Sometimes the default size for your root volume in an EC2 instance isn’t good enough. As there is no clear documentation on the best way to do this for CloudFormation managed instances, I’m describing my methods for increasing the size here.

The use case

When creating an EC2 instance it is usually set to use the default 8GB EBS root volume, but now we require more space than this. We can of course attach additional mount points, but in order to keep it simple we prefer not to do so.

That means we need to increase the size of the root volume. First off, you need to realise that this cannot be done without any downtime. If your servers are loadbalanced, make sure to only update them one at a time to prevent downtime.

There are explanations of how to increase the size of volumes, but these are not compatible with CloudFormation and your root device might be replaced the next time you do an update. Obviously, that is not a desired outcome.

The CloudFormation solution

For this, we will use the BlockDeviceMappings property for AWS::EC2::Instance. This setting allows you to set the size of your EBS volumes. Using the DSL this looks like the following snippet

The below instruction will destroy your current root volume so don’t run it until after you’ve read the rest.

Resource("TestServer") {
    Type "AWS::EC2::Instance"
    Property("ImageId", FnFindInMap( "AWSRegionArch2AMI", Ref("AWS::Region"),"AMI"))
    Property("BlockDeviceMappings", [
        {
            "DeviceName" => "/dev/sda1",
            "Ebs" => {
                "VolumeSize" => "120",
                "VolumeType" => "gp2"
            }
        }
    ])
}

Updating your instance after adding the BlockDeviceMappings property this to your configuration will give you a new instance, with a new root volume at the specified size. In this case, a 120GB SSD¹ volume. If that is all you need, you’re done. If you wish to retain the data you need to do a bit more work before running the above.

Unfortunately, you cannot change the Snapshot Id for a root volume so we need another solution than taking a snapshot of the volume. What we’ll do is create an AMI based on the whole instance instead.

First we will want to stop the instance, to ensure there can’t be any data loss.

aws ec2 stop-instances --instance-ids "i-123456" --profile demo

And then we’ll create the actual image.

aws ec2 create-image --instance-id i-123456 --name "test-bigger" --description "Bigger disk for Test server" --profile demo

The output of this command is similar to this.

{
    "ImageId": "ami-123456"
}

Now all that you need to do is update the ImageId in your CloudFormation template, add the BlockDeviceMappings section and update your stack to have your current instance replaced with a new one that has the increased disk size.

Without CloudFormation

If you’re not using CloudFormation you can still use the above method of creating an AMI and then launching a new instance using this. However, as your instance will likely be connected to other things like an Elastic IP it might be easier instead to replace the root volume. You can do this with the following commands.

Let’s start with making sure we have all the information we need. We need the volume id of the root device, which we can find using the below command.

aws ec2 describe-instances --instance-id i-123456 --profile demo

Next, we stop the instance.

aws ec2 stop-instances --instance-ids i-123456 --profile demo

Then we create a snapshot of the root volume. The output of the command to create this snapshot also will provide the snapshot id.

aws ec2 create-snapshot --volume-id vol-123456 --description "backup for test server" --profile demo

Now we can create a volume based on this snapshot. When doing so we provide the size in GiB, the id of the snapshot, the availability zone, and the volume type.

aws ec2 create-volume --size 100 --snapshot-id snap-123456 --availability-zone us-east-1a --volume-type gp2 --profile demo

Now all we need to do is swap the the volumes and start the server again.

aws ec2 detach-volume --volume-id vol-123456 --profile demo
aws ec2 attach-volume --volume-id vol-234567 --instance-id i-123456 --device "/dev/sda1" --profile demo
aws ec2 start-instances --instance-ids i-123456 --profile demo

And that’s all. The EC2 instance is now using the new volume and you can remove the old volume whenever you wish.