This is a CLI tool that waits for an event before continuing. Simples. But it does it cross platform and as a single dependency that can be downloaded into your container or environment.

Typically, you would use this to wait on another resource (such as an HTTP resource) to become available before continuing - or timeout and exit with an error.

At the moment, you can wait for a few different kinds of thing. They are:

• HTTP or HTTPS success response or any expected response following regular expressions
• TCP or GRPC connection
• DNS IP resolve address change

Using the go command:

go install github.com/dnnrly/wait-for/cmd/wait-for@latest

If you don't have Go installed (in a Docker container, for example) then you can take advantage of the pre-built versions. Check out the releases and check out the links for direct downloads. You can download and unpack a release like so:

wget https://github.com/dnnrly/wait-for/releases/download/v0.0.5/wait-for_0.0.5_linux_386.tar.gz
gunzip wait-for_0.0.5_linux_386.tar.gz
tar -xfv wait-for_0.0.5_linux_386.tar

In your Dockerfile, you can do this:

ADD https://github.com/dnnrly/wait-for/releases/download/v0.0.1/wait-for_0.0.5_linux_386.tar.gz wait-for.tar.gz
RUN gunzip wait-for.tar.gz && tar -xf wait-for.tar

Feel free to choose from any of the other releases though.

$ wait-for http://your-service-here:8080/health https://another-service/

$ wait-for -status=[0-2]{3} http://your-service-here:8080/health 

$ wait-for grpc-server:8092 other-grpc-server:9091

$ wait-for dns:google.com

This will wait for the list of IP addresses bound to that DNS name to be updated, regardless of order. You can use this to wait for a DNS update such as failover or other similar operations.

$ wait-for preconfigured-service

By default, wait-for will look for a file in the current directory called .wait-for.yml. In this, you can define the names of services that you would like to wait on.

wait-for:
  preconfigured-service:
    type: http
    target: http://the-service:8080/health?reload=true
    interval: 5s
    timeout: 60s
    http-client-timeout: 3s
  another-service:
    type: http
    target: https://another-one
  grpcService:
    type: grpc
    target: localhost:9092
  snmp-service:
    type: tcp
    target: snmp-trap-dns:514
  dns-thing:
    type: dns
    target: your.r53-entry.com

You can use wait-for to do some of the orchestration for you in your compose file. A good example would be something like this:

version: '3'
services:
  web:
    build: .
    ports:
      - "8080"
    command: sh -c 'wait-for tcp:db:5432 && ./your-api
    depends_on:
      - db
  db:
    image: "postgres:13-alpine"
    command: "-c log_statement=all"
    environment:
      POSTGRES_DB: weallvote-api
      POSTGRES_USER: ${POSTGRES_USER:-postgres}
      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-postgres}

To build the tool so that you can run it locally, you can use the following command.

$ make build

You can run the tests as the build system would, using the following command:

$ make test

You can also run the Go tests in the 'usual' way with the following command:

$ go test ./...

There is a suite of GoDog tests that check that the built tooling works as expected.

$ make acceptance-test

Depending on how your system is set up, it might not be possible for you to open up the necessary ports to run the acceptance tests. To get around this you can run those same tests in Docker

$ make acceptance-test-docker