AWS Lightsail default ssh userid

To ssh into AWS ec2 instances the default user id is usually ‘ec2-user’ (see my ec2 ssh checklist here).

Lightsail vps instances appear to use different default userids, depending on the OS. For example, for an Ubuntu Lightsail instance the default ssh userid is ‘ubuntu’:

ssh -i path-to-your-ssh-pen-file ubuntu@your-instance-ip

ssh into a running EC2 instance being used for an AWS Beanstalk deployment

The EB cli provides an ‘eb ssh env-name’ command to provide ssh access into a running EC2 instance. If the environment wasn’t setup to enable ssh, you’ll see this:

> eb ssh test
ERROR: This environment is not set up for SSH. Use "eb ssh --setup" to set up SSH for the environment.

After running ‘eb ssh –setup’ you’ll be prompted to confirm the environment name, and then select a previously configured SSH keypair to use to provide auth when logging on to the instance.

If an EC2 instance is already running in the env, this setup step configures new EC2 instances to enable ssh, and then conducts a rolling update to take down any previously started EC2 instances and replace them with new instances.

Deploying a Docker container to AWS Elastic Beanstalk

In my previous post, I looked at using the EB cli to deploy a Spring Boot app to Beanstalk. If you have an app that you have packaged in a Docker container, you can prepare this for deployment to Beanstalk using the EB cli command:

$ eb init -p docker application-name

This is described in the docs here.

This command inits the app for deployment, creating a default .elasticbeanstalk/config.yml file that looks like this:

branch-defaults:
default:
environment: null
group_suffix: null
global:
application_name: beanstalk-docker-with-mounted-volume
branch: null
default_ec2_keyname: null
default_platform: Docker
default_region: us-west-2
include_git_submodules: true
instance_profile: null
platform_name: null
platform_version: null
profile: null
repository: null
sc: null
workspace_type: Application

Next create a Beanstalk environment for deploying your app:

$ eb create environment-name

This will take a few minutes on your first deploy as it provisions everything required for running your app on Beanstalk, including an Auto Scaling Group and an EC2 instance.

Deploying a Spring Boot app to AWS Elastic Beanstalk with the eb cli

Deploying a Spring Boot app to AWS Elastic Beanstalk is relatively easy using the eb cli tool, if you’re prepared to accept some defaults as part of the deploy. It’s probably possible to configure/customize IAM roles, Security Groups etc, but accepting the defaults is an easy way to deploy during development.

To initialize a new Elastic Beanstalk deployment, run in the root of your Spring Boot source folder:

eb init

Before creating the environment and deploying your app, edit the .elasticbeanstalk/config.yml that the previous step creates, to configure the built app jar to be deployed, by adding this section:

deploy:
artifact: target/your-app-jar-0.0.1-SNAPSHOT.jar

To create a single dev/test env with no load balancing:

eb create --single

To tear down the deployed ec2 instance running your Spring Boot app:

eb terminate

Service Port Configuration

By default, Beanstalk services are deployed on port 5000. Since Spring Boot apps are on port 8080 by default, the quickest way to configure Spring Boot to accept incoming requests on port 5000 is to edit the ‘Updates, monitoring and logging’ section and add a new env var for Spring Boot to reconfigure to use port 5000 instead (this is described here)

Added new env var SERVER_PORT=5000: