sql
• https://github.com/ioguix/pgsql-bloat-estimation/blob/master/btree/btree_bloat.sql
• https://github.com/reorg/pg_repack
• https://hakibenita.com/postgresql-unused-index-size
• https://github.com/ioguix/pgsql-bloat-estimation
• https://www.percona.com/blog/pg_repack-rebuild-postgresql-database-objects-online
• https://www.percona.com/blog/pg_repack-rebuild-postgresql-database-objects-online
• https://github.com/andyatkinson/pg_scripts/blob/main/top_updated_tables.sql
• https://docs.crunchybridge.com/insights-metrics/bloat-and-vacuum#

pgcompacttable

• https://github.com/dataegret/pgcompacttable It works by reordering tuples from the end of a table towards the front of a table, which allows the table size to shrink.

reindex

• https://github.com/ghdrako/doc_snipets/wiki/postgres%E2%80%90vacume%E2%80%90pg_repack%E2%80%90cloudsql-pg_squeeze/

In PostgreSQL 12, REINDEX gained support for the CONCURRENTLY option. This works similarly to how pg_repack does, but it’s native to PostgreSQL. By being native, it continues to receive improvements in newer versions.

REINDEX (VERBOSE)  INDEX  CONCURRENTLY index_trips_on_driver_id;

Providing a table name as an option to REINDEX performs a reindex for all indexes that are on the specified table.

REINDEX (VERBOSE)  TABLE  CONCURRENTLY trips;