Validation Quiz - MODULE 13
This 30-question quiz allows you to validate your knowledge on CloudWatch and observability before moving on to the practical lab.
Question 1
What is CloudWatch?
A) A weather service
B) AWS's monitoring and observability service
C) A backup service
D) A CDN
See the answer
Answer: B
Amazon CloudWatch is AWS's monitoring and observability service. It collects metrics, logs, and events from AWS resources and applications, allowing you to monitor, alert, and respond to changes.
Question 2
What are the three pillars of observability?
A) CPU, RAM, Disk
B) Metrics, Logs, Traces
C) EC2, S3, RDS
D) Cost, Performance, Security
See the answer
Answer: B
The three pillars are: Metrics (numerical data measuring performance), Logs (textual records of events), and Traces (tracking requests across services). Together, they provide a complete view of the system.
Question 3
What is a CloudWatch metric?
A) A log file
B) A time series of data points representing a measurement
C) An event
D) A dashboard
See the answer
Answer: B
A metric is a time series of numerical values. Examples: CPUUtilization, NetworkIn, RequestCount. Each metric has a namespace (AWS/EC2), dimensions (InstanceId), and datapoints (timestamp + value).
Question 4
What is a dimension in CloudWatch?
A) A physical measurement
B) A name/value pair that uniquely identifies a metric
C) A time unit
D) A type of graph
See the answer
Answer: B
A dimension is a name/value pair that categorizes the metric. Example: InstanceId=i-12345 for CPUUtilization. A metric with different dimensions is a different metric. Maximum 30 dimensions per metric.
Question 5
What is the default granularity of standard EC2 metrics?
A) 1 second
B) 1 minute
C) 5 minutes
D) 1 hour
See the answer
Answer: C
Standard EC2 metrics are collected every 5 minutes (free). Detailed monitoring (paid) collects every minute. Some custom metrics can go up to 1 second (high-resolution).
Question 6
How to get EC2 metrics every minute?
A) It's automatic
B) Enable "Detailed Monitoring" on the instance
C) Use CloudWatch Logs
D) Via Lambda
See the answer
Answer: B
Detailed Monitoring collects EC2 metrics every minute instead of 5 minutes. There is an additional cost. Enable it at launch or via the console/CLI for existing instances.
Question 7
What is the CloudWatch agent?
A) An AWS employee
B) Software to install for collecting system metrics and custom logs
C) A managed service
D) A type of alarm
See the answer
Answer: B
The CloudWatch agent installs on EC2 (or on-premises) to collect additional system metrics (memory, disk, processes) and log files. Standard EC2 metrics (CPU, network) are collected without an agent.
Question 8
Which metrics are NOT collected by EC2 without the CloudWatch agent?
A) CPUUtilization
B) Memory usage and disk space
C) NetworkIn/Out
D) StatusCheckFailed
See the answer
Answer: B
Without the agent, EC2 collects: CPU, network, status checks. The agent is needed for: memory usage, disk space, processes, and application logs. The hypervisor cannot see inside the OS.
Question 9
What is a CloudWatch alarm?
A) A siren
B) A metric monitoring that triggers actions based on thresholds
C) An error log
D) A dashboard
See the answer
Answer: B
An alarm monitors a metric and changes state (OK, ALARM, INSUFFICIENT_DATA) based on configured thresholds. It can trigger actions: SNS notifications, Auto Scaling, EC2 actions, Lambda.
Question 10
What are the three states of a CloudWatch alarm?
A) On, Off, Standby
B) OK, ALARM, INSUFFICIENT_DATA
C) Green, Yellow, Red
D) Low, Medium, High
See the answer
Answer: B
OK: the metric is within normal limits. ALARM: the threshold is exceeded. INSUFFICIENT_DATA: not enough data to determine the state (missing metric or new alarm).
Question 11
How to configure an alarm for CPU > 80% for 5 minutes?
A) Threshold 80%, period 5 minutes, 1 datapoint
B) Threshold 80%, period 1 minute, 5 datapoints out of 5
C) It's not possible
D) Via Lambda
See the answer
Answer: B
Configure: threshold 80%, evaluation period 1 minute, number of datapoints in alarm 5, over a window of 5 periods. The alarm goes to ALARM if 5 consecutive points exceed 80% (i.e., 5 minutes).
Question 12
What is a "Composite Alarm"?
A) An alarm made of plastic
B) An alarm that combines multiple alarms with logical operators (AND, OR)
C) A duplicated alarm
D) A silent alarm
See the answer
Answer: B
A Composite Alarm evaluates multiple child alarms with AND, OR, NOT operators. Example: alert only if CPU > 80% AND Memory > 90%. Reduces false positives and allows smarter alerts.
Question 13
What action can be triggered by a CloudWatch alarm?
A) Only emails
B) SNS notification, Auto Scaling action, EC2 action, Lambda
C) Only Auto Scaling
D) No action possible
See the answer
Answer: B
Possible actions: send to an SNS topic (which can trigger email, SMS, Lambda), trigger Auto Scaling actions (scale out/in), EC2 actions (stop, terminate, reboot, recover), and Systems Manager Automation.
Question 14
What is CloudWatch Logs?
A) Paper logs
B) Service to collect, store, and analyze application and system logs
C) A type of metric
D) A dashboard
See the answer
Answer: B
CloudWatch Logs collects and stores logs from many sources: Lambda, EC2 (via agent), ECS, VPC Flow Logs, Route 53, etc. It allows searching, analyzing, and creating alerts based on logs.
Question 15
What is a Log Group in CloudWatch?
A) A group of employees
B) A container that groups log streams from the same source or application
C) A type of metric
D) An alarm
See the answer
Answer: B
A Log Group groups related log streams (e.g., /aws/lambda/my-function). It defines retention, encryption, and permissions. Each log stream in the group represents a specific source (instance, container).
Question 16
What is a Log Stream?
A) A video stream
B) A sequence of log events from the same source
C) A type of notification
D) A metric
See the answer
Answer: B
A Log Stream is a sequence of log events sharing the same source. Example: a Lambda container creates a stream, each EC2 instance creates its stream. Streams are grouped in Log Groups.
Question 17
What is the default retention of CloudWatch logs?
A) 7 days
B) 30 days
C) Never expire
D) 1 year
See the answer
Answer: C
By default, logs are never deleted (Never Expire). This is important as it can generate significant storage costs. Configure appropriate retention (1 day to 10 years) according to your needs.
Question 18
How to search in CloudWatch logs?
A) Impossible
B) Filter patterns, CloudWatch Logs Insights, or export to S3/Athena
C) Via Excel
D) By email
See the answer
Answer: B
Use: Filter Patterns for simple searches, CloudWatch Logs Insights for complex SQL-like queries, or export to S3 and use Athena. Logs Insights is powerful for real-time analysis.
Question 19
What is CloudWatch Logs Insights?
A) A premium version of CloudWatch
B) An interactive query service on logs with a dedicated query language
C) A dashboard
D) An agent
See the answer
Answer: B
Logs Insights allows querying logs with a specific query language. You can filter, aggregate, and visualize results. Example: fields @timestamp, @message | filter @message like /ERROR/ | sort @timestamp desc
Question 20
What is a Metric Filter in CloudWatch Logs?
A) A physical filter
B) A rule that extracts metrics from patterns in logs
C) An alarm filter
D) A type of log
See the answer
Answer: B
A Metric Filter scans logs for specific patterns and increments a CloudWatch metric. Example: count occurrences of "ERROR" in logs to create an "ErrorCount" metric usable in an alarm.
Question 21
What is a CloudWatch dashboard?
A) A car dashboard
B) A customizable page displaying metric and alarm widgets
C) A PDF report
D) A list of logs
See the answer
Answer: B
A CloudWatch dashboard is a customizable page with widgets: metric graphs, alarm status, text, logs. Create operational dashboards to visualize system health at a glance.
Question 22
What is a runbook?
A) A race book
B) A document describing standard operational procedures for incident response
C) A Python script
D) A dashboard
See the answer
Answer: B
A runbook documents the steps to diagnose and resolve specific incidents. It includes: symptoms, possible causes, verification steps, corrective actions, and escalation. Essential for consistent incident response.
Question 23
What is AWS X-Ray?
A) A medical X-ray
B) A distributed tracing service for analyzing requests across services
C) A security scanner
D) A type of metric
See the answer
Answer: B
AWS X-Ray traces requests end-to-end across distributed services (Lambda, API Gateway, EC2, etc.). It shows a service map, latencies, and helps identify bottlenecks and errors.
Question 24
How to enable X-Ray for Lambda?
A) Automatic
B) Enable "Active tracing" in the Lambda configuration
C) Install an agent
D) Via IAM
See the answer
Answer: B
Enable "Active tracing" in the Lambda function configuration. Ensure the role has xray:PutTraceSegments and xray:PutTelemetryRecords permissions. Use the X-Ray SDK in the code for more detailed traces.
Question 25
What is CloudWatch Events (EventBridge)?
A) An event calendar
B) An event bus to react to state changes of AWS resources
C) A type of log
D) An alarm
See the answer
Answer: B
CloudWatch Events (now Amazon EventBridge) captures AWS state change events (instance terminated, S3 file created, etc.) and routes them to targets (Lambda, SNS, SQS). Enables reactive automation.
Question 26
How to be alerted when an EC2 instance is terminated?
A) Check manually
B) Create an EventBridge rule for the EC2 Instance State-change event with state "terminated"
C) Via logs
D) Impossible
See the answer
Answer: B
Create an EventBridge rule: source AWS/ec2, event type "EC2 Instance State-change Notification", detail-state "terminated". The target can be SNS (notification), Lambda (automated action), or others.
Question 27
What is Container Insights?
A) An analysis of physical containers
B) A CloudWatch feature to monitor ECS and EKS clusters
C) A type of container
D) A generic dashboard
See the answer
Answer: B
Container Insights collects metrics and logs from containerized environments (ECS, EKS, Kubernetes on EC2). It provides metrics at the cluster, service, task/pod level, and allows container troubleshooting.
Question 28
How to export CloudWatch logs for long-term archiving?
A) Impossible
B) Export to S3 (manual or automated), or Kinesis Data Firehose continuously
C) Via email
D) To DynamoDB
See the answer
Answer: B
Export to S3: one-time manual export, or configure a Subscription with Kinesis Data Firehose for continuous streaming. S3 is cheaper for long-term storage. Use S3 lifecycle for Glacier archiving.
Question 29
What is an anomaly detection alarm?
A) An alarm to detect security anomalies
B) An alarm that uses machine learning to detect unusual behaviors
C) A defective alarm
D) A manual alarm
See the answer
Answer: B
Anomaly detection uses ML to learn the normal behavior of a metric and alert when values fall outside the expected band. Useful when fixed thresholds are not suitable (variable load by time).
Question 30
What is the best practice for organizing CloudWatch alarms?
A) Create as many alarms as possible
B) Organize by criticality (P1, P2, P3), use tags, and document actions
C) One alarm for everything
D) No alarm
See the answer
Answer: B
Best practices: categorize by severity (P1 critical, P2 important, P3 informational), use naming conventions, document associated runbooks, route to the right teams via SNS, and avoid noisy alarms (alert fatigue).
Result
Count your correct answers:
- 27-30: Excellent - You master CloudWatch
- 22-26: Good - Some points to review
- 17-21: Average - Review the course before the lab
- Less than 17: Insufficient - Retake the course from the beginning
Proceed to the Monitoring Lab once you have reached at least 22 correct answers.