Manual Pages  — NATD

(If you need NAT on a PPP link, ppp(8) provides the -nat option that gives most of the natd functionality, and uses the same libalias(3) library.)

The natd utility normally runs in the background as a daemon. It is passed raw IP packets as they travel into and out of the machine, and will possibly change these before re-injecting them back into the IP packet stream.

It changes all packets destined for another host so that their source IP address is that of the current machine. For each packet changed in this manner, an internal table entry is created to record this fact. The source port number is also changed to indicate the table entry applying to the packet. Packets that are received with a target IP of the current host are checked against this internal table. If an entry is found, it is used to determine the correct target IP address and port to place in the packet.

The following command line options are available:

-log | -l
	Log various aliasing statistics and information to the file /var/log/alias.log. This file is truncated each time natd is started.
-deny_incoming | -d
	Do not pass incoming packets that have no entry in the internal translation table. If this option is not used, then such a packet will be altered using the rules in -target_address below, and the entry will be made in the internal translation table.
-log_denied
	Log denied incoming packets via syslog(3) (see also -log_facility).
-log_facility facility_name
	Use specified log facility when logging information via syslog(3). Argument facility_name is one of the keywords specified in syslog.conf(5).
-use_sockets | -s
	Allocate a socket(2) in order to establish an FTP data or IRC DCC send connection. This option uses more system resources, but guarantees successful connections when port numbers conflict.
-same_ports | -m
	Try to keep the same port number when altering outgoing packets. With this option, protocols such as RPC will have a better chance of working. If it is not possible to maintain the port number, it will be silently changed as per normal.
-verbose | -v
	Do not call daemon(3) on startup. Instead, stay attached to the controlling terminal and display all packet alterations to the standard output. This option should only be used for debugging purposes.
-unregistered_only | -u
	Only alter outgoing packets with an unregistered source address. According to RFC 1918, unregistered source addresses are 10.0.0.0/8, 172.16.0.0/12 and 192.168.0.0/16.
-redirect_port proto
	targetIP: targetPORT[-targetPORT ] [aliasIP: ] aliasPORT[-aliasPORT ] [remoteIP[: remotePORT[-remotePORT] ]] Redirect incoming connections arriving to given port(s) to another host and port(s). Argument proto is either tcp or udp, targetIP is the desired target IP address, targetPORT is the desired target port number or range, aliasPORT is the requested port number or range, and aliasIP is the aliasing address. Arguments remoteIP and remotePORT can be used to specify the connection more accurately if necessary. If remotePORT is not specified, it is assumed to be all ports. Arguments targetIP , aliasIP and remoteIP can be given as IP addresses or as hostnames. The targetPORT , aliasPORT and remotePORT ranges need not be the same numerically, but must have the same size. When targetPORT , aliasPORT or remotePORT specifies a singular value (not a range), it can be given as a service name that is searched for in the services(5) database. For example, the argument     tcp inside1:telnet 6666 means that incoming TCP packets destined for port 6666 on this machine will be sent to the telnet port on the inside1 machine.     tcp inside2:2300-2399 3300-3399 will redirect incoming connections on ports 3300-3399 to host inside2, ports 2300-2399. The mapping is 1:1 meaning port 3300 maps to 2300, 3301 maps to 2301, etc.
-redirect_proto proto localIP [
	publicIP [remoteIP] ] Redirect incoming IP packets of protocol proto (see protocols(5)) destined for publicIP address to a localIP address and vice versa. If publicIP is not specified, then the default aliasing address is used. If remoteIP is specified, then only packets coming from/to remoteIP will match the rule.
-redirect_address localIP publicIP
	Redirect traffic for public IP address to a machine on the local network. This function is known as static NAT. Normally static NAT is useful if your ISP has allocated a small block of IP addresses to you, but it can even be used in the case of single address:     redirect_address 10.0.0.8 0.0.0.0 The above command would redirect all incoming traffic to machine 10.0.0.8. If several address aliases specify the same public address as follows redirect_address 192.168.0.2 public_addr redirect_address 192.168.0.3 public_addr redirect_address 192.168.0.4 public_addr the incoming traffic will be directed to the last translated local address (192.168.0.4), but outgoing traffic from the first two addresses will still be aliased to appear from the specified public_addr.
-redirect_port proto
	targetIP: targetPORT[, targetIP: targetPORT[, ...amp; ] ] [aliasIP: ] aliasPORT [remoteIP [:remotePORT] ]
-redirect_address
	localIP[, localIP[, ...amp; ]] publicIP These forms of -redirect_port and -redirect_address are used to transparently offload network load on a single server and distribute the load across a pool of servers. This function is known as LSNAT (RFC 2391). For example, the argument     tcp www1:http,www2:http,www3:http www:http means that incoming HTTP requests for host www will be transparently redirected to one of the www1, www2 or www3, where a host is selected simply on a round-robin basis, without regard to load on the net.
-dynamic
	If the -n or -interface option is used, natd will monitor the routing socket for alterations to the interface passed. If the interface's IP address is changed, natd will dynamically alter its concept of the alias address.
-in_port | -i port
	Read from and write to divert(4) port port, treating all packets as "incoming".
-out_port | -o port
	Read from and write to divert(4) port port, treating all packets as "outgoing".
-port | -p port
	Read from and write to divert(4) port port, distinguishing packets as "incoming" or "outgoing" using the rules specified in divert(4). If port is not numeric, it is searched for in the services(5) database. If this option is not specified, the divert port named natd will be used as a default.
-alias_address | -a address
	Use address as the aliasing address. Either this or the -interface option must be used (but not both), if the -proxy_only option is not specified. The specified address is usually the address assigned to the "public" network interface. All data passing out will be rewritten with a source address equal to address. All data coming in will be checked to see if it matches any already-aliased outgoing connection. If it does, the packet is altered accordingly. If not, all -redirect_port, -redirect_proto and -redirect_address assignments are checked and actioned. If no other action can be made and if -deny_incoming is not specified, the packet is delivered to the local machine using the rules specified in -target_address option below.
-t | -target_address address
	Set the target address. When an incoming packet not associated with any pre-existing link arrives at the host machine, it will be sent to the specified address. The target address may be set to 255.255.255.255, in which case all new incoming packets go to the alias address set by -alias_address or -interface. If this option is not used, or called with the argument 0.0.0.0, then all new incoming packets go to the address specified in the packet. This allows external machines to talk directly to internal machines if they can route packets to the machine in question.
-interface | -n interface
	Use interface to determine the aliasing address. If there is a possibility that the IP address associated with interface may change, the -dynamic option should also be used. If this option is not specified, the -alias_address option must be used. The specified interface is usually the "public" (or "external") network interface.
-config | -f file
	Read configuration from file. A file should contain a list of options, one per line, in the same form as the long form of the above command line options. For example, the line     alias_address 158.152.17.1 would specify an alias address of 158.152.17.1. Options that do not take an argument are specified with an argument of yes or no in the configuration file. For example, the line     log yes is synonymous with -log. Options can be divided to several sections. Each section applies to own natd instance. This ability allows the configuration of one natd process for several NAT instances. The first instance that always exists is a "default" instance. Each another instance should begin with     instance instance_name At the next should be placed a configuration option. Example:     amp;# default instance     port 8668     alias_address 158.152.17.1     amp;# second instance     instance dsl1     port 8888     alias_address 192.168.0.1 Trailing spaces and empty lines are ignored. A ‘amp;#’ sign will mark the rest of the line as a comment.
-instance instancename
	This option switches command line options processing to configure instance instancename (creating it if necessary) till the next -instance option or end of command line. It is easier to set up multiple instances in the configuration file specified with the -config option rather than on a command line.
-globalport port
	Read from and write to divert(4) port port, treating all packets as "outgoing". This option is intended to be used with multiple instances: packets received on this port are checked against internal translation tables of every configured instance. If an entry is found, packet is aliased according to that entry. If no entry was found in any of the instances, packet is passed unchanged, and no new entry will be created. See the section MULTIPLE INSTANCES for more details.
-reverse
	This option makes natd reverse the way it handles "incoming" and "outgoing" packets, allowing it to operate on the "internal" network interface rather than the "external" one. This can be useful in some transparent proxying situations when outgoing traffic is redirected to the local machine and natd is running on the internal interface (it usually runs on the external interface).
-proxy_only
	Force natd to perform transparent proxying only. Normal address translation is not performed.
-proxy_rule
	[type encode_ip_hdr | encode_tcp_stream] port xxxx server a.b.c.d:yyyy Enable transparent proxying. Outgoing TCP packets with the given port going through this host to any other host are redirected to the given server and port. Optionally, the original target address can be encoded into the packet. Use encode_ip_hdr to put this information into the IP option field or encode_tcp_stream to inject the data into the beginning of the TCP stream.
-punch_fw
	basenumber:count This option directs natd to "punch holes" in an ipfirewall(4) based firewall for FTP/IRC DCC connections. This is done dynamically by installing temporary firewall rules which allow a particular connection (and only that connection) to go through the firewall. The rules are removed once the corresponding connection terminates. A maximum of count rules starting from the rule number basenumber will be used for punching firewall holes. The range will be cleared for all rules on startup. This option has no effect when the kernel is in security level 3, see init(8) for more information.
-skinny_port port
	This option allows you to specify the TCP port used for the Skinny Station protocol. Skinny is used by Cisco IP phones to communicate with Cisco Call Managers to set up voice over IP calls. By default, Skinny aliasing is not performed. The typical port value for Skinny is 2000.
-log_ipfw_denied
	Log when a packet cannot be re-injected because an ipfw(8) rule blocks it. This is the default with -verbose.
-pid_file | -P file
	Specify an alternate file in which to store the process ID. The default is /var/run/natd.pid.
-exit_delay ms
	Specify delay in ms before daemon exit after signal. The default is 10000.

1. Build a custom kernel with the following options:

   options IPFIREWALL
   options IPDIVERT
   

   Refer to the handbook for detailed instructions on building a custom kernel.

2. Ensure that your machine is acting as a gateway. This can be done by specifying the line

       gateway_enable=YES

   in the /etc/rc.conf file or using the command

       sysctl net.inet.ip.forwarding=1

3. If you use the -interface option, make sure that your interface is already configured. If, for example, you wish to specify ‘tun0’ as your interface, and you are using ppp(8) on that interface, you must make sure that you start ppp prior to starting natd.

Running natd is fairly straight forward. The line

    natd -interface ed0

should suffice in most cases (substituting the correct interface name). Please check rc.conf(5) on how to configure it to be started automatically during boot. Once natd is running, you must ensure that traffic is diverted to natd:

1. You will need to adjust the /etc/rc.firewall script to taste. If you are not interested in having a firewall, the following lines will do:

   /sbin/ipfw -f flush
   /sbin/ipfw add divert natd all from any to any via ed0
   /sbin/ipfw add pass all from any to any
   

   The second line depends on your interface (change ‘ed0’ as appropriate).

   You should be aware of the fact that, with these firewall settings, everyone on your local network can fake his source-address using your host as gateway. If there are other hosts on your local network, you are strongly encouraged to create firewall rules that only allow traffic to and from trusted hosts.

   If you specify real firewall rules, it is best to specify line 2 at the start of the script so that natd sees all packets before they are dropped by the firewall.

   After translation by natd, packets re-enter the firewall at the rule number following the rule number that caused the diversion (not the next rule if there are several at the same number).

2. Enable your firewall by setting

       firewall_enable=YES

   in /etc/rc.conf. This tells the system startup scripts to run the /etc/rc.firewall script. If you do not wish to reboot now, just run this by hand from the console. NEVER run this from a remote session unless you put it into the background. If you do, you will lock yourself out after the flush takes place, and execution of /etc/rc.firewall will stop at this point - blocking all accesses permanently. Running the script in the background should be enough to prevent this disaster.

          net 1.2.3.0/24
1.2.3.1 ------------------ sis0
(router)                (1.2.3.4)
                                         net 10.0.0.0/24
                                  sis1 ------------------- 10.0.0.2
                               (10.0.0.1)
          net 2.3.4.0/24
2.3.4.1 ------------------ sis2
(router)                (2.3.4.5)

In addition to these instances' private divert(4) sockets, a third socket called the "globalport" is created; packets sent to natd via this one will be matched against all instances and translated if an existing entry is found, and unchanged if no entry is found. The following lines are placed into /etc/natd.conf:

log
deny_incoming
verbose

instance default
interface sis0
port 1000
redirect_port tcp 10.0.0.2:122 122

instance sis2
interface sis2
port 2000
redirect_port tcp 10.0.0.2:122 122

globalport 3000

And the following ipfw(8) rules are used:

ipfw -f flush

ipfw add      allow ip from any to any via sis1

ipfw add      skipto 1000 ip from any to any in via sis0
ipfw add      skipto 2000 ip from any to any out via sis0
ipfw add      skipto 3000 ip from any to any in via sis2
ipfw add      skipto 4000 ip from any to any out via sis2

ipfw add 1000 count ip from any to any

ipfw add      divert 1000 ip from any to any
ipfw add      allow ip from any to any

ipfw add 2000 count ip from any to any

ipfw add      divert 3000 ip from any to any

ipfw add      allow ip from 1.2.3.4 to any
ipfw add      skipto 5000 ip from 2.3.4.5 to any

ipfw add      prob .5 skipto 4000 ip from any to any

ipfw add      divert 1000 ip from any to any
ipfw add      allow ip from any to any

ipfw add 3000 count ip from any to any

ipfw add      divert 2000 ip from any to any
ipfw add      allow ip from any to any

ipfw add 4000 count ip from any to any

ipfw add      divert 2000 ip from any to any

ipfw add 5000 fwd 2.3.4.1 ip from 2.3.4.5 to not 2.3.4.0/24
ipfw add      allow ip from any to any

Here the packet from internal network to Internet goes out via ‘sis0’ (rule number 2000) and gets caught by the globalport socket (3000). After that, either a match is found in a translation table of one of the two instances, or the packet is passed to one of the two other divert(4) ports (1000 or 2000), with equal probability. This ensures that load balancing is done on a per-flow basis (i.e., packets from a single TCP connection always flow through the same interface). Translated packets with source IP of a non-default interface (‘sis2’) are forwarded to the appropriate router on that interface.