12

We recently got the following notification regarding one of our instances:

Retiring: This instance is scheduled for retirement after December 14, 2015 at 7:00:00 PM UTC-5.

Instances that are scheduled for retirement will be shut down after the specified date. We recommend that you launch replacement instances and start migrating to them.

Availability Zone: us-east-1c
Event type: instance-stop
Event status: Scheduled
Description: The instance is running on degraded hardware
Start time: December 14, 2015 at 7:00:00 PM UTC-5

Further details show that the event type is instance-stop. Per the documentation:

Instance stop: The instance will be stopped and started to migrate it to a new host computer. Applies only to instances backed by Amazon EBS.

Instance retirement: The instance will be terminated.

Reboot: Either the instance will be rebooted (instance reboot) or the host computer for the instance will be rebooted (system reboot).

System maintenance: The instance might be temporarily affected by network maintenance or power maintenance.

Instance stop. Does this mean that we do not have to do anything? If not, what do we have to do to avoid service disruption?

PeterKA
  • 21,234
  • 4
  • 21
  • 44

1 Answers1

20

Following solution works if your instance is not in an Auto Scaling Group.

I am assuming you are running the instance in a VPC. You cannot avoid service disruption unless you have some kind of HA. You have 2 options if the root device type is ebs.

1) You can create an AMI of your instance. Launch a new instance from that AMI and then stop the old instance.

2) Easiest option which I always prefer: Stop your instance and start it again. Thats it. New physical hardware will be chosen automatically. If your old instance had a public IP, your new instance will get a new public IP. But if your instance had elastic IP associated with it, you do not have to do anything more.

helloV
  • 42,534
  • 4
  • 100
  • 125
  • Thanks so much for your time and help @helloV. I already feel like things are looking up. :) For option 2, for how long would I have to leave the instance in the stopped state? I can confirm it has an elastic IP address. – PeterKA Dec 14 '15 at 06:17
  • 2
    You don't have to wait after the instance is stopped. Once the instance state changes to stopped, you can start the instance immediately or any time later. – helloV Dec 14 '15 at 06:19
  • Please bear with me a little longer. This is all new to me. What does this mean? **Note that when your instances are stopped: *Any data on the ephemeral storage of your instances will be lost.*** – PeterKA Dec 14 '15 at 06:23
  • 3
    The data that is not stored on EBS volume (ephemeral) will be lost. Since your instance's root device type is EBS, you need not worry. If you are not sure I suggest you create an AMI of your instance before stopping your instance. After you stop and start your instance, if you find some data is missing, you can always launch a new instance from your AMI. http://stackoverflow.com/questions/11566223/what-data-is-stored-in-ephemeral-storage-of-amazon-ec2-instance – helloV Dec 14 '15 at 06:32
  • You are the best! I used option 2 and it worked just fine. I am deeply humbled. – PeterKA Dec 14 '15 at 06:40
  • Do you think this is acceptable to get an email you can easily miss that your server failed because of Amazon fault. And your server can be down for a long time until you realize it yourself by incident? And they don't even try to work it out with you? No alerts, no contact. And where is the cloud advantage of duplicating the server in multiple places to prevent that? – Ronen Festinger Jun 21 '17 at 11:01
  • FYI, my website didn't come back up because I had to go in and associate my Elastic IP to the new instance. It didn't do it automatically. – adam0101 Nov 07 '19 at 18:43
  • @adam0101 not sure what you mean by `new instance`. If you stopped and started the same instance, the elastic IP should remain with the instance. Something could be different in your setup. – helloV Nov 07 '19 at 21:29
  • @helloV, when I shut down my instance and restarted it, I had two entries in my instance list for a brief time. The one I shutdown which instantly was marked at "terminated" and a new one with the same name which showed as "running". I'm guessing they switched me to new hardware. My website did not come back up and I noticed that the properties of my elastic IP had been wiped out. Once I reassociated it with the instance, the website came back up. – adam0101 Nov 08 '19 at 18:42
  • @adam0101 your instance is most likely in an autoscaling group (ASG) which explains the behavior you are described. Based on the ASG policy, the behavior varies. Best to check with your AWS admin. – helloV Nov 08 '19 at 22:13