Difference between revisions of "Matching packet metainformation"

From nftables wiki
Jump to navigation Jump to search
m (list pkttype options)
 
(24 intermediate revisions by one other user not shown)
Line 1: Line 1:
The meta selectors allows you to match -- [[Setting_packet_metainformation |and in some cases, set]] -- packet metainformation. That is, information the local host has about the packet (such as how / when it was received) that is not necessarily carried in the packet itself.
The meta selectors allows you to match -- [[Setting_packet_metainformation |and in some cases, set]] -- packet metainformation. That is, information the local host has about the packet (such as how / when it was received) that is not necessarily carried in the packet itself.


= The meta selectors =
= Matching by packet info =


The following meta selectors are available:
The following ''meta'' selectors match packets by information carried by the packet itself:


{| class="wikitable"
{| class="wikitable"
!colspan="5"|Meta Selectors
!colspan="5"|''meta'' Packet Info Selectors
|- style="vertical-align:bottom;"
|- style="vertical-align:bottom;"
! Keyword
! Keyword
Line 15: Line 15:


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''mark''
| ''pkttype''
| [[Setting_packet_metainformation |Y]]
| [[Setting_packet_metainformation |Y]]
| packet mark
| packet type (''unicast'', ''broadcast'', ''multicast'', ''other'')
| mark
| pkt_type
|
 
|- style="vertical-align:top;"
| ''length''
|
| packet length in bytes
| integer (32 bit)
|
 
|- style="vertical-align:top;"
| ''protocol''
|
| packet protocol / EtherType protocol value
| ether_type
| as in skb->protocol
 
|- style="vertical-align:top;"
| ''nfproto''
|
|
| netfilter packet protocol family
| integer (32 bit)
| like ipv4, ipv6, etc...; useful only in inet table
|- style="vertical-align:top;"
| ''l4proto''
|
| layer 4 protocol
| integer (8 bit)
| like tcp, udp, etc...; skips ipv6 extension headers
|}
= Matching by interface =
The following ''meta'' selectors match packets based on incoming or outgoing interfaces:
{| class="wikitable"
!colspan="5"|''meta'' Interface Selectors
|- style="vertical-align:bottom;"
! Keyword
! [[Setting_packet_metainformation |Settable]]
! style="text-align:left;" | Description
! style="text-align:left;" | [[Data_types|Data Type]]
! style="text-align:left;" | Notes


|- style="vertical-align:top;"
|- style="vertical-align:top;"
Line 26: Line 69:
| input interface index
| input interface index
| iface_index
| iface_index
|
| Faster than ''iifname'' as it only has to compare a 32-bit unsigned integer instead of a string.
The interface index is dynamically allocated, so don't use this for interfaces that are dynamically created and destroyed, eg. ''ppp0''.


|- style="vertical-align:top;"
|- style="vertical-align:top;"
Line 40: Line 84:
| input interface type
| input interface type
| iface_type
| iface_type
|
|- style="vertical-align:top;"
| ''iifkind''
|
| input interface kind name
| ifkind
|
|- style="vertical-align:top;"
| ''iifgroup''
|
| input interface group
| devgroup
|
|


Line 47: Line 105:
| output interface index
| output interface index
| iface_index
| iface_index
|
| Faster than ''oifname'' as it only has to compare a 32-bit unsigned integer instead of a string.
The interface index is dynamically allocated, so don't use this for interfaces that are dynamically created and destroyed, eg. ''ppp0''.


|- style="vertical-align:top;"
|- style="vertical-align:top;"
Line 64: Line 123:


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''skuid''
| ''oifkind''
|
|
| UID associated with originating socket
| output interface kind name
| uid
| ifkind
|
|


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''skgid''
| ''oifgroup''
|
|
| GID associated with originating socket
| output interface group
| gid
| devgroup
|
 
|- style="vertical-align:top;"
| ''nftrace''
| [[Setting_packet_metainformation |Y]]
| [[Ruleset_debug/tracing|nftrace debugging]] bit
|
|
 
|- style="vertical-align:top;"
| ''rtclassid''
|
| routing realm
| realm
|
 
|- style="vertical-align:top;"
| ''ibriport''
|
| input bridge port
|
|
 
|- style="vertical-align:top;"
| ''obriport''
|
| output bridge port
|  
|
|


Line 110: Line 141:
| input bridge interface name
| input bridge interface name
| ifname
| ifname
|
| equivalent to obsolete ''ibriport'' keyword


|- style="vertical-align:top;"
|- style="vertical-align:top;"
Line 117: Line 148:
| output bridge interface name
| output bridge interface name
| ifname
| ifname
|
|equivalent to obsolete ''oibriport'' keyword


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''pkttype''
| ''ibrvproto''
| [[Setting_packet_metainformation |Y]]
| packet type
| pkt_type
|
|
| input bridge vlan protocol
| ether_type
|


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''cpu''
| ''ibrpvid''
|
| CPU number processing the packet
| integer (32 bit)
|
|
| input bridge port pvid
| integer (16 bit)
|


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''iifgroup''
| ''sdif''
|
|  
| input interface group
| slave device interface index
| devgroup
| integer
|
|  


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''oifgroup''
| ''sdifname''
|
| output interface group
| devgroup
|
|
| slave device interface name
| ifname
|


|- style="vertical-align:top;"
|}
| ''cgroup''
|
| control group ID
| integer (32 bit)
|
 
|- style="vertical-align:top;"
| ''ipsec''
|
| true if packet was ipsec encrypted
| boolean (1 bit)
|


|- style="vertical-align:top;"
An example rule that uses ''iifname'' to accept all traffic entering the loopback pseudodevice ''lo'':
| ''time''
|
| timestamp of packet reception
| integer (32 bit) or string
|


|- style="vertical-align:top;"
<source lang="bash">
| ''day''
% nft add rule filter input meta iifname lo accept
|
</source>
| day of week
| integer (32 bit) or string
|


|- style="vertical-align:top;"
= Matching by packet mark, routing class and realm =
| ''hour''
|
| hour of day
| string
|


|- style="vertical-align:top;"
{| class="wikitable"
| ''length''
!colspan="5"|''meta'' Packet Mark & Routing Selectors
|
|- style="vertical-align:bottom;"
| packet length in bytes
! Keyword
| integer (32 bit)
! [[Setting_packet_metainformation |Settable]]
|
! style="text-align:left;" | Description
! style="text-align:left;" | [[Data_types|Data Type]]
! style="text-align:left;" | Notes


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''protocol''
| ''mark''
|
| [[Setting_packet_metainformation |Y]]
| packet protocol / EtherType protocol value
| packet mark
| ether_type
| mark
| as in skb->protocol
 
|- style="vertical-align:top;"
| ''nfproto''
|
|
| netfilter packet protocol family
| integer (32 bit)
| like ipv4, ipv6, etc...; useful only in inet table
|- style="vertical-align:top;"
| ''l4proto''
|
| layer 4 protocol
| integer (8 bit)
| like tcp, udp, etc...; skips ipv6 extension headers


|- style="vertical-align:top;"
|- style="vertical-align:top;"
Line 215: Line 209:
| tc packet priority
| tc packet priority
| tc_handle
| tc_handle
|
| [[Classification_to_tc_structure_example |detailed usage example]]


|- style="vertical-align:top;"
|- style="vertical-align:top;"
| ''random''
| ''rtclassid''
|
|
| pseudo-random number
| routing realm
| integer (32 bit)
| realm
|  
| Routing realm references:
<ul>
<li>[http://linux-ip.net/gl/ip-cref/ip-cref-node172.html linux-ip.net]
<li>[http://www.policyrouting.org/PolicyRoutingBook/ONLINE/CH07.web.html policyrouting.org]
</ul>


|- style="vertical-align:top;"
|}
| ''secmark''
| [[Setting_packet_metainformation |Y]]
| packet secmark
|
|


|- style="vertical-align:top;"
You can match packets whose mark is 123 with the following rule:
| ''ibrvproto''
|
| bridge protocol
|
|


|- style="vertical-align:top;"
<source lang="bash">
| ''ibrpvid''
nft add rule filter output meta mark 123 counter
|
</source>
| bridge pvid
|
|


|}
* Since nftables v0.7 you can match the packet priority, the tc classid:


= Matching by interface =
<source>
% nft add rule filter forward meta priority abcd:1234
</source>


You can use one of the following selectors to match the interface name:
* Packet without set priority can be matched using meta priority none
<source>
% nft add rule filter forward meta priority none
</source>


* ''iifname'', to match the input network interface name.
See also: [[Matching routing information|''nexthop'' and ''fib'' selectors]]
* ''oifname'', to match the output network interface name.
* ''iif'', to match the interface index of the network interface name. This is faster than ''iifname'' as it only has to compare a 32-bits unsigned integer instead of a string. The interface index is dynamically allocated, so don't use this for interfaces that are dynamically created and destroyed, eg. ''ppp0''.
* ''oif'', like ''iif'' but it matches the output network interface index.


An example usage of the interface name is the following:


<source lang="bash">
= Matching by socket UID / GID =
% nft add rule filter input meta oifname lo accept
</source>


This rule accepts all traffic for the loopback pseudodevice ''lo''.
{| class="wikitable"
!colspan="5"|''meta'' UID / GID Selectors
|- style="vertical-align:bottom;"
! Keyword
! [[Setting_packet_metainformation |Settable]]
! style="text-align:left;" | Description
! style="text-align:left;" | [[Data_types|Data Type]]
! style="text-align:left;" | Notes


= Matching by packet mark =
|- style="vertical-align:top;"
| ''skuid''
|
| UID associated with originating socket
| uid
|


You can match packets whose mark is 123 with the following rule:
|- style="vertical-align:top;"
| ''skgid''
|
| GID associated with originating socket
| gid
|


<source lang="bash">
|}
nft add rule filter output meta mark 123 counter
</source>
 
= Matching by socket UID =


You can use your user name to match traffic, eg.
You can use your user name to match traffic, eg.
Line 310: Line 307:
'''Important''': Beware if you test this with ''ping'', it is usually installed with suid so that traffic will match the root user (uid=0).
'''Important''': Beware if you test this with ''ping'', it is usually installed with suid so that traffic will match the root user (uid=0).


= Matching by tc priority =


* Since nftables v0.7 you can match the packet priority, the tc classid:
= Matching by time =
 
{| class="wikitable"
!colspan="5"|''meta'' Time Selectors
|- style="vertical-align:bottom;"
! Keyword
! [[Setting_packet_metainformation |Settable]]
! style="text-align:left;" | Description
! style="text-align:left;" | [[Data_types|Data Type]]
! style="text-align:left;" | Notes
 
|- style="vertical-align:top;"
| ''time''
|
| timestamp of packet reception
| time
| Can specify as:
* integer: ns since epoch, or
* string: date in ISO format.
 
|- style="vertical-align:top;"
| ''day''
|
| day of week
| day
| Can specify as:
* integer: 0 = Sunday to 6 = Saturday, or
* case-insensitive string: "Monday", "tuesday", etc. Unique abbreviations also work: "fri", "Sat".
 
|- style="vertical-align:top;"
| ''hour''
|
| hour of day
| hour
| 24-hour "HH:MM:SS", with seconds optional.
 
|}
 
 
= Matching by security selectors =
 
{| class="wikitable"
!colspan="5"|''meta'' Security Selectors
|- style="vertical-align:bottom;"
! Keyword
! [[Setting_packet_metainformation |Settable]]
! style="text-align:left;" | Description
! style="text-align:left;" | [[Data_types|Data Type]]
! style="text-align:left;" | Notes
 
|- style="vertical-align:top;"
| ''cpu''
|
| CPU number processing the packet
| integer (32 bit)
|
 
|- style="vertical-align:top;"
| ''cgroup''
|
| socket control group ID
| integer (32 bit)
|
 
|- style="vertical-align:top;"
| ''secmark''
| [[Setting_packet_metainformation |Y]]
| packet secmark
| integer (32 bit)
|
 
|- style="vertical-align:top;"
| ''ipsec''
|
| true if packet was ipsec encrypted
| boolean (1 bit)
| equivalent to obsolete ''secpath'' keyword
 
|}
 
 
= Matching by miscellaneous selectors =
 
In addition to those in the above subsections, the following miscellaneous meta selectors are available:
 
{| class="wikitable"
!colspan="5"|''meta'' Miscellaneous Selectors
|- style="vertical-align:bottom;"
! Keyword
! [[Setting_packet_metainformation |Settable]]
! style="text-align:left;" | Description
! style="text-align:left;" | [[Data_types|Data Type]]
! style="text-align:left;" | Notes
 
|- style="vertical-align:top;"
| ''nftrace''
| [[Setting_packet_metainformation |Y]]
| [[Ruleset_debug/tracing|nftrace debugging]] bit
| boolean (1 bit)
|


<source>
|- style="vertical-align:top;"
% nft add rule filter forward meta priority abcd:1234
| ''random''
</source>
|
| pseudo-random number
| integer (32 bit)
|


* Packet without set priority can be matched using meta priority none
|}
<source>
% nft add rule filter forward meta priority none
</source>

Latest revision as of 16:33, 28 March 2024

The meta selectors allows you to match -- and in some cases, set -- packet metainformation. That is, information the local host has about the packet (such as how / when it was received) that is not necessarily carried in the packet itself.

Matching by packet info

The following meta selectors match packets by information carried by the packet itself:

meta Packet Info Selectors
Keyword Settable Description Data Type Notes
pkttype Y packet type (unicast, broadcast, multicast, other) pkt_type
length packet length in bytes integer (32 bit)
protocol packet protocol / EtherType protocol value ether_type as in skb->protocol
nfproto netfilter packet protocol family integer (32 bit) like ipv4, ipv6, etc...; useful only in inet table
l4proto layer 4 protocol integer (8 bit) like tcp, udp, etc...; skips ipv6 extension headers

Matching by interface

The following meta selectors match packets based on incoming or outgoing interfaces:

meta Interface Selectors
Keyword Settable Description Data Type Notes
iif input interface index iface_index Faster than iifname as it only has to compare a 32-bit unsigned integer instead of a string.

The interface index is dynamically allocated, so don't use this for interfaces that are dynamically created and destroyed, eg. ppp0.

iifname input interface name ifname
iiftype input interface type iface_type
iifkind input interface kind name ifkind
iifgroup input interface group devgroup
oif output interface index iface_index Faster than oifname as it only has to compare a 32-bit unsigned integer instead of a string.

The interface index is dynamically allocated, so don't use this for interfaces that are dynamically created and destroyed, eg. ppp0.

oifname output interface name ifname
oiftype output interface type iface_type
oifkind output interface kind name ifkind
oifgroup output interface group devgroup
ibrname input bridge interface name ifname equivalent to obsolete ibriport keyword
obrname output bridge interface name ifname equivalent to obsolete oibriport keyword
ibrvproto input bridge vlan protocol ether_type
ibrpvid input bridge port pvid integer (16 bit)
sdif slave device interface index integer
sdifname slave device interface name ifname

An example rule that uses iifname to accept all traffic entering the loopback pseudodevice lo:

% nft add rule filter input meta iifname lo accept

Matching by packet mark, routing class and realm

meta Packet Mark & Routing Selectors
Keyword Settable Description Data Type Notes
mark Y packet mark mark
priority Y tc packet priority tc_handle detailed usage example
rtclassid routing realm realm Routing realm references:

You can match packets whose mark is 123 with the following rule:

nft add rule filter output meta mark 123 counter
  • Since nftables v0.7 you can match the packet priority, the tc classid:
% nft add rule filter forward meta priority abcd:1234
  • Packet without set priority can be matched using meta priority none
% nft add rule filter forward meta priority none

See also: nexthop and fib selectors


Matching by socket UID / GID

meta UID / GID Selectors
Keyword Settable Description Data Type Notes
skuid UID associated with originating socket uid
skgid GID associated with originating socket gid

You can use your user name to match traffic, eg.

% nft add rule filter output meta skuid pablo counter

Or the 32-bits unsigned integer (UID) in case there is no entry in /etc/passwd for a given user.

% nft add rule filter output meta skuid 1000 counter

Let's just generate some HTTP traffic to test this rule:

% wget --spider http://www.google.com

Then, if you check the counters, you can verify that the packets are matching that rule.

% nft list table filter
table ip filter {
        chain output {
                 type filter hook output priority 0;
                 skuid pablo counter packets 7 bytes 510
        }

        chain input {
                 type filter hook input priority 0;
        }
}

Important: Beware if you test this with ping, it is usually installed with suid so that traffic will match the root user (uid=0).


Matching by time

meta Time Selectors
Keyword Settable Description Data Type Notes
time timestamp of packet reception time Can specify as:
  • integer: ns since epoch, or
  • string: date in ISO format.
day day of week day Can specify as:
  • integer: 0 = Sunday to 6 = Saturday, or
  • case-insensitive string: "Monday", "tuesday", etc. Unique abbreviations also work: "fri", "Sat".
hour hour of day hour 24-hour "HH:MM:SS", with seconds optional.


Matching by security selectors

meta Security Selectors
Keyword Settable Description Data Type Notes
cpu CPU number processing the packet integer (32 bit)
cgroup socket control group ID integer (32 bit)
secmark Y packet secmark integer (32 bit)
ipsec true if packet was ipsec encrypted boolean (1 bit) equivalent to obsolete secpath keyword


Matching by miscellaneous selectors

In addition to those in the above subsections, the following miscellaneous meta selectors are available:

meta Miscellaneous Selectors
Keyword Settable Description Data Type Notes
nftrace Y nftrace debugging bit boolean (1 bit)
random pseudo-random number integer (32 bit)