Postgres 9.4 is adding a new in-core GiST operator class for the [inet and cidr] (https://www.postgresql.org/docs/9.4/static/datatype-net-types.html#DATATYPE-INET) datatypes. It has been introduced by this commit:

Working with logical replication and a cluster of nodes heavily linked among each other can result in particularly disturbing results. If a logical change is received on a node that itself regenerates additional changes, it is really easy to finish with an uncontrolable chain reaction.

These last couple of days I have been developing for studies a background worker able to fetch changes from a logical decoder already developed and presented on this blog called [decoder_raw] (https://github.com/michaelpq/pg_plugins/tree/master/decoder_raw) able to generate raw queries based on the logical changes decoded on server from its WAL. This receiver, called [receiver_raw] (https://github.com/michaelpq/pg_plugins/tree/master/receiver_raw) runs as a background worker that connects to a node running decoder_raw using a replication connection and then fetches decoded logical changes from it, which are in this case ready-to-be-applied raw queries. Those changes are then applied on the database one by one using the [SPI] (https://www.postgresql.org/docs/devel/static/spi.html) through a loop process that sleeps during a customizable amount of nap time. Note that receiver_raw is actually designed to be performant as it applies one single batch of changes using only one transaction in a single loop process. That’s the best doable particularly for long transactions.

When developing a logical change receiver with the new [logical decoding] (https://www.postgresql.org/docs/devel/static/logicaldecoding.html) facility of Postgres 9.4, there are a couple of new commands and a certain libpq protocol to be aware of before beginning any development for a logical replication receiver (on top of knowing some [basics] (/postgresql-2/postgres-9-4-feature-highlight-basics-logical-decoding/) and what is an [output decoder plugin] (/postgresql-2/postgres-9-4-feature-highlight-output-plugin-logical-replication/)). Here is an exhaustive list of the commands to know.

Today here is a highlight of a new Postgres 9.4 feature interesting for developers and companies doing packaging of Postgres on Windows as it makes possible the installation of client-only binaries and libraries using MSVC. It has been introduced by this commit:

Continuing on the series of posts about logical replication and after looking at [some basics] (/postgresql-2/postgres-9-4-feature-highlight-basics-logical-decoding/) and the concept of [REPLICA IDENTITY] (/postgresql-2/postgres-9-4-feature-highlight-replica-identity-logical-replication/) for a table, now is time to begin more serious things by having a look at how to develop an output plugin for the logical decoding facility.

Among the many things to say about logical replication features added in PostgreSQL 9.4, REPLICA IDENTITY is a new table-level parameter that can be used to control the information written to WAL to identify tuple data that is being deleted or updated (an update being a succession of an insert and a delete in MVCC).

The second huge feature coming in PostgreSQL 9.4 with [jsonb] (postgresql-2/postgres-9-4-feature-highlight-indexing-jsonb/) is called [logical decoding] (https://www.postgresql.org/docs/devel/static/logicaldecoding.html). In short, it is a new plugin facility that can be used to decode changes that happen on a database and stream them to external sources. It can be used for many things like replication, auditing or even online upgrade solutions.

Except if you have been cut from the Internet the last week, you have already heard of Heartbleed. This good candidate for the “Bug of the Year 2014” price is already costing a better-not-to-count amount of money in maintenance and development for many companies around the world.

PostgreSQL 9.4 is shipping with a new feature called [jsonb] (https://www.postgresql.org/docs/devel/static/datatype-json.html), which is a new data type able to store JSON data supporting GIN indexing (!). In short, this feature, one of the most important of the upcoming release, if not the most important, puts Postgres directly in good position in the field of document-oriented database systems.