Serverless Laravel applications

This guide helps you run Laravel applications on AWS Lambda using Bref. These instructions are kept up to date to target the latest Laravel version.

A demo application is available on GitHub at


First, make sure you have followed the Installation guide to create an AWS account and install the necessary tools.

Next, in an existing Laravel project, install Bref and the Laravel-Bref package.

composer require bref/bref bref/laravel-bridge --update-with-dependencies

Then let's create a serverless.yml configuration file:

php artisan vendor:publish --tag=serverless-config

How it works

By default, the Laravel-Bref package will automatically configure Laravel to work on AWS Lambda.

If you are curious, the package will automatically:


We do not want to deploy "dev" caches that were generated on our machine (because paths will be different on AWS Lambda). Let's clear them before deploying:

php artisan config:clear

When running in AWS Lambda, the Laravel application will automatically cache its configuration when booting. You don't need to run php artisan config:cache before deploying.

Let's deploy now:

serverless deploy

When finished, the deploy command will show the URL of the application.

Deploying for production

At the moment, we deployed our local installation to Lambda. When deploying for production, we probably don't want to deploy:

  • development dependencies,
  • our local .env file,
  • or any other dev artifact.

Follow the deployment guide for more details.


In case your application is showing a blank page after being deployed, have a look at the logs.

Laravel Artisan

As you may have noticed, we define a function named "artisan" in serverless.yml. That function is using the Console runtime, which lets us run Laravel Artisan on AWS Lambda.

For example, to execute an artisan command on Lambda for the above configuration, run the below command.

serverless bref:cli --args="<artisan command and options>"

For more details follow the "Console" guide.


To deploy Laravel websites, assets need to be served from AWS S3. The easiest approach is to use the Server-side website construct of the Lift plugin.

This will deploy a Cloudfront distribution that will act as a proxy: it will serve static files directly from S3 and will forward everything else to Lambda. This is very close to how traditional web servers like Apache or Nginx work, which means your application doesn't need to change! For more details, read the official documentation.

First install the plugin:

serverless plugin install -n serverless-lift

Then add this configuration to your serverless.yml file:

service: laravel
    # ...

    # ...

    - ./vendor/bref/bref
    - serverless-lift

        type: server-side-website
            '/js/*': public/js
            '/css/*': public/css
            '/favicon.ico': public/favicon.ico
            '/robots.txt': public/robots.txt
            # add here any file or directory that needs to be served from S3

Before deploying, compile your assets:

npm run prod

Now deploy your website using serverless deploy. Lift will create all required resources and take care of uploading your assets to S3 automatically.

For more details, see the Websites section of this documentation and the official Lift documentation.

Assets in templates

Assets referenced in templates should be via the asset() helper:

<script src="{{ asset('js/app.js') }}"></script>

If your templates reference some assets via direct path, you should edit them to use the asset() helper:

- <img src="/images/logo.png">
+ <img src="{{ asset('images/logo.png') }}">

File storage on S3

Laravel has a filesystem abstraction that lets us easily change where files are stored. When running on Lambda, you will need to use the s3 adapter to store files on AWS S3.

To do this, set FILESYSTEM_DISK: s3 either in serverless.yml or your production .env file. We can also create an S3 bucket via serverless.yml directly:

# ...
    # ...
        # environment variable for Laravel
        AWS_BUCKET: !Ref Storage
                # Allow Lambda to read and write files in the S3 buckets
                -   Effect: Allow
                    Action: s3:*
                        - !Sub '${Storage.Arn}' # the storage bucket
                        - !Sub '${Storage.Arn}/*' # and everything inside

        # Create our S3 storage bucket using CloudFormation
            Type: AWS::S3::Bucket

That's it! The AWS credentials (AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY and AWS_SESSION_TOKEN) are set automatically in AWS Lambda, you don't have to define them.

Public files

Laravel has a special disk called public: this disk stores files that we want to make public, like uploaded photos, generated PDF files, etc.

Again, those files cannot be stored on Lambda, i.e. they cannot be stored in the default storage/app/public directory. You need to store those files on S3.

Do not run php artisan storage:link in AWS Lambda: it is now useless, and it will fail because the filesystem is read-only in Lambda.

To store public files on S3, you could replace the disk in the code:

- Storage::disk('public')->put('avatars/1', $fileContents);
+ Storage::disk('s3')->put('avatars/1', $fileContents);

but doing this will not let your application work locally. A better solution, but more complex, involves making the public disk configurable. Let's change the configuration in config/filesystems.php:

    | Default Public Filesystem Disk

+   'public' => env('FILESYSTEM_DISK', 'public_local'),


    'disks' => [

        'local' => [
            'driver' => 'local',
            'root' => storage_path('app'),

-        'public' => [
+        'public_local' => [
            'driver' => 'local',
            'root' => storage_path('app/public'),
            'url' => env('APP_URL').'/storage',
            'visibility' => 'public',

        's3' => [
            'driver' => 's3',
            'key' => env('AWS_ACCESS_KEY_ID'),
            'secret' => env('AWS_SECRET_ACCESS_KEY'),
            'token' => env('AWS_SESSION_TOKEN'),
            'region' => env('AWS_DEFAULT_REGION'),
            'bucket' => env('AWS_BUCKET'),
            'url' => env('AWS_URL'),

+        's3_public' => [
+            'driver' => 's3',
+            'key' => env('AWS_ACCESS_KEY_ID'),
+            'secret' => env('AWS_SECRET_ACCESS_KEY'),
+            'token' => env('AWS_SESSION_TOKEN'),
+            'region' => env('AWS_DEFAULT_REGION'),
+            'bucket' => env('AWS_PUBLIC_BUCKET'),
+            'url' => env('AWS_URL'),
+        ],


You can now configure the public disk to use S3 by changing serverless.yml or your production .env:


Laravel Queues

To run Laravel Queues on AWS Lambda using Amazon SQS, we don't want to run the php artisan queue:work command. Instead, we create a function that is invoked immediately when there are new jobs to process.

To create the SQS queue (and the permissions for the Lambda functions to read/write to it), we can either do that manually, or use serverless.yml.

To make things simpler, we will use the Serverless Lift plugin to create and configure the SQS queue.

First install the Lift plugin:

serverless plugin install -n serverless-lift

Then use the Queue construct in serverless.yml:

    # ...
        # ...
        SQS_QUEUE: ${construct:jobs.queueUrl}

    # ...

        type: queue
            handler: Bref\LaravelBridge\Queue\QueueHandler
            runtime: php-81
            timeout: 60 # seconds

We define Laravel environment variables in provider.environment (this could also be done in the deployed .env file):

  • QUEUE_CONNECTION: sqs enables the SQS queue connection
  • SQS_QUEUE: ${construct:jobs.queueUrl} passes the URL of the created SQS queue

If you want to create the SQS queue manually, you will need to set these variables. AWS credentials (AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY) are automatically set up with the appropriate permissions for Laravel to use the SQS queue.

That's it! Anytime a job is pushed to Laravel Queues, it will be sent to SQS, and SQS will invoke our "worker" function so that it is processed.


In the example above, we set the full SQS queue URL in the SQS_QUEUE variable.

If you only set the queue name (which is also valid), you need to set the SQS_PREFIX environment variable too. For example: SQS_PREFIX: "https://sqs.${aws:region}${aws:accountId}".

How it works

When integrated with AWS Lambda, SQS has a built-in retry mechanism and storage for failed messages. These features work slightly differently than Laravel Queues. The "Bref for Laravel" integration does not use these SQS features.

Instead, "Bref for Laravel" makes all the feature of Laravel Queues work out of the box, just like on any server. Read more in the Laravel Queues documentation.

Note: the "Bref-Laravel bridge" v1 used to do the opposite. We changed that behavior in Bref v2 in order to make the experience smoother for Laravel users.

Laravel Octane

To run the HTTP application with Laravel Octane instead of PHP-FPM, change the following options in the web function:

        handler: Bref\LaravelBridge\Http\OctaneHandler
        runtime: php-81
            BREF_LOOP_MAX: 250
        # ...

Keep the following details in mind:

  • Laravel Octane does not need Swoole or RoadRunner on AWS Lambda, so it is not possible to use Swoole-specific features.
  • Octane keeps Laravel booted in a long-running process, beware of memory leaks.
  • BREF_LOOP_MAX specifies the number of HTTP requests handled before the PHP process is restarted (and the memory is cleared).

Persistent database connections

You can keep database connections persistent across requests to make your application even faster. To do so, set the OCTANE_PERSIST_DATABASE_SESSIONS environment variable:

      handler: Bref\LaravelBridge\Http\OctaneHandler
      runtime: php-81
          BREF_LOOP_MAX: 250
        # ...

Note that if you are using PostgreSQL (9.6 or newer), you need to set idle_in_transaction_session_timeout either in your RDS database's parameter group, or on a specific database itself.

ALTER DATABASE SET idle_in_transaction_session_timeout = '10000' -- 10 seconds in ms


By default, the Bref bridge will move Laravel's storage and cache directories to /tmp. This is because all the filesystem except /tmp is read-only.

Note that the /tmp directory isn't shared across Lambda instances. If you Lambda function scales up, the cache will be empty in new instances (or after a deployment).

If you want the cache to be shared across all Lambda instances, for example if your application caches a lot of data or if you use it for locking mechanisms (like API rate limiting), you can instead use Redis or DynamoDB.

DynamoDB is the easiest to set up and is "pay per use". Redis is a bit more complex as it requires a VPC and managing instances, but offers slightly faster response times.

Using DynamoDB

To use DynamoDB as a cache store, change this configuration in config/cache.php:

    # config/cache.php
    'dynamodb' => [
        'driver' => 'dynamodb',
        'key' => env('AWS_ACCESS_KEY_ID'),
        'secret' => env('AWS_SECRET_ACCESS_KEY'),
        'region' => env('AWS_DEFAULT_REGION', 'us-east-1'),
        'table' => env('DYNAMODB_CACHE_TABLE', 'cache'),
        'endpoint' => env('DYNAMODB_ENDPOINT'),
+       'attributes' => [
+           'key' => 'id',
+           'expiration' => 'ttl',
+       ]

Then follow this section of the documentation to deploy your DynamoDB table using the Serverless Framework.

Maintenance mode

Similar to the php artisan down command, you may put your app into maintenance mode. All that's required is setting the MAINTENANCE_MODE environment variable:

        MAINTENANCE_MODE: ${param:maintenance, null}

You can then deploy:

# Full deployment (goes through CloudFormation):
serverless deploy --param="maintenance=1"

# Or quick update of the functions config only:
serverless deploy function --function=web --update-config --param="maintenance=1"
serverless deploy function --function=artisan --update-config --param="maintenance=1"
serverless deploy function --function=<function-name> --update-config --param="maintenance=1"

To take your app out of maintenance mode, redeploy without the --param="maintenance=1" option.

Laravel Passport

Laravel Passport has a passport:install command. However, this command cannot be run in Lambda because it needs to write files to the storage/ directory.

Instead, here is what you need to do:

  • Run php artisan passport:keys locally to generate key files.

    This command will generate the storage/oauth-private.key and storage/oauth-public.key files, which need to be deployed.

    Depending on how you deploy your application (from your machine, or from CI), you may want to whitelist them in serverless.yml:

              - ...
              # Exclude the 'storage' directory
              - '!storage/**'
              # Except the public and private keys required by Laravel Passport
              - 'storage/oauth-private.key'
              - 'storage/oauth-public.key'
  • You can now deploy the application:

    serverless deploy
  • Finally, you can create the tokens (which is the second part of the passport:install command):

    serverless bref:cli --args="passport:client --personal --name 'Laravel Personal Access Client'"
    serverless bref:cli --args="passport:client --password --name 'Laravel Personal Access Client'"

All these steps were replacements of running the passport:install command from the Passport documentation.