require 'sidekiq'

require '../../lib/datadog/statsd'

# How to test these examples:

# - Run a redis server on port 6379:

# $ docker run --rm --name redis -d -p6379:6379 redis

# - Start sidekiq with this file:

# $ sidekiq -d -r ./sidekiq_example.rb

# - In another shell, perform the jobs using:

# $ echo "ExampleWorkerInstance.perform_async"| irb -r ./sidekiq_example.rb

# or

# $ echo "ExampleEphemeralInstance.perform_async"| irb -r ./sidekiq_example.rb

# This Sidekiq worker uses a DogStatsD instance created every time this worker job

# is executed. Because of that, it is important to close this created client to

# free the resources it is using (a socket and a thread).

# Closing the client will also invoke "flush(sync: true)" to ensure metrics are

# flushed even if the client instance internal buffer is not full.

class ExampleEphemeralInstance

include Sidekiq::Worker

def perform()

client = Datadog::Statsd.new('localhost', 8125)

client.increment('example_metric.sample', tags: ['environment:dev'])

# flush all metrics created during the job execution and free resources used by

# this ephemeral dogstatsd client instance:

client.close()

puts("Metrics flushed and client closed")

end

end

# This Sidekiq worker is using a single global instance created for the worker.

# It is instantiated when the worker is instantiated. It is important to flush the

# metrics at the end of the job execution in order to push them to the Datadog intake.

class ExampleWorkerInstance

include Sidekiq::Worker

STATSD_CLIENT = Datadog::Statsd.new('localhost', 8125)

def perform()

STATSD_CLIENT.increment('example_metric.sample', tags: ['environment:dev'])

STATSD_CLIENT.flush(sync: true) # flush all metrics created during the job execution

puts("Metrics flushed")

end

end