Lesson 1: The AWS Ops Toolbox on One Page

Why start with a map?

When something breaks at 2 a.m., you don't want to be choosing between nine AWS services that all have "Cloud" or "Manager" in their name. Experienced operators settle one question before they open any console: what kind of job is this? Is it measuring, following a request, building infrastructure, fixing servers, keeping track of what you own, paying for it, or getting advice? Once you know the job, you only have two or three services to choose from.

Think of this lesson as the label on each drawer of a toolbox. Later levels open the drawers one at a time. Here we only learn which tool lives in which drawer, which labels have changed, and which tools AWS is retiring.

The seven groups

Gotcha: some services do more than one job. Compute Optimizer gives advice, but its recommendations feed Cost Optimization Hub, so it also matters for cost. Application Signals sits inside CloudWatch but does the tracing job. Put a service in the group for the question it answers first, not the console it lives in.

Current and former names

Old names hang around in blog posts, IAM prefixes and Terraform modules, so you need to recognise them:

• AWS Systems Manager was Amazon EC2 Systems Manager, and before that the agent was called Simple Systems Manager. The new name came in November 2017. That's why the CLI is still aws ssm and the agent is still "SSM Agent".
• AWS Health Dashboard replaced both the Personal Health Dashboard (now the Your account health view) and the Service Health Dashboard (now the Service health view) in February 2022. The API is still aws health.
• Amazon Managed Grafana was called Amazon Managed Service for Grafana during its preview.

What is retiring

The X-Ray service itself is not retiring. Only the AWS-specific SDKs and the daemon that send data to it are. Existing users of Incident Manager, Change Manager and AppRegistry can keep using them in accounts that already have them, but don't design a new estate around any of the three.

Walkthrough: find out what you already use

Before choosing anything, check which ops services your account already pays for. Cost Explorer can group spend by service:

# Last month's spend, grouped by service (Cost Explorer API runs in us-east-1)
aws ce get-cost-and-usage \
  --time-period Start=2026-09-01,End=2026-10-01 \
  --granularity MONTHLY \
  --metrics UnblendedCost \
  --group-by Type=DIMENSION,Key=SERVICE \
  --query 'ResultsByTime[0].Groups[?contains(Keys[0], `CloudWatch`) || contains(Keys[0], `X-Ray`) || contains(Keys[0], `Config`)].[Keys[0], Metrics.UnblendedCost.Amount]' \
  --output table

The output looks roughly like this (example figures):

-----------------------------------
|       GetCostAndUsage           |
+---------------------+-----------+
|  AWS Config         |  41.20    |
|  AWS X-Ray          |  3.85     |
|  AmazonCloudWatch   |  612.47   |
+---------------------+-----------+

Note that each call to the Cost Explorer API is billed, so don't put this in a loop. Next, look for code still using the X-Ray daemon, which loses support in February 2027:

# Search ECS task definitions in this Region for the X-Ray daemon image
for td in $(aws ecs list-task-definitions --query 'taskDefinitionArns[]' --output text); do
  aws ecs describe-task-definition --task-definition "$td" \
    --query 'taskDefinition.containerDefinitions[].image' --output text \
    | grep -q 'xray' && echo "Uses X-Ray daemon: $td"
done

Every hit goes on the migration list for OpenTelemetry.

Name the group before the service

Put each need into a group first:

• "Alert me when p99 latency goes above 500 ms" → monitoring (CloudWatch alarm), or tracing if you also need to know which downstream call is slow.
• "Patch 400 EC2 instances every second Tuesday" → configuration and patching (Systems Manager Patch Manager).
• "Every new account must start with logging and guardrails" → governance (Control Tower over Organizations).
• "Tell finance before the dev account goes over $2,000" → cost (Budgets).
• "Is the slowdown us or AWS?" → advice and health (AWS Health).

Best practice: write the group next to every requirement in your backlog. Teams that skip this step tend to buy a tool that does a lot of everything and fits none of their needs well.

Key takeaways

• Seven groups: monitoring and logging, tracing, IaC, configuration and patching, inventory and governance, cost, advice and health.
• Place a service by the question it answers first, not by the console it sits in.
• Systems Manager was EC2 Systems Manager (hence ssm). AWS Health Dashboard replaced the Personal and Service Health Dashboards in 2022.
• Retiring: X-Ray SDKs and daemon (maintenance from 25 Feb 2026, end of support 25 Feb 2027; move to OpenTelemetry). Incident Manager and Change Manager closed to new customers from 7 Nov 2025. AppRegistry closed to new customers from 30 Jul 2026. Evidently ended on 16 Oct 2025.
• Name the group first, then pick from the two or three services in it.