Why start with a map
AWS has well over a dozen analytics and data services. Several have been renamed, and a few are quietly closing. Before you can choose between them, you need to see them all on one page, sorted by the job each one does. Think of it as a floor plan of a building that keeps renovating: the rooms stay where they are, but the signs on the doors change, and some doors now say "existing tenants only".
In this level you build that floor plan. You will also learn two habits that will stop you making embarrassing recommendations: translating old names into new ones, and checking a service's lifecycle status before you propose it.
The map: seven groups, one job per service
| Group | Service (current name) | The job in one sentence |
|---|---|---|
| Warehouses and lakehouses | Amazon Redshift (provisioned and Serverless) | A columnar SQL data warehouse for fast, repeated BI and reporting queries. |
| Amazon S3 Tables | S3 storage built for Apache Iceberg tables, with table maintenance managed for you. | |
| Query engines | Amazon Athena | Serverless SQL straight over files in S3, with no cluster to run. |
| Data integration and pipelines | AWS Glue (ETL jobs) | Serverless Spark and Python jobs that extract, transform and load data. |
| Amazon Data Firehose | Delivers streaming data into S3, Redshift, OpenSearch and other destinations, with no code. | |
| Streaming and messaging | Amazon Kinesis Data Streams | A durable, ordered stream you write to and read from with your own code. |
| Amazon MSK (Managed Streaming for Apache Kafka) | Managed Apache Kafka, for teams that already use Kafka APIs and tools. | |
| Amazon Managed Service for Apache Flink | Runs Flink applications that process streams in real time. | |
| Big data processing | Amazon EMR (on EC2, Serverless, on EKS) | Spark, Hive, Trino and other open-source frameworks, from fully managed to fully in your control. |
| Catalogues and governance | AWS Glue Data Catalog | The shared metadata store (databases, tables, schemas) that Athena, EMR and Redshift read. |
| AWS Lake Formation | Fine-grained permissions (down to column and row) over data registered in the catalogue. | |
| Amazon SageMaker Unified Studio and Amazon DataZone | A single workspace and business data catalogue for finding, sharing and governing data. Unified Studio is built on DataZone. | |
| Search and BI | Amazon OpenSearch Service | Search and log analytics, with dashboards. |
| Amazon Quick (BI part: Quick Sight) | Dashboards and BI for business users, plus AI agents. |
Gotcha: Glue appears twice on purpose. Glue ETL is a pipeline tool. The Glue Data Catalog is metadata that almost every other service on this page depends on. Treat them as two separate decisions.
Current and former names
Old blog posts, Stack Overflow answers and Terraform modules still use the old names, and so do the APIs. Keep this translation table to hand:
| Current name | Former name(s) | What still shows the old name |
|---|---|---|
| Amazon Managed Service for Apache Flink | Amazon Kinesis Data Analytics (renamed 30 Aug 2023) | CLI aws kinesisanalyticsv2, Terraform aws_kinesisanalyticsv2_application |
| Amazon Data Firehose | Amazon Kinesis Data Firehose (renamed Feb 2024) | CLI aws firehose, Terraform aws_kinesis_firehose_delivery_stream |
| Amazon OpenSearch Service | Amazon Elasticsearch Service (renamed 8 Sep 2021) | Legacy CLI aws es, Terraform aws_elasticsearch_domain |
| Amazon Quick, with Quick Sight as its BI part | Amazon QuickSight, then Amazon Quick Suite (launched 9 Oct 2025) | CLI aws quicksight, Terraform aws_quicksight_* |
Pitfall: "Kinesis Data Analytics for SQL applications" was not simply renamed. It was a separate product, and it was discontinued. New applications were blocked from 15 Oct 2025, and remaining applications were deleted on 27 Jan 2026. If an old runbook shows SQL-on-Kinesis, it needs migrating to Managed Service for Apache Flink (or Flink Studio), not just a new label.
Lifecycle status: closed, retiring or gone
AWS uses three lifecycle terms, defined on its AWS Lifecycle Changes page (aws.amazon.com/products/lifecycle redirects there):
- Maintenance: no new customers. Existing users carry on, but the service gets no new features.
- Sunset: existing users should migrate. There is a shutdown date, typically about 12 months away.
- Full shutdown: the service is gone.
Items relevant to this map:
| Service / feature | Status | Date | Look at instead |
|---|---|---|---|
| AWS Data Pipeline | Closed to new customers | 25 Jul 2024 | Glue, Step Functions, MWAA |
| Amazon CloudSearch | Closed to new customers | 25 Jul 2024 | OpenSearch Service |
| AWS IoT Analytics | Closed to new customers 24 Jul 2024; support ended | 15 Dec 2025 | Kinesis/Firehose + S3 + Athena |
| Timestream for LiveAnalytics | Closed to new customers | 20 Jun 2025 | Timestream for InfluxDB, among others |
| AWS Glue for Ray jobs | Closed to new customers | 30 Apr 2026 | Ray on Amazon EKS |
Best practice: "Closed to new customers" applies per account. A colleague whose account already uses Data Pipeline can keep creating pipelines. Your new account cannot. A demo that works in their account proves nothing about yours.
Walkthrough: check before you recommend
AWS doesn't offer an API that reports lifecycle status, so the check takes two steps. First, read the lifecycle page and the service's own "availability change" or "end of support" doc page. Second, confirm the service is offered in your Region. That second step can be done from the CLI, using AWS's public Systems Manager parameters:
# Step 1 (manual): open https://docs.aws.amazon.com/general/latest/gr/service-lifecycle.html
# and search the service's docs for "availability change" or "end of support".
# Step 2: in which Regions is Athena offered? (public parameter, no setup needed)
aws ssm get-parameters-by-path \
--path /aws/service/global-infrastructure/services/athena/regions \
--query 'Parameters[].Value' \
--region eu-west-2 # query from a Region that supports this path
# Sample output (truncated):
# ["eu-west-1", "eu-west-2", "us-east-1", ...]
# Step 3: list every service AWS offers in your Region, sorted, then scan it
aws ssm get-parameters-by-path \
--path /aws/service/global-infrastructure/regions/eu-west-2/services \
--query 'Parameters[].Name | sort(@)'
These commands only read public data, so nothing is created and there is nothing to clean up.
Trade-off: The Region list tells you a service exists in a Region, not that your account can sign up for it. A maintenance-mode service can still appear there. That is why step 1 matters.
Your deliverable
Copy the two tables above into your team wiki as analytics-map.md. Then add one column, "Status checked on (date)". Statuses change: about 20 AWS services and features (none of them on this map) moved to maintenance on 30 July 2026 alone. A map with no date on it goes stale without anyone noticing.
Key takeaways
- Sort services by the job they do: warehouse/lakehouse, query, pipelines, streaming, big data processing, catalogue/governance, BI/search.
- Translate old names: Kinesis Data Analytics → Managed Service for Apache Flink; Kinesis Data Firehose → Data Firehose; Elasticsearch Service → OpenSearch Service; QuickSight → Quick Suite → Amazon Quick (BI part: Quick Sight). The CLI and Terraform names often keep the old ones.
- Maintenance means no new customers, sunset means a shutdown date, and full shutdown means the service is gone.
- Data Pipeline, Timestream for LiveAnalytics and Glue for Ray are closed to new customers. IoT Analytics support has already ended.
- Always check the lifecycle page and Region availability, and put the date you checked on the map.