Semantic Logger Cheatsheet

semantic_logger_cheatsheet.rb

# Use blocks to prevent unnecessary evaluation
logger.debug do 
  count = results.inject(0) {|sum, i| i+sum }  "A total of #{count} were processed" 
end 

# Measure the duration of any block of code 
count = logger.benchmark_info('Counting users') do  
  User.where('created_at <= ?', date).count
end 


# Only log when the block takes longer than 50ms 
value = logger.benchmark_warning('Memcache call was slow',  min_duration: 50) do  
  memcache.get('key') 
end


# Add tags to any block of code
logger.tagged('127.0.0.1', 'jbloggs') do  
  logger.info 'Started processing normally'  
  logger.debug 'Calling supplier'  
  logger.trace 'Received Response', response 
end


# Add the ip-address and a tracking number to every Rails message:
config.log_tags = [
  :remote_ip, 
  lambda do |request| 
    request.headers["Tracking-Number"] = random_tracking_number
  end 
] 


# Log low level messages as trace 
logger.trace('Received', raw_response)


# Semantic hashes
logger.error( 
  "Oops external call failed",  
  result: :failed,  
  reason_code: -10 
)


# Threads
# Name different threads
Thread.current.name = 'Main' 
logger = SemanticLogger['Job'] 
logger.info "Handing work to processing thread"   
Thread.new do 
   Thread.current.name = 'Processor'  
   logger = SemanticLogger['User']  
   logger.info "Started processing" 
end


# Log rescued exceptions
begin 
  google.search('semantic logger')  
rescue StandardError => exc 
  logger.error 'Failed calling supplier', exc 
  'No data available' 
end 

# Metrics

# Increment google search count metric
logger.info( 
  message: 'Called Google',   
  metric: 'google/search' 
)

count = logger.benchmark_info('Counting users’,
  metric: ‘user/count’) do
  User.where('created_at <= ?', date).count
end