AWS Certified Solutions Architect – Associate Application Integration, Analytics & Operational Excellence Syllabus

In Kinesis Data Streams, what is a shard and what is its throughput limit?

A shard is a unit of capacity. Each shard ingests up to 1 MB/sec or 1,000 records/sec, and outputs up to 2 MB/sec. Throughput scales by adding shards.

What is the key difference between Kinesis Data Streams and Kinesis Data Firehose?

Data Streams is real-time (sub-second), requires you to manage consumers and scaling (shards), and stores data 24h–365 days. Firehose is near-real-time (buffered, ~60s min), fully managed/auto-scaling, with no storage and direct delivery to destinations.

What is the default and maximum data retention period in Kinesis Data Streams?

Default retention is 24 hours; it can be extended up to 365 days.

What is Amazon Athena and what does it query?

Athena is a serverless, interactive query service that runs standard SQL directly on data in Amazon S3. You pay per amount of data scanned per query; no infrastructure to manage.

How can you reduce cost and improve performance of Amazon Athena queries?

Use columnar compressed formats (Parquet/ORC), partition data, and compress files — these reduce the amount of data scanned, lowering both cost and query time.

What is AWS Glue and what is the Glue Data Catalog?

AWS Glue is a serverless ETL service for discovering, preparing, and transforming data. The Glue Data Catalog is a central metadata repository (tables/schemas) used by Athena, Redshift Spectrum, and EMR; Glue Crawlers auto-populate it.

What is Amazon Redshift and what architecture does it use?

Redshift is a fully managed, petabyte-scale data warehouse using columnar storage and Massively Parallel Processing (MPP). It is optimized for OLAP/analytics queries, not high-concurrency OLTP transactions.

What is Amazon Redshift Spectrum?

Spectrum lets you run Redshift SQL queries directly against exabytes of data in S3 without loading it into Redshift tables, using the Glue Data Catalog for metadata.

What is the difference between OLTP and OLAP, and which AWS services fit each?

OLTP = transactional, many small reads/writes (use RDS/DynamoDB). OLAP = analytical, complex queries over large datasets (use Redshift/Athena/EMR).

What is Amazon EMR and what frameworks does it run?

EMR (Elastic MapReduce) is a managed big-data cluster platform running open-source frameworks like Apache Hadoop, Spark, Hive, HBase, Presto, and Flink for large-scale data processing.

What is Amazon QuickSight?

QuickSight is a serverless, ML-powered business intelligence (BI) service for creating interactive dashboards and visualizations. SPICE is its in-memory calculation engine for fast queries.

What is Amazon OpenSearch Service used for?

It is a managed service for running OpenSearch/Elasticsearch for search, log analytics, full-text search, and observability, often paired with Kibana/OpenSearch Dashboards and Kinesis/Logstash for ingestion.

What three telemetry types does Amazon CloudWatch collect, and what is a custom metric?

Metrics, Logs, and Events/Alarms. A custom metric is application-defined data you publish to CloudWatch (e.g., memory usage), since metrics like RAM are not collected by default on EC2.

What is a CloudWatch alarm and what states can it have?

An alarm watches a metric over a period and triggers actions based on thresholds. States: OK, ALARM, and INSUFFICIENT_DATA. Actions can include SNS notifications, Auto Scaling, or EC2 actions.

What is the difference between CloudWatch basic and detailed monitoring for EC2?

Basic monitoring sends metrics every 5 minutes (free). Detailed monitoring sends metrics every 1 minute (additional cost).

What does AWS CloudTrail record, and what is it used for?

CloudTrail records API calls and account activity (who, what, when, from where) for governance, compliance, and auditing. Events can be sent to S3 and CloudWatch Logs; it is enabled by default for 90 days of management events.

What is the key distinction between CloudWatch and CloudTrail?

CloudWatch monitors performance/operational metrics and logs ('how is it performing'). CloudTrail audits API/user activity ('who did what'). CloudWatch = performance, CloudTrail = audit trail.

What is AWS X-Ray used for?

X-Ray provides distributed tracing to analyze and debug applications, especially microservices. It builds a service map showing request flows, latency, and errors across components to find bottlenecks.

What is AWS Systems Manager (SSM) and name two key capabilities.

Systems Manager is a unified operations service for managing AWS and on-prem resources. Key features include Parameter Store (config/secrets), Session Manager (shell access without SSH/bastion), Patch Manager, Run Command, and Automation.

What is the difference between SSM Parameter Store and AWS Secrets Manager?

Parameter Store stores config/secrets (free standard tier) with no built-in rotation. Secrets Manager (paid) adds automatic secret rotation, native integration with RDS, and cross-account access for secrets management.

What does AWS Trusted Advisor check, and what are its five categories?

Trusted Advisor inspects your environment and gives best-practice recommendations across five categories: Cost Optimization, Performance, Security, Fault Tolerance, and Service Limits (Service Quotas).

What is the AWS Health Dashboard?

It provides alerts and guidance on AWS service events. The Service Health Dashboard shows general AWS status, while the Personal Health Dashboard (Your Account) shows events specifically affecting your resources.

Define RTO (Recovery Time Objective).

RTO is the maximum acceptable amount of time to restore service after a disaster — i.e., how long the system can be down before recovery must complete.

Define RPO (Recovery Point Objective).

RPO is the maximum acceptable amount of data loss measured in time — i.e., how far back the last usable backup must be (how much recent data you can afford to lose).

What are the four AWS disaster recovery strategies, ordered from highest RTO/RPO to lowest?

1) Backup & Restore (cheapest, highest RTO/RPO), 2) Pilot Light, 3) Warm Standby, 4) Multi-Site / Active-Active (most expensive, near-zero RTO/RPO).

Describe the Pilot Light DR strategy.

Core critical components (e.g., databases replicating data) run continuously in the recovery region, while the rest of the infrastructure is provisioned and scaled up only when disaster strikes. Lower cost than warm standby, faster than backup/restore.

Describe the Warm Standby DR strategy.

A scaled-down but fully functional copy of the production environment always runs in the recovery region. On disaster, it is scaled up to handle full load — faster recovery than Pilot Light, higher cost.

How do you achieve high availability for an application across an AWS Region?

Deploy resources across multiple Availability Zones (Multi-AZ), front them with an Elastic Load Balancer, and use Auto Scaling so the system tolerates the loss of an entire AZ.

What design principles help build fault-tolerant architectures on AWS?

Eliminate single points of failure (multi-AZ, redundancy), decouple components (SQS/SNS), automate recovery (Auto Scaling/health checks), use managed/replicated data stores, and design for graceful degradation.

What are the six pillars of the AWS Well-Architected Framework?

Operational Excellence, Security, Reliability, Performance Efficiency, Cost Optimization, and Sustainability.

Name key cost optimization practices under the AWS Well-Architected Framework.

Right-size resources, use the appropriate pricing model (Savings Plans/Reserved/Spot Instances), adopt serverless/managed services, implement S3 lifecycle/storage tiering, and monitor spend with Cost Explorer/Budgets to eliminate idle resources.