Sidekiq SQS Processor

A Ruby gem that integrates Amazon SQS with Sidekiq for efficient and scalable message processing. Process SQS messages asynchronously with the power and monitoring capabilities of Sidekiq.

Features

• Automatic SQS polling - Continuously polls SQS queues for new messages
• Parallel processing - Process messages across multiple Sidekiq workers
• SNS integration - Auto-unwraps SNS notifications
• Configurable polling strategies - Choose between continuous or scheduled polling
• Automatic retries - Leverage Sidekiq's retry mechanism with SQS visibility timeout
• Easy message routing - Route messages to different workers based on content
• Rails integration - Simple setup with generators for Rails applications
• Comprehensive monitoring - Track processing stats and worker health

Installation

Add this line to your application's Gemfile:

gem 'sidekiq-sqs-processor'

And then execute:

$ bundle install

Or install it yourself:

$ gem install sidekiq_sqs_processor

Dependencies

This gem requires the following dependencies in your application:

• Sidekiq (~> 7.0)
• aws-sdk-sqs (~> 1.0)
• Rails (>= 6.0)

Rails Setup

If you're using Rails, run the built-in generator to create the initializer and example worker:

$ rails generate sidekiq_sqs_processor:install

Basic Configuration

Rails Applications

After running the installer, edit the generated initializer at config/initializers/sidekiq_sqs_processor.rb:

SidekiqSqsProcessor.configure do |config|
  # AWS region for SQS queues
  config.aws_region = 'us-east-1'

  # SQS queue URLs to poll
  config.queue_urls = [
    'https://sqs.us-east-1.amazonaws.com/123456789012/my-queue'
  ]

  # Polling configuration
  config.polling_type = :continuous

  # Map queue URLs to worker classes
  config.queue_workers = {
    'https://sqs.us-east-1.amazonaws.com/123456789012/queue1' => 'Queue1Worker',
    'https://sqs.us-east-1.amazonaws.com/123456789012/queue2' => 'Queue2Worker'
  }

  # Optional: Set custom logger
  config.logger = Rails.logger

  # Optional: Set custom error handler
  config.error_handler = ->(error, context) do
    # Handle errors (e.g., report to monitoring service)
    Rails.logger.error("SQS Error: #{error.message}\nContext: #{context.inspect}")
  end
end

Non-Rails Applications

For non-Rails applications, add this to your initialization code:

require 'sidekiq_sqs_processor'

SidekiqSqsProcessor.configure do |config|
  # AWS region for SQS queues
  config.aws_region = 'us-east-1'

  # SQS queue URLs to poll
  config.queue_urls = [
    'https://sqs.us-east-1.amazonaws.com/123456789012/my-queue'
  ]

  # Start polling when Sidekiq starts
  Sidekiq.configure_server do |sidekiq_config|
    sidekiq_config.on(:startup) do
      SidekiqSqsProcessor.start_continuous_poller
    end

    sidekiq_config.on(:shutdown) do
      SidekiqSqsProcessor.stop_continuous_poller
    end
  end
end

Creating Workers

Using the Generator (Rails)

$ rails generate sidekiq_sqs_processor:worker UserNotification

This will create:

• app/workers/user_notification_worker.rb
• spec/workers/user_notification_worker_spec.rb (if using RSpec)
• or test/workers/user_notification_worker_test.rb (if using Test::Unit)

Manual Worker Creation

Create a class that inherits from SidekiqSqsProcessor::BaseWorker:

class MyWorker < SidekiqSqsProcessor::BaseWorker
  # Override the process_message method to handle your SQS messages
  def process_message(message_body)
    # message_body is already parsed from JSON if it was a JSON string

    # Your processing logic here
    User.find_by(id: message_body['user_id'])&.notify(message_body['message'])

    # Return a result (optional)
    { status: 'success' }
  end
end

Usage Examples

Basic Message Processing

class OrderProcessor < SidekiqSqsProcessor::BaseWorker
  def process_message(message_body)
    if message_body.is_a?(Hash) && message_body['order_id']
      # Process an order
      process_order(message_body['order_id'], message_body['items'])
    else
      logger.error("Invalid order message format: #{message_body.inspect}")
    end
  end

  private

  def process_order(order_id, items)
    # Your order processing logic
    Order.process(order_id, items)
  end
end

Handling Different Message Types

class EventProcessor < SidekiqSqsProcessor::BaseWorker
  def process_message(message_body)
    case message_body['event_type']
    when 'user_created'
      create_user(message_body['data'])
    when 'order_placed'
      process_order(message_body['data'])
    when 'payment_received'
      process_payment(message_body['data'])
    else
      logger.warn("Unknown event type: #{message_body['event_type']}")
    end
  end

  # ... processing methods ...
end

Handling SNS Messages

SNS messages are automatically detected and unwrapped, so the message_body will contain the inner message content:

class NotificationProcessor < SidekiqSqsProcessor::BaseWorker
  def process_message(message_body)
    # For an SNS message, the SNS envelope has been removed
    # and message_body is the parsed content of the SNS Message field

    logger.info("Processing notification: #{message_body.inspect}")
    # ... your processing logic ...
  end
end

Direct Message Enqueueing

You can also enqueue messages directly to a worker without going through SQS:

# Enqueue a message to a specific worker
SidekiqSqsProcessor.enqueue_message(
  MyWorker,
  { order_id: 123, items: ['item1', 'item2'] }
)

Configuration Options

AWS Configuration

config.aws_region = 'us-east-1'              # AWS region
config.aws_access_key_id = 'YOUR_KEY'        # Optional - uses env vars by default
config.aws_secret_access_key = 'YOUR_SECRET' # Optional - uses env vars by default

SQS Configuration

config.queue_urls = ['https://sqs.region.amazonaws.com/account/queue']
config.visibility_timeout = 300    # 5 minutes
config.wait_time_seconds = 20      # 20 seconds (long polling)
config.max_number_of_messages = 10 # Max messages per receive call

Polling Configuration

config.polling_type = :continuous  # :continuous or :scheduled
config.polling_frequency = 60      # Only used for scheduled (seconds)
config.polling_enabled = true      # Enable/disable polling
config.poll_on_startup = true      # Start polling when app starts

Sidekiq Configuration

config.worker_queue_name = 'sqs_default' # Default Sidekiq queue
config.worker_retry_count = 5            # Default retry count

Error Handling

# Custom error handler
config.error_handler = ->(error, context) do
  # Send to your error tracking service
  Sentry.capture_exception(error, extra: context)
end

Best Practices

Message Structure

For best routing and processing, use structured JSON messages:

{
  "type": "order_placed",
  "data": {
    "order_id": "12345",
    "customer_id": "67890",
    "items": [
      {"id": "item1", "quantity": 2},
      {"id": "item2", "quantity": 1}
    ]
  },
  "metadata": {
    "source": "web",
    "timestamp": "2025-04-21T10:15:30Z"
  }
}

Worker Organization

• Create separate workers for different message types or domains
• Use class inheritance for common processing logic
• Keep workers small and focused

Error Handling

• Use the error_handler config option for global error reporting
• Implement custom error handling in your workers for specific cases
• Let Sidekiq handle retries for transient failures

Visibility Timeout

• Set your SQS visibility timeout longer than your Sidekiq job timeout
• A good rule of thumb: visibility_timeout = (average_processing_time * 5) + max_retry_delay

Message Size Limits

Remember that SQS has a 256KB message size limit. For larger data:

• Store the data externally (e.g., S3) and include a reference in the message
• Split large datasets across multiple messages

Development

After checking out the repo, run bin/setup to install dependencies. Then, run rake spec to run the tests. You can also run bin/console for an interactive prompt that will allow you to experiment.

To install this gem onto your local machine, run bundle exec rake install.

About BigBinary

wheel is maintained by BigBinary. BigBinary is a software consultancy company. We build web and mobile applications using Ruby on Rails, React.js and React Native.