AWS Health notifications, explained

AWS Health is what tells you about the things specific to your own account — an EC2 instance scheduled for retirement, a maintenance window, a service incident — and most of it arrives by mail as well. Forward that mail to Sabaki and it comes back explained, with the deadline pulled out and the whole thing filed under the right customer and the right account.

Last updated: 2026-09-22

What an AWS Health mail is

AWS Health (the AWS Health Dashboard) reports the events and the scheduled work that affect the resources in your own account, separately from the overall state of AWS. The common ones: an EC2 instance retired because the hardware hosting it is degrading, scheduled maintenance on EC2 or RDS, an incident in a region you use, and notice that a certificate or an API is going away.

All of it shows up in the AWS Health Dashboard and, at the same time, lands in the account's contact mailbox. The sender is no-reply-aws@amazon.com, and the subject usually carries the account ID.

Where these mails are hard to read

In a Health mail, the part that matters — by when, what to do, and what happens if you do not — sits halfway down the body. In a retirement notice, for instance, the text goes on to say that after the deadline the instance will be stopped, that with an EBS root volume a stop and a start move it to a different host, and that the public IP may change. Skim past that and you find out on the morning of the deadline, when the instance has already stopped.

Holding several customers' accounts adds the work of turning the account ID in the subject into a customer. The same mail arrives once per customer, so checking every time whose is whose is not realistic.

What Sabaki does with them

Sabaki works on the Health notifications that arrive by mail. Forward them to that customer's forwarding address and it checks the DKIM signature to confirm they came from AWS, pulls the account ID and the instance ID out of the body to file them under the customer, explains what is going to happen, what it affects and what to do, and extracts the deadline. There is nothing to configure in the AWS Health API or in EventBridge; Sabaki does not integrate with either of them directly today.

Example: the mail that arrives, and what Sabaki shows

The subject and body imitate a real notification. Account IDs and dates are fictional.

The mail that arrives (English)
From
Amazon Web Services <no-reply-aws@amazon.com>
Subject
Amazon EC2 Instance Retirement [AWS Account: 123456789012]

Hello,

EC2 has detected degradation of the underlying hardware hosting your Amazon EC2 instance (instance-ID: i-0f1e2d3c4b5a69788) in the ap-northeast-1 region. Due to this degradation, your instance could already be unreachable. After 2026-10-12 00:00 UTC your instance, which has an EBS volume as the root device, will be stopped.

You can see more information on your instances that are scheduled for retirement in the AWS Management Console.

What Sabaki shows (English)
Action requiredCompany A (online shop)123456789012 (Company A, production)Due: 2026-10-12 09:00 (UTC+9)

The hardware hosting EC2 instance i-0f1e2d3c4b5a69788 in the production account is degrading, and the instance will be stopped from 09:00 (UTC+9) on 12 October. Stop it and start it again before then to move it to another host.

This is a retirement notice caused by a failing host. The instance may already have stopped responding, so the first thing to check is whether it is still serving. The root device is EBS, so stopping the instance and starting it again brings it up on a different host with its data intact. If you are not using an Elastic IP the public IP will change, so note down the DNS records and anything else pointed at it first. Past the deadline, AWS stops it for you.

Within AWS Health, something that needs no action — an AWS incident notice, say — is shown as “Info”, while something that needs acting on immediately, such as a warning that an account is about to be suspended, is shown as “Critical”.

From the notice to the message you send on

The explanation spells out the options you have, so it doubles as the draft of the message you send your customer. Put Slack on the other end and the summary, the deadline and the explanation are posted to that customer's channel, so whoever is on it reads before moving. What was done stays in the response ledger, and a notification still not handled past its deadline is obvious at a glance in the inbox.

Where the AWS Health Dashboard still earns its place

Some Health events never arrive by mail, so there is no reason to stop checking the Dashboard. What Sabaki takes off you is the rest of it: reading the mail that does arrive, working out whose it is and what it is about, explaining it, and keeping a record of what was done.

Forward one AWS notification and it comes back explained

The Free plan covers one account and one seat, with no credit card. Nothing changes on the AWS side, and no access keys are registered.

Related questions

Related guides

If this helps, pass it on to someone who needs it.