72 releases
0.29.3 | Nov 13, 2024 |
---|---|
0.28.2 | Jun 13, 2024 |
0.26.1 | Jan 29, 2024 |
0.25.0 | Oct 24, 2023 |
0.9.0 | Mar 29, 2023 |
#60 in Database interfaces
856 downloads per month
Used in pgx_pgmq
82KB
1K
SLoC
Postgres Message Queue (PGMQ)
PGMQ is a lightweight, distributed message queue. It's like AWS SQS and RSMQ but native to Postgres.
Message queues allow you to decouple and connect microservices. Send, store, and receive messages between components scalably, without dropping messages or needing other services to be available.
PGMQ was created by Tembo. Our goal is to make the full Postgres ecosystem accessible to everyone. We're building a radically simplified Postgres platform designed to be developer-first and easily extensible. PGMQ is a part of that project.
This project contains two APIs, a pure Rust client side library and the Rust SDK wrapped around the Postgres extension.
The Rust client for the Postgres extension
. This gives the you the an ORM-like experience with the Postgres extension and makes managing connection pools, transactions, and serialization/deserialization much easier.
use pgmq::PGMQueueExt;
The pure Rust client
. This provides minimal functionality but can be used on any existing Postgres instance.
use pgmq::PGMQueue;
Quick start
- First, you will need Postgres. We use a container in this example.
docker run -d --name pgmq-postgres -e POSTGRES_PASSWORD=postgres -p 5432:5432 quay.io/tembo/pg16-pgmq:latest
-
If you don't have Docker installed, it can be found here.
-
Make sure you have the Rust toolchain installed:
-
Clone the project and run the basic example:
git clone https://github.com/tembo-io/pgmq.git
cd pgmq-rs
cargo run --example basic
Minimal example at a glance
use pgmq::{errors::PgmqError, Message, PGMQueue};
use serde::{Deserialize, Serialize};
use serde_json::Value;
#[tokio::main]
async fn main() -> Result<(), PgmqError> {
// Initialize a connection to Postgres
println!("Connecting to Postgres");
let queue: PGMQueue = PGMQueue::new("postgres://postgres:postgres@0.0.0.0:5432".to_owned())
.await
.expect("Failed to connect to postgres");
// Create a queue
println!("Creating a queue 'my_queue'");
let my_queue = "my_example_queue".to_owned();
queue.create(&my_queue)
.await
.expect("Failed to create queue");
// Structure a message
#[derive(Serialize, Debug, Deserialize)]
struct MyMessage {
foo: String,
}
let message = MyMessage {
foo: "bar".to_owned(),
};
// Send the message
let message_id: i64 = queue
.send(&my_queue, &message)
.await
.expect("Failed to enqueue message");
// Use a visibility timeout of 30 seconds
// Once read, the message will be unable to be read
// until the visibility timeout expires
let visibility_timeout_seconds: i32 = 30;
// Read a message
let received_message: Message<MyMessage> = queue
.read::<MyMessage>(&my_queue, Some(&visibility_timeout_seconds))
.await
.unwrap()
.expect("No messages in the queue");
println!("Received a message: {:?}", received_message);
assert_eq!(received_message.msg_id, message_id);
// archive the messages
let _ = queue.archive(&my_queue, &received_message.msg_id)
.await
.expect("Failed to archive message");
println!("archived the messages from the queue");
Ok(())
}
Transactions
You can execute all of PGMQ's operations within a transaction along with other database operations. See the transaction example or run the example with:
cargo run --example transactions
Sending messages
You can send one message at a time with queue.send()
or several with queue.send_batch()
.
These methods can be passed any type that implements serde::Serialize
. This means you can prepare your messages as JSON or as a struct.
Reading messages
Reading a message will make it invisible (unavailable for consumption) for the duration of the visibility timeout (vt). No messages are returned when the queue is empty or all messages are invisible.
Messages can be parsed as serde_json::Value or into a struct. queue.read()
returns an Result<Option<Message<T>>, PgmqError>
where T
is the type of the message on the queue. It returns an error when there is an issue parsing the message (PgmqError::JsonParsingError
) or if PGMQ is unable to reach postgres (PgmqError::DatabaseError
).
Note that when parsing into a struct
(say, you expect MyMessage{foo: "bar"}
), and the data of the message does not correspond to the struct definition (e.g. {"hello": "world"}
), an error will be returned and unwrapping this result the way it is done for demo purposes in the example above will cause a panic, so you will rather want to handle this case properly.
Read a single message with queue.read()
or as many as you want with queue.read_batch()
.
Archive or Delete a message
Remove the message from the queue when you are done with it. You can either completely .delete()
, or .archive()
the message. Archived messages are deleted from the queue and inserted to the queue's archive table. Deleted messages are just deleted.
Read messages from the queue archive with SQL:
SELECT *
FROM pgmq.a_{your_queue_name};
License: PostgreSQL
Dependencies
~39–52MB
~898K SLoC