EC2 Automatic-recovery

1 Introduction

1.1 Why we need self-healing feature for EC2?

Based on the SLA provided by Amazon, the availability of the EC2 and EBS is between 99% to 99.5%, it is a far cry for SLA with 99.999% availability we had seem for on-premise data centers. I am not saying AWS is unreliable, this is what it is for any public cloud, you just need to set up your environment to anticipate failure, even for on-premise data center.

1.2 What are the solutions?

There are two ways to handle EC2 failure automatically:

• Setup up auto-scaling with one instance in the group
• Setup EC2 auto-recovery

Auto-Recovery is much cheaper (or even no-cost if you use 5 minutes sample time for cloud watch) and easier to set up then using auto-scaling group.

1.3 What is auto-recovery?

Auto Recovery is an Amazon EC2 feature that is designed to increase instance availability. This feature can be used to recover instances automatically when the system shows some unhealthy condition due to impairment.

1.4 What might trigger an auto-recovery?

The following is some of the problems that will trigger activate auto-recovery:

• Loss of network connectivity
• Loss of system power
• Software issues on the physical host
• Hardware issues on the physical host that impact network reachability

1.5 How EC2 Auto-recovery works?

Auto-recovery is not set up by default, so you must set up first, I will describe the process in later sections.

Assuming you had already set up auto-recovery on your EC2 instance, auto-recovery will works in the following manner:

1. The EC2 instance fails for one of the previously described reasons.
2. CloudWatch monitor the EC2 instance and detect failure.
3. A CloudWatch triggers the recovery of the EC2 instance.
4. A new EC2 instance will be launch on new hardware to replace the failed one. The new EC2 instance is a clone of the failed EC2 instance. The ID, the private and public IP addresses will stay the same. As long as data is stored on EBS volumes, no data is lost.

The following diagram illustrate the process:

1.6 Restrictions

The recover action is supported only on instances with the following characteristics:

• Use a C3, C4, M3, M4, R3, R4, T2, or X1 instance type
• Run in a VPC (not EC2-Classic)
• Use shared tenancy (the tenancy attribute is set to default)
• Use EBS volumes, including encrypted EBS volumes (not instance store volumes)
• If your instance has a public IPv4 address, it retains the public IPv4 address after recovery.

2 How to Enable Automatic-recovery of EC2 instance

AWS Cloudwatch is used to monitor the instances and auto-recovery in case of any type of impairment of the system. Whenever status check fail occurs, the recovery action takes place.

Following is the steps to setup EC2 auto-recovery:

• Login to AWS console, go to EC2 dashboard,

1. select the EC2 instance you want to setup auto-recovery.
2. Click on the  monitoring tab
3. “Create Alarm.”

      

• On the Create Alarm screen

1. On “Take the Action” checkbox choose “Recover this instance”.
2. Make it  “1 consecutive period(s) of 5 minutes” to avoid detail CloudWatch charges. (Note: if this instance is critical and the cost become not important, we should use “2 consecutive periods of 1 minutes)
3. Click “Create” button.

      

• Click close

Now, the instance has been set up to Auto-Recovery. The instance will be automatically recovered in case of any system impairment.

3 Reference

• http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/ec2-instance-recover.html
• https://cloudonaut.io/high-availability-is-a-no-brainer-ec2-auto-recovery/