Performing Cloud Incident Containment Procedures

Performing Cloud Incident Containment Procedures

Overview

Cloud incident containment requires cloud-native approaches that differ significantly from traditional on-premises response. Containment procedures must leverage platform-specific controls including security groups, IAM policies, network ACLs, and service-level isolation to restrict compromised resources while preserving forensic evidence. According to the 2025 Unit 42 Global Incident Response Report, responding to cloud incidents requires understanding shared responsibility models, ephemeral infrastructure, and API-driven operations. Effective containment involves credential revocation, resource isolation, evidence snapshot creation, and automated response playbook execution.

When to Use

• When conducting security assessments that involve performing cloud incident containment procedures
• When following incident response procedures for related security events
• When performing scheduled security testing or auditing activities
• When validating security controls through hands-on testing

Common Misconfigurations & Verification

• Containment that destroys evidence: terminating/stopping an instance before snapshotting wipes the instance-store volume and RAM. Always create-snapshot/New-AzSnapshot/gcloud compute disks snapshot and capture memory before isolation, and prefer quarantine SG/NSG over stop. Tag and Object-Lock the snapshot so cleanup automation can't delete it.
• Token revocation that isn't: disabling an IAM user's access key or AccountEnabled=$false does not kill already-issued STS/refresh tokens — they remain valid until expiry. For AWS attach the aws:TokenIssueTime deny condition (and update the role trust policy) ; for Azure run Revoke-AzureADUserAllRefreshToken; for GCP the SA key delete leaves active OAuth tokens until TTL. Verify with CloudTrail/Activity/Audit logs showing the principal's calls now AccessDenied.
• Isolation gaps: an EC2/VM SG/NSG change leaves existing established connections alive in some stacks, and removing one of several attached SGs/NICs leaves another path open. Confirm all NICs/SGs, NACLs, and public IPs are covered; for serverless, set Lambda reserved concurrency to 0 and delete event-source mappings.
• Cross-account / persistence missed: attackers leave behind backdoor IAM roles, added trust relationships, new access keys, and Lambda/EventBridge triggers. Containing the obvious identity but not auditing iam:CreateRole/CreateAccessKey/AttachRolePolicy since the compromise window lets them walk back in.
• Concrete verification the action worked: replay a known-bad API call and confirm it now fails; check the instance is unreachable on its prior listeners; confirm the snapshot exists with COMPLIANCE-mode lock; and verify CloudTrail/Activity logs are still flowing to write-protected storage (containment must not break logging).

Prerequisites

• Familiarity with incident response concepts and tools
• Access to a test or lab environment for safe execution
• Python 3.8+ with required dependencies installed
• Appropriate authorization for any testing activities

AWS Containment Procedures

1. Credential Compromise Containment

# Disable compromised IAM user access keys
aws iam update-access-key --user-name compromised-user \
  --access-key-id AKIA... --status Inactive

# List and disable all access keys for user
aws iam list-access-keys --user-name compromised-user
aws iam delete-access-key --user-name compromised-user --access-key-id AKIA...

# Attach deny-all policy to compromised user
aws iam put-user-policy --user-name compromised-user \
  --policy-name DenyAll \
  --policy-document '{
    "Version": "2012-10-17",
    "Statement": [{
      "Effect": "Deny",
      "Action": "*",
      "Resource": "*"
    }]
  }'

# Revoke all active sessions for IAM role
aws iam put-role-policy --role-name compromised-role \
  --policy-name RevokeOldSessions \
  --policy-document '{
    "Version": "2012-10-17",
    "Statement": [{
      "Effect": "Deny",
      "Action": "*",
      "Resource": "*",
      "Condition": {
        "DateLessThan": {"aws:TokenIssueTime": "'$(date -u +%Y-%m-%dT%H:%M:%SZ)'"}
      }
    }]
  }'

# Invalidate temporary credentials by updating role trust policy
aws iam update-assume-role-policy --role-name compromised-role \
  --policy-document '{"Version":"2012-10-17","Statement":[]}'

2. EC2 Instance Isolation

# Create quarantine security group (no inbound, no outbound)
aws ec2 create-security-group --group-name quarantine-sg \
  --description "Quarantine - No traffic allowed" --vpc-id vpc-xxxxx

# Remove all rules from quarantine SG (default allows outbound)
aws ec2 revoke-security-group-egress --group-id sg-quarantine \
  --ip-permissions '[{"IpProtocol":"-1","FromPort":-1,"ToPort":-1,"IpRanges":[{"CidrIp":"0.0.0.0/0"}]}]'

# Take forensic snapshot BEFORE containment
aws ec2 create-snapshot --volume-id vol-xxxxx \
  --description "Forensic snapshot - IR Case 2025-001" \
  --tag-specifications 'ResourceType=snapshot,Tags=[{Key=IR-Case,Value=2025-001}]'

# Apply quarantine security group to compromised instance
aws ec2 modify-instance-attribute --instance-id i-xxxxx \
  --groups sg-quarantine

# Tag instance as compromised
aws ec2 create-tags --resources i-xxxxx \
  --tags Key=IR-Status,Value=Contained Key=IR-Case,Value=2025-001

# Capture memory (if SSM agent available)
aws ssm send-command --instance-ids i-xxxxx \
  --document-name "AWS-RunShellScript" \
  --parameters 'commands=["dd if=/dev/mem of=/tmp/memory.dump bs=1M"]'

3. S3 Bucket Containment

# Block all public access
aws s3api put-public-access-block --bucket compromised-bucket \
  --public-access-block-configuration \
  BlockPublicAcls=true,IgnorePublicAcls=true,BlockPublicPolicy=true,RestrictPublicBuckets=true

# Apply deny policy to bucket
aws s3api put-bucket-policy --bucket compromised-bucket \
  --policy '{
    "Version": "2012-10-17",
    "Statement": [{
      "Sid": "DenyAllExceptForensics",
      "Effect": "Deny",
      "NotPrincipal": {"AWS": "arn:aws:iam::ACCOUNT:role/IR-Forensics"},
      "Action": "s3:*",
      "Resource": ["arn:aws:s3:::compromised-bucket","arn:aws:s3:::compromised-bucket/*"]
    }]
  }'

# Enable versioning to preserve evidence
aws s3api put-bucket-versioning --bucket compromised-bucket \
  --versioning-configuration Status=Enabled

# Enable Object Lock for evidence preservation
aws s3api put-object-lock-configuration --bucket evidence-bucket \
  --object-lock-configuration '{
    "ObjectLockEnabled": "Enabled",
    "Rule": {"DefaultRetention": {"Mode": "COMPLIANCE", "Days": 365}}
  }'

4. Lambda Function Containment

# Set reserved concurrency to 0 (stops all invocations)
aws lambda put-function-concurrency --function-name compromised-function \
  --reserved-concurrent-executions 0

# Remove all event source mappings
aws lambda list-event-source-mappings --function-name compromised-function
aws lambda delete-event-source-mapping --uuid mapping-uuid

Azure Containment Procedures

1. Identity Containment

# Revoke all user sessions
Revoke-AzureADUserAllRefreshToken -ObjectId "user-object-id"

# Disable user account
Set-AzureADUser -ObjectId "user-object-id" -AccountEnabled $false

# Reset user password
Set-AzureADUserPassword -ObjectId "user-object-id" -Password (
  ConvertTo-SecureString "TempP@ss!" -AsPlainText -Force
) -ForceChangePasswordNextLogin $true

# Block sign-in via Conditional Access (emergency policy)
# Create policy blocking user from all cloud apps

# Revoke Azure AD application consent
Remove-AzureADServiceAppRoleAssignment -ObjectId "sp-object-id" \
  -AppRoleAssignmentId "assignment-id"

2. VM Isolation

# Create Network Security Group with deny-all rules
$nsg = New-AzNetworkSecurityGroup -ResourceGroupName "rg" -Location "eastus" `
  -Name "quarantine-nsg" `
  -SecurityRules @(
    New-AzNetworkSecurityRuleConfig -Name "DenyAllInbound" -Protocol * `
      -Direction Inbound -Priority 100 -SourceAddressPrefix * `
      -SourcePortRange * -DestinationAddressPrefix * `
      -DestinationPortRange * -Access Deny,
    New-AzNetworkSecurityRuleConfig -Name "DenyAllOutbound" -Protocol * `
      -Direction Outbound -Priority 100 -SourceAddressPrefix * `
      -SourcePortRange * -DestinationAddressPrefix * `
      -DestinationPortRange * -Access Deny
  )

# Take disk snapshot for forensics
$vm = Get-AzVM -ResourceGroupName "rg" -Name "compromised-vm"
$snapshotConfig = New-AzSnapshotConfig -SourceUri $vm.StorageProfile.OsDisk.ManagedDisk.Id `
  -Location "eastus" -CreateOption Copy
New-AzSnapshot -ResourceGroupName "rg" -SnapshotName "forensic-snap" -Snapshot $snapshotConfig

# Apply quarantine NSG to VM NIC
$nic = Get-AzNetworkInterface -ResourceGroupName "rg" -Name "compromised-nic"
$nic.NetworkSecurityGroup = $nsg
Set-AzNetworkInterface -NetworkInterface $nic

3. Storage Account Containment

# Remove network access
Update-AzStorageAccountNetworkRuleSet -ResourceGroupName "rg" `
  -Name "storageaccount" -DefaultAction Deny

# Regenerate access keys
New-AzStorageAccountKey -ResourceGroupName "rg" -Name "storageaccount" -KeyName key1
New-AzStorageAccountKey -ResourceGroupName "rg" -Name "storageaccount" -KeyName key2

# Revoke all SAS tokens (by rotating keys)
# Enable immutability for evidence preservation

GCP Containment Procedures

1. IAM Containment

# Remove all IAM bindings for compromised service account
gcloud projects get-iam-policy PROJECT_ID --format=json > policy.json
# Edit policy.json to remove compromised account bindings
gcloud projects set-iam-policy PROJECT_ID policy.json

# Disable service account
gcloud iam service-accounts disable SA_EMAIL

# Delete service account keys
gcloud iam service-accounts keys list --iam-account SA_EMAIL
gcloud iam service-accounts keys delete KEY_ID --iam-account SA_EMAIL

2. Compute Instance Isolation

# Create forensic snapshot
gcloud compute disks snapshot compromised-disk \
  --snapshot-names forensic-snap-$(date +%Y%m%d) \
  --zone us-central1-a

# Apply firewall rule to deny all traffic
gcloud compute firewall-rules create quarantine-deny-all \
  --network default --action DENY --rules all \
  --target-tags quarantine --priority 0

# Tag compromised instance
gcloud compute instances add-tags compromised-instance \
  --tags quarantine --zone us-central1-a

# Remove external IP
gcloud compute instances delete-access-config compromised-instance \
  --access-config-name "External NAT" --zone us-central1-a

Evidence Preservation Best Practices

1. Always snapshot before containment - Create disk/volume snapshots before network isolation
2. Preserve CloudTrail/Activity Logs - Copy logs to write-protected storage
3. Document all actions - Timestamp every containment step taken
4. Use break-glass procedures - Pre-establish emergency access for IR team
5. Maintain forensic chain of custody - Hash all evidence artifacts

MITRE ATT&CK Cloud Techniques

References

• Sygnia: Cloud Incident Response Best Practices
• Unit 42: Responding to Cloud Incidents
• Wiz: Cloud Incident Response Checklist
• Microsoft Cloud Security Benchmark - IR