Applies To:
Show Versions
BIG-IP LTM
- 13.1.5, 13.1.4, 13.1.3, 13.1.1, 13.1.0
Overview: The carrier-grade NAT (CGNAT) module
Translation address persistence
The CGNAT module can assign the same external (translation) address to all connections originated by the same internal client. For example, providing endpoint-independent address mapping.
Automatic external inbound connection handling
CGNAT can accept inbound external connections to active translation address/port combinations to facilitate endpoint-independent filtering as described in section 5 of RFC 4787. This is also known as a full-cone NAT.
More efficient logging
CGNAT supports log messages that map external addresses and ports back to internal clients for both troubleshooting and compliance with law enforcement/legal constraints.
Network address and port translation
Network address and port translation (NAPT) mode provides standard address and port translation allowing multiple clients in a private network to access remote networks using the single IP address assigned to their router.
Deterministic assignment of translation addresses
Deterministic mode is an option used to assign translation address, and is port-based on the client address/port and destination address/port. It uses reversible mapping to reduce logging, while maintaining the ability for translated IP address to be discovered for troubleshooting and compliance with regulations. Deterministic mode also provides an option to configure backup-members.
Port block allocation of translation addresses
Port block allocation (PBA) mode is an option that reduces logging, by logging only the allocation and release of a block of ports. When a subscriber sends a translation request, the BIG-IP system services the request from a block of ports that is assigned to a single IP address, and only logs the allocation and release of that block of ports. The BIG-IP system applies subsequent requests from the service provider to that block of ports until all ports are used.
Licensing
Designed for service providers, the CGNAT module is offered as a stand-alone license or as an add-on license for Local Traffic Manager™ (LTM®) and Policy Enforcement Manager™ (PEM).
Task summary
About ALG Profiles
Application Layer Gateway (ALG) profiles provide the CGNAT with protocol and service functionality that modifies the necessary application protocol header and payload, thus allowing these protocols to seamlessly traverse the NAT. FTP, RTSP, SIP, and PPTP profiles that are supported with ALG profiles, and added to the CGNAT configuration as needed.
An FTP, RTSP, or SIP profile can use an Automap, NAPT, DNAT, or PBA address translation mode when providing necessary logging.
About CGNAT translation address persistence and inbound connections
The BIG-IP® system enables you to manage RFC-defined behavior for translation address persistence and inbound connections.
Translation Address Persistence
When you configure an LSN pool, the CGNAT Persistence Mode setting assigns translation endpoints in accordance with the selected configuration mode: NAPT, Deterministic NAT (DNAT), or Port Block Allocation (PBA). It is important to note that this CGNAT translation address persistence is different from the persistence used in the BIG-IP Local Traffic Manager™ (LTM®) load balancing. CGNAT translation address persistence uses a selected translation address, or endpoint, across multiple connections from the same subscriber address, or endpoint.
The BIG-IP system provides three Persistence Mode settings (None, Address, and Address Port) for each configuration mode.
| Persistence Mode | Description |
|---|---|
| None | Translation addresses are not preserved for the subscriber. Each outbound connection might receive a different translation address. This setting provides the lowest overhead and highest performance. |
| Address | CGNAT preserves the translation address for the subscriber. When a connection is
established, CGNAT determines if this subscriber already has a translation address. If the
subscriber already has a translation address, then CGNAT uses the translation address stored
in the persistence record, and locates a port for that connection. If no port is available,
then CGNAT selects a different address. This setting provides greater overhead on each
connection and less performance.
Note: DNAT reserves both addresses and ports for
a subscriber; however, persistence might still be of value when a subscriber's
deterministic mappings span two translation addresses. In this instance, persistence
prefers the same address each time.
|
| Address Port | CGNAT preserves the translation address and port of the subscriber's connection, so that the endpoint can be reused on subsequent connections. This setting provides Endpoint Independent Mapping (EIM) behavior. Additionally, like the Address setting for Persistence Mode, this setting provides greater overhead on each connection and less performance. |
Inbound Connections
The Inbound Connections setting determines whether the Large Scale NAT (LSN) allows connections to be established inbound to the LSN subscriber or client. This setting provides greater overhead, including a lookup on inbound entries for each connection to prevent endpoint overloading, and a reduction in the use of the translation space.
When you disable inbound connections, the BIG-IP system provides greater efficiency in address space utilization by allowing endpoint overloading, where two different subscribers can use the same translation address and port, as long as each subscriber connects to a different host.
When you enable inbound connections, the BIG-IP system restricts the use of a translation address and port to a single subscriber, and ensures that only one subscriber address and port uses a translation endpoint.
About IPv6 prefixes
IPv6 128-bit addresses include a network prefix in the leftmost fields, and subnet in the remaining fields. For example, an IPv6 address of 2001:0db8:0000:0000:0000:0000:0000:0000 with a 32-bit prefix equates to a network of 2001:0db8, written as 2001:db8::/32. A network written as 2001:db8::/32 omits leading zeros in four-digit groups, uses :: to indicate collapsed zero groups, and uses /32 to indicate the 32-bit prefix.
About IPv4 prefixes
IPv4 32-bit addresses include a network prefix in the leftmost fields, and a host identifier in the remaining fields. For example, an address of 192.168.1.0/24 includes the prefix of the IPv4 network starting at the given address, having 24 bits allocated for the network prefix, and the remaining 8 bits reserved for host addressing.
