Life goes on. We continue to acquaint you with the most interesting PostgreSQL news.
Postgres Pro Enterprise 11.17.1 and Postgres Pro Standard 11.17.1
IN Postgres Pro Enterprise 11.17.1 Fixed bugs that users pointed out. Serious improvements were made in the multimaster extension:
– Now it is recommended to use it in a configuration with three nodes, one of which is voting. More details here;
– fixed the problem of bloating WAL by clearing the synchronization points of the remote node and correcting the calculation of the minimum required LSN. Previously, WAL bloating sometimes occurred when a node was removed from a cluster;
– fixed the problem with the return of the node to the cluster after a long shutdown of this node;
Fixed a bug at the synchronization point during initialization of the multimaster module, which occurred in the event of a failure before the first synchronization.
In addition, the mechanism of the built-in connection pool has been improved. Compared with the previous version, the following innovations appeared in it:
– the dedicated_users parameter, which allows you to specify a list of users for whom dedicated service processes will be used in the connection pool mode;
– Separate serving processes can now accept connections from different users, so that all connections to the same database will belong to the same shared pool.
There are improvements common to Postgres Pro Enterprise 11.17.1 and Postgres Pro Standard 11.17.1. For example, the pg_probackup utility has been updated to version 2.2.7, and mamonsu has been updated to version 2.4.4.
For these and other release innovations, see the Release Notes from PPE documentation and PPS.
Postgres Pro Standard 12.2.1
Differences of this version from PostgreSQL 12 and from Postgres Pro Standard 11.17.1 can be traced by the corresponding Release Notes.
pgAdmin 4 v 4.19.
In that release Python 3.8 support has appeared, debugger operation has been improved, when a connection is interrupted by pgAdmin, it is guaranteed that transactions that are not yet committed will be rolled back. Fixed 18 bugs.
The first stable release after moving to Git, supports the PostgreSQL 13 development branch. Optimized for PostgreSQL 12.2, GEOS 3.8.0, SFCGAL 1.3.7, GDAL
3.0.4, PROJ 6.3.1, protobuf-c 1.3.3, json-c 0.13.1. About release herePDF documentation, and download from here.
PostGIS-only tile server (map server) written in Go. How to install and configure it is written here.
In that release Yandex PostgreSQL puller, which is being developed by Yandex, there are settings for better compatibility with pgbouncer. Replication storage types are no longer supported.
PAF, that is, the PostgreSQL Automatic Failover, a popular Pacemaker-corosync-based failover cluster solution, is only compatible with Pacemaker 1.1.13 on corosync 2.x. PAF now supports PostgreSQL 12. You can read about this release. here.
They claim that this release has improved a custom conflict resolution mechanism, optimized performance and improved compatibility with PostgreSQL 11 and 12. You can download here.
Now you can copy tables and sequences under a different name to a node. Technical details and downloads can be found. from here.
Postgresql Anonymizer 0.6
This extension hides or replaces personal information or commercial data from users from PostgreSQL. The project was led by Damen Kloshar, and later gave it under the wing of Dalibo. on the dalibo github. You can read about the release here.
The patch for compressing duplicates (deduplication) in B-Tree indices by Anastasia Lubennikova (Postgres Professional) was communicated Peter Geigan. Nastya managed to report on the innovation at the conference PGConf India 2020.
Charting the spread of COVID-19 using data
TimescaleDB Covers tellhow to use their extension and PostGIS to analyze the spread of the virus, how to formulate a query such as “how many neighbors within a 1 mile radius fell ill yesterday.”
Microsoft Azure Welcomes PostgreSQL Committers
Microsoft is not just buying Postgres companies. Predatory MS PR services lure leading open source developers. Posted by interview with Thomas Munro, Jeff Davis, and Andres Freund.
Robert Haas explainswhy even a well-configured automatic cleaning might not save you from problems when a lot of transactions hung. Robert analyzes the reasons, talks about monitoring.
Parallelism, what next?
Amit Kapila tells how parallel query processing has developed in PostgreSQL since version 9.6. As well as what to expect in version 13, and plans for the future.
PostgreSQL Person of the Week: Joe Conway; Bruce momjian
Short interview with Joe Conway. Classic (20 years with Postgres), author of PL / R talks about his life, work in Crunchy Data. Believes in a brighter future for Postgres.
Bruce (24 years with Postgres) even more concisebut he places important accents: his favorite extension is PostGIS, the main achievements of PG 12 are REINDEX CONCURRENTLY, multi-column statistics, Cte inlining, and the most needed PG in the future is sharding.
Horizontal Scalability / Sharding in PostgreSQL Core – Missing pieces of the puzzle
Sharding theme expands EnterpriseDB veteran Ahsan Hadi, now vice president of development at HighGo Software. He talks about sharding on FDW, about Global Transaction Manager, about global snapshots and much more.
And again about recursive queries
Not only us, but also Postgres Weekly advises reading (or re-reading) an article by Yegor Rogov from the Postgres Professional education department. The article is not new, but the reason (publication of the translation) is good. Here is the link to the original.
Postgres at a low level
And another recommendation for a second (for someone first) viewing. On the this video Dmitry Dolgov from Zalando for 37 minutes talks about the architecture of the lower levels of Postgres, the use of strace and gdb.
Postgres Subtransactions and Performance
About subtransactions in PL / pgSQL, about some performance and reliability problems that may arise at the same time, about diagnostics they write on the Cybertec blog.
Handling NULL Values in PostgreSQL
On the Percona website disassemble features of working with NULL in C / C ++, in Java, and, of course, in Postgres (there is nothing about other procedural languages that work with Postgres – it’s a pity).
Title articles completely: Joe bot, an SQL query optimization assistant, updated to version 0.5.0.. Describes a chatbot project for Slack running on top of the Database Lab (from Postgres.ai).
While we will not notify about cancellations and transfers. Information is too volatile.
subscribe to the channel postgresso!