E.20. Postgres Pro Enterprise 11.9.1

Release date: 2020-08-27

E.20.1. Overview

This release is based on PostgreSQL 11.9 and Postgres Pro Enterprise 11.8.1. All changes inherited from PostgreSQL 11.9 are listed in PostgreSQL 11.9 Release Notes. As compared with Postgres Pro Enterprise 11.8.1, this version also provides the following changes:

  • Disabling password reuse now correctly works with SCRAM authentication. Previously SCRAM-encrypted password was reused even when it was not supposed to.

  • Fixed bug in pgpro_scheduler that prevented stopping of the running jobs on schedule.stop().

  • Fixed the GUC parameter group naming. A Compressed File System parameter group name has been omitted previously which caused other group names to be shifted.

  • Improved the rum extension:

    • Improved full text search for tsvector values with the lexeme weights specified. When executing queries with the weight-based constraints, the index search results do not need to be rechecked in table, so these queries perform much faster.

    • Fixed the error which causes result loss when the negation operator is used in such queries.

  • Removing multimaster temporary schemas (mtm_tmp_X_Y) that can be left over after a node crash. This ensures that such orphaned schemas are not accumulated on crashes.

  • Removed CFS map recovery procedure executed at server start. It could create race condition and did not provide any benefit.

  • Introduced new multimaster extension version to enable functionality implemented in Postgres Pro Enterprise version 11.8.1. Previously this functionality was not enabled on Postgres Pro Enterprise upgrade.

  • Upgraded pg_probackup to latest version 2.4.2:

    • New options and flags can now be used to add flexibility to delete, backup, restore, archive-push and set-backup commands.

    • Incremental restore and support for multi-timeline incremental chains have been added.

    • Postgres Pro parameters slot_name and primary_conninfo can be used during restore.

    • archive-push and archive-get commands considerably reworked.

    • Improvements have been achieved in speed and memory consumption.

    See pg_probackup documentation for details.

  • Ended support for PTRACK versions lower than 2.0 (1.5 - 1.7) since they contain a critical bug that can cause silent corruption in PTRACK backups that were taken under load. PTRACK 1.x API is retained for backward compatibility although all the functionality is no longer available. It is recommended that you avoid restoring PTRACK backups taken earlier, but if such a need arises, validate any restored cluster instance using the command: pg_probackup checkdb --amcheck.

  • Added the optional tune argument to the pg-setup initdb command. This argument allows you to choose one of the predefined customized configurations for your database cluster.

  • Ended support for SUSE Linux Enterprise Server 11.

  • Fixed race conditions in BRIN index that caused errors:

    • "failed to find parent tuple for heap-only tuple ...".

      The error could occur when the brin_summarize_new_values() function and HOT updates were executed simultaneously in concurrent transactions.

    • "corrupted BRIN index: inconsistent range map".

      The error could occur when BRIN index's desummarization and a bitmap scan were executed simultaneously in concurrent transactions.

  • Upgraded mamonsu for Linux systems to version 2.5.1. Now it is based on Python 3. Version 2.3.4 is still provided for Windows systems.

E.20.2. Migration to Version 11.9.1

If you are upgrading from a Postgres Pro Enterprise release based on the same PostgreSQL major version, it is enough to install the new version into your current installation directory.

While functions numeric_eq, numeric_ne, numeric_gt, numeric_ge, numeric_lt, and numeric_le are actually leakproof, they were not marked as such in Postgres Pro Enterprise 11.6.1 or lower, which could lead to incorrect query optimization. In particular, it negatively affected query execution if row-level security policy was in use. Version 11.7.1 repairs this issue for new installations by correcting the initial catalog data, but existing installations will still have incorrect markings unless you update pg_proc entries for these functions. You can run pg_upgrade to upgrade your server instance to a version containing the corrected initial data, or manually correct these entries in each database of the installation using the ALTER FUNCTION command. For example:

ALTER FUNCTION pg_catalog.numeric_eq LEAKPROOF

Version 11.7.1 also fixes SIMILAR TO and POSIX regular expressions that use character classes for icu collations, so you may need to check for objects that use such regular expressions.

Starting from Postgres Pro Enterprise 11.6.1, the ICU library upgrade does not interfere with the server start. Before connecting to a database using ICU as the default collation, Postgres Pro compares this collation version to the one provided by the ICU library and displays a warning if the collation versions do not match; you may need to rebuild the objects that depend on the default collation if you think the collation change may affect the sort order of your data. To suppress these warnings, you can use the ALTER COLLATION "default" REFRESH VERSION command, as explained in ALTER COLLATION.

To migrate from PostgreSQL, as well as Postgres Pro Standard or Postgres Pro Enterprise based on a previous PostgreSQL major version, see the migration instructions for version 11.