Home / server / Questions / 698007
In Process
ed.
Asked: 2015-06-11 08:02:39 +0800 CST

Launch instances via Elastic Beanstalk without Elastic IP configured

  • 772

I'm creating a tool to deploy Docker images to EC2 via ElasticBeanstalk, so the dev team can quickly demo their work without having to merge. I don't want Elastic IP to be enabled for these apps as they are short lived.

The reason I'm asking this is because if my tool builds more than 5 apps (and it has quite quickly) any new app will fail because no Elastic IP can be assigned to the new app. This is because an AWS account typically has a max of 5 Elastic IPs available. I can go to the aws web console and choose 'Disassociate Elastic IP Address' (see below) on an instance, after which the ec2 instance will get a new ip (after a few minutes). This is what I want, but I want to do it programmatically, and would prefer to boot the app without it instead of doing it once the instance is created.

disassociate menu option

Is there a configuration option I can use to use to disable Elastic IP for the new instance?

I'm using the node aws-sdk, but any tips in any language will do.

I'm creating m1.small single instances.

If this isn't possible when launching, I'll have to disassociate the Elastic IP from the instance using the EC2.disassociateAddress function.

elastic-beanstalk

4 Answers

  1. When using the environment type "Single instance", you always get an EIP.

    From Beanstalk developer guide, Environment Types:

    A single-instance environment contains one Amazon EC2 instance with an Elastic IP address.

    Disabling the "Associate Public IP Address" option does not have any effect.

    Switch to "Load-balancing, Autoscaling" Environment to get by without an EIP.

    • 13

  2. I don't want Elastic IP to be enabled for these apps as they are short lived.

    If you're in a VPC's public subnet (which is likely), one is required for that instance to be able to communicate with the Internet. Per http://docs.aws.amazon.com/AmazonVPC/latest/UserGuide/VPC_Internet_Gateway.html:

    To enable an instance in your public subnet to communicate with the Internet, it must have a public IP address or an Elastic IP address that's associated with a private IP address on your instance. Your instance is only aware of the private (internal) IP address space defined within the VPC and subnet. The Internet gateway logically provides the one-to-one NAT on behalf of your instance, so that when traffic leaves your VPC subnet and goes to the Internet, the reply address field is set to the public IP address or Elastic IP address of your instance, and not its private IP address. Conversely, traffic that's destined for public IP address or Elastic IP address of your instance has its destination address translated into the instance's private IP address before the traffic is delivered to the VPC.

    The automatically-assigned EIPs that come with a new instance don't cost anything, and they automatically go away when the instance is decommissioned.

    • 1

  3. There is an elastic beanstalk environment option that might be what you are after.

    http://docs.aws.amazon.com/elasticbeanstalk/latest/dg/command-options-general.html#command-options-general-elbloadbalancer

    Namespace: aws:ec2:vpc AssociatePublicIpAddress : false

    If you set it then all the nodes created by this environment will not have a public IP. Inbound traffic will have to use a load balancer. The VPC and subnets will need to be accociated with a NAT gateway or an internet gateway to get outbound internet access. If you do not have outbound internet access then the Elastic Beanstalk build will fail (It won't be able to reach AWS services).

    • 1

  4. After you Disassociate the Elastic IP Address, you have about 10 seconds to 'release' it. If you release it, it will not come back.

    • 1