TasksMaster Bot

TasksMaster bot - simple serverless bot which you can use as template for you own bot

He contains base setup of enviroment and tools to start develop telegram bot on lambda functions with DynamoDB from zero (minimal) configuration

Contains:

• Node enviroment setup
• Typescript setup
• Webpack serverless setup - to compile typescript and pack for deploy node_modules
• Enviroment variables - setup for work with enviroment variables, you can setup their local and in produciton
• Encripted file variables - setup for store variables in encripted file
• Tests - you can develop you function by run test of functions locally
• DynamoDB - setup for star work with dynamodb
• Offline - allow you start you server local and develop without deploys (you can curl your functions)
• DynamoDB Local - setup for local instance of DynamoDB will allow you develop on local machine with database
• Telegraf - library for simplify work with Telegram API

Servless from scratch tutorial

This tutorial use make util (it available for linux and windows) to create one file which commands you can use, but if you prefer type commands by self or not want to use make you can read all command in Makefile. Format of file trying to be self decriptive and easy to understand for new commers.

Fast setup

Just go step by step and you will deploy you function to AWS

git clone https://github.com/LeoVS09/serverless-aws-node-bootstrap.git
cd ./serverless-aws-node-bootstrap

# Genereta enviroment template
make setup
# Write you secrets to '.env' file

# Then start docker container with installed dependencies
make console

# Write you project name, app, org into serverless.yml
# app and org you can find into you dashboard.serverless.com

Then create a bot from Telegram, sending this message to @BotFather

/newbot

Put received token to secrets.dev.yml

TELEGRAM_TOKEN: YOUT_TELEGRAM_BOT_TOKEN

Then deploy you function to AWS

make deploy
# will run serverless deploy

Configure webhook to telegram

make TOKEN={token} telegram-webhook {end-poinnt}
# curl --request POST --url https://api.telegram.org/bot{token}/setWebhook --header 'content-type: application/json' --data '{"url": "{end-poinnt}"}'

Setup local enviroment

This command will generate configuration enviroment .env file which will be used by docker and serveless

make setup

In docker isolated linux development

This bootstrap allow (but not require) develop all code inside docker container

This will achive you prevent problems:

• If you develop on windows, but want use linux enviromnt in production, you can develop in linux container
• If you not want install all required tool and packages in global enviroment you machine you can use predefinet container
• If multiple new team members will work on you package you not need to explain what need to install on their machines

Start docker container, sync all local files, get console inside

make console

Rebuild local docker image

make docker-build

Deploy you serverless functions

Commands to deploy you function

Note login to serverless when deploy first time Deploy to aws cloud by serverless

make deploy
# serverless deploy

Login to serverless

make login
# serverless login

Create new function

You can create new function by sls create command

This command will generate for you new handler file, add new function to serverless.yml config and add intial test

You can use predefined make command for it

make FN=newFunction HANDL=api/functionc/index create 
# sls create function -f newFunction --handler api/functionc/index

Serverless tips

You can deploy faster by update only codee and dependencies of individual function Example for hello function

make deploy-fn hello
# sls deploy function -f hello

Get logs of deployed hello function

make logs hello
# serverless logs -t -f hello

Invoke function in cloud and print their log

make invoke hello
# serverless invoke --log --function=hello

Invoke function locally and print logs

make local hello
# serverless invoke local --log --function=hello

Enviroment variables

You can setup you enviroment variables by set them in serverless.yml file

    enviroment:
      SECRET_FUNCTION_TOKEN: ${env:SECRET_FUNCTION_TOKEN}
      STAGE: ${self:provider.stage}

where ${env:} local enviroment variables which must be passed you can define them in AWS cloud aws docs

For local development we must paste them in KEY:VALUES pairs

serverless invoke local --function=hello --log -e SECRET_FUNCTION_TOKEN=VALUE OTHER_ENVIROMENT_VARIBLE=ANOTHER_VALUE

For setup it from file we maked dev.env file which will be readed by make command read-local-enviroment

So you can just run it by

make local-env hello
# . ./dev.env && serverless invoke local --log -e SECRET_FUNCTION_TOKEN="$$SECRET_FUNCTION_TOKEN" --function=hello

Encription

If you want to store you project in open source, you might be able to store secret tokens and keys. You store them in file secrets.dev.yml (for developement as example) and encrit to save it in version control (git)

make setup command was create for you example file secrets.dev.yml.

SECRET_SERVICE_KEY: SOME_SECRET_KEY

This file will be used to ush secret variables to enviroment variables of you function

custom:
  secrets: ${file(secrets.${opt:stage, self:provider.stage}.yml)}
# ...
enviroment:
    SECRET_SERVICE_KEY: ${self:custom.secrets.SECRET_SERVICE_KEY}

And when you will run serverless deploy it will push variables from file

If you can store and push variables for defferent stages by add stage parameter to deploy

# will read secrets from secrets.prod.yml file
make deploy --stage prod 
# serverless deploy --stage prod

Also you can encript file by password

# Will generate secrets.dev.yml.encripted
make encript-dev "Password"
# serverless encrypt --stage dev --password "Password"

And add `secrets.dev.yml.encripted file to git

After new checkout this file must be decripted for deploy ny command

# Will decript file secrets.dev.yml.encripted to secrets.dev.yml
make decript-dev "Password"
# serverless decrypt --stage dev --password "Password"

Tests

Test generation providet by serverless-mocha-plugin

Currently cannot have any way to use typescript with jest powered by serverless-jest-plugin or jest-environment-serverless. Use mocha until jest support will be realised to 1 version.

Before test any function which depend on DynamoDB install and start local inctance of DynamoDB

You can invoke test by command

make test
# sls webpack -o testBuild
# sls invoke test --root testBuild/service
#	rm -rf testBuild

or run tests to one function by

make test-fn hello 
# sls invoke test -f hello

also you can add test to existing function

make create-test hello
# sls create test -f hello

and when you create new handler it will create test for you automatically

make FN=newFunction HANDL=api/functionc/index create 
# sls create test -f newFunction --handler api/functionc/index

Local development

If you want use AWS SDK or work with dynamodb on you machine you need serverless-offline and/or serverleess-dynamodb-local

serverless-offline - allow you run command

# Will start local inctance of serverless which you can curl (default on 3000 port)
make offline
# sls offline start

serverless-dynamodb-client will automatically handle enviroment and try to connect to local instance when run locally

Run offline not in docker

You need disable option which allow connect outside of container

# serverless.yml
custom:
  serverless-offline:
    # disable if you run offline outside of container
    host: "0.0.0.0" # allow connect outside 

Run DynamoDB not in container

By default in this bootstrap is included serverleess-dynamodb-local and dynamo container which will started automatically when you run make console and will connect on start

But if you want just start local instance DynamoDB on you machine or inside your conteiner without any other Firstly you must install dynamodb on you machine

make dynamodb-install:
# sls dynamodb install

Then you disable options which tell serverleess-dynamodb-local connect to container

# serverless.yml
custom:
  dynamodb:
    start:
      # Please disable nearest options to start and connect in local machine
      host: dynamo # or the name of your Dynamo docker container
      port: "8000" # the port of our Dynamo docker container
      noStart: true

Then type this command to start dynamodb locally

# Will start local instance of DynamoDB (on 8000 port by default)
make dynamodb-start
# sls dynamodb start

Seed Database

You can add seed option which will inject data on start of inctance DynamoDB

# serverless.yml
custom:
  dynamodb:
    start:
      seed: true

    seed:
      domain:
        sources:
          - table: domain-widgets
            sources: [./domainWidgets.json]
          - table: domain-fidgets
            sources: [./domainFidgets.json]
      test:
        sources:
          - table: users
            rawsources: [./fake-test-users.json]
          - table: subscriptions
            sources: [./fake-test-subscriptions.json]

Which you can use on seed command to inject data

make dynamodb-seed domain,test
# sls dynamodb seed --seed=domain,test

Or seed on start of inctance

make dynamodb-start --seed=domain,test
# sls dynamodb start --seed=domain,test

Not forget to add fake data .json files before make seed

Auto-scaling

AWS Lambda

By default, AWS Lambda limits the total concurrent executions across all functions within a given region to 100. The default limit is a safety limit that protects you from costs due to potential runaway or recursive functions during initial development and testing. To increase this limit above the default, follow the steps in To request a limit increase for concurrent executions.

DynamoDB

When you create a table, you specify how much provisioned throughput capacity you want to reserve for reads and writes. DynamoDB will reserve the necessary resources to meet your throughput needs while ensuring consistent, low-latency performance. You can change the provisioned throughput and increasing or decreasing capacity as needed.

# serverless.yml
# This options control DynamoDB limits
  ProvisionedThroughput:
    ReadCapacityUnits: 1
    WriteCapacityUnits: 1

In case you expect a lot of traffic fluctuation we recommend to checkout this guide on how to auto scale DynamoDB