Configuring Sites

Prev Next

Configure a Site to group aspects of an appliance's behavior. Sites are used with Gateways, LogForwarders, Metrics Aggregators, and Connectors.

New systems include a default Site. This Site is tagged builtin and has entitlement-based routing (EBR) enabled.


Prerequisites

Gather the following information before you configure a Site:

  • The network subnets for each Site (if using site-based routing).

  • Whether more than one Gateway will be assigned to this Site. See the High availability section for more information.

  • An understanding of how and where user traffic will be routed.

  • Name resolver details, such as DNS resolver IP address or name resolver configuration details. See the DNS and name resolution section for more information.

  • An available and valid license for the new Site.

Configure the following before adding a Site:

  • Trusted certificates. Add any certificates required to use name resolvers to Trusted Certificates (System > Trusted Certificates).


Adding a Site

To add or edit a Site:

  1. Go to Sites (System > Sites).

  2. Select +Add and complete the following fields:

    • Name. The name of the Site as it appears in the system and in the client's list of entitlements. The name can match your physical site name (such as data center R02), but it refers to the AppGate ZTNA Site you are configuring. You can have more than one AppGate ZTNA Site per physical site.

    • Notes. Optional. Enter any notes for the Site.

    • Tags. Click +Add to add tags to the Site.

  3. Continue to complete the fields in the following sections:

Site Data

  • Geolocation. Enter the geolocation. You can copy this value from apps such as Google Maps.

NOTE

If you use the Use for nearest Site selection option, you must enter this information.

  • Description. A short description of the Site that will appear in the client's Site Details.

    The following HTML tags are supported: 'a', 'b', 'i', 'strike', 's', 'strong', 'small', 'del', 'sub', 'sup', 'br', 'code', 'ol', 'ul', 'li', 'dl', 'dt', 'dd'

Tunneling

These fields define tunnel configuration for sending user traffic to and from the Site's Gateways. See the Tunnel architectures section for more information.

  • Tunnel Protocol (VPN). The tunnel can use either TCP or UDP. The choice of protocol does not affect tunneled traffic from the Gateway to protected resources.

    • TLSv1.3. Use TCP-based tunneling, as defined in the appliance's System TLS Connection fields.

    • QUIC. UDP-based tunneling is not recommended. It offers no advantages over TCP-based tunneling and may limit functionality in some cases.

You may need to review Secure Tunnel Settings in the Gateway appliance configuration.

  • Tunnel Traffic.

  • Disable source NAT on Gateways. When source NAT is disabled, the Site sees tunneled traffic originating from the user's assigned IP pool address. Source NAT (SNAT) is used by default. It should be used when protected networks do not allow client tun IP addresses, or if connecting multiple Gateways to the same network without using IP Pool Mappings. You cannot use reverse IP access (such as TCP down) in entitlements when source NAT is in use.

When using SNAT, the relationship between the user's identity and local network traffic is lost. In some security scenarios, you must maintain this relationship to establish a detailed audit trail. A Site option lets you enable NAT-IP and NAT-port logs, which adds two additional records to the IP access audit log.

NOTE

Without source NAT, you may need to configure the routing in your network (for IP Pool ranges to the Gateways).

  • IP Pool Mappings. An alternative IP pool can be used to map the client's IP address for this Site. Mapping is done by Translation (10.1.1.17>10.1.100.17) or Allocation (10.1.1.17>10.1.100.1).

This field is optional and must be set only to avoid IP routing issues in certain situations. See IP pool mapping for more information. There is also an example in Routing client traffic.

  • Log NAT-IP and NAT-port. This adds two extra audit log fields, which can affect performance. Enable this setting only when needed. When Gateways use source NAT, two additional fields (nat_src_ip and nat_src_port) are added to the ip_access audit log. This applies only in the case of ALLOW rules and only for TCP and UDP protocols. See audit log detail for more information.

  • Alternative default route for all user traffic. Diverts all user tunneled traffic to its own default gateway. Normally, when using Route all traffic through tunnel, the default gateway is the same as the appliance’s. When set, this IP address is used for tunneled user traffic, except traffic destined for the Gateway's local subnet. That traffic is routed to the appliance’s default Gateway only if the appliance is configured with a static IP address (not DHCP).

Appliance-generated traffic continues to use the appliance’s default Gateway (for DNS requests, scripts making external calls, and similar traffic).

  • IPv4. IPv4 address of the default Gateway for user tunneled traffic.

  • IPv6. IPv6 address of the default Gateway for user tunneled traffic.

See the example in Routing client traffic.

  • Terminate tunneled HTTPS traffic using this PKCS #12 file. To decrypt HTTPS streams and reverse-proxy traffic to protected hosts, a PKCS #12 file is required.

The URL access feature operates as a reverse proxy. The appliance requires a PKCS#12 file to terminate HTTP(S) connections, decrypt and URL-filter the traffic, and reestablish an HTTPS connection to protected hosts. URL access supports multiple PKCS#12 files, each containing a certificate signed by a trusted CA and the private key. A wildcard certificate is an alternative to multiple certificates, but limit its scope due to the security implications if it is compromised.

  • PKCS #12 Files. Upload one or more PKCS #12 files that contain the relevant certificates and private key to allow AppGate ZTNA to intercept encrypted traffic.

  • File. Select the PKCS #12 file to upload.

  • Password. The password for the PKCS #12 file.

  • Verify Upstream. All upstream servers are checked for valid certificates that match their hostnames by default. See Adding third-party certificates for more details.

To configure URL filtering, once a PKCS #12 file is in place, select the HTTP up action type and optionally select some HTTP methods when configuring your entitlement.

  • Change IP access audit log interval. Change the default time interval between repeating IP access audit log records.

  • Time Interval (seconds). Set the time interval between repeating IP access audit log entries. Default is every 120s. Setting it to 0s will not log IP accesses.

This is an important setting for system reliability. By default, the system produces audit log records for established TCP connections at a repeat interval of 30 seconds. For example, 5,000 users with 10 TCP connections each over 10 hours produces 15 million audit log entries. This can significantly affect disk usage, so increase the interval unless audit requirements dictate otherwise.

The 0-second setting (do not log) is useful when Alternative default route for all user traffic is enabled, as the downstream appliance can handle logging. It is also useful when Route all traffic through tunnel is enabled, as this can result in many connections—for example, web pages with many embedded links.

  • Advanced Tunnel Options.

  • Tunnel traffic to a fallback Site when the primary Site is unavailable. The client tries the fallback Site when it cannot connect to this Site. Provide extra resources on the Gateways; insufficient RAM and CPU will cause the fallback Site to become overloaded with users from the failed Site. See High availability for details.

  • Fallback Site. Select an alternative fallback Site. Access to entitlements must be available on the fallback Site.

  • Use for nearest Site selection. The entitlements' Site can be overridden with this geolocated Site. This must be enabled in the policy’s Override Site setting.

  • Use local client hostname/IP to connect to Gateways. When a client is on the local LAN, it connects using internal hostnames or IPs instead of external hostnames.

  • Associated Public IPs/Networks. The Controller specifies the use of the Local Client Hostname/IP when users connect from any of these public IPs or networks. Multiple IPs may be specified, or a range specified using CIDR format.

NOTE

If no Local Client Hostname/IP has been defined in the Gateway’s Secure Tunnel Settings, the Client Hostname/IP will be used.

Client Routing

These fields define client routes, directing user traffic to one of the Site's Gateways. See Routing client traffic for more information.

  • Type of Routing.

  • Entitlement-based routing. In entitlement-based routing (EBR), the Gateway sends resolved entitlements to the client. These entitlements are used to configure required routes to Sites. EBR is enabled by default.

  • Subnet-based routing. In subnet-based routing (SBR), the client uses static subnets to configure routes to Sites. SBR can be used with EBR. Add a subnet and a description to identify its use. If you add 0.0.0.0/0 to capture all traffic to this Site, it is translated into two routes on the client: 0.0.0.0/1 and 128.0.0.0/1.

  • Route all traffic through tunnel (Default Gateway). All other client traffic is routed to this Site in the Collective, which serves as the default Gateway. Add suitable entitlements for this Site.

List of network actions allowing various protocols and ports for traffic management.

  • Excluded Subnets. Use this field to exclude subnets from the default Gateway Site, such as local network IP addresses. Use the same format as for SBR.

NOTE

If Advanced Ringfence is enabled, excluded subnets will be honored and not blocked.

DNS Resolving

These fields configure DNS resolving for the Site. See the DNS and name resolution section for more information.

NOTE

In the case of cloud environments, a DNS server may not be required as cloud resolvers use cloud APIs. These return the required IP addresses directly.

  • Host Resolving. The hosts file on Gateway appliances are used to resolve hostnames within entitlements. Add hostnames to the Hosts File in the Miscellaneous tab of the appliance configuration.

  • DNS Resolvers. A limited default DNS resolver is included to allow Gateways to resolve user entitlements. As you add more, you can set the order in which they are tried using the up and down arrows. Use of the dns:// syntax is supported in entitlement actions. See Hostnames (DNS) for more information about this syntax.

  • Perform AAAA lookups. Resolves IPv6 addresses in addition to IPv4. AAAA lookups are disabled by default to improve performance.

  • Update Interval (seconds). DNS resolver polling frequency (seconds) to get the latest IP address information. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Match Domains (Zone). When a match domain is set, this DNS resolver is only used for matching DNS names, eliminating unnecessary DNS lookups. When no match domain is set, this DNS resolver is used for all DNS lookups. The match domain is also used as the zone when Zone transfers is enabled.

  • DNS Servers. Enter one or more IP addresses for DNS servers. If left empty, the appliances' DNS server settings are used.

  • Zone transfers (to Gateway). Delegate DNS to the Gateway for zones defined in Match Domains. Zone transfers are recommended because the DNS resolver makes a lot of DNS requests.

NOTE

Zone transfers usually require that the feature is enabled on the DNS server for specific IP addresses; in this case, the Gateways.

  • Client DNS auto-configuration. Uses this resolver's DNS settings for clients when no other DNS policy exists. A hidden DNS policy is created with a fallback Site enabled, along with a suitable hidden entitlement. This is assigned automatically when a user has at least one other entitlement for this Site and no other DNS policy assigned.

DNS Forwarding

These fields configure DNS forwarding for the Site. The DNS forwarder provides an alternative means of resolving DNS queries. See the DNS and name resolution section for more information.

NOTE

On mobile operating systems, the DNS forwarder is supported only when used with Route all traffic through tunnel (Default Gateway). This is due to API limitations of mobile operating systems.

DNS forwarding supports the use of the *.company.com syntax (domain:// syntax is deprecated) in entitlement actions. See IP/hostnames for more information about this syntax. IPv4 and IPv6 are optional, but at least one must be completed.

Complete the following fields to configure DNS forwarding for the Site:

  • DNS Forwarder IPv4 address. A random IP address is assigned (100.127.x.y) to the DNS forwarder. This address should then be used in the DNS policy for the clients. This can be changed if you need to assign a different address.

  • DNS Forwarder IPv6 address. Enter an IPv6 address for the DNS forwarder. This address should then be used in the DNS policy for the clients. Add access entitlements in your DNS policy for the IP addresses from above if not using client DNS auto-configuration.

  • DNS Servers. Enter one or more IP addresses for the DNS servers the DNS forwarder will use. These must be recursive, forwarding unknown DNS requests. If left empty, the appliances' DNS server settings are used.

  • Client DNS.

    • Match Domains. The match domains combined with configured DNS servers are used to generate the client DNS configuration. A hidden DNS policy is created with a fallback Site enabled, along with a suitable hidden entitlement. This is assigned automatically when a user has at least one other entitlement for this Site and no other DNS policy assigned.

  • Resolved IPs - Allow Destinations. Resolved IP addresses will used only if they fall within these allowed destination ranges. By default, all IPs are allowed.

Cloud Resolvers

These fields configure cloud resolvers for a Site. See Cloud API resolvers for more information. Use the up and down arrows to set the order within the same resolver type.

List of GCP Resolvers with update intervals displayed for each resolver.

NOTE

If multiple resolvers are configured (of the same type), they are tried one by one until a specified name matches (tag name, security group name, and so on). No other resolvers are tried for the matched names. Make sure the names used are different for each resolver.

Select the tab for the type of resolver you want to configure:

AWS

Complete the following fields to enter parameters to resolve aws:// resource names. See Cloud API resolvers for more information about where to obtain the required parameters.

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the API to get the latest IP address information for hostnames. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Access Method. Select the AWS API access method. If using the instance's IAM role, you will need to create a valid access policy for it to work.

    • Use instance’s IAM role. IAM roles are the preferred access method for the AWS resolver. API calls made by this AppGate ZTNA instance inherit the rights of the referenced role.

    • Use access key. Use this mode only when you have remote Gateways outside of AWS that protect AWS resources.

      • Access Key ID. Enter an AWS Access Key ID.

      • Secret Access Key. Enter the Secret Access Key for the Key ID.

  • Regions. If your Site and VPC span multiple regions, you can configure one AWS resolver to use across all regions instead of defining a separate resolver for each. Use this section to list the codes of regions provided by your AWS account. The resolver queries each region in turn. For information about AWS region codes, see Regions and Zones.

  • Partition. The default AWS partition is aws. There are also aws-us-gov (GovCloud) and aws-cn (China).

  • Amazon Services. Users can enable EC2, EKS, or RDS services and can also enable an API Gateway. By default, the EC2 service will be selected. Enable the  RDS and EKS services only when needed, to avoid permissions warnings and unnecessary downloads.

  • Optional Settings.

    • Use assumed roles. When selected, assumed roles are used. Assumed roles are useful when resources exist in other accounts or regions. For configuration guidance, see Cloud API resolvers.

      • Assumed Roles. Enter as many assumed roles as required.

      • Resolve with master credentials. If selected, the resolver uses any assumed IAM roles and the master account (jump account) IAM role.

    • Disable VPC auto discovery. By default, the name resolver automatically discovers available VPC IDs and resolves names across all of them.

      • VPCs IDs. If not using VPC Auto Discovery, you can add VPC IDs the name resolver will use.

    • Use HTTPS proxy. If selected, all traffic to the AWS API goes through this proxy.

      • Address. Specify an HTTPS proxy in the format: https://<server>:<port>

Azure

Complete the following fields to enter parameters to resolve azure:// resource names:

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the Azure API to get the latest IP address information for hostnames. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Access Method. Select Use Managed Identities if all Gateways on the Sites have Azure system assigned managed identities.

    • Use Managed Identities. See Cloud API resolvers for more information about configuring system assigned managed identities.

    • Use App Registration. See Cloud API resolvers for more information about where to obtain the required parameters.

  • Tenant ID. Enter the directory (tenant) ID obtained in the Azure portal App registration summary screen.

  • Client ID. Enter the application (client) ID be obtained in the Azure portal App registration summary screen.

  • Secret. For Azure App registrations, create a new client secret and copy the value.

  • Disable Subscription ID Auto Discovery. Select to disable auto-discovery of the IDs specified. Select +Add to enter the ID(s).

Claroty CTD

You must have the Operational Technology license to use the Claroty CTD resolver. Complete the following fields to enter parameters to resolve claroty-ctd:// resource names:

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the Claroty API to get the latest IP address information for hostnames. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Hostname. Enter the Claroty CTD server host name.

  • Port. Enter the Claroty CTD server port. The default is 443.

  • Username. Enter the username for the account used to connect to Claroty CTD.

  • Password. Enter the account password. Select Refer to local secret to use a secret from the Secrets manager.

ColorTokens

Complete the following fields to enter parameters to resolve colortokens:// resource names.

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the ColorTokens API to download the latest asset information. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds. The minimum is 10 seconds; lower values are raised to 10.

  • Cluster Name. Enter the hostname (FQDN) of your ColorTokens cluster, as shown in the ColorTokens web console. The Gateway connects to https://<Cluster Name>/api.

  • Tenant ID. Enter the tenant ID obtained in the ColorTokens web console.

  • User ID. Enter the user ID obtained in the ColorTokens web console.

  • Fingerprint. Enter the fingerprint obtained from the ColorTokens web console.

  • Private Key. Copy and paste the RSA private key (PEM format) of the ColorTokens API key, or upload the key using the Choose a file… button. Tenant

Google Cloud (GCP)

Complete the following fields to enter parameters to resolve gcp:// resource names. See Cloud API resolvers for more information about where to obtain the required parameters.

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the API to get the latest IP address information for hostnames. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Projects Filter. Enter project names or labels. For example: name:uk*, label:devops, label.type:production.

  • Instances Filter. Enter a filter expression. It must specify the field name, a comparison operator, and the filtering value. For example: name:saturn.

  • Forwarding Rules Filter. Enter a filter expression relating to GCP load balancers. For example: name:testlb, region:"${REGION_PROJECT_URL}europe-*".

Illumio

Complete the following fields to enter parameters to resolve illumio:// resource names. To use Illumio REST APIs, you must be an authorized Illumio user and have session credentials to sign in to the Policy Compute Engine (PCE).

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the API to get the latest IP address information for hostnames. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Hostname. Enter the hostname of the Illumio instance.

  • Port. Enter the port number. The default value is 8443.

  • Organization ID. Enter the organization ID used with REST API calls. On-premises PCE deployments have an organization ID of 1. The Illumio Secure Cloud (SaaS) service gives each customer a unique organization ID.

  • Username. Enter the username for a service account with rights to make PCE API queries.

  • Password. Enter the password for the service account.

VMware vSphere

Complete the following fields to enter parameters to resolve esx:// resource names. This resolver may require uploading a certificate to Trusted Certificates (System > Trusted Certificates) if vSphere uses a self-signed certificate. Upload the root CA certificate.

  • Name. Enter a name for the resolver.

  • Update Interval (seconds). Resolver polling frequency. Enter the number of seconds before the Gateway queries the API to get the latest IP address information for hostnames. This affects how quickly the Gateway responds to changes. Defaults to 60 seconds.

  • Hostname. Enter the hostname of the ESXi/vCenter server.

  • Username. Enter the username for a service account with read-only rights to folders and instances.

  • Password. Enter the password for the service account.


Managing Sites

Site Health Details

Select the status in the Health column to open the Site Health Details window, which displays the following:

  • Site health. Current health status of the Site.

  • Gateways. Health status of Gateways associated with the Site.

  • Sessions. Displays the number of Relayed Sessions, Direct Sessions, or Direct Sessions via NAT Traversal, depending on the Site.

  • Connection Brokers. Connection Brokers associated with the Site and their health status.

  • View appliances related to this Site. This link takes you to a filtered list of appliances associated with the Site.

Testing name resolvers

Use the action button—the three dots icon (Three circular shapes stacked vertically on a dark background, selected to access a menu.) to the right of each line item—to access the following option:

  • Test Name Resolvers. Each Site has a test button. After you configure resolvers for all Sites, you can test them. The syntax for testing the resolver is the same syntax you use when you configure entitlement actions.

Actions menu

Select or create a Site to access the Actions menu on the Add/Edit Site page. The Actions menu allows you to Clone or Delete a Site.