Skip to main content

Recommendations on sending firewall logs to Security Data Lake for MDR

GravityZone Security Data Lake extends Bitdefender MDR by ingesting and normalizing additional telemetry sources. This provides analysts with expanded visibility and enriched datasets for investigation and detection workflows, enabling the MDR team to:

  • Use third-party telemetry data to perform deeper investigations and hunt threats more effectively.

  • Improve investigation speed through enriched log context, risk scoring, and historical context.

  • Reduce false positives by applying asset-based risk prioritization and filtering out noise.

  • Enhance response effectiveness by providing analysts with greater context and supporting evidence.

GravityZone Security Data Lake for MDR currently supports ingestion of firewall telemetry to extend the data available to the SOC, with additional telemetry categories planned for the future. The following firewalls are currently supported:

Optimal logging scope for firewalls

Firewall integrations should provide detailed network connection telemetry, including action taken, connection outcome, and associated threat context. This data supports the MDR team's detection of lateral movement, data exfiltration, command-and-control activity, and potential account compromise.

To ensure consistent normalization and effective alerting, you must collect and forward the following log categories:

  1. Traffic / connection logs (Allow and Deny)

    • Session start and end events

    • Policy action (accept, deny, drop, reset)

    • Source and destination IP addresses

    • Source and destination ports

    • Protocol

    • Rule or policy name / ID

    • Interface or zone

    • Bytes and packets transferred

    These logs provide the baseline dataset required for connection-level detection, anomaly identification, and data exfiltration analysis.

  2. Threat / security logs

    • IPS / IDS alerts

    • Malware / antivirus detections

    • Vulnerability signatures

    • URL filtering events

    • Botnet / C2 detections

    • Application control blocks

    Each event should include severity, signature or threat ID, threat name, category, and disposition (blocked or allowed).

  3. System / administrative / configuration logs

    • Admin login success and failure

    • Privilege changes

    • Configuration or policy changes

    • Commit events

    • High availability failover events

    • Time changes

    • Logging subsystem errors

    These logs enable administrative activity monitoring and configuration integrity alerts.

Note

While sending all the log categories listed above ensures optimal detection and alerting, it can also generate a high volume of traffic and related infrastructure costs. For example, Allow traffic/connection logs are useful in detecting data exfiltration attempts but also record high volumes of benign traffic, so omitting these logs may help reduce costs. We recommend that you choose a mix of logs that balances detection coverage with ingestion, storage, and processing costs.

It is also strongly recommended to collect and forward the following log categories if they are available:

  • VPN logs (tunnel up / down, authentication, assigned IPs, user attribution)

  • NAT translation logs (original and translated IP / port mappings)

  • DHCP logs (if provided by the firewall)

  • DNS logs (if the firewall acts as DNS resolver)

Field-level requirements for proper normalization

For accurate mapping onto the Security Data Lake Information Model, logs must consistently include the following fields:

  1. Connection context

    • Timestamp (synchronized with NTP and aligned in UTC)

    • Device identifier (hostname or serial number)

    • Log subtype (traffic, threat, system)

    • Source IP

    • Destination IP

    • Source port

    • Destination port

    • Protocol

  2. Decision and policy context

    • Action (allow, deny, drop, reset)

    • Rule or policy name / ID

    • Zone or interface

    • Direction (inbound / outbound), if available

  3. Volume metrics

    • Bytes transferred

    • Packets transferred

    • Session duration

  4. Threat context (for security logs)

    • Signature or threat ID

    • Threat name

    • Severity

    • Category

    • Outcome (blocked or allowed)

  5. Identity context (if available)

    • Username

    • Application

    • URL or domain

    • NAT - original and translated fields

Vendor-specific notes

  • Palo Alto: Ensure traffic, threat, and system logs are forwarded. These log categories map directly to the Palo Alto Illuminate content pack.

  • FortiGate: Send logs in Syslog or raw / plaintext format. Do not use CEF. Include both traffic and security event logs.

  • pfSense / OPNsense: Forward firewall filter logs, authentication logs (GUI / SSH), and optionally DNS and DHCP logs.

  • Other supported vendors (for example, Checkpoint, Cisco ASA, Juniper SRX, SonicWall): Forward traffic, threat, and administrative log categories via supported input types.