Firewall configuration /etc/config/firewall

OpenWrt's firewall management application firewall is mainly configured through /etc/config/firewall.

Most of the information in this wiki will focus on the configuration files and content. The LuCI and UCI interfaces are user abstractions, ultimately modifying the configuration files.

  • The main firewall config file is /etc/config/firewall, and this is edited to modify the firewall settings
    • :!: Create a backup of the firewall config prior to making changes
      • Should changes cause a loss-of-connectivity to the router, you will need to access it in failsafe mode to restore the backup
    • Once the settings are changed, and after double checking changes, reload the firewall via /etc/init.d/firewall reload
      • This is a simple shell script calling fw4 reload, and will print diagnostics to the console as it parses the new firewall configuration. Check for errors!
  • Any line using # in the first character is not parsed
    • Comments are utilized to describe, explain, or quickly comment out, a section
  • OpenWrt 22.03 and later ships with firewall4 by default, which uses nftables as a backend. It accepts the same UCI configuration syntax as fw3.
  • The UCI firewall configuration in /etc/config/firewall covers a reasonable subset of NetFilter rules, but not all of them
    • To provide more functionality, include mechanisms are available.
      • You can either include a shell script with nftables commands, or include nftables snippets at different locations.
  • Whenever possible, use the firewall UCI config /etc/config/firewall

LuCI is a good mechanism to view and modify the firewall configuration.

  • It is located under Network → Firewall and maps closely to the configuration file sections.
  • It takes a little longer to modify the firewall configuration, but has a higher level of organization than the config files.

Make changes and reload using the Save & Apply button.

  • :!: LuCI will remove all comment [#] lines from /etc/config/firewall!

UCI is a low-level abstraction to the configuration files and can be accessed remotely through SSH.

uci add firewall rule
uci set firewall.@rule[-1].name='Reject VPN to LAN traffic'
uci set firewall.@rule[-1].src='vpn'
uci set firewall.@rule[-1].dest='lan'
uci set firewall.@rule[-1].proto='all'
uci set firewall.@rule[-1].target='REJECT'
uci commit firewall
service firewall restart

These would be presumed to be the final rules (each proto creates a rule) in the VPNLAN forward chain, as all packets from VPN will be rejected.

Show firewall configuration:

# uci show firewall
firewall.@rule[20]=rule
firewall.@rule[20].name='Reject VPN to LAN traffic'
firewall.@rule[20].src='vpn'
firewall.@rule[20].dest='lan'
firewall.@rule[20].proto='all'
firewall.@rule[20].target='REJECT'
...

UCI is useful to view the firewall configuration, but not to do any meaningful modifications for the following reasons:

  • Essential prior knowledge of where a firewall rule needs to go into the rule array in order to make it work.
  • uci does not recognize content within the /etc/firewall.user script.
  • uci commit is necessary to save the changes, but still needs /etc/init.d/firewall reload to reload new tables.

Below is an overview of the section types that may be defined in the firewall configuration.

  • A minimal firewall configuration for a router usually consists of one defaults section, at least two zones (lan and wan), and one forwarding to allow traffic from lan to wan.
  • The forwarding section is not strictly required when there are no more than two zones, as the rule can then be set as the 'global default' for that zone.

The defaults section declares global firewall settings which do not belong to specific zones:

config defaults
	option	input			'ACCEPT'
	option	output			'ACCEPT'
	option	forward			'REJECT'
	option	custom_chains		'1'
	option	drop_invalid		'1'
	option	synflood_protect	'1'
	option	synflood_rate		'25/s'
	option	synflood_burst		'50'
	option	tcp_ecn			'1'
	option	tcp_syncookies		'1'
	option	tcp_window_scaling	'1'

Options

Name Type Required Default Description
input string no ACCEPT Set policy for the INPUT chain of the filter table.
output string no ACCEPT Set policy for the OUTPUT chain of the filter table.
forward string no REJECT Set policy for the FORWARD chain of the filter table.
drop_invalid boolean no 0 Drop invalid packets (e.g. not matching any active connection).
syn_flood boolean no 0 Enable SYN flood protection (obsoleted by synflood_protect setting).
synflood_protect boolean no 0 Enable SYN flood protection.
synflood_rate string no 25/s Set rate limit (packets/second) for SYN packets above which the traffic is considered a flood.
synflood_burst string no 50 Set burst limit for SYN packets above which the traffic is considered a flood if it exceeds the allowed rate.
tcp_syncookies boolean no 1 Enable the use of SYN cookies.
tcp_ecn integer no 0 0 Disable, 1 Enable, 2 Enable when requested for ingress (but disable for egress) Explicit Congestion Notification. Affects only traffic originating from the router itself. Implemented upstream in Linux Kernel. See kernel docs.
tcp_window_scaling boolean no 1 Enable TCP window scaling.
accept_redirects boolean no 0 Accepts redirects. Implemented upstream in Linux Kernel. See kernel docs.
accept_source_route boolean no 0 Implemented upstream in Linux Kernel. See kernel docs.
custom_chains boolean no 1 Enable generation of custom rule chain hooks for user generated rules. User rules would be typically stored in firewall.user but some packages e.g. BCP38 also make use of these hooks.
disable_ipv6 boolean no 0 Disable IPv6 firewall rules. (not supported by fw4)
flow_offloading boolean no 0 Enable software flow offloading for connections. (decrease cpu load / increase routing throughput)
flow_offloading_hw boolean no 0 Enable hardware flow offloading for connections. (depends on flow_offloading and hw capability)
tcp_reject_code reject_code no 0 Defined in firewall3/options.h. Seems to determine method of packet rejection; (tcp reset, or drop, vs ICMP Destination Unreachable, or closed)
any_reject_code reject_code no 1 Defined in firewall3/options.h. Seems to determine method of packet rejection; (tcp reset, or drop, vs ICMP Destination Unreachable, or closed)
auto_helper bool no 1 Enable Conntrack helpers
auto_includes bool no 1 (fw4 only, 22.03 and later) Enable automatic nftables includes under /usr/share/nftables.d/

A zone section groups one or more interfaces and serves as a source or destination for forwardings, rules and redirects.

config zone
	option	name		'wan'
	option	network		'wan wan6'
	option	input		'REJECT'
	option	output		'ACCEPT'
	option	forward		'REJECT'
	option	masq		'1'
	option	mtu_fix		'1'
  • MASQUERADE (NAT) of outgoing traffic (WAN) is controlled on a per-zone basis on the outgoing interface.
  • INPUT rules for a zone describe what happens to traffic trying to reach the router itself through an interface in that zone.
  • OUTPUT rules for a zone describe what happens to traffic originating from the router itself going through an interface in that zone.
  • FORWARD rules for a zone describe what happens to traffic passing between different interfaces belonging in the same zone.

Options

Name Type Required Default Description
name zone name yes (none) Unique zone name. 11 characters is the maximum working firewall zone name length.
network list no (none) List of interfaces attached to this zone. If omitted and neither extra* options, subnets nor devices are given, the value of name is used by default. Alias interfaces defined in the network config cannot be used as valid 'standalone' networks. Use list syntax.
masq boolean no 0 Specifies whether outgoing zone IPv4 traffic should be masqueraded. This is typically enabled on the wan zone.
masq6 boolean no 0 Specifies whether outgoing zone IPv6 traffic should be masqueraded. This is typically enabled on the wan zone. Available with fw4. Requires sourcefilter=0 for DHCPv6 interfaces with missing GUA prefix.
masq_src list of subnets no 0.0.0.0/0 Limit masquerading to the given source subnets. Negation is possible by prefixing the subnet with !; multiple subnets are allowed.
masq_dest list of subnets no 0.0.0.0/0 Limit masquerading to the given destination subnets. Negation is possible by prefixing the subnet with !; multiple subnets are allowed.
masq_allow_invalid boolean no 0 Do not add DROP INVALID rules, if masquerading is used. The DROP rules are supposed to prevent NAT leakage (see commit in firewall3).
mtu_fix boolean no 0 Enable MSS clamping for outgoing zone traffic.
input string no From defaults section Policy (ACCEPT, REJECT, DROP) for incoming zone traffic.
output string no From defaults section Policy (ACCEPT, REJECT, DROP) for outgoing zone traffic.
forward string no From defaults section Policy (ACCEPT, REJECT, DROP) for forwarded zone traffic.
family string no (auto) Specifies the address family (ipv4, ipv6 or any) for which the rules are generated. If unspecified, matches the address family of other options in this section and defaults to any.
log int no 0 Bit field to enable logging in the filter and/or mangle tables, bit 0 = filter, bit 1 = mangle.
log_limit string no 10/minute Limits the amount of log messages per interval.
device list no (none) List of L3 network interface names attached to this zone, e.g. tun+ or ppp+ to match any TUN or PPP interface. This is specifically suitable for undeclared interfaces which lack built-in netifd support such as OpenVPN. Otherwise network is preferable and device should be avoided.
subnet list no (none) List of IP subnets attached to this zone.
extra string no (none) Extra arguments passed directly to iptables. Note that these options are passed to both source and destination classification rules, therefor direction-specific options like --dport should not be used here - in this case the extra_src and extra_dest options should be used instead.
extra_src string no Value of extra Extra arguments passed directly to iptables for source classification rules.
extra_dest string no Value of extra Extra arguments passed directly to iptables for destination classification rules.
custom_chains bool no 1 Enable generation of custom rule chain hooks for user generated rules. Has no effect if disabled (0) in the defaults section (see above).
enabled bool no yes if set to 0, zone is disabled
auto_helper bool no 1 for non-masq zone Add CT helpers for zone
helper cthelper no (none) List of helpers to add to zone

The forwarding sections control the traffic flow between zones.

config forwarding
	option	src		'lan'
	option	dest		'wan'
  • Only one direction is covered by a forwarding rule. To allow bidirectional traffic flows between two zones, two forwardings are required, with src and dest reversed in each.

Options

Name Type Required Default Description
name forward name no (none) Unique forwarding name.
src zone name yes (none) Specifies the traffic source zone. Refers to one of the defined zone names. For typical port forwards this usually is 'wan'.
dest zone name yes (none) Specifies the traffic destination zone. Refers to one of the defined zone names.
family string no any Specifies the address family (ipv4, ipv6 or any) for which the rules are generated.
enabled bool no yes If set to 0, forward is disabled.
ipset string no (none) If specified, match traffic against the given ipset. The match can be inverted by prefixing the value with an exclamation mark.

:!: The rules generated for this section rely on the state match which needs connection tracking to work.

  • At least one of the src or dest zones needs to have connection tracking enabled through the masq option.

The rule section is used to define basic accept, drop, or reject rules to allow or restrict access to specific ports or hosts.

config rule
	option	name		'Reject LAN to WAN for custom IP'
	option	src		'lan'
	option	src_ip		'192.168.1.2'
	option	src_mac		'00:11:22:33:44:55'
	option	src_port	'80'
	option	dest		'wan'
	option	dest_ip		'194.25.2.129'
	option	dest_port	'120'
	option	proto		'tcp'
	option	target		'REJECT'

Below example is based on a inter zone forward case (where zone forward is set to reject) where you have one firewall zone called lan with two interfaces. In one interface you have a server with IP address 172.30.100.1 and the other interface is the default lan interface with 192.168.1.0/24 IP range. This configuration case will allow IPv4 tcp traffic from all IP addresses in the default lan interface to specifically connect only to the server IP address and to the server port 22.

config rule
	option name             'forward ssh to server'
	option family           'ipv4'
	option src              'lan'
	option src_ip           '192.168.1.0/24'
	option dest             'lan'
	option dest_ip          '172.30.100.1'
	option proto            'tcp'
	option dest_port        '22'
	option target           'ACCEPT'
  • In OpenWrt firewall, the src and dest are tied to the target:
  • If src and dest are given, the rule matches forwarded traffic
  • If only src is given, the rule matches incoming traffic
  • If only dest is given, the rule matches outgoing traffic
  • If neither src nor dest are given, the rule defaults to an outgoing traffic rule
  • IP address for src_ip and dest_ip can be a specific IP address or use CIDR notations to define a complete interface group of IP addresses as a source or destination, for instance 192.168.1.0/24.
  • Port ranges are specified as start-stop, for instance 6666-6670.

Options

Name Type Required Default Description
name string no (none) Name of rule
src zone name no (none) Specifies the traffic source zone. Refers to one of the defined zone names, or * for any zone. If omitted, the rule applies to output traffic.
src_ip ip address no (none) Match incoming traffic from the specified source IP address, CIDR notations can be used, see note above.
src_mac mac address no (none) Match incoming traffic from the specified MAC address
src_port port or range no (none) Match incoming traffic from the specified source port or port range, if relevant proto is specified. Multiple ports can be specified like '80 443 465' 1.
proto protocol name or number no tcp udp Match incoming traffic using the given protocol. Can be one (or several when using list syntax) of tcp, udp, udplite, icmp, esp, ah, sctp, or all or it can be a numeric value, representing one of these protocols or a different one. A protocol name from /etc/protocols is also allowed. The number 0 is equivalent to all.
icmp_type list of type names or numbers no any For protocol icmp select specific ICMP types to match. Values can be either exact ICMP type numbers or type names (see below).
dest zone name no (none) Specifies the traffic destination zone. Refers to one of the defined zone names, or * for any zone. If specified, the rule applies to forwarded traffic; otherwise, it is treated as input rule.
dest_ip ip address no (none) Match incoming traffic directed to the specified destination IP address, CIDR notations can be used, see note above. With no dest zone, this is treated as an input rule!
dest_port port or range no (none) Match incoming traffic directed at the given destination port or port range, if relevant proto is specified. Multiple ports can be specified like '80 443 465' 1.
ipset string no (none) If specified, match traffic against the given ipset. The match can be inverted by prefixing the value with an exclamation mark. You can specify the direction as 'setname src' or 'setname dest'. The default if neither src nor dest are added is to assume src
mark mark/mask no (none) If specified, match traffic against the given firewall mark, e.g. 0xFF to match mark 255 or 0x0/0x1 to match any even mark value. The match can be inverted by prefixing the value with an exclamation mark, e.g. !0x10 to match all but mark #16.
start_date date (yyyy-mm-dd) no (always) If specifed, only match traffic after the given date (inclusive).
stop_date date (yyyy-mm-dd) no (always) If specified, only match traffic before the given date (inclusive).
start_time time (hh:mm:ss) no (always) If specified, only match traffic after the given time of day (inclusive).
stop_time time (hh:mm:ss) no (always) If specified, only match traffic before the given time of day (inclusive).
weekdays list of weekdays no (always) If specified, only match traffic during the given week days, e.g. sun mon thu fri to only match on sundays, mondays, thursdays and Fridays. The list can be inverted by prefixing it with an exclamation mark, e.g. ! sat sun to always match but on Saturdays and sundays.
monthdays list of dates no (always) If specified, only match traffic during the given days of the month, e.g. 2 5 30 to only match on every 2nd, 5th and 30rd day of the month. The list can be inverted by prefixing it with an exclamation mark, e.g. ! 31 to always match but on the 31st of the month.
utc_time boolean no 0 Treat all given time values as UTC time instead of local time.
target string yes DROP Firewall action (ACCEPT, REJECT, DROP, MARK, NOTRACK) for matched traffic.
set_mark mark/mask yes for target MARK (none) Zeroes out the bits given by mask and ORs value into the packet mark. If mask is omitted, 0xFFFFFFFF is assumed.
set_xmark Zeroes out the bits given by mask and XORs value into the packet mark. If mask is omitted, 0xFFFFFFFF is assumed.
family string no (auto) Specifies the address family (ipv4, ipv6 or any) for which the rules are generated. If unspecified, matches the address family of other options in this section and defaults to any.
limit string no (none) Maximum average matching rate; specified as a number, with an optional /second, /minute, /hour or /day suffix. Examples: 3/minute, 3/min or 3/m.
limit_burst integer no 5 Maximum initial number of packets to match, allowing a short-term average above limit.
extra string no (none) Extra arguments to pass to iptables. Useful mainly to specify additional match options, such as -m policy --dir in for IPsec.
enabled boolean no yes Enable or disable rule.
device string no FIXME FIXME
direction direction no FIXME FIXME direction_out
set_helper cthelper no FIXME FIXME
helper cthelper no FIXME FIXME

ICMP name types

address-mask-reply host-redirect pong time-exceeded
address-mask-request host-unknown port-unreachable timestamp-reply
any host-unreachable precedence-cutoff timestamp-request
communication-prohibited ip-header-bad protocol-unreachable TOS-host-redirect
destination-unreachable network-prohibited redirect TOS-host-unreachable
echo-reply network-redirect required-option-missing TOS-network-redirect
echo-request network-unknown router-advertisement TOS-network-unreachable
fragmentation-needed network-unreachable router-solicitation ttl-exceeded
host-precedence-violation parameter-problem source-quench ttl-zero-during-reassembly
host-prohibited ping source-route-failed ttl-zero-during-transit

Port forwardings (DNAT) are defined by redirect sections. Port Redirects are also commonly known as “port forwarding” or “virtual servers”.

  • All incoming traffic on the specified source zone which matches the given rules will be directed to the specified internal host.
  • Port ranges are specified as start-stop, for instance 6666-6670.

Destination NAT

See also: IPv6 port forwarding

config redirect
	option	name		'DNAT WAN to LAN for SSH'
	option	src		'wan'
	option	src_dport	'19900'
	option	dest		'lan'
	option	dest_ip		'192.168.1.1'
	option	dest_port	'22'
	option	proto		'tcp'
	option	target		'DNAT'

:!: If a src_dport is not included in the config section, packets matching the other config options, on any port, will be forwarded to the destination port specified in that config section. This could pose a security risk to the application running on the destination port the config section opens. One way to test for this issue, is to use Gibson Research Corporation's ShieldsUP! service, and probe the desired ports on your router. The response could be open, closed, or stealth (drop). In cases of open or closed ports, packets are reaching a destination host, and are sending ack/reply packets back. Whereas stealthed ports drop packets; from the perspective of the probing system (Gibson Research), that system cannot definitively know if those packets may, or may not be reaching the destination host.

Source NAT

Masquerade is the most common form of SNAT, changing the source of traffic to WAN to the router's public IP. SNAT can also be done manually:

config redirect
	option	name		'SNAT DMZ 192.168.1.250 to WAN 1.2.3.4 for ICMP'
	option	src		'dmz'
	option	src_ip		'192.168.1.250'
	option	src_dip		'1.2.3.4'
	option	dest		'wan'
	option	proto		'icmp'
	option	target		'SNAT'

Options

See also: List of SNAT options @ OpenWrt SNAPSHOT

Name Type Required Default Description
name string no string Name of redirect
src zone name yes for DNAT target (none) Specifies the traffic source zone. Refers to one of the defined zone names. For typical port forwards this usually is wan.
src_ip ip address no (none) Match incoming traffic from the specified source IP address.
src_dip ip address yes for SNAT target (none) For DNAT, match incoming traffic directed at the given destination IP address. For SNAT rewrite the source address to the given address.
src_mac mac address no (none) Match incoming traffic from the specified MAC address.
src_port port or range no (none) Match incoming traffic originating from the given source port or port range on the client host.
src_dport port or range no (none) For DNAT, match incoming traffic directed at the given destination port or port range on this host. For SNAT rewrite the source ports to the given value.
proto protocol name or number no tcp udp Match incoming traffic using the given protocol. Can be one (or several when using list syntax) of tcp, udp, udplite, icmp, esp, ah, sctp, or all or it can be a numeric value, representing one of these protocols or a different one. A protocol name from /etc/protocols is also allowed. The number 0 is equivalent to all.
dest zone name yes for SNAT target (none) Specifies the traffic destination zone. Refers to one of the defined zone names. Irrelevant for DNAT target.
dest_ip ip address no (none) For DNAT, redirect matches incoming traffic to the specified internal host. For SNAT, it matches traffic directed at the given address. For DNAT, if the dest_ip is not specified, the rule is translated in a redirect rule, otherwise it is a DNAT rule.
dest_port port or range no (none) For DNAT, redirect matched incoming traffic to the given port on the internal host. For SNAT, match traffic directed at the given ports. Only a single port or range can be specified, not disparate ports as with Rules (below).
ipset string no (none) If specified, match traffic against the given ipset. The match can be inverted by prefixing the value with an exclamation mark. Unsupported in firewall4.
mark string no (none) If specified, match traffic against the given firewall mark, e.g. 0xFF to match mark 255 or 0x0/0x1 to match any even mark value. The match can be inverted by prefixing the value with an exclamation mark, e.g. !0x10 to match all but mark #16.
start_date date (yyyy-mm-dd) no (always) If specifed, only match traffic after the given date (inclusive).
stop_date date (yyyy-mm-dd) no (always) If specified, only match traffic before the given date (inclusive).
start_time time (hh:mm:ss) no (always) If specified, only match traffic after the given time of day (inclusive).
stop_time time (hh:mm:ss) no (always) If specified, only match traffic before the given time of day (inclusive).
weekdays list of weekdays no (always) If specified, only match traffic during the given week days, e.g. sun mon thu fri to only match on Sundays, Mondays, Thursdays and Fridays. The list can be inverted by prefixing it with an exclamation mark, e.g. ! sat sun to always match but on Saturdays and sundays.
monthdays list of dates no (always) If specified, only match traffic during the given days of the month, e.g. 2 5 30 to only match on every 2nd, 5th and 30rd day of the month. The list can be inverted by prefixing it with an exclamation mark, e.g. ! 31 to always match but on the 31st of the month.
utc_time boolean no 0 Treat all given time values as UTC time instead of local time.
target string no DNAT NAT target (DNAT or SNAT) to use when generating the rule.
family string no (auto) Specifies the address family (ipv4, ipv6 or any) for which the rules are generated. If unspecified, matches the address family of other options in this section and defaults to ipv4.
reflection boolean no 1 Activate NAT reflection for this redirect - applicable to DNAT targets.
reflection_src string no internal The source address to use for NAT-reflected packets if reflection is 1. This can be internal or external, specifying which interface’s address to use. Applicable to DNAT targets.
reflection_zone list of zone names no (none) List of zones for which reflection should be enabled. Applicable to DNAT targets.
limit string no (none) Maximum average matching rate; specified as a number, with an optional /second, /minute, /hour or /day suffix. Examples: 3/second, 3/sec or 3/s.
limit_burst integer no 5 Maximum initial number of packets to match, allowing a short-term average above limit.
extra string no (none) Extra arguments to pass to iptables. Useful mainly to specify additional match options, such as -m policy --dir in for IPsec.
enabled string no 1 or yes Enable the redirect rule or not.
helper cthelper no FIXME FIXME

See also: IP set examples

fw4 supports referencing or creating IP sets to simplify matching of large address or port lists without the need for creating one rule per item to match. fw4 supports fewer options (see below).

Options (fw3)

Name Type Required Default Description
enabled boolean no 1 Allows to disable the declaration of the ipset without the need to delete the section.
external string no (none) If the external option is set to a name, the firewall will simply reference an already existing ipset pointed to by the name. If the external option is unset, the firewall will create the ipset on start and destroy it on stop.
name string yes if external is unset
no if external is set
(none) if external is unset
value of external if external is set
Specifies the firewall internal name of the ipset which is used to reference the set in rules or redirects.
family string no ipv4 Specifies the address family (ipv4 or ipv6) for which the IP set is created. Only applicable to storage types hash and list, the bitmap type implies ipv4.
storage string no varies Specifies the storage method (bitmap, hash or list) used by the ipset, the default varies depending on the used datatypes (see match option below). In most cases the storage method can be automatically inferred from the datatype combination but in some cases multiple choices are possible (e.g. bitmap:ip vs. hash:ip). :!: This is only required by fw3 and must be removed from the fw4 configuration.
match list of direction/type tuples yes (none) Specifies the matched data types (ip, port, mac, net or set) and their direction (src or dest). The direction is joined with the datatype by an underscore to form a tuple, e.g. src_port to match source ports or dest_net to match destination CIDR ranges. When using ipsets matching on multiple elements, e.g. hash:ip,port, specify the packet fields to match on in quotes or comma-separated (i.e. “match dest_ip dest_port”).
iprange IP range yes for storage type bitmap with datatype ip (none) Specifies the IP range to cover, see ipset(8). Only applicable to the hash storage type.
portrange Port range yes for storage type bitmap with datatype port (none) Specifies the port range to cover, see ipset(8). Only applicable to the hash storage type.
netmask integer no 32 If specified, network addresses will be stored in the set instead of IP host addresses. Value must be between 1 and 32, see ipset(8). Only applicable to the bitmap storage type with match ip or the hash storage type with match ip.
maxelem integer no 65536 Limits the number of items that can be added to the set, only applicable to the hash and list storage types.
hashsize integer no 1024 Specifies the initial hash size of the set, only applicable to the hash storage type.
timeout integer no 0 Specifies the default timeout for entries added to the set. A value of 0 means enabling the timeout capability flag on a set, but do not put the timeout to entries.
entry setentry no (none) The IP address, CIDR, or MAC. Each list entry is a single CIDR, or IP etc when not using ranges or masks etc above.
loadfile string no (none) A path URL on the openwrt filesystem to a file containing a list of CIDRs.

Options (fw4)

Name Type Required Default Description
enabled boolean no 1 Allows to disable the declaration of the ipset without the need to delete the section.
comment boolean no (none) Seems like a bug: should be a string for user defined comment.
name string yes (none) Specifies the firewall internal name of the ipset which is used to reference the set in rules or redirects.
family string no ipv4 Specifies the address family (ipv4 or ipv6) for which the IP set is created.
match list of ipsettypes yes (none) Specifies the matched data types (ip, port, mac, net or set) and their direction (src or dest). The direction is joined with the datatype by an underscore to form a tuple, e.g. src_port to match source ports or dest_net to match destination CIDR ranges.
maxelem integer no 65536 Limits the number of items that can be added to the set.
timeout integer no 0 Specifies the default timeout for entries added to the set. A value of 0 means enabling the timeout capability flag on a set, but do not put the timeout to entries.
entry List of entry no The IP address, CIDR, or MAC. Each list entry is a single CIDR, or IP etc.
loadfile string no (none) A path URL on the openwrt filesystem to a file containing a list of CIDRs.

IP set types

prefix_suffix

Prefix, aka direction Suffix aka datatype Notes
src_* Source matching of suffix
dest_* Destination matching of suffix
_ip IP addresses
_port TCP/UDP ports
_mac MAC addresses
_net Subnets
_set Unsupported in firewall4

Storage / Match options

The order of datatype matches is significant

Family Storage Match Notes
ipv4 bitmap ip Requires iprange option
ipv4 bitmap ip mac Requires iprange option
ipv4 bitmap port Requires portrange option
any hash ip -
any hash net -
any hash ip port -
any hash net port -
any hash ip port ip -
any hash ip port net -
- list set Meta type to create a set-of-sets

fw4 has several ways to include custom firewall rules. In all cases, the custom rules need to be written in nftables-style.

Config include section with nftables snippets

One way is to include nftables snippets. There are several possible positions in the nftables structure.

For example, to add custom logging to the input_wan chain:

# /etc/config/firewall
config include
	option	type		'nftables'
	option	path		'/etc/my_custom_firewall_rule.nft'
	option	position	'chain-pre'
	option	chain		'input_wan'
 
# /etc/my_custom_firewall_rule.nft
tcp dport 0-1023 log prefix "Inbound WAN connection attempt to low TCP port: "

To add one or more custom chains:

config include
	option	type		'nftables'
	option	path		'/etc/my_custom_firewall_chain.nft'
	option	position	'table-post'

Supported options:

Name Type Required Default Description
enabled boolean no 1 Allows to disable the corresponding include without having to delete the section
type string no script Specifies the type of the include, either script for compatibility with fw3 (shell script, see below) or nftables for nftables snippets
path file name yes - Specifies the filename to include
position string yes (if type is nftables) table-post Specifies the position at which the rules will be inserted (see below for allowed values)
chain string yes (if position matches chain-*) - Specifies the chain in which the rules will be inserted

The possible positions for nftables snippets are:

Position Meaning
ruleset-pre At the very beginning, before the fw4 table definition
ruleset-post At the very end, after the fw4 table definition
table-pre At the beginning of the fw4 table, before any chain definition
table-post At the end of the fw4 table, after all chains definition
chain-pre At the beginning of $chain (defined in option chain), before rules in this chain
chain-post At the end of $chain (defined in option chain), after rules in this chain

Run fw4 print to understand the table / chain / rules structure.

Note that -pre can also be written as -prepend, and -post can also be written as -postpend or -append.

Config include section with shell script

Custom rule inclusion through a shell script works similarly as fw3, but the script should use nftables. Option fw4_compatible is required when the path is /etc/firewall.user to indicate to fw4 that the script is compatible with nftables.

config include
	option	enabled		1
	option	type		'script'
	option	path		'/etc/firewall.user'
	option	fw4_compatible	1

Options family and reload from fw3 are no longer supported with fw4.

Default drop-in includes

fw4 includes /etc/nftables.d/*.nft by default, at the beginning of the fw4 table (equivalent to the table-pre position)

It means that custom chains can be created by adding a file ending in .nft in the /etc/nftables.d/ directory.

Drop-in includes for package authors

For package authors that need custom firewall rules, it is possible to add nftables snippets in the following directories, depending on the desired position:

Path Position
/usr/share/nftables.d/ruleset-pre/*.nft Included at the very beginning, before the fw4 table definition
/usr/share/nftables.d/ruleset-post/*.nft Included at the very end, after the fw4 table definition
/usr/share/nftables.d/table-pre/*.nft Included at the beginning of the fw4 table, before any chain definition
/usr/share/nftables.d/table-post/*.nft Included at the end of the fw4 table, after all chains definition
/usr/share/nftables.d/chain-pre/${chain}/*.nft Included at the beginning of ${chain} (a valid fw4 chain name), before rules in this chain
/usr/share/nftables.d/chain-post/${chain}/*.nft Included at the end of ${chain} (a valid fw4 chain name), after rules in this chain

Snippets need to be readable files and their name must end with .nft.

It is possible to include custom firewall scripts by specifying one or more include sections in the firewall configuration:

config include
	option	path		'/etc/firewall.user'

The /etc/firewall.user script is empty by default.

Options

Name Type Required Default Description
enabled boolean no 1 Allows to disable the corresponding include without having to delete the section.
type string no script Specifies the type of the include, can be script for traditional shell script includes or restore for plain files in iptables-restore format.
path file name yes /etc/firewall.user Specifies a shell script to execute on boot or firewall restarts.
family string no any Specifies the address family (ipv4, ipv6 or any) for which the include is called.
reload boolean no 0 Specifies whether the include should be called on reload. This is only needed if the include injects rules into internal chains.

Includes of type script may contain arbitrary commands, for example advanced nftables rules or tc commands required for traffic shaping.

This website uses cookies. By using the website, you agree with storing cookies on your computer. Also you acknowledge that you have read and understand our Privacy Policy. If you do not agree leave the website.More information about cookies
  • Last modified: 2024/11/24 12:25
  • by charles_harris