Skip to content

Latest commit

 

History

History
157 lines (141 loc) · 4.73 KB

logging.md

File metadata and controls

157 lines (141 loc) · 4.73 KB

#Logging By default µServerless uses logback to configure logger behavior. as of version 0.0.2 it structures log messages as JSON using the logstash-logback-encoder. In addition to the default logged values

An log.info like log.info("This is my message") Will result in the following log message:

*** INFO: {
  "@timestamp": "2018-11-11T16:52:47.502-08:00",
  "@version": "1",
  "message": "This is my message",
  "logger_name": "com.example.App.Function",
  "thread_name": "ScalaTest-run-running-LambdaHandler",
  "level": "INFO",
  "level_value": 20000,
  // Custom values added to each message by the default configuration
  "type": "log",
  "stage": "test",
  "function": "lambda-function-name",
  "lambda_version": "$LATEST"
}

Adding custom properties to the log message

You may want to attach additional information to your log message in oder to be able to find it better if you forward these emails to a custom db like elastic search. To add custom properties to include the use the keyValue from the logstash StructuredArguments:

import net.logstash.logback.argument.StructuredArguments._
...

log.info("This is my message", keyValue("Foo", "bar"), keyValue("BAZ", "BLAH"))

This will result in the following message:

*** INFO: {
  "@timestamp": "2018-11-11T16:52:47.502-08:00",
  "@version": "1",
  "message": "This is my message",
  "logger_name": "com.example.App.Function",
  "thread_name": "ScalaTest-run-running-LambdaHandler",
  "level": "INFO",
  "level_value": 20000,
  
  // Values added to each message by the default configuration
  "type": "log",
  "stage": "test",
  "function": "lambda-function-name",
  "lambda_version": "$LATEST",
  
  // Custom values
  "Foo": "bar",
  "BAZ": "BLAH"
}

Error logging:

Error logs are handled by a custom log appender. This is to enable µServerless to build a custom error payload when exception occur that containing the stack trace.

*** ERROR : {
  "message": "Origin 'http://baz.com' is not authorized",
  "exceptionClass": "io.onema.userverless.exception.HandleRequestException",
  "stackTrace": [
    {
      "fileName": "ApiGatewayHandler.scala",
      "lineNumber": 95,
      "className": "io.onema.userverless.function.ApiGatewayHandler$Cors",
      "methodName": "cors()"
    },
    {
      "fileName": "ApiGatewayHandler.scala",
      "lineNumber": 83,
      "className": "io.onema.userverless.function.ApiGatewayHandler$Cors",
      "methodName": "cors$()"
    },
    ...
  ],
  "@timestamp": "2018-11-11T16:52:47.781-0800",
  "type": "exception",
  "stage": "test",
  "function": "lambda-function-name"",
  "lambdaVersion": "$LATEST""
}

If you use the error method directly, you will only see a message with the following format:

*** ERROR: %msg%n

Metrics

Metrics is a way to log a Statsd compatible message. There are two types of metrics available: timer and count.

Timers are available as a code-block (thunk):

import io.onema.userverless.monitoring.LogMetrics.time

time("NameOfMyEvent") {
  val thisIsMyCode = "Foo"
}

This will result in the following message:

*** METRIC: {
  "@timestamp": "2018-11-11T16:52:47.748-08:00",
  "@version": "1",
  "message": "NameOfMyEvent:0|ms|@1|#stage:test",
  "logger_name": "io.onema.userverless.monitoring.Metrics",
  "thread_name": "ScalaTest-run-running-ApiGatewayHandlerWithCorsTest",
  "level": "INFO",
  "level_value": 20000,
  "metric_name": "NameOfMyEvent",
  "metric_type": "time",
  "type": "metric",
  "stage": "test",
  "function": "lambda-function-name"",
  "lambda_version": "$LATEST""
}

A count can be added like such:

import io.onema.userverless.monitoring.LogMetrics.count

count("NameOfCountEvent")

This will result in a log containing the following message: NameOfMyEvent:1|c|@1|#stage:test"

NOTE:

The metrics are only logged and are not submitted to CloudWatch. To get the metrics reported to CloudWatch you need to use the Overwatch app

Adding custom tags to your metrics

To submit custom tags in your metrics, pass key value pairs to the functions:

count("CounterName", ("MyKey", "MyValue"), ("AnotherKey", "AnotherValue"))
time("TimerName", ("MyKey", "MyValue"), ("AnotherKey", "AnotherValue"))

Selecting the log level

By default the log level is set to DEBUG, you can overwrite this by setting the LOG_LEVEL environment variable to one of the log levels:

  • DEBUG
  • INFO
  • WARN
  • ERROR

Override log configuration

Simply add a src/main/resources/logback.xml with your custom configuration and add a custom merging strategy to your build.sbt:

assemblyMergeStrategy in assembly := {
  case "logback.xml" => MergeStrategy.first
  case x =>
    val oldStrategy = (assemblyMergeStrategy in assembly).value
    oldStrategy(x)
}