E.19. Postgres Pro Enterprise 11.10.1

Release date: 2020-11-30

E.19.1. Overview

This release is based on PostgreSQL 11.10 and Postgres Pro Enterprise 11.9.1. All changes inherited from PostgreSQL 11.10 are listed in PostgreSQL 11.10 Release Notes. As compared with Postgres Pro Enterprise 11.9.1, this version also provides the following changes:

  • Ended support for CentOS 6.

  • Fixed a bug in the pg_variables module. Now you can create a variable in a transaction after removal of a variable with the same name.

  • Upgraded mamonsu to version 2.6.2.

  • Upgraded pgbouncer to version 1.15.

  • Improved the precision of pgbench when the --default-isolation-level option is used. The default isolation level is now set in the environment variable PGOPTIONS so that TPS values are computed consistently, with or without the --connect option.

  • Fixed CFS to dramatically speed up pg_basebackup for a database that contains a lot of small tables. Previously, nearly 1 MB of zeroes was produced for each relation to be copied.

  • Fixed a bug that permitted duplication of rows in a table with a unique index. Only tables with more than one unique index were prone to this issue. The duplicates could appear if insertions were done under a heavy concurrent load and concurrently with DDL operations on the table that triggered invalidation of the relation cache.

  • Fixed a bug in the function for estimating the selectivity by index when evaluating a condition. Previously, the server could crash while executing this function if the type cache invalidation occurred at an unsuitable moment.

  • Updated pgpro_scheduler:

    • Now pgpro_scheduler waits in the suspended state on a standby server to be started when the standby is promoted to a master server. Previously it didn't work on a standby server either and besides generated a lot of error messages.

    • Enhanced the scheduling scale used by pgpro_scheduler — jobs can now be scheduled to a precision of seconds. Also some predefined keywords can now be used instead of crontab strings.

  • Updated multimaster:

    • Increased the maximum length of a transaction identifier that is specified in PREPARE TRANSACTION. Now the identifier can be up to 200 bytes long, while previously it was limited to 31 bytes.

    • Fixed a bug that could lead to the mtm-logrep-receiver process crash in case of failure when starting a dynamic worker. This failure could occur if the max_worker_processes parameter was not increased appropriately.

  • The use of pg_upgrade for upgrading a PostgreSQL or Postgres Pro Standard instance to Postgres Pro Enterprise 11 is no longer supported. If you need to upgrade from these versions, use a dump/restore.

E.19.2. Migration to Version 11.10.1

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.