INFRASTRUCTURE LEARNINGINFRASTRUCTURE LEARNING
インフラ大全インフラ大全IT INFRASTRUCTURE / LEARNING
Infrastructure learning guide
Find a lesson or lab/

Esc to close · Search terms stay on your device

LEARNING日本語

Topic lessonTopic: AWS security services explained: prevention, findings and response

AWS security services explained: prevention, findings and response

Combine prevention that reduces risk, detection that identifies signs of trouble, and response that investigates and limits impact. Compare four representative AWS services using a fictional reporting server. Keep the presence of a finding separate from evidence that compromise or recovery has been established.

What you will learn

Choose functions and evidence for prevention, detection and response; explain what to pass to the responsible team without treating a finding as proof of compromise or completed remediation.

Prerequisite

Know that AWS resources support applications and data, and permissions control operations on them. No AWS account or key is needed. Read fictional summaries without submitting logs, credentials or real security data.

Prevention, detection and response answer different questions

Prevention reduces the likelihood or impact of problems, for example by withholding unnecessary permissions. Detection observes activity or configuration and identifies signs of trouble. Response investigates those signs and their impact, then has authorized people take appropriate protective or recovery actions and verify the outcome. Prevention does not remove the need for detection and response.

An Amazon GuardDuty finding, as used in this lesson, reports a potential security issue. An incident involves a violation, or an imminent threat of violating security policies or practices. A generated finding differs from a substantiated conclusion about compromise, damage or cause. Handle suspicions under the organization's response procedures while keeping unconfirmed conclusions unconfirmed.

A fictional reporting server receives a suspicious-activity report

A fictional organization processes internal reports on server R. An analyst receives these three summaries about the same R in learning account A and Region A. A Region is a geographical area for resources. F1, T1 and C1 are reading examples, not reproductions of product screens or real logs.

Three fictional summaries: supplied facts and remaining gaps
ItemWhat this example suppliesWhat it does not establish alone
F1: threat findingSigns of suspicious activity involving server R during the period.Compromise, impact scope or completed response.
T1: activity recordOne request to change R's traffic configuration was denied during the same period.Outcomes of other requests or a causal link to F1.
C1: configuration recordsThe supplied configuration items for R show no difference at the two compared points.Unrecorded state, intervening or later changes, or internal application state.

Connect controls, evidence, threat indicators and response

AWS Identity and Access Management (IAM) is a representative access-control service. AWS CloudTrail records activity, AWS Config provides resource configuration and history, and Amazon GuardDuty detects potential threats. They provide different information. These four illustrate roles; they are not an inventory of every AWS security service.

Connect functions: prevention still needs observation and response decisions
Connect functions: prevention still needs observation and response decisions1: Prevent: limit access — Keep observing even when controls exist — Detect: assess signs. 2: Detect: assess signs — Investigate indicators; owners decide on action — Respond: act safely12Prevent: limitaccessDetect: assesssignsRespond: act safelyConnect functions: prevention still needs observation and response decisions1: Prevent: limit access — Keep observing even when controls exist — Detect: assess signs. 2: Detect: assess signs — Investigate indicators; owners decide on action — Respond: act safely12Prevent: limitaccessDetect: assesssignsRespond: actsafely
  1. Prevent: limit access → Detect: assess signsKeep observing even when controls exist
  2. Detect: assess signs → Respond: act safelyInvestigate indicators; owners decide on action
Prevent: limit access
Use controls such as IAM to consider who may do what.
Detect: assess signs
Distinguish GuardDuty findings from activity and configuration evidence.
Respond: act safely
Check authority, evidence and impact; verify the result of any action.

The diagram connects functions. Keep observing despite access controls, investigate indicators and bring them into an owner's response decision. Arrows do not specify log delivery, network paths or configured automatic remediation. Real work is not a one-pass sequence: responding can include revisiting permissions.

Start with the question, not a service name as proof of safety
Representative functionQuestion it supportsKeep separate
IAM: access controlWhich identity may perform which actions on which resources?Sign-in differs from permission for an operation. Permission does not prove that activity was legitimate.
CloudTrail: activity recordsWhich operation was requested and which outcome was recorded?An operation name is not proof of success or cause. Check event categories, period and destination.
Config: configuration and evaluationWhat configuration was recorded? Does it meet selected criteria?Recording, rule evaluation and remediation outcomes are different; recording and criteria have scope.
GuardDuty: threat detectionWhich potential security issue was identified?A finding alone does not prove compromise or recovery. Check subject, features and data-source conditions.

Do not assign each service to exactly one stage. Activity and configuration records can support both discovering an issue and investigating a response. Combining Config rule evaluation with remediation still requires separate checks of actions, permissions and results. A displayed feature name does not establish your organization's enablement, collection scope, costs or success.

Change a condition and choose the next conclusion

Choose an option to see the result and explanation.

F1 arrives. What is the appropriate initial reading?
Read the answer and reason

Answer: Treat it as a potential issue; verify the subject, content and time with the responsible team · Use the finding as an investigative lead and separate evidence from unknowns. Do not ignore a serious suspicion; follow the organization's response process.

Add two facts: T1 was denied, and C1 has no difference at the two points. Is F1 now a confirmed false positive?
Read the answer and reason

Answer: No; check each item's scope and seek additional evidence · A denied request, supplied configuration and a suspicious-activity indicator are different information. Keep cause and impact unverified while aligning subject and collection scope.

Change the question: how was R's traffic configuration recorded earlier? Which evidence should you examine first?
Read the answer and reason

Answer: Config history aligned to the resource, time and recording scope · For a configuration question, establish recording conditions and read that history. Who changed it and whether traffic succeeded remain separate questions.

An additional evaluation shows R does not meet a selected Config rule. What is still required?
Read the answer and reason

Answer: Assign ownership and an appropriate action; check authority, impact and results if it is performed · Do not infer success from evaluation. Manual or automated actions need appropriate configuration, execution and verification of the required state.

No finding is displayed for another period. What follows?
Read the answer and reason

Answer: Safety remains unverified; check subject, period, features and collection conditions · First establish what could be observed. Use other required records and checks in the assessment.

Prepare a useful handoff, then choose further reading

  1. Retain the necessary subject, period, sources and observed facts. Do not submit real data or secrets to this site.
  2. Separate hypotheses, impact and collection gaps from facts when handing over. Assess urgency under your organization's procedures.
  3. If a change is needed, review authority, business impact, evidence preservation and a way to recover.
  4. Record actions and outcomes, business/data recovery and remaining unknowns. Closing a notification alone does not resolve the issue.

Continue to IAM for permissions, CloudTrail for activity records, or Config for configuration and evaluation. Use the related lessons at the end to revisit responsibility and investigation when needed.

Check, then explain in your own words

  1. State the prevention, detection and response questions for server R.
  2. Separate what F1's indicator, T1's denial and C1's unchanged values establish from what remains unknown.
  3. Explain the next decision using evidence, ownership and action results rather than feature names or notifications alone.
Check your understanding: Someone says, 'we have findings and records.' What should you check before concluding compromise or recovery?

Align subject, time, content and collection scope; separate observations, hypotheses and unknowns. Establish the responsible team's investigation or action and verify its results.

A potential-issue report, activity/configuration records, and actual compromise, impact or recovery are different facts.

Sources and conditions checked

The official references used in this lesson are collected here with their scope, conditions and corresponding sections. Independent expert review and live AWS comparison have not been performed.