Network Exposure

This section provides a detailed overview of network exposure capabilities in Tenable One Cloud Exposure.

Network exposure in Cloud Exposure is evaluated through a verified reachability model. The model attempts to reach resources, and if successful, the resources is classified as public.

This model integrates the following components:

  • Network Evaluation — Static analysis of cloud configurations to identify potentially internet-facing resources.
  • Network Scanner — The authoritative component that performs active connection attempts to validate exposure.
  • Network Endpoints — Records of the specific host and port combinations for each confirmed exposed resource.

Together, these components reduce false positives and provide a clear picture of real external risk.

Network Evaluation

Network Evaluation is the initial component Cloud Exposure uses to assess potential exposure. It analyzes cloud resource configurations collected during the orchestration cycle to determine whether a resource appears internet-facing.

The role of Network Evaluation depends on the resource type:

  • Direct Resources — For resources like EC2 instances or RDS databases, Network Evaluation acts as a discovery mechanism to identify potential exposures for verification by the Network Scanner. For these resources, a public status is only confirmed after a successful scan. See the list of supported resources.
  • Indirect Resources — For resources behind load balancers or API gateways, Network Evaluation remains the primary method for determining if a resource is public.

Because it relies on configuration data, Network Evaluation runs automatically with each cycle and requires no additional setup. However, this method cannot account for external security layers such as WAFs, third-party firewalls, or managed proxies.

For example, an EC2 instance is launched with a public IP and a security group that allows inbound traffic on port 22 from 0.0.0.0/0. During the next orchestration cycle, Cloud Exposure retrieves this configuration and flags the instance as potentially exposed.

If the instance is protected by a firewall or proxy that blocks the connection, the Network Scanner will fail to establish a handshake. In this case, the instance is marked as not public. The resource is only classified as public if the scanner successfully verifies reachability.

Network Evaluation Coverage and Limitations

Network Evaluation supports all integrated cloud providers. It evaluates a range of resources that may be internet-facing, including:

  • Virtual machines and instances with public IPs.
  • Load balancers and gateways.
  • Other networking components that control access to cloud workloads.

Limitations:

  • IPv4 Support — Only IPv4 is supported; IPv6 is not currently supported.
  • Container Networking — Kubernetes services and container networking objects are not currently included in evaluation.
  • Third-Party Controls — Network Evaluation cannot detect access controlled by third-party firewalls or managed proxies. These are instead addressed by the Network Scanner through verified reachability.

Network Scanner

The Network Scanner is the authoritative component of network exposure analysis, once enabled. After Network Evaluation flags resources as potentially internet-facing, the Network Scanner validates those resources by attempting direct connections from the internet.

The scanner operates at two layers:

  • Transport layer — Attempts a TCP connection to confirm the resource is reachable from the internet. See supported resources.
  • Application layer — For supported compute resources on confirmed HTTP and HTTPS endpoints, the scanner sends an HTTP GET request and captures the response — including HTTP status code, SSL/TLS protocol version, a screenshot of the landing page, and the HTML response. See supported resources.

Only direct resources with an IP address and port are scanned. Indirect resources, such as those behind load balancers (ALB, NLB) or API Gateways, are not currently supported and continue to rely on Network Evaluation for exposure status.

For example, an EC2 instance is flagged because its security group allows inbound traffic on port 443 from 0.0.0.0/0. The Network Scanner attempts to connect on port 443. If successful, Cloud Exposure creates a Network Endpoint and classifies the resource as Public. Because the EC2 instance is also supported for application layer scanning, the Scanner sends a GET request and records the HTTP status, SSL/TLS version, landing page screenshot, and HTML response.

Impact on Finding Severity

The Network Scanner directly affects the severity of findings.

For example, a public resource such as an EC2 instance was classified as Public. The following table shows how the severity is affected:

Severity Conditions
Critical Never
High Contains a Network Endpoint + Vulnerability severity is Critical + Permission severity is High
Medium One of the following:
- Contains a Network Endpoint + Vulnerability severity is Critical
or
- Contains a Network Endpoint + Permission severity is High
Low Determined as internet-facing by Network Evaluation

Supported Resources

The following direct resource types support Network Scanner transport-layer and/or application-layer scanning:

Provider Resource Name Transport-Layer Scanning Application-Layer Scanning
AWS DocumentDB Instance ✔
EC2 Instance ✔ ✔
ECS Service ✔ ✔
Elastic Beanstalk Environment ✔
Classic Load Balancer ✔ ✔
Network Load Balancer ✔ ✔
Application Load Balancer ✔ ✔
Neptune Instance ✔
RDS Instance ✔
Redshift Cluster ✔
Azure Application Gateway ✔
Network Load Balancer ✔ ✔
Virtual Machine ✔ ✔
Virtual Machine Scale Set Instance ✔ ✔
App Service ✔
GCP Load Balancer ✔
VM Instance ✔ ✔
Cloud SQL Instance ✔
Cloud Run Service ✔
App Engine ✔
OCI Compute Instance ✔ ✔
Network Load Balancer ✔ ✔
Load Balancing Load Balancer ✔ ✔
Note: Indirect resources, including those sitting behind load balancers (ALB, NLB, Application Gateway) or API gateways, are not supported by the Network Scanner. These resources use Network Evaluation (static configuration analysis) to determine if they are public.

The Network Scanner originates traffic from the following public IP addresses:

IP Address
3.147.184.122
3.19.220.21
15.205.53.205 (United States Gov Cloud)
3.30.64.189 (United States Gov Cloud)

Network Endpoints

Network Endpoints are created when the Network Scanner confirms that a resource flagged by Network Evaluation is reachable from the internet.

Each Network Endpoint represents a unique combination of a destination host (IP address or DNS name) and port. For supported compute resources, confirmed HTTP and HTTPS endpoints also include Application layer data captured by the scanner. These endpoints provide the deterministic evidence required by the verified reachability model to classify a direct resource as Public.

Keep in mind:

  • Direct resources only — Endpoints are created only for direct resources with an IP address and port.
  • Classification trigger — For supported direct resources, the existence of a Network Endpoint is the only way a resource is classified as Public.
  • Recent results — Only the most recent scan results are shown on the Network Endpoints page; endpoints are not currently stored historically.
  • Finding impact — While endpoints are scan results and not findings themselves, they determine whether the severity of network-based findings is affected.

How Network Endpoints Are Created

When Network Evaluation flags a resource as potentially internet-facing, the Network Scanner attempts direct connections.

If the connection attempt succeeds, Cloud Exposure creates a Network Endpoint for the specific host and port. This confirmed reachability automatically updates the resource status to Public and adjusts the severity of any relevant finding.

For supported compute resources on HTTP and HTTPS endpoints, the scanner also sends a GET request and records the Application layer response data.

For example, an EC2 instance may be flagged because its security group allows inbound traffic on port 443 from 0.0.0.0/0. If the scanner connects successfully, Cloud Exposure records <instance-IP>:443 as a Network Endpoint. This confirms the resource is both configured as exposed and actively reachable, resulting in a Public classification.

Note: If you want to prevent specific resources from being scanned or appearing as endpoints, you can manage them via Exclusions.

Port Scanning Logic

The Network Scanner tests resources on a list of Common Service Ports.

  • If the total number of open ports on a resource (well-known and other) is less than or equal to the number of well-known ports, the scanner tests all of them.
  • If the total number of open ports exceeds the number of well-known ports, the scanner tests only the well-known ports.

This approach ensures broad coverage without excessive scanning overhead.

Common Service Ports

Application Protocol Port Protocol Type
AzureSqlManagedInstance 3342 Tcp
Cassandra 7000 Tcp
CassandraSsl 7001 Tcp
Dns 53 Tcp
Elasticsearch 9200 Tcp
Ftp 21 Tcp
Geneve 6081 Tcp
Grafana 3000 Tcp
Http
  • 80
  • 591
  • 7990
  • 8000
  • 8001
  • 8008
  • 8080
  • 8081
  • 8082
  • 8088
  • 8800
  • 8808
  • 8880
  • 8881
  • 8882
  • 8888
  • 9000
  • 9090
  • 10000
Tcp
Https
  • 443
  • 1443
  • 2443
  • 3443
  • 4443
  • 5443
  • 6443
  • 7443
  • 8443
  • 11434
Tcp
Imap 143 Tcp
Imaps 993 Tcp
IpSecKeyExchange 500 Tcp
Istio
  • 15006
  • 15010
  • 15012
  • 15021
Tcp
Kubernetes
  • 10250
  • 10255
  • 6443
Tcp
Ldap 389 Tcp
Memcached 11211 Tcp
MongoDb 27017 Tcp
Mssql 1433 Tcp
MysqlAurora 3306 Tcp
Nfs 2049 Tcp
OracleRds 1521 Tcp
Pop3 110 Tcp
Pop3S 995 Tcp
Postgresql 5432 Tcp
Rdp 3389 Tcp
Redis 6379 Tcp
Redshift 5439 Tcp
Sftp 990 Tcp
Smb 445 Tcp
Smtp 25 Tcp
Smtps 465 Tcp
Ssh 22 Tcp
Telnet 23 Tcp
UPnP 5000 Tcp
VNC 5900 Tcp
WinrmHttp 5985 Tcp
WinrmHttps 5986 Tcp