GuardDuty

GuardDuty

See -> Amazon GuardDuty Best Practices

• Analyzes selected logs to produce observable records of suspicious activities (findings)
  • VPC Flow Logs (does not require you to enable VPC Flow Logs)
  • CloudTrail (does not require you to create a trail)
  • DNS queries (from VPC DNS resolvers)
  • ECS Runtime Monitoring
• Regional: can aggregate via CloudWatch Events to push to a central store
• Based on threat intelligence (IP addresses and domain-based lists) and machine learning
  • Receives feeds from 3^rd parties like Proofpoint, CrowdStrike, and AWS Security for known malicious domains/IP addresses
• Sample malicious behaviour flagged
  • Unusual API calls, calls from a known malicious IP
  • Attempts to disable CloudTrail logging
  • Unauthorized deployments
  • Compromised instances
  • Reconnaissance by attackers
  • Port scanning, failed logins

Components

Component	Description
Detector	Consumes information and generates findings within a specific AWS account and region Suspension: GuardDuty allows you to disable or suspend the detector in an account, on a per-region basis Suspending a detector stops the detection of new findings but keeps information about previously detected findings Disabling a detector stops the detection and deletes all related findings
Finding	Attributes ID time of the finding severity finding type affected resources action details Naming convention: name contains, in order Threat purpose (objective of the attack) Resource type affected by the suspicious activity Threat family name and variant (optional) Artifact (type of resource owned by the attacker)
Master Account	The master account receives findings from other (member) accounts Has the capability to: manage (enable, disable, or suspend) the detectors manage the findings workflow (archive and create suppression rules) configure threat lists for member accounts

Delivery of Findings

Type	Description
Archive	You can automatically send findings to an archive by creating suppression rules Each suppression rule is represented by a filter When a finding matches the filter, the finding is automatically marked as archived
GuardDuty Console	Last 90 days
S3	GuardDuty will export active findings (not suppressed) within 5m of its first occurrence If an active finding receives recurrent events, you can configure how frequently those events are reported (every 15m, 1h, 6h) Exported files of findings are encrypted by KMS
CloudWatch Events	Send new findings to CloudWatch Events every 5mins, and updated findings every 6 hours (default)

AWS Security Incident Response

• AWS Security Incident Response combines the power of automated monitoring and investigation, accelerated communication and coordination, and direct 24/7 access to the AWS Customer Incident Response Team (CIRT) to quickly prepare for, respond to, and recover from security events
• New AWS Security Incident Response helps organizations respond to and recover from security events

---

1. AWS Security Services Best Practices ↩