Serverless Local Schedule

This plugin allows you to specify a timezone on your lambdas and step functions triggered by AWS CloudWatch Events.

Install

npm i -D @distinction-dev/serverless-local-schedule

or

yarn add -D @distinction-dev/serverless-local-schedule

Add the plugin to your serverless.yml file

plugins:
  - "@distinction-dev/serverless-local-schedule"

How it works

When you specify a schedule as an event in your serverless config, this plugin generates 6 different crons to deal with DST changes during that year like so:-

functions:
  hello:
    handler: handler.hello
    events:
      - schedule:
          rate: cron(0 10 * * ? *)
          timezone: America/New_York

functions:
  hello:
    handler: handler.hello
    events:
      - schedule:
          rate: cron(0 15 * 1-2,12 ? *) # full non-DST months
      - schedule:
          rate: cron(0 15 1-10 3 ? *) # non-DST portion of March
      - schedule:
          rate: cron(0 14 11-31 3 ? *) # DST portion of March
      - schedule:
          rate: cron(0 14 * 4-10 ? *) # full DST months
      - schedule:
          rate: cron(0 14 1-3 11 ? *) # DST portion of November
      - schedule:
          rate: cron(0 15 4-31 11 ? *) # non-DST portion of November

Usage

Lambda Functions

functions:
  hello:
    handler: handler.hello
    events:
      - schedule:
          rate: cron(0 10 * * ? *)
          timezone: America/New_York

It works by converting that into 6 different schedules, effectively the same as having the following configuration:

functions:
  hello:
    handler: handler.hello
    events:
      - schedule:
          rate: cron(0 15 * 1-2,12 ? *) # full non-DST months
      - schedule:
          rate: cron(0 15 1-10 3 ? *) # non-DST portion of March
      - schedule:
          rate: cron(0 14 11-31 3 ? *) # DST portion of March
      - schedule:
          rate: cron(0 14 * 4-10 ? *) # full DST months
      - schedule:
          rate: cron(0 14 1-3 11 ? *) # DST portion of November
      - schedule:
          rate: cron(0 15 4-31 11 ? *) # non-DST portion of November

Step Functions

stepFunctions:
  stateMachines:
    helloStepFunc:
      events:
        - schedule:
            rate: cron(0 10 * * ? *)
            timezone: America/New_York

It works by converting that into 6 different schedules, effectively the same as having the following configuration:

stepFunctions:
  stateMachines:
    helloStepFunc:
      events:
        - schedule:
            rate: cron(0 15 * 1-2,12 ? *) # full non-DST months
        - schedule:
            rate: cron(0 15 1-10 3 ? *) # non-DST portion of March
        - schedule:
            rate: cron(0 14 11-31 3 ? *) # DST portion of March
        - schedule:
            rate: cron(0 14 * 4-10 ? *) # full DST months
        - schedule:
            rate: cron(0 14 1-3 11 ? *) # DST portion of November
        - schedule:
            rate: cron(0 15 4-31 11 ? *) # non-DST portion of November

Extra points to consider

• To see what crontabs you'll get for your timezone, try https://distinction-dev.github.io/local-crontab/

• The - schedule: cron(* * * * ? *) short syntax isn't supported.

• Unfortunately you cannot specify day of the week in the cron expression i.e. cron(0 7 ? * MON-FRI *). This is because to support the split months (March & November in the US), the plugin has to specify a day of month (EG: November 1-3 in 2018), so you cannot specify a DOW other than ? unfortunately. Recommended workaround for this is to move the day of week check into your code so it's just a no-op on non weekdays for instance.

Debugging

A launch.json file for vscode is provided which will run sls package in the test service.

It will be useful if you would like to test and see how the plugin works or debug with your crontab or event configuration

History

Originally developed by Capital One, now maintained by Distinction Dev

This was maintained by Serverless Inc for some time and then forked over by Distinction Dev

Capital One considers itself the bank a technology company would build. It's delivering best-in-class innovation so that its millions of customers can manage their finances with ease. Capital One is all-in on the cloud and is a leader in the adoption of open source, RESTful APIs, microservices and containers. We build our own products and release them with a speed and agility that allows us to get new customer experiences to market quickly. Our engineers use artificial intelligence and machine learning to transform real-time data, software and algorithms into the future of finance, reimagined.