nft — Administration tool for packet filtering and classification


nft [[ -nNscae ]] [ -I directory ] [[ -f filename ] | [ -i ] | [ cmd ...]]

nft -h

nft -v


nft is used to set up, maintain and inspect packet filtering and classification rules in the Linux kernel.


For a full summary of options, run nft --help.

-h, --help

Show help message and all options.

-v, --version

Show version.

-n, --numeric

Show data numerically. When used once (the default behaviour), skip lookup of addresses to symbolic names. Use twice to also show Internet services (port numbers) numerically. Use three times to also show protocols and UIDs/GIDs numerically.

-N, --reversedns

Translate IP addresses to names. Usually requires network traffic for DNS lookup.

-s, --stateless

Omit stateful information of rules and stateful objects.

-c, --check

Check commands validity without actually applying the changes.

-a, --handle

Show rule handles in output.

-e, --echo

When inserting items into the ruleset using add, insert or replace commands, print notifications just like nft monitor.

-I, --includepath directory

Add the directory directory to the list of directories to be searched for included files. This option may be specified multiple times.

-f, --file filename

Read input from filename.

-i, --interactive

Read input from an interactive readline CLI.

Input file format

Lexical conventions

Input is parsed line-wise. When the last character of a line, just before the newline character, is a non-quoted backslash (\), the next line is treated as a continuation. Multiple commands on the same line can be separated using a semicolon (;).

A hash sign (#) begins a comment. All following characters on the same line are ignored.

Identifiers begin with an alphabetic character (a-z,A-Z), followed zero or more alphanumeric characters (a-z,A-Z,0-9) and the characters slash (/), backslash (\), underscore (_) and dot (.). Identifiers using different characters or clashing with a keyword need to be enclosed in double quotes (").

Include files

include "filename"

Other files can be included by using the include statement. The directories to be searched for include files can be specified using the -I/--includepath option. You can override this behaviour either by prepending ./ to your path to force inclusion of files located in the current working directory (ie. relative path) or / for file location expressed as an absolute path.

If -I/--includepath is not specified, then nft relies on the default directory that is specified at compile time. You can retrieve this default directory via -h/--help option.

Include statements support the usual shell wildcard symbols (*,?,[]). Having no matches for an include statement is not an error, if wildcard symbols are used in the include statement. This allows having potentially empty include directories for statements like include "/etc/firewall/rules/*". The wildcard matches are loaded in alphabetical order. Files beginning with dot (.) are not matched by include statements.

Symbolic variables

define variable = expr


Symbolic variables can be defined using the define statement. Variable references are expressions and can be used initialize other variables. The scope of a definition is the current block and all blocks contained within.

Example 1. Using symbolic variables

define int_if1 = eth0
define int_if2 = eth1
define int_ifs = { $int_if1, $int_if2 }

filter input iif $int_ifs accept

Address families

Address families determine the type of packets which are processed. For each address family the kernel contains so called hooks at specific stages of the packet processing paths, which invoke nftables if rules for these hooks exist.


IPv4 address family.


IPv6 address family.


Internet (IPv4/IPv6) address family.


ARP address family, handling IPv4 ARP packets.


Bridge address family, handling packets which traverse a bridge device.


Netdev address family, handling packets from ingress.

All nftables objects exist in address family specific namespaces, therefore all identifiers include an address family. If an identifier is specified without an address family, the ip family is used by default.

IPv4/IPv6/Inet address families

The IPv4/IPv6/Inet address families handle IPv4, IPv6 or both types of packets. They contain five hooks at different packet processing stages in the network stack.

Table 1. IPv4/IPv6/Inet address family hooks

prerouting All packets entering the system are processed by the prerouting hook. It is invoked before the routing process and is used for early filtering or changing packet attributes that affect routing.
input Packets delivered to the local system are processed by the input hook.
forward Packets forwarded to a different host are processed by the forward hook.
output Packets sent by local processes are processed by the output hook.
postrouting All packets leaving the system are processed by the postrouting hook.

ARP address family

The ARP address family handles ARP packets received and sent by the system. It is commonly used to mangle ARP packets for clustering.

Table 2. ARP address family hooks

input Packets delivered to the local system are processed by the input hook.
output Packets send by the local system are processed by the output hook.

Bridge address family

The bridge address family handles ethernet packets traversing bridge devices.

The list of supported hooks is identical to IPv4/IPv6/Inet address families above.

Netdev address family

The Netdev address family handles packets from ingress.

Table 3. Netdev address family hooks

ingress All packets entering the system are processed by this hook. It is invoked before layer 3 protocol handlers and it can be used for early filtering and policing.


{[list] | [flush]}ruleset [family]

{export} [ruleset] {format}

The ruleset keyword is used to identify the whole set of tables, chains, etc. currently in place in kernel. The following ruleset commands exist:


Print the ruleset in human-readable format.


Clear the whole ruleset. Note that unlike iptables, this will remove all tables and whatever they contain, effectively leading to an empty ruleset - no packet filtering will happen anymore, so the kernel accepts any valid packet it receives.


Print the ruleset in machine readable format. The mandatory format parameter may be either xml or json.

It is possible to limit list and flush to a specific address family only. For a list of valid family names, see ADDRESS FAMILIES above.

Note that contrary to what one might assume, the output generated by export is not parseable by nft -f. Instead, the output of list command serves well for that purpose.


{[add] | [delete] | [list] | [flush]}table [family] {table}

Tables are containers for chains, sets and stateful objects. They are identified by their address family and their name. The address family must be one of ip, ip6, inet, arp, bridge, netdev. The inet address family is a dummy family which is used to create hybrid IPv4/IPv6 tables. The meta expression nfproto keyword can be used to test which family (ipv4 or ipv6) context the packet is being processed in. When no address family is specified, ip is used by default.


Add a new table for the given family with the given name.


Delete the specified table.


List all chains and rules of the specified table.


Flush all chains and rules of the specified table.


{[add] | [create]}chain [family] table chain [ { {type} {hook} [device] {priority ;} [policy ;] } ]

{[delete] | [list] | [flush]}chain [family] {table} {chain}

{rename}chain [family] {table} {chain} {newname}

Chains are containers for rules. They exist in two kinds, base chains and regular chains. A base chain is an entry point for packets from the networking stack, a regular chain may be used as jump target and is used for better rule organization.


Add a new chain in the specified table. When a hook and priority value are specified, the chain is created as a base chain and hooked up to the networking stack.


Similar to the add command, but returns an error if the chain already exists.


Delete the specified chain. The chain must not contain any rules or be used as jump target.


Rename the specified chain.


List all rules of the specified chain.


Flush all rules of the specified chain.

For base chains, type, hook and priority parameters are mandatory.

Table 4. Supported chain types

filterallallStandard chain type to use in doubt.
natip, ip6prerouting, input, output, postroutingChains of this type perform Network Address Translation based on conntrack entries. Only the first packet of a connection actually traverses this chain - its rules usually define details of the created conntrack entry (NAT statements for instance).
routeip, ip6outputIf a packet has traversed a chain of this type and is about to be accepted, a new route lookup is performed if relevant parts of the IP header have changed. This allows to e.g. implement policy routing selectors in nftables.

Apart from the special cases illustrated above (e.g. nat type not supporting forward hook or route type only supporting output hook), there are two further quirks worth noticing:

  • netdev family supports merely a single combination, namely filter type and ingress hook. Base chains in this family also require the device parameter to be present since they exist per incoming interface only.
  • arp family supports only input and output hooks, both in chains of type filter.

The priority parameter accepts a signed integer value which specifies the order in which chains with same hook value are traversed. The ordering is ascending, i.e. lower priority values have precedence over higher ones.

Base chains also allow to set the chain's policy, i.e. what happens to packets not explicitly accepted or refused in contained rules. Supported policy values are accept (which is the default) or drop.


[[add] | {insert}]rule [family] {table} {chain} [position position] {statement...}

{replace}rule [family] {table} {chain} {handle handle} {statement...}

{delete}rule [family] {table} {chain} {handle handle}

Rules are constructed from two kinds of components according to a set of grammatical rules: expressions and statements.


Add a new rule described by the list of statements. The rule is appended to the given chain unless a position is specified, in which case the rule is appended to the rule given by the position.


Similar to the add command, but the rule is prepended to the beginning of the chain or before the rule at the given position.


Similar to the add command, but the rule replaces the specified rule.


Delete the specified rule.


{add} set [family] {table} {set} { {type} [flags] [timeout] [gc-interval] [elements] [size] [policy] }

{[delete] | [list] | [flush]} set [family] {table} {set}

{[add] | [delete]} element [family] {table} {set} { {elements} }

Sets are elements containers of an user-defined data type, they are uniquely identified by an user-defined name and attached to tables.


Add a new set in the specified table.


Delete the specified set.


Display the elements in the specified set.


Remove all elements from the specified set.

add element

Comma-separated list of elements to add into the specified set.

delete element

Comma-separated list of elements to delete from the specified set.

Table 5. Set specifications

typedata type of set elementsstring: ipv4_addr, ipv6_addr, ether_addr, inet_proto, inet_service, mark
flagsset flagsstring: constant, interval, timeout
timeouttime an element stays in the setstring, decimal followed by unit. Units are: d, h, m, s
gc-intervalgarbage collection interval, only available when timeout or flag timeout are activestring, decimal followed by unit. Units are: d, h, m, s
elementselements contained by the setset data type
sizemaximun number of elements in the setunsigned integer (64 bit)
policyset policystring: performance [default], memory


{add} map [family] {table} {map} { {type} [flags] [elements] [size] [policy] }

{[delete] | [list] | [flush]} map [family] {table} {map}

{[add] | [delete]} element [family] {table} {map} { {elements} }

Maps store data based on some specific key used as input, they are uniquely identified by an user-defined name and attached to tables.


Add a new map in the specified table.


Delete the specified map.


Display the elements in the specified map.


Remove all elements from the specified map.

add element

Comma-separated list of elements to add into the specified map.

delete element

Comma-separated list of element keys to delete from the specified map.

Table 6. Map specifications

typedata type of map elementsstring ':' string: ipv4_addr, ipv6_addr, ether_addr, inet_proto, inet_service, mark, counter, quota. Counter and quota can't be used as keys
flagsmap flagsstring: constant, interval
elementselements contained by the mapmap data type
sizemaximun number of elements in the mapunsigned integer (64 bit)
policymap policystring: performance [default], memory

Stateful objects

{[add] | [delete] | [list] | [reset]} type [family] {table} {object}

Stateful objects are attached to tables and are identified by an unique name. They group stateful information from rules, to reference them in rules the keywords "type name" are used e.g. "counter name".


Add a new stateful object in the specified table.


Delete the specified object.


Display stateful information the object holds.


List-and-reset stateful object.


ct {helper} {type} {type} {protocol} {protocol} [l3proto] [family]

Ct helper is used to define connection tracking helpers that can then be used in combination with the "ct helper set" statement. type and protocol are mandatory, l3proto is derived from the table family by default, i.e. in the inet table the kernel will try to load both the ipv4 and ipv6 helper backends, if they are supported by the kernel.

Table 7. conntrack helper specifications

typename of helper typequoted string (e.g. "ftp")
protocollayer 4 protocol of the helperstring (e.g. tcp)
l3protolayer 3 protocol of the helperaddress family (e.g. ip)

Example 2. defining and assigning ftp helper

Unlike iptables, helper assignment needs to be performed after the conntrack lookup has completed, for example with the default 0 hook priority.

table inet myhelpers {
  ct helper ftp-standard {
     type "ftp" protocol tcp
  chain prerouting {
      type filter hook prerouting priority 0;
      tcp dport 21 ct helper set "ftp-standard"


counter [packets bytes]

Table 8. Counter specifications

packetsinitial count of packetsunsigned integer (64 bit)
bytesinitial count of bytesunsigned integer (64 bit)


quota [[over] | [until]] [used]

Table 9. Quota specifications

quotaquota limit, used as the quota nameTwo arguments, unsigned interger (64 bit) and string: bytes, kbytes, mbytes. "over" and "until" go before these arguments
usedinitial value of used quotaTwo arguments, unsigned interger (64 bit) and string: bytes, kbytes, mbytes


Expressions represent values, either constants like network addresses, port numbers etc. or data gathered from the packet during ruleset evaluation. Expressions can be combined using binary, logical, relational and other types of expressions to form complex or relational (match) expressions. They are also used as arguments to certain types of operations, like NAT, packet marking etc.

Each expression has a data type, which determines the size, parsing and representation of symbolic values and type compatibility with other expressions.

describe command

describe {expression}

The describe command shows information about the type of an expression and its data type.

Example 3. The describe command

$ nft describe tcp flags
payload expression, datatype tcp_flag (TCP flag) (basetype bitmask, integer), 8 bits

pre-defined symbolic constants:
fin                           	0x01
syn                           	0x02
rst                           	0x04
psh                           	0x08
ack                           	0x10
urg                           	0x20
ecn                           	0x40
cwr                           	0x80

Data types

Data types determine the size, parsing and representation of symbolic values and type compatibility of expressions. A number of global data types exist, in addition some expression types define further data types specific to the expression type. Most data types have a fixed size, some however may have a dynamic size, f.i. the string type.

Types may be derived from lower order types, f.i. the IPv4 address type is derived from the integer type, meaning an IPv4 address can also be specified as an integer value.

In certain contexts (set and map definitions) it is necessary to explicitly specify a data type. Each type has a name which is used for this.

Integer type

Table 10. 

NameKeywordSizeBase type

The integer type is used for numeric values. It may be specified as decimal, hexadecimal or octal number. The integer type doesn't have a fixed size, its size is determined by the expression for which it is used.

Bitmask type

Table 11. 

NameKeywordSizeBase type

The bitmask type (bitmask) is used for bitmasks.

String type

Table 12. 

NameKeywordSizeBase type

The string type is used to for character strings. A string begins with an alphabetic character (a-zA-Z) followed by zero or more alphanumeric characters or the characters /, -, _ and .. In addition anything enclosed in double quotes (") is recognized as a string.

Example 4. String specification

# Interface name
filter input iifname eth0

# Weird interface name
filter input iifname "(eth0)"

Link layer address type

Table 13. 

NameKeywordSizeBase type
Link layer addresslladdrvariableinteger

The link layer address type is used for link layer addresses. Link layer addresses are specified as a variable amount of groups of two hexadecimal digits separated using colons (:).

Example 5. Link layer address specification

# Ethernet destination MAC address
filter input ether daddr 20:c9:d0:43:12:d9

IPv4 address type

Table 14. 

NameKeywordSizeBase type
IPv4 addressipv4_addr32 bitinteger

The IPv4 address type is used for IPv4 addresses. Addresses are specified in either dotted decimal, dotted hexadecimal, dotted octal, decimal, hexadecimal, octal notation or as a host name. A host name will be resolved using the standard system resolver.

Example 6. IPv4 address specification

# dotted decimal notation
filter output ip daddr

# host name
filter output ip daddr localhost

IPv6 address type

Table 15. 

NameKeywordSizeBase type
IPv6 addressipv6_addr128 bitinteger

The IPv6 address type is used for IPv6 addresses. FIXME

Example 7. IPv6 address specification

# abbreviated loopback address
filter output ip6 daddr ::1

Boolean type

Table 16. 

NameKeywordSizeBase type
Booleanboolean1 bitinteger

The boolean type is a syntactical helper type in user space. It's use is in the right-hand side of a (typically implicit) relational expression to change the expression on the left-hand side into a boolean check (usually for existence).

The following keywords will automatically resolve into a boolean type with given value:

Table 17. 


Example 8. Boolean specification

The following expressions support a boolean comparison:

Table 18. 

fibCheck route existence.
exthdrCheck IPv6 extension header existence.
tcp optionCheck TCP option header existence.

# match if route exists
filter input fib daddr . iif oif exists

# match only non-fragmented packets in IPv6 traffic
filter input exthdr frag missing

# match if TCP timestamp option is present
filter input tcp option timestamp exists

ICMP Type type

Table 19. 

NameKeywordSizeBase type
ICMP Typeicmp_type8 bitinteger

The ICMP Type type is used to conveniently specify the ICMP header's type field.

The following keywords may be used when specifying the ICMP type:

Table 20. 


Example 9. ICMP Type specification

# match ping packets
filter output icmp type { echo-request, echo-reply }

ICMP Code type

Table 21. 

NameKeywordSizeBase type
ICMP Codeicmp_code8 bitinteger

The ICMP Code type is used to conveniently specify the ICMP header's code field.

The following keywords may be used when specifying the ICMP code:

Table 22. 


ICMPv6 Type type

Table 23. 

NameKeywordSizeBase type
ICMPv6 Typeicmpv6_type8 bitinteger

The ICMPv6 Type type is used to conveniently specify the ICMPv6 header's type field.

The following keywords may be used when specifying the ICMPv6 type:

Table 24. 


Example 10. ICMPv6 Type specification

# match ICMPv6 ping packets
filter output icmpv6 type { echo-request, echo-reply }

ICMPv6 Code type

Table 25. 

NameKeywordSizeBase type
ICMPv6 Codeicmpv6_code8 bitinteger

The ICMPv6 Code type is used to conveniently specify the ICMPv6 header's code field.

The following keywords may be used when specifying the ICMPv6 code:

Table 26. 


ICMPvX Code type

Table 27. 

NameKeywordSizeBase type
ICMPvX Codeicmpx_code8 bitinteger

The ICMPvX Code type abstraction is a set of values which overlap between ICMP and ICMPv6 Code types to be used from the inet family.

The following keywords may be used when specifying the ICMPvX code:

Table 28. 


Conntrack types

This is an overview of types used in ct expression and statement:

Table 29. 

NameKeywordSizeBase type
conntrack statect_state4 bytebitmask
conntrack directionct_dir8 bitinteger
conntrack statusct_status4 bytebitmask
conntrack event bitsct_event4 bytebitmask
conntrack labelct_label128 bitbitmask

For each of the types above, keywords are available for convenience:

Table 30. conntrack state (ct_state)


Table 31. conntrack direction (ct_dir)


Table 32. conntrack status (ct_status)


Table 33. conntrack event bits (ct_event)


Possible keywords for conntrack label type (ct_label) are read at runtime from /etc/connlabel.conf.

Primary expressions

The lowest order expression is a primary expression, representing either a constant or a single datum from a packet's payload, meta data or a stateful module.

Meta expressions

meta {[length] | [nfproto] | [l4proto] | [protocol] | [priority]}

[meta] {[mark] | [iif] | [iifname] | [iiftype] | [oif] | [oifname] | [oiftype] | [skuid] | [skgid] | [nftrace] | [rtclassid] | [ibriport] | [obriport] | [pkttype] | [cpu] | [iifgroup] | [oifgroup] | [cgroup] | [random]}

A meta expression refers to meta data associated with a packet.

There are two types of meta expressions: unqualified and qualified meta expressions. Qualified meta expressions require the meta keyword before the meta key, unqualified meta expressions can be specified by using the meta key directly or as qualified meta expressions.

Table 34. Meta expression types

lengthLength of the packet in bytesinteger (32 bit)
nfprotoreal hook protocol family, useful only in inet tableinteger (32 bit)
protocolEthertype protocol valueether_type
priorityTC packet prioritytc_handle
markPacket markmark
iifInput interface indexiface_index
iifnameInput interface namestring
iiftypeInput interface typeiface_type
oifOutput interface indexiface_index
oifnameOutput interface namestring
oiftypeOutput interface hardware typeiface_type
skuidUID associated with originating socketuid
skgidGID associated with originating socketgid
rtclassidRouting realmrealm
ibriportInput bridge interface namestring
obriportOutput bridge interface namestring
pkttypepacket typepkt_type
cpucpu number processing the packetinteger (32 bits)
iifgroupincoming device groupdevgroup
oifgroupoutgoing device groupdevgroup
cgroupcontrol group idinteger (32 bits)
randompseudo-random numberinteger (32 bits)

Table 35. Meta expression specific types

iface_index Interface index (32 bit number). Can be specified numerically or as name of an existing interface.
ifname Interface name (16 byte string). Does not have to exist.
iface_type Interface type (16 bit number).
uid User ID (32 bit number). Can be specified numerically or as user name.
gid Group ID (32 bit number). Can be specified numerically or as group name.
realm Routing Realm (32 bit number). Can be specified numerically or as symbolic name defined in /etc/iproute2/rt_realms.
devgroup_type Device group (32 bit number). Can be specified numerically or as symbolic name defined in /etc/iproute2/group.
pkt_type Packet type: Unicast (addressed to local host), Broadcast (to all), Multicast (to group).

Example 11. Using meta expressions

# qualified meta expression
filter output meta oif eth0

# unqualified meta expression
filter output oif eth0

fib expressions

fib {[saddr] | [daddr] [[mark] | [iif] | [oif]]} {[oif] | [oifname] | [type]}

A fib expression queries the fib (forwarding information base) to obtain information such as the output interface index a particular address would use. The input is a tuple of elements that is used as input to the fib lookup functions.

Table 36. fib expression specific types

oifOutput interface indexinteger (32 bit)
oifnameOutput interface namestring
typeAddress typefib_addrtype

Example 12. Using fib expressions

# drop packets without a reverse path
filter prerouting fib saddr . iif oif missing drop

# drop packets to address not configured on ininterface
filter prerouting fib daddr . iif type != { local, broadcast, multicast } drop

# perform lookup in a specific 'blackhole' table (0xdead, needs ip appropriate ip rule)
filter prerouting meta mark set 0xdead fib daddr . mark type vmap { blackhole : drop, prohibit : jump prohibited, unreachable : drop }

Routing expressions

rt {[classid] | [nexthop]}

A routing expression refers to routing data associated with a packet.

Table 37. Routing expression types

classidRouting realmrealm
nexthopRouting nexthopipv4_addr/ipv6_addr
mtuTCP maximum segment size of routeinteger (16 bit)

Table 38. Routing expression specific types

realm Routing Realm (32 bit number). Can be specified numerically or as symbolic name defined in /etc/iproute2/rt_realms.

Example 13. Using routing expressions

# IP family independent rt expression
filter output rt classid 10

# IP family dependent rt expressions
ip filter output rt nexthop
ip6 filter output rt nexthop fd00::1
inet filter output rt ip nexthop
inet filter output rt ip6 nexthop fd00::1

Payload expressions

Payload expressions refer to data from the packet's payload.

Ethernet header expression

ether [ethernet header field]

Table 39. Ethernet header expression types

daddrDestination MAC addressether_addr
saddrSource MAC addressether_addr

VLAN header expression

vlan [VLAN header field]

Table 40. VLAN header expression

idVLAN ID (VID)integer (12 bit)
cfiCanonical Format Indicatorinteger (1 bit)
pcpPriority code pointinteger (3 bit)

ARP header expression

arp [ARP header field]

Table 41. ARP header expression

htypeARP hardware typeinteger (16 bit)
hlenHardware address leninteger (8 bit)
plenProtocol address leninteger (8 bit)

IPv4 header expression

ip [IPv4 header field]

Table 42. IPv4 header expression

versionIP header version (4)integer (4 bit)
hdrlengthIP header length including optionsinteger (4 bit) FIXME scaling
dscpDifferentiated Services Code Pointdscp
ecnExplicit Congestion Notificationecn
lengthTotal packet lengthinteger (16 bit)
idIP IDinteger (16 bit)
frag-offFragment offsetinteger (16 bit)
ttlTime to liveinteger (8 bit)
protocolUpper layer protocolinet_proto
checksumIP header checksuminteger (16 bit)
saddrSource addressipv4_addr
daddrDestination addressipv4_addr

ICMP header expression

icmp [ICMP header field]

Table 43. ICMP header expression

typeICMP type fieldicmp_type
codeICMP code fieldinteger (8 bit)
checksumICMP checksum fieldinteger (16 bit)
idID of echo request/responseinteger (16 bit)
sequencesequence number of echo request/responseinteger (16 bit)
gatewaygateway of redirectsinteger (32 bit)
mtuMTU of path MTU discoveryinteger (16 bit)

IPv6 header expression

ip6 [IPv6 header field]

Table 44. IPv6 header expression

versionIP header version (6)integer (4 bit)
dscpDifferentiated Services Code Pointdscp
ecnExplicit Congestion Notificationecn
flowlabelFlow labelinteger (20 bit)
lengthPayload lengthinteger (16 bit)
nexthdrNexthdr protocolinet_proto
hoplimitHop limitinteger (8 bit)
saddrSource addressipv6_addr
daddrDestination addressipv6_addr

ICMPv6 header expression

icmpv6 [ICMPv6 header field]

Table 45. ICMPv6 header expression

typeICMPv6 type fieldicmpv6_type
codeICMPv6 code fieldinteger (8 bit)
checksumICMPv6 checksum fieldinteger (16 bit)
parameter-problempointer to probleminteger (32 bit)
packet-too-bigoversized MTUinteger (32 bit)
idID of echo request/responseinteger (16 bit)
sequencesequence number of echo request/responseinteger (16 bit)
max-delaymaximum response delay of MLD queriesinteger (16 bit)

TCP header expression

tcp [TCP header field]

Table 46. TCP header expression

sportSource portinet_service
dportDestination portinet_service
sequenceSequence numberinteger (32 bit)
ackseqAcknowledgement numberinteger (32 bit)
doffData offsetinteger (4 bit) FIXME scaling
reservedReserved areainteger (4 bit)
flagsTCP flagstcp_flag
windowWindowinteger (16 bit)
checksumChecksuminteger (16 bit)
urgptrUrgent pointerinteger (16 bit)

UDP header expression

udp [UDP header field]

Table 47. UDP header expression

sportSource portinet_service
dportDestination portinet_service
lengthTotal packet lengthinteger (16 bit)
checksumChecksuminteger (16 bit)

UDP-Lite header expression

udplite [UDP-Lite header field]

Table 48. UDP-Lite header expression

sportSource portinet_service
dportDestination portinet_service
checksumChecksuminteger (16 bit)

SCTP header expression

sctp [SCTP header field]

Table 49. SCTP header expression

sportSource portinet_service
dportDestination portinet_service
vtagVerfication Taginteger (32 bit)
checksumChecksuminteger (32 bit)

DCCP header expression

dccp [DCCP header field]

Table 50. DCCP header expression

sportSource portinet_service
dportDestination portinet_service

Authentication header expression

ah [AH header field]

Table 51. AH header expression

nexthdrNext header protocolinet_proto
hdrlengthAH Header lengthinteger (8 bit)
reservedReserved areainteger (16 bit)
spiSecurity Parameter Indexinteger (32 bit)
sequenceSequence numberinteger (32 bit)

Encrypted security payload header expression

esp [ESP header field]

Table 52. ESP header expression

spiSecurity Parameter Indexinteger (32 bit)
sequenceSequence numberinteger (32 bit)

IPcomp header expression

comp [IPComp header field]

Table 53. IPComp header expression

nexthdrNext header protocolinet_proto
cpiCompression Parameter Indexinteger (16 bit)

Extension header expressions

Extension header expressions refer to data from variable-sized protocol headers, such as IPv6 extension headers and TCPs options.

nftables currently supports matching (finding) a given ipv6 extension header or TCP option.

hbh {[nexthdr] | [hdrlength]}

frag {[nexthdr] | [frag-off] | [more-fragments] | [id]}

rt {[nexthdr] | [hdrlength] | [type] | [seg-left]}

dst {[nexthdr] | [hdrlength]}

mh {[nexthdr] | [hdrlength] | [checksum] | [type]}

tcp option {[eol] | [noop] | [maxseg] | [window] | [sack-permitted] | [sack] | [sack0] | [sack1] | [sack2] | [sack3] | [timestamp]} [tcp_option_field]

The following syntaxes are valid only in a relational expression with boolean type on right-hand side for checking header existence only:

exthdr {[hbh] | [frag] | [rt] | [dst] | [mh]}

tcp option {[eol] | [noop] | [maxseg] | [window] | [sack-permitted] | [sack] | [sack0] | [sack1] | [sack2] | [sack3] | [timestamp]}

Table 54. IPv6 extension headers

hbhHop by Hop
rtRouting Header
fragFragmentation header
dstdst options
mhMobility Header

Table 55. TCP Options

KeywordDescriptionTCP option fields
eolEnd of option listkind
noop1 Byte TCP No-op optionskind
maxsegTCP Maximum Segment Sizekind, length, size
windowTCP Window Scalingkind, length, count
sack-permittedTCP SACK permittedkind, length
sackTCP Selective Acknowledgement (alias of block 0)kind, length, left, right
sack0TCP Selective Acknowledgement (block 0)kind, length, left, right
sack1TCP Selective Acknowledgement (block 1)kind, length, left, right
sack2TCP Selective Acknowledgement (block 2)kind, length, left, right
sack3TCP Selective Acknowledgement (block 3)kind, length, left, right
timestampTCP Timestampskind, length, tsval, tsecr

Example 14. finding TCP options

filter input tcp option sack-permitted kind 1 counter

Example 15. matching IPv6 exthdr

ip6 filter input frag more-fragments 1 counter

Conntrack expressions

Conntrack expressions refer to meta data of the connection tracking entry associated with a packet.

There are three types of conntrack expressions. Some conntrack expressions require the flow direction before the conntrack key, others must be used directly because they are direction agnostic. The packets, bytes and avgpkt keywords can be used with or without a direction. If the direction is omitted, the sum of the original and the reply direction is returned. The same is true for the zone, if a direction is given, the zone is only matched if the zone id is tied to the given direction.

ct {[state] | [direction] | [status] | [mark] | [expiration] | [helper] | [label] | [l3proto] | [protocol] | [bytes] | [packets] | [avgpkt] | [zone]}

ct {[original] | [reply]} {[l3proto] | [protocol] | [proto-src] | [proto-dst] | [bytes] | [packets] | [avgpkt] | [zone]}

ct {[original] | [reply]} {[ip] | [ip6]} {[saddr] | [daddr]}

Table 56. Conntrack expressions

stateState of the connectionct_state
directionDirection of the packet relative to the connectionct_dir
statusStatus of the connectionct_status
markConnection markmark
expirationConnection expiration timetime
helperHelper associated with the connectionstring
labelConnection tracking label bit or symbolic name defined in connlabel.conf in the nftables include pathct_label
l3protoLayer 3 protocol of the connectionnf_proto
saddrSource address of the connection for the given directionipv4_addr/ipv6_addr
daddrDestination address of the connection for the given directionipv4_addr/ipv6_addr
protocolLayer 4 protocol of the connection for the given directioninet_proto
proto-srcLayer 4 protocol source for the given directioninteger (16 bit)
proto-dstLayer 4 protocol destination for the given directioninteger (16 bit)
packetspacket count seen in the given direction or sum of original and replyinteger (64 bit)
bytesbytecount seen, see description for packets keywordinteger (64 bit)
avgpktaverage bytes per packet, see description for packets keywordinteger (64 bit)
zoneconntrack zoneinteger (16 bit)

A description of conntrack-specific types listed above can be found sub-section CONNTRACK TYPES above.


Statements represent actions to be performed. They can alter control flow (return, jump to a different chain, accept or drop the packet) or can perform actions, such as logging, rejecting a packet, etc.

Statements exist in two kinds. Terminal statements unconditionally terminate evaluation of the current rule, non-terminal statements either only conditionally or never terminate evaluation of the current rule, in other words, they are passive from the ruleset evaluation perspective. There can be an arbitrary amount of non-terminal statements in a rule, but only a single terminal statement as the final statement.

Verdict statement

The verdict statement alters control flow in the ruleset and issues policy decisions for packets.

{[accept] | [drop] | [queue] | [continue] | [return]}

{[jump] | [goto]} {chain}


Terminate ruleset evaluation and accept the packet.


Terminate ruleset evaluation and drop the packet.


Terminate ruleset evaluation and queue the packet to userspace.


Continue ruleset evaluation with the next rule. FIXME


Return from the current chain and continue evaluation at the next rule in the last chain. If issued in a base chain, it is equivalent to accept.

jump chain

Continue evaluation at the first rule in chain. The current position in the ruleset is pushed to a call stack and evaluation will continue there when the new chain is entirely evaluated of a return verdict is issued.

goto chain

Similar to jump, but the current position is not pushed to the call stack, meaning that after the new chain evaluation will continue at the last chain instead of the one containing the goto statement.

Example 16. Verdict statements

# process packets from eth0 and the internal network in from_lan
# chain, drop all packets from eth0 with different source addresses.

filter input iif eth0 ip saddr jump from_lan
filter input iif eth0 drop

Payload statement

The payload statement alters packet content. It can be used for example to set ip DSCP (differv) header field or ipv6 flow labels.

Example 17. route some packets instead of bridging

# redirect tcp:http from to local machine for routing instead of bridging
# assumes 00:11:22:33:44:55 is local MAC address.
bridge input meta iif eth0 ip saddr tcp dport 80 meta pkttype set unicast ether daddr set 00:11:22:33:44:55

Example 18. Set IPv4 DSCP header field

ip forward ip dscp set 42

Extension header statement

The extension header statement alters packet content in variable-sized headers. This can currently be used to alter the TCP Maximum segment size of packets, similar to TCPMSS.

Example 19. change tcp mss

tcp flags syn tcp option maxseg size set 1360
# set a size based on route information:
tcp flags syn tcp option maxseg size set rt mtu

Log statement

log [prefix quoted_string] [level syslog-level] [flags log-flags]

log [group nflog_group] [prefix quoted_string] [queue-threshold value] [snaplen size]

The log statement enables logging of matching packets. When this statement is used from a rule, the Linux kernel will print some information on all matching packets, such as header fields, via the kernel log (where it can be read with dmesg(1) or read in the syslog). If the group number is specified, the Linux kernel will pass the packet to nfnetlink_log which will multicast the packet through a netlink socket to the specified multicast group. One or more userspace processes may subscribe to the group to receive the packets, see libnetfilter_queue documentation for details. This is a non-terminating statement, so the rule evaluation continues after the packet is logged.

Table 57. log statement options

prefixLog message prefixquoted string
syslog-levelSyslog level of loggingstring: emerg, alert, crit, err, warn [default], notice, info, debug
groupNFLOG group to send messages tounsigned integer (16 bit)
snaplenLength of packet payload to include in netlink messageunsigned integer (32 bit)
queue-thresholdNumber of packets to queue inside the kernel before sending them to userspaceunsigned integer (32 bit)

Table 58. log-flags

tcp sequenceLog TCP sequence numbers.
tcp optionsLog options from the TCP packet header.
ip optionsLog options from the IP/IPv6 packet header.
skuidLog the userid of the process which generated the packet.
etherDecode MAC addresses and protocol.
allEnable all log flags listed above.

Example 20. Using log statement

# log the UID which generated the packet and ip options
ip filter output log flags skuid flags ip options

# log the tcp sequence numbers and tcp options from the TCP packet
ip filter output log flags tcp sequence,options

# enable all supported log flags
ip6 filter output log flags all

Reject statement

reject [ [with] {[icmp] | [icmp6] | [icmpx]} [type] {[icmp_type] | [icmp6_type] | [icmpx_type]} ]

reject [ [with] {tcp} {reset} ]

A reject statement is used to send back an error packet in response to the matched packet otherwise it is equivalent to drop so it is a terminating statement, ending rule traversal. This statement is only valid in the input, forward and output chains, and user-defined chains which are only called from those chains.

The different ICMP reject variants are meant for use in different table families:

Table 59. 


For a description of the different types and a list of supported keywords refer to DATA TYPES section above. The common default reject value is port-unreachable.

Counter statement

A counter statement sets the hit count of packets along with the number of bytes.

counter {packets number } {bytes number }

Conntrack statement

The conntrack statement can be used to set the conntrack mark and conntrack labels.

ct {[mark] | [event] | [label] | [zone]} [set]value

The ct statement sets meta data associated with a connection. The zone id has to be assigned before a conntrack lookup takes place, i.e. this has to be done in prerouting and possibly output (if locally generated packets need to be placed in a distinct zone), with a hook priority of -300.

Table 60. Conntrack statement types

eventconntrack event bitsbitmask, integer (32 bit)
helpername of ct helper object to assign to the connectionquoted string
markConnection tracking markmark
labelConnection tracking labellabel
zoneconntrack zoneinteger (16 bit)

Example 21. save packet nfmark in conntrack

ct mark set meta mark

Example 22. set zone mapped via interface

table inet raw {
  chain prerouting {
      type filter hook prerouting priority -300;
      ct zone set iif map { "eth1" : 1, "veth1" : 2 }
  chain output {
      type filter hook output priority -300;
      ct zone set oif map { "eth1" : 1, "veth1" : 2 }

Example 23. restrict events reported by ctnetlink

ct event set new,related,destroy

Meta statement

A meta statement sets the value of a meta expression. The existing meta fields are: priority, mark, pkttype, nftrace.

meta {[mark] | [priority] | [pkttype] | [nftrace]} [set]value

A meta statement sets meta data associated with a packet.

Table 61. Meta statement types

priorityTC packet prioritytc_handle
markPacket markmark
pkttypepacket typepkt_type
nftraceruleset packet tracing on/off. Use monitor trace command to watch traces0, 1

Limit statement

limit [rate] [over]packet_number [/] {[second] | [minute] | [hour] | [day]} [burst packet_number packets]

limit [rate] [over]byte_number {[bytes] | [kbytes] | [mbytes]} [/] {[second] | [minute] | [hour] | [day] | [week]} [burst byte_number bytes]

A limit statement matches at a limited rate using a token bucket filter. A rule using this statement will match until this limit is reached. It can be used in combination with the log statement to give limited logging. The over keyword, that is optional, makes it match over the specified rate.

Table 62. limit statement values

packet_numberNumber of packetsunsigned integer (32 bit)
byte_numberNumber of bytesunsigned integer (32 bit)

NAT statements

snat [to address [:port]] [persistent, random, fully-random]

snat [to address - address [:port - port]] [persistent, random, fully-random]

dnat [to address [:port]] [persistent, random, fully-random]

dnat [to address [:port - port]] [persistent, random, fully-random]

masquerade [to [:port]] [persistent, random, fully-random]

masquerade [to [:port - port]] [persistent, random, fully-random]

redirect [to [:port]] [persistent, random, fully-random]

redirect [to [:port - port]] [persistent, random, fully-random]

The nat statements are only valid from nat chain types.

The snat and masquerade statements specify that the source address of the packet should be modified. While snat is only valid in the postrouting and input chains, masquerade makes sense only in postrouting. The dnat and redirect statements are only valid in the prerouting and output chains, they specify that the destination address of the packet should be modified. You can use non-base chains which are called from base chains of nat chain type too. All future packets in this connection will also be mangled, and rules should cease being examined.

The masquerade statement is a special form of snat which always uses the outgoing interface's IP address to translate to. It is particularly useful on gateways with dynamic (public) IP addresses.

The redirect statement is a special form of dnat which always translates the destination address to the local host's one. It comes in handy if one only wants to alter the destination port of incoming traffic on different interfaces.

Note that all nat statements require both prerouting and postrouting base chains to be present since otherwise packets on the return path won't be seen by netfilter and therefore no reverse translation will take place.

Table 63. NAT statement values

addressSpecifies that the source/destination address of the packet should be modified. You may specify a mapping to relate a list of tuples composed of arbitrary expression key with address value.ipv4_addr, ipv6_addr, eg. abcd::1234, or you can use a mapping, eg. meta mark map { 10 :, 20 : }
portSpecifies that the source/destination address of the packet should be modified.port number (16 bits)

Table 64. NAT statement flags

persistentGives a client the same source-/destination-address for each connection.
randomIf used then port mapping will be randomized using a random seeded MD5 hash mix using source and destination address and destination port.
fully-randomIf used then port mapping is generated based on a 32-bit pseudo-random algorithm.

Example 24. Using NAT statements

# create a suitable table/chain setup for all further examples
add table nat
add chain nat prerouting { type nat hook prerouting priority 0; }
add chain nat postrouting { type nat hook postrouting priority 100; }

# translate source addresses of all packets leaving via eth0 to address
add rule nat postrouting oif eth0 snat to

# redirect all traffic entering via eth0 to destination address
add rule nat prerouting iif eth0 dnat to

# translate source addresses of all packets leaving via eth0 to whatever
# locally generated packets would use as source to reach the same destination
add rule nat postrouting oif eth0 masquerade

# redirect incoming TCP traffic for port 22 to port 2222
add rule nat prerouting tcp dport 22 redirect to :2222

Queue statement

This statement passes the packet to userspace using the nfnetlink_queue handler. The packet is put into the queue identified by its 16-bit queue number. Userspace can inspect and modify the packet if desired. Userspace must then drop or reinject the packet into the kernel. See libnetfilter_queue documentation for details.

queue [num queue_number] [bypass]

queue [num queue_number_from - queue_number_to] [bypass,fanout]

Table 65. queue statement values

queue_numberSets queue number, default is 0.unsigned integer (16 bit)
queue_number_fromSets initial queue in the range, if fanout is used.unsigned integer (16 bit)
queue_number_toSets closing queue in the range, if fanout is used.unsigned integer (16 bit)

Table 66. queue statement flags

bypassLet packets go through if userspace application cannot back off. Before using this flag, read libnetfilter_queue documentation for performance tuning recomendations.
fanoutDistribute packets between several queues.

Additional commands

These are some additional commands included in nft.


The monitor command allows you to listen to Netlink events produced by the nf_tables subsystem, related to creation and deletion of objects. When they occur, nft will print to stdout the monitored events in either XML, JSON or native nft format.

To filter events related to a concrete object, use one of the keywords 'tables', 'chains', 'sets', 'rules', 'elements' , 'ruleset'.

To filter events related to a concrete action, use keyword 'new' or 'destroy'.

Hit ^C to finish the monitor operation.

Example 25. Listen to all events, report in native nft format

% nft monitor

Example 26. Listen to added tables, report in XML format

% nft monitor new tables xml

Example 27. Listen to deleted rules, report in JSON format

% nft monitor destroy rules json

Example 28. Listen to both new and destroyed chains, in native nft format

% nft monitor chains

Example 29. Listen to ruleset events such as table, chain, rule, set, counters and quotas, in native nft format

% nft monitor ruleset

Error reporting

When an error is detected, nft shows the line(s) containing the error, the position of the erroneous parts in the input stream and marks up the erroneous parts using carrets (^). If the error results from the combination of two expressions or statements, the part imposing the constraints which are violated is marked using tildes (~).

For errors returned by the kernel, nft can't detect which parts of the input caused the error and the entire command is marked.

Example 30. Error caused by single incorrect expression

<cmdline>:1:19-22: Error: Interface does not exist
filter output oif eth0

Example 31. Error caused by invalid combination of two expressions

<cmdline>:1:28-36: Error: Right hand side of relational expression (==) must be constant
filter output tcp dport == tcp dport
                        ~~ ^^^^^^^^^

Example 32. Error returned by the kernel

<cmdline>:0:0-23: Error: Could not process rule: Operation not permitted
filter output oif wlan0

Exit status

On success, nft exits with a status of 0. Unspecified errors cause it to exit with a status of 1, memory allocation errors with a status of 2, unable to open Netlink socket with 3.

See Also

iptables(8), ip6tables(8), arptables(8), ebtables(8), ip(8), tc(8)

There is an official wiki at:


nftables was written by Patrick McHardy and Pablo Neira Ayuso, among many other contributors from the Netfilter community.


Copyright © 2008-2014 Patrick McHardy 
Copyright © 2013-2016 Pablo Neira Ayuso 

nftables is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License version 2 as published by the Free Software Foundation.

This documentation is licenced under the terms of the Creative Commons Attribution-ShareAlike 4.0 license, CC BY-SA 4.0.