Postgres Message Queue (PGMQ)

A lightweight message queue. Like AWS SQS and RSMQ but on Postgres.

Documentation: https://tembo.io/pgmq/

Source: https://github.com/tembo-io/pgmq

Features

Lightweight - No background worker or external dependencies, just Postgres functions packaged in an extension
Guaranteed "exactly once" delivery of messages to a consumer within a visibility timeout
API parity with AWS SQS and RSMQ
Messages stay in the queue until explicitly removed
Messages can be archived, instead of deleted, for long-term retention and replayability

Supported on Postgres 14-17.

Postgres Message Queue (PGMQ)

Installation

The fastest way to get started is by running the Tembo Docker image, where PGMQ comes pre-installed in Postgres.

docker run -d --name pgmq-postgres -e POSTGRES_PASSWORD=postgres -p 5432:5432 tembo.docker.scarf.sh/tembo/pg17-pgmq:latest

If you'd like to build from source, you can follow the instructions in CONTRIBUTING.md.

Updating

To update PGMQ versions, follow the instructions in UPDATING.md.

Client Libraries

Community

SQL Examples

# Connect to Postgres
psql postgres://postgres:postgres@0.0.0.0:5432/postgres

-- create the extension in the "pgmq" schema
CREATE EXTENSION pgmq;

Creating a queue

Every queue is its own table in the pgmq schema. The table name is the queue name prefixed with q_. For example, pgmq.q_my_queue is the table for the queue my_queue.

-- creates the queue
SELECT pgmq.create('my_queue');

 create
-------------

(1 row)

Send two messages

-- messages are sent as JSON
SELECT * from pgmq.send(
  queue_name  => 'my_queue',
  msg         => '{"foo": "bar1"}'
);

The message id is returned from the send function.

 send
-----------
         1
(1 row)

-- Optionally provide a delay
-- this message will be on the queue but unable to be consumed for 5 seconds
SELECT * from pgmq.send(
  queue_name => 'my_queue',
  msg        => '{"foo": "bar2"}',
  delay      => 5
);

 send
-----------
         2
(1 row)

Read messages

Read 2 message from the queue. Make them invisible for 30 seconds. If the messages are not deleted or archived within 30 seconds, they will become visible again and can be read by another consumer.

SELECT * FROM pgmq.read(
  queue_name => 'my_queue',
  vt         => 30,
  qty        => 2
);

 msg_id | read_ct |          enqueued_at          |              vt               |     message
--------+---------+-------------------------------+-------------------------------+-----------------
      1 |       1 | 2023-08-16 08:37:54.567283-05 | 2023-08-16 08:38:29.989841-05 | {"foo": "bar1"}
      2 |       1 | 2023-08-16 08:37:54.572933-05 | 2023-08-16 08:38:29.989841-05 | {"foo": "bar2"}

If the queue is empty, or if all messages are currently invisible, no rows will be returned.

SELECT * FROM pgmq.read(
  queue_name => 'my_queue',
  vt         => 30,
  qty        => 1
);

 msg_id | read_ct | enqueued_at | vt | message
--------+---------+-------------+----+---------

Pop a message

-- Read a message and immediately delete it from the queue. Returns an empty record if the queue is empty or all messages are invisible.
SELECT * FROM pgmq.pop('my_queue');

 msg_id | read_ct |          enqueued_at          |              vt               |     message
--------+---------+-------------------------------+-------------------------------+-----------------
      1 |       1 | 2023-08-16 08:37:54.567283-05 | 2023-08-16 08:38:29.989841-05 | {"foo": "bar1"}

Archive a message

Archiving a message removes it from the queue and inserts it to the archive table.

-- Archive message with msg_id=2.
SELECT pgmq.archive(
  queue_name => 'my_queue',
  msg_id     => 2
);

 archive
--------------
 t
(1 row)

Or archive several messages in one operation using msg_ids (plural) parameter:

First, send a batch of messages

SELECT pgmq.send_batch(
  queue_name => 'my_queue',
  msgs       => ARRAY['{"foo": "bar3"}','{"foo": "bar4"}','{"foo": "bar5"}']::jsonb[]
);

 send_batch 
------------
          3
          4
          5
(3 rows)

Then archive them by using the msg_ids (plural) parameter.

SELECT pgmq.archive(
  queue_name => 'my_queue',
  msg_ids    => ARRAY[3, 4, 5]
);

 archive 
---------
       3
       4
       5
(3 rows)

Archive tables can be inspected directly with SQL. Archive tables have the prefix a_ in the pgmq schema.

SELECT * FROM pgmq.a_my_queue;

 msg_id | read_ct |          enqueued_at          |          archived_at          |              vt               |     message     
--------+---------+-------------------------------+-------------------------------+-------------------------------+-----------------
      2 |       0 | 2024-08-06 16:03:41.531556+00 | 2024-08-06 16:03:52.811063+00 | 2024-08-06 16:03:46.532246+00 | {"foo": "bar2"}
      3 |       0 | 2024-08-06 16:03:58.586444+00 | 2024-08-06 16:04:02.85799+00  | 2024-08-06 16:03:58.587272+00 | {"foo": "bar3"}
      4 |       0 | 2024-08-06 16:03:58.586444+00 | 2024-08-06 16:04:02.85799+00  | 2024-08-06 16:03:58.587508+00 | {"foo": "bar4"}
      5 |       0 | 2024-08-06 16:03:58.586444+00 | 2024-08-06 16:04:02.85799+00  | 2024-08-06 16:03:58.587543+00 | {"foo": "bar5"}

Delete a message

Send another message, so that we can delete it.

SELECT pgmq.send('my_queue', '{"foo": "bar6"}');

 send
-----------
        6
(1 row)

Delete the message with id 6 from the queue named my_queue.

SELECT pgmq.delete('my_queue', 6);

 delete
-------------
 t
(1 row)

Drop a queue

Delete the queue my_queue.

SELECT pgmq.drop_queue('my_queue');

 drop_queue
-----------------
 t
(1 row)

Configuration

Partitioned Queues

You will need to install pg_partman if you want to use pgmq partitioned queues.

pgmq queue tables can be created as a partitioned table by using pgmq.create_partitioned(). pg_partman handles all maintenance of queue tables. This includes creating new partitions and dropping old partitions.

Partitions behavior is configured at the time queues are created, via pgmq.create_partitioned(). This function has three parameters:

queue_name: text: The name of the queue. Queues are Postgres tables prepended with q_. For example, q_my_queue. The archive is instead prefixed by a_, for example a_my_queue.

partition_interval: text - The interval at which partitions are created. This can be either any valid Postgres Duration supported by pg_partman, or an integer value. When it is a duration, queues are partitioned by the time at which messages are sent to the table (enqueued_at). A value of 'daily' would create a new partition each day. When it is an integer value, queues are partitioned by the msg_id. A value of '100' will create a new partition every 100 messages. The value must agree with retention_interval (time based or numeric). The default value is daily. For archive table, when interval is an integer value, then it will be partitioned by msg_id. In case of duration it will be partitioned on archived_at unlike queue table.

retention_interval: text - The interval for retaining partitions. This can be either any valid Postgres Duration supported by pg_partman, or an integer value. When it is a duration, partitions containing data greater than the duration will be dropped. When it is an integer value, any messages that have a msg_id less than max(msg_id) - retention_interval will be dropped. For example, if the max msg_id is 100 and the retention_interval is 60, any partitions with msg_id values less than 40 will be dropped. The value must agree with partition_interval (time based or numeric). The default is '5 days'. Note: retention_interval does not apply to messages that have been deleted via pgmq.delete() or archived with pgmq.archive(). pgmq.delete() removes messages forever and pgmq.archive() moves messages to the corresponding archive table forever (for example, a_my_queue).

In order for automatic partition maintenance to take place, several settings must be added to the postgresql.conf file, which is typically located in the postgres DATADIR. pg_partman_bgw.interval in postgresql.conf. Below are the default configuration values set in Tembo docker images.

Add the following to postgresql.conf. Note, changing shared_preload_libraries requires a restart of Postgres.

pg_partman_bgw.interval sets the interval at which pg_partman conducts maintenance. This creates new partitions and dropping of partitions falling out of the retention_interval. By default, pg_partman will keep 4 partitions "ahead" of the currently active partition.

shared_preload_libraries = 'pg_partman_bgw' # requires restart of Postgres
pg_partman_bgw.interval = 60
pg_partman_bgw.role = 'postgres'
pg_partman_bgw.dbname = 'postgres'

Visibility Timeout (vt)

pgmq guarantees exactly once delivery of a message within a visibility timeout. The visibility timeout is the amount of time a message is invisible to other consumers after it has been read by a consumer. If the message is NOT deleted or archived within the visibility timeout, it will become visible again and can be read by another consumer. The visibility timeout is set when a message is read from the queue, via pgmq.read(). It is recommended to set a vt value that is greater than the expected time it takes to process a message. After the application successfully processes the message, it should call pgmq.delete() to completely remove the message from the queue or pgmq.archive() to move it to the archive table for the queue.