sbin/route/route.8

.\" Copyright (c) 1983, 1991, 1993
.\"	The Regents of the University of California.  All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\"    notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\"    notice, this list of conditions and the following disclaimer in the
.\"    documentation and/or other materials provided with the distribution.
.\" 3. Neither the name of the University nor the names of its contributors
.\"    may be used to endorse or promote products derived from this software
.\"    without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\"     @(#)route.8	8.3 (Berkeley) 3/19/94
.\"
.Dd June 16, 2023
.Dt ROUTE 8
.Os
.Sh NAME
.Nm route
.Nd manually manipulate the routing tables
.Sh SYNOPSIS
.Nm
.Op Fl j Ar jail
.Op Fl dnqtv
.Ar command
.Oo
.Op Ar modifiers
.Ar args
.Oc
.Sh DESCRIPTION
The
.Nm
utility is used to manually manipulate the network
routing tables.
It normally is not needed, as a
system routing table management daemon, such as
.Xr routed 8 ,
should tend to this task.
.Pp
The
.Nm
utility supports a limited number of general options,
but a rich command language, enabling the user to specify
any arbitrary request that could be delivered via the
programmatic interface discussed in
.Xr route 4 .
.Pp
The following options are available:
.Bl -tag -width indent
.It Fl 4
Specify
.Cm inet
address family as family hint for subcommands.
.It Fl 6
Specify
.Cm inet6
address family as family hint for subcommands.
.It Fl d
Run in debug-only mode, i.e., do not actually modify the routing table.
.It Fl n
Bypass attempts to print host and network names symbolically
when reporting actions.
(The process of translating between symbolic
names and numerical equivalents can be quite time consuming, and
may require correct operation of the network; thus it may be expedient
to forget this, especially when attempting to repair networking operations).
.It Fl t
Run in test-only mode.
.Pa /dev/null
is used instead of a socket.
.It Fl v
(verbose) Print additional details.
.It Fl q
Suppress all output from the
.Cm add , change , delete ,
and
.Cm flush
commands.
.It Fl j Ar jail
Run inside a jail.
.El
.Pp
The
.Nm
utility provides the following commands:
.Pp
.Bl -tag -width Fl -compact
.It Cm add
Add a route.
.It Cm flush
Remove all routes.
.It Cm delete
Delete a specific route.
.It Cm del
Another name for the
.Cm delete
command.
.It Cm change
Change aspects of a route (such as its gateway).
.It Cm get
Lookup and display the route for a destination.
.It Cm monitor
Continuously report any changes to the routing information base,
routing lookup misses, or suspected network partitionings.
.It Cm show
Another name for the
.Cm get
command.
.El
.Pp
The monitor command has the syntax:
.Pp
.Bd -ragged -offset indent -compact
.Nm
.Op Fl n
.Cm monitor Op Fl fib Ar number
.Ed
.Pp
The flush command has the syntax:
.Pp
.Bd -ragged -offset indent -compact
.Nm
.Op Fl n
.Cm flush Oo Ar family Oc Op Fl fib Ar number
.Ed
.Pp
If the
.Cm flush
command is specified,
.Nm
will ``flush'' the routing tables of all gateway entries.
When the address family may is specified by any of the
.Fl inet6 ,
or
.Fl inet
modifiers, only routes having destinations with addresses in the
delineated family will be deleted.
Additionally,
.Fl 4
or
.Fl 6
can be used as aliases for
.Fl inet
and
.Fl inet6
modifiers.
When a
.Fl fib
option is specified, the operation will be applied to
the specified FIB
.Pq routing table .
.Pp
The add command has the following syntax:
.Pp
.Bd -ragged -offset indent -compact
.Nm
.Op Fl n
.Cm add
.Op Fl net No \&| Fl host
.Ar destination gateway
.Op Ar netmask
.Op Fl fib Ar number
.Ed
.Pp
and the other commands have the following syntax:
.Pp
.Bd -ragged -offset indent -compact
.Nm
.Op Fl n
.Ar command
.Op Fl net No \&| Fl host
.Ar destination
.Op Ar gateway Op Ar netmask
.Op Fl fib Ar number
.Ed
.Pp
where
.Ar destination
is the destination host or network,
.Ar gateway
is the next-hop intermediary via which packets should be routed.
Routes to a particular host may be distinguished from those to
a network by interpreting the Internet address specified as the
.Ar destination
argument.
The optional modifiers
.Fl net
and
.Fl host
force the destination to be interpreted as a network or a host, respectively.
Otherwise, if the
.Ar destination
has a
.Dq local address part
of
INADDR_ANY
.Pq Li 0.0.0.0 ,
or if the
.Ar destination
is the symbolic name of a network, then the route is
assumed to be to a network; otherwise, it is presumed to be a
route to a host.
Optionally, the
.Ar destination
could also be specified in the
.Ar net Ns / Ns Ar bits
format.
.Pp
For example,
.Li 128.32
is interpreted as
.Fl host Li 128.0.0.32 ;
.Li 128.32.130
is interpreted as
.Fl host Li 128.32.0.130 ;
.Fl net Li 128.32
is interpreted as
.Li 128.32.0.0 ;
.Fl net Li 128.32.130
is interpreted as
.Li 128.32.130.0 ;
and
.Li 192.168.64/20
is interpreted as
.Fl net Li 192.168.64 Fl netmask Li 255.255.240.0 .
.Pp
A
.Ar destination
of
.Ar default
is a synonym for the default route.
For
.Li IPv4
it is
.Fl net Fl inet Li 0.0.0.0 ,
and for
.Li IPv6
it is
.Fl net Fl inet6 Li :: .
.Pp
If the destination is directly reachable
via an interface requiring
no intermediary system to act as a gateway, the
.Fl interface
modifier should be specified;
the gateway given is the address of this host on the common network,
indicating the interface to be used for transmission.
Alternately, if the interface is point to point the name of the interface
itself may be given, in which case the route remains valid even
if the local or remote addresses change.
.Pp
The optional
.Fl netmask
modifier is intended
to achieve the effect of an OSI ESIS
redirect with the netmask option,
or to manually add subnet routes with
netmasks different from that of the implied network interface
(as would otherwise be communicated using the OSPF or ISIS routing protocols).
One specifies an additional ensuing address parameter
(to be interpreted as a network mask).
The implicit network mask generated in the AF_INET case
can be overridden by making sure this option follows the destination parameter.
.Pp
For
.Dv AF_INET6 ,
the
.Fl prefixlen
qualifier
is available instead of the
.Fl mask
qualifier because non-continuous masks are not allowed in IPv6.
For example,
.Fl prefixlen Li 32
specifies that a network mask of
.Li ffff:ffff:0000:0000:0000:0000:0000:0000
will be used.
The default prefixlen is 64.
However, it is assumed to be 0 if
.Cm default
is specified for
.Ar destination .
Note that the qualifier works only for
.Dv AF_INET6
address family.
.Pp
Routes have associated flags which influence operation of the protocols
when sending to destinations matched by the routes.
These flags may be set (or sometimes cleared)
by indicating the following corresponding modifiers:
.Bd -literal
-xresolve  RTF_XRESOLVE   - emit mesg on use (for external lookup)
-iface    ~RTF_GATEWAY    - destination is directly reachable
-static    RTF_STATIC     - manually added route
-nostatic ~RTF_STATIC     - pretend route added by kernel or daemon
-reject    RTF_REJECT     - emit an ICMP unreachable when matched
-blackhole RTF_BLACKHOLE  - silently discard pkts (during updates)
-proto1    RTF_PROTO1     - set protocol specific routing flag #1
-proto2    RTF_PROTO2     - set protocol specific routing flag #2
.Ed
.Pp
The optional modifiers
.Fl rtt ,
.Fl rttvar ,
.Fl sendpipe ,
.Fl recvpipe ,
.Fl mtu ,
.Fl hopcount ,
.Fl expire ,
and
.Fl ssthresh
provide initial values to quantities maintained in the routing entry
by transport level protocols, such as TCP or TP4.
These may be individually locked by preceding each such modifier to
be locked by
the
.Fl lock
meta-modifier, or one can
specify that all ensuing metrics may be locked by the
.Fl lockrest
meta-modifier.
.Pp
Note that
.Fl expire
accepts expiration time of the route as the number of seconds since the
Epoch
.Pq see Xr time 3 .
When the first character of the number is
.Dq +
or
.Dq - ,
it is interpreted as a value relative to the current time.
.Pp
The optional modifier
.Fl fib Ar number
specifies that the command will be applied to a non-default FIB.
The
.Ar number
must be smaller than the
.Va net.fibs
.Xr sysctl 8
MIB.
When this modifier is not specified,
or a negative number is specified,
the default FIB shown in the
.Va net.my_fibnum
.Xr sysctl 8
MIB will be used.
.Pp
The
.Ar number
allows multiple FIBs by a comma-separeted list and/or range
specification.
The
.Qq Fl fib Li 2,4,6
means the FIB number 2, 4, and 6.
The
.Qq Fl fib Li 1,3-5,6
means the 1, 3, 4, 5, and 6.
.Pp
In a
.Cm change
or
.Cm add
command where the destination and gateway are not sufficient to specify
the route (as in the ISO case where several interfaces may have the
same address), the
.Fl ifp
or
.Fl ifa
modifiers may be used to determine the interface or interface address.
.Pp
All symbolic names specified for a
.Ar destination
or
.Ar gateway
are looked up first as a host name using
.Xr gethostbyname 3 .
If this lookup fails,
.Xr getnetbyname 3
is then used to interpret the name as that of a network.
.Pp
The
.Nm
utility uses a routing socket and the new message types
.Dv RTM_ADD , RTM_DELETE , RTM_GET ,
and
.Dv RTM_CHANGE .
As such, only the super-user may modify
the routing tables.
.Pp
.Fx provides support for scalable multipath routing.
It is activated by default, but can be turned off by setting the
.Va net.route.multipath
.Xr sysctl 8
MIB to 0.
.Pp
There are multiple route lookup algorithms available.
They can be configured by setting
.Va net.route.algo.inet.algo
for IPv4 and
.Va net.route.algo.inet6.algo
for IPv6
.Xr sysctl 8
MIBs.
.Pp
A list of available algorithms can be obtained by accessing the
following
.Xr sysctl 8
MIBs
.Va net.route.algo.inet.algo_list
for IPv4 and
.Va net.route.algo.inet6.algo_list
for IPv6.
.Pp
The following algorithms are available:
.Bl -tag -width radix_lockless
.It radix
Base system radix backend.
.It bsearch
Lockless binary search in a special IP array, tailored for a small FIB
with <16 routes.
This algorithm is only available for IPv4.
.It radix_lockless
Lockless immutable radix, re-created on every rtable change,
tailored for a small FIB with <1000 routes.
.It dpdk_lpm
DPDK DIR24-8-based lookups, lockless datastructure, optimized
for large FIBs.
DIR24-8 relies on a large flat lookup table (64 MB with IPv4) which is
directly indexed by the more significant portion of the lookup key.
In order to use the dpdk_lpm algorithm one or both of the
following kernel modules must be loaded via
.Xr loader.conf 5 :
.Bl -tag -width dpdk_lpm6.ko -compact
.It dpdk_lpm4.ko
DPDK implementation for IPv4.
.It dpdk_lpm6.ko
DPDK implementation for IPv6.
.El
.It dxr
IPv4 only, lockless, compressed lookup structure
(below 2.5 Bytes per IPv4 prefix for large BGP FIBs)
which easily fits into modern CPU cache hierarchies,
lookup throughput scales linearly with CPU cores.
Loadable as a kernel module at runtime or via
.Xr loader.conf 5 :
.Bl -tag -width fib_dxr.ko -compact
.It fib_dxr.ko
.El
.El
.Pp
The algorithms are selected automatically based on the size of the routing
table of the system.
They can be changed, but not every algorithm performs best for every
FIB size.
.Sh EXIT STATUS
.Ex -std
.Sh EXAMPLES
Add a default route to the network routing table.
This will send all packets for destinations not available in the routing table
to the default gateway at 192.168.1.1:
.Pp
.Dl route add -net 0.0.0.0/0 192.168.1.1
.Pp
A shorter version of adding a default route can also be written as:
.Pp
.Dl route add default 192.168.1.1
.Pp
Add a static route to the 172.16.10.0/24 network via the 172.16.1.1 gateway:
.Pp
.Dl route add -net 172.16.10.0/24 172.16.1.1
.Pp
Change the gateway of an already established static route in the routing table:
.Pp
.Dl route change -net 172.16.10.0/24 172.16.1.2
.Pp
Display the route for a destination network:
.Pp
.Dl route show 172.16.10.0
.Pp
Delete a static route from the routing table:
.Pp
.Dl route delete -net 172.16.10.0/24 172.16.1.2
.Pp
Remove all routes from the routing table:
.Pp
.Dl route flush
.Pp
The routing table can be listed with
.Xr netstat 1 .
.Sh DIAGNOSTICS
.Bl -diag
.It "add [host \&| network ] %s: gateway %s flags %x"
The specified route is being added to the tables.
The
values printed are from the routing table entry supplied
in the
.Xr ioctl 2
call.
If the gateway address used was not the primary address of the gateway
(the first one returned by
.Xr gethostbyname 3 ) ,
the gateway address is printed numerically as well as symbolically.
.It "delete [ host \&| network ] %s: gateway %s flags %x"
As above, but when deleting an entry.
.It "%s %s done"
When the
.Cm flush
command is specified, each routing table entry deleted
is indicated with a message of this form.
.It "Network is unreachable"
An attempt to add a route failed because the gateway listed was not
on a directly-connected network.
The next-hop gateway must be given.
.It "not in table"
A delete operation was attempted for an entry which
was not present in the tables.
.It "routing table overflow"
An add operation was attempted, but the system was
low on resources and was unable to allocate memory
to create the new entry.
.It "gateway uses the same route"
A
.Cm change
operation resulted in a route whose gateway uses the
same route as the one being changed.
The next-hop gateway should be reachable through a different route.
.El
.Sh SEE ALSO
.Xr netstat 1 ,
.Xr netintro 4 ,
.Xr route 4 ,
.Xr loader.conf 5 ,
.Xr arp 8 ,
.Xr routed 8
.Sh HISTORY
The
.Nm
utility appeared in
.Bx 4.2 .
.Sh BUGS
The first paragraph may have slightly exaggerated
.Xr routed 8 Ns 's
abilities.
.Pp
Currently, routes with the
.Dv RTF_BLACKHOLE
flag set need to have the gateway set to an instance of the
.Xr lo 4
driver, using the
.Fl iface
option, for the flag to have any effect; unless IP fast forwarding
is enabled, in which case the meaning of the flag will always
be honored.