[![Join the chat at https://gitter.im/arnaud-lb/php-rdkafka](https://badges.gitter.im/arnaud-lb/php-rdkafka.svg)](https://gitter.im/arnaud-lb/php-rdkafka?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
PHP-rdkafka is a **stable**, **production-ready**, **long term support**, and **fast** Kafka client for PHP based on [librdkafka](https://github.com/edenhill/librdkafka).
It supports PHP 7, PHP 8, PHP 5 (in older versions), all librdkafka versions since 0.11, all Kafka versions since 0.8. This makes it **easy to deploy** the extension in production.
The goal of the extension is to be a low-level un-opinionated librdkafka binding focused on production and long term support.
The high level and low level *consumers*, *producer*, and *metadata* APIs are supported.
Documentation is available [here](https://arnaud-lb.github.io/php-rdkafka/phpdoc/book.rdkafka.html).
<li>True Serverless Kafka with per-request-pricing</li>
<li>Managed Apache Kafka, works with all Kafka clients</li>
<li>Built-in REST API designed for serverless and edge functions</li>
</ul>
[Start for free in 30 seconds!](https://upstash.com/?utm_source=php-rdkafka)
</td>
</tr>
</table>
php-rdkafka supports Ukraine. Proceeds from our generous sponsors are currently donated to the [Support Ukraine collective](https://opencollective.com/support-ukraine).
Configuration parameters used below can be found in [Librdkafka Configuration reference](https://github.com/edenhill/librdkafka/blob/master/CONFIGURATION.md)
### Producing
#### Creating a producer
For producing, we first need to create a producer, and to add brokers (Kafka
servers) to it:
``` php
<?php
$conf=newRdKafka\Conf();
$conf->set('log_level',(string)LOG_DEBUG);
$conf->set('debug','all');
$rk=newRdKafka\Producer($conf);
$rk->addBrokers("10.0.0.1:9092,10.0.0.2:9092");
```
#### Producing messages
> **Warning** Make sure that your producer follows proper shutdown (see below) to not lose messages.
Next, we create a topic instance from the producer:
``` php
<?php
$topic=$rk->newTopic("test");
```
From there, we can produce as much messages as we want, using the produce
The first argument is the partition. RD_KAFKA_PARTITION_UA stands for
*unassigned*, and lets librdkafka choose the partition.
The second argument are message flags and should be either 0
or `RD_KAFKA_MSG_F_BLOCK` to block produce on full queue.
The message payload can be anything.
#### Proper shutdown
This should be done prior to destroying a producer instance
to make sure all queued and in-flight produce requests are completed
before terminating. Use a reasonable value for `$timeout_ms`.
> **Warning** Not calling flush can lead to message loss!
```php
$rk->flush($timeout_ms);
```
In case you don't care about sending messages that haven't been sent yet,
you can use `purge()` before calling `flush()`:
```php
// Forget messages that are not fully sent yet
$rk->purge(RD_KAFKA_PURGE_F_QUEUE);
$rk->flush($timeout_ms);
```
### High-level consuming
The RdKafka\KafkaConsumer class supports automatic partition assignment/revocation. See the example [here](https://arnaud-lb.github.io/php-rdkafka-doc/phpdoc/rdkafka.examples.html#example-1).
### Low-level consuming (legacy)
> **Note** The low-level consumer is a legacy API, please prefer using the high-level consumer
We first need to create a low level consumer, and to add brokers (Kafka
servers) to it:
``` php
<?php
$conf=newRdKafka\Conf();
$conf->set('log_level',(string)LOG_DEBUG);
$conf->set('debug','all');
$rk=newRdKafka\Consumer($conf);
$rk->addBrokers("10.0.0.1,10.0.0.2");
```
Next, create a topic instance by calling the `newTopic()` method, and start
consuming on partition 0:
``` php
<?php
$topic=$rk->newTopic("test");
// The first argument is the partition to consume from.
// The second argument is the offset at which to start consumption. Valid values
librdkafka will buffer up to 1GB of messages for each consumed partition by default. You can lower memory usage by reducing the value of the ``queued.max.messages.kbytes`` parameter on your consumers.
### topic.metadata.refresh.sparse and topic.metadata.refresh.interval.ms
Each consumer and producer instance will fetch topics metadata at an interval defined by the ``topic.metadata.refresh.interval.ms`` parameter. Depending on your librdkafka version, the parameter defaults to 10 seconds, or 600 seconds.
librdkafka fetches the metadata for all topics of the cluster by default. Setting ``topic.metadata.refresh.sparse`` to the string ``"true"`` makes sure that librdkafka fetches only the topics he uses.
Setting ``topic.metadata.refresh.sparse`` to ``"true"``, and ``topic.metadata.refresh.interval.ms`` to 600 seconds (plus some jitter) can reduce the bandwidth a lot, depending on the number of consumers and topics.
### internal.termination.signal
This setting allows librdkafka threads to terminate as soon as librdkafka is done with them. This effectively allows your PHP processes / requests to terminate quickly.
When enabling this, you have to mask the signal like this:
``` php
<?php
// once
pcntl_sigprocmask(SIG_BLOCK,array(SIGIO));
// any time
$conf->set('internal.termination.signal',SIGIO);
```
### socket.blocking.max.ms (librdkafka < 1.0.0)
> Maximum time a broker socket operation may block. A lower value improves responsiveness at the expense of slightly higher CPU usage.
Reducing the value of this setting improves shutdown speed. The value defines the maximum time librdkafka will block in one iteration of a read loop. This also defines how often the main librdkafka thread will check for termination.
### queue.buffering.max.ms
This defines the maximum and default time librdkafka will wait before sending a batch of messages. Reducing this setting to e.g. 1ms ensures that messages are sent ASAP, instead of being batched.
This has been seen to reduce the shutdown time of the rdkafka instance, and of the PHP process / request.
## Performance / Low-latency settings
Here is a configuration optimized for low latency. This allows a PHP process / request to send messages ASAP and to terminate quickly.
``` php
<?php
$conf=new\RdKafka\Conf();
$conf->set('socket.timeout.ms',50);// or socket.blocking.max.ms, depending on librdkafka version
if(function_exists('pcntl_sigprocmask')){
pcntl_sigprocmask(SIG_BLOCK,array(SIGIO));
$conf->set('internal.termination.signal',SIGIO);
}else{
$conf->set('queue.buffering.max.ms',1);
}
$producer=new\RdKafka\Producer($conf);
$consumer=new\RdKafka\Consumer($conf);
```
It is advised to call poll at regular intervals to serve callbacks. In `php-rdkafka:3.x`
poll was also called during shutdown, so not calling it in regular intervals might
lead to a slightly longer shutdown. The example below polls until there are no more events in the queue:
The source of the documentation can be found [here](https://github.com/arnaud-lb/php-rdkafka-doc)
## Asking for Help
If the documentation is not enough, feel free to ask a questions on the php-rdkafka channels on [Gitter](https://gitter.im/arnaud-lb/php-rdkafka) or [Google Groups](https://groups.google.com/forum/#!forum/php-rdkafka).
## Stubs
Because your IDE is not able to auto discover php-rdkadka api you can consider usage of external package providing a set of stubs for php-rdkafka classes, functions and constants: [kwn/php-rdkafka-stubs](https://github.com/kwn/php-rdkafka-stubs)
## Contributing
If you would like to contribute, thank you :)
Before you start, please take a look at the [CONTRIBUTING document](https://github.com/arnaud-lb/php-rdkafka/blob/master/CONTRIBUTING.md) to see how to get your changes merged in.
## Credits
Documentation copied from [librdkafka](https://github.com/edenhill/librdkafka).
Authors: see [contributors](https://github.com/arnaud-lb/php-rdkafka/graphs/contributors).
## License
php-rdkafka is released under the [MIT](https://github.com/arnaud-lb/php-rdkafka/blob/master/LICENSE) license.