.\"	$OpenBSD: carp.4,v 1.28 2007/09/18 09:18:04 mpf Exp $
.\"
.\" Copyright (c) 2003, Ryan McBride.  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.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE PROJECT 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 PROJECT 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.
.\"
.Dd $Mdocdate: September 18 2007 $
.Dt CARP 4
.Os
.Sh NAME
.Nm carp
.Nd Common Address Redundancy Protocol
.Sh SYNOPSIS
.Cd "pseudo-device carp"
.Sh DESCRIPTION
The
.Nm
interface is a pseudo-device which implements and controls the
CARP protocol.
.Nm
allows multiple hosts on the same local network to share a set of IP addresses.
Its primary purpose is to ensure that these
addresses are always available, but in some configurations
.Nm
can also provide load balancing functionality.
.Pp
A
.Nm
interface can be created at runtime using the
.Ic ifconfig carp Ns Ar N Ic create
command or by setting up a
.Xr hostname.if 5
configuration file for
.Xr netstart 8 .
.Pp
To use
.Nm ,
the administrator needs to configure at minimum
a common virtual host ID (VHID) and
virtual host IP address on each machine which is to take part in the virtual
group.
Additional parameters can also be set on a per-interface basis:
.Cm advbase
and
.Cm advskew ,
which are used to control how frequently the host sends advertisements when it
is the master for a virtual host, and
.Cm pass
which is used to authenticate carp advertisements.
Finally
.Cm carpdev
is used to specify which interface the
.Nm
device attaches to.
If unspecified, the kernel attempts to set it by looking for
another interface with the same subnet.
These configurations can be done using
.Xr ifconfig 8 ,
or through the
.Dv SIOCSVH
ioctl.
.Pp
.Nm
can also be used in conjunction with
.Xr ifstated 8
to respond to changes in CARP state;
however, for most uses this will not be necessary.
See the manual page for
.Xr ifstated 8
for more information.
.Pp
Additionally, there are a number of global parameters which can be set using
.Xr sysctl 8 :
.Bl -tag -width xxxxxxxxxxxxxxxxxxxxxxxxxx
.It net.inet.carp.allow
Accept incoming
.Nm
packets.
Enabled by default.
.It net.inet.carp.preempt
Allow virtual hosts to preempt each other.
It is also used to failover
.Nm
interfaces as a group.
When the option is enabled and one of the
.Nm
enabled physical interfaces
goes down,
.Cm advskew
is changed to 240 on all
.Nm
interfaces.
See also the first example.
Disabled by default.
.It net.inet.carp.log
Log bad
.Nm
packets.
Disabled by default.
.It net.inet.carp.arpbalance
Balance local traffic using ARP.
Disabled by default.
.El
.Sh LOAD BALANCING
.Nm
provides two mechanisms to load balance incoming traffic
over a group of
.Nm
hosts:
ARP balancing and IP balancing.
.Pp
Which one to use mainly depends on the network environment
.Nm
is being used in.
ARP balancing has limited abilities for load balancing the
incoming connections between hosts in an Ethernet network.
It only works for clients in the local network, because
ARP balancing spreads the load by varying ARP replies
based on the source MAC address of the host sending the query.
Therefore it cannot balance traffic that crosses a router, because the
router itself will always be balanced to the same virtual host.
.Pp
IP balancing is not dependent on ARP and therefore also works
for traffic that comes over a router.
This method should work in all environments and can
also provide more fine grained load balancing than ARP balancing.
The downside of IP balancing is that it requires the traffic
that is destined towards the load balanced IP addresses
to be received by all
.Nm
hosts.
While this is always the case when connected to a hub,
it has to play some tricks in switched networks, which
will result in a higher network load.
.Pp
A rule of thumb might be to use ARP balancing if there
are many hosts on the same network segment and
to use IP balancing for all other cases.
.Pp
The configuration of ARP and IP load balancing is quite similar:
a load balancing group is created out of multiple
.Nm
interfaces by configuring them with the same IP addresses,
but to different VHIDs.
All
.Nm
nodes in the cluster are configured identically, except
for a different
.Cm advskew
to control which interfaces on a host will be the designated master.
See the
.Sx EXAMPLES
section for a practical example of load balancing.
.Ss ARP BALANCING
For load balancing, several
.Nm
interfaces are configured to the same IP address, but to different VHIDs.
Once an ARP request is received, the CARP protocol will use a hashing
function against the source MAC address in the ARP request to determine
which VHID the request belongs to.
If the corresponding
.Nm
interface is in master state, the ARP request will be answered, otherwise
it will be ignored.
.Pp
The ARP load balancing has some limitations.
Firstly, ARP balancing only works on the local network segment.
It cannot balance traffic that crosses a router, because the
router itself will always be balanced to the same virtual host.
Secondly, ARP load balancing can lead to asymmetric routing
of incoming and outgoing traffic, thus combining it with
.Xr pfsync 4
requires special care, because this can create a race condition between
balanced routers and the host they are serving.
ARP balancing can be safely used with pfsync if the
.Xr pf 4
ruleset translates the source address to an unshared address on the
outgoing interface using a NAT rule.
This requires multiple CARP groups with
.Em different
IP addresses on the outgoing interface, configured so that each host is the
master of one group.
.Ss IP BALANCING
IP load balancing works by utilizing the network itself to distribute
incoming traffic to all
.Nm
nodes in the cluster.
Each packet is filtered on the incoming
.Nm
interface so that only one node in the cluster accepts the
packet.
All the other nodes will just silently drop it.
The filtering function uses a hash over the source and destination
address of the IPv4 or IPv6 packet and compares the result against the
state of the
.Nm
load balancing group.
.Pp
A load balancing group consists of two or more
.Nm
interfaces per host which are configured with common IP addresses
but different VHIDs.
IP balancing is activated by setting the
.Cm link0
flag on the first interface of the group.
In most cases it is recommended to also enable the
.Cm link1
flag.
This flag enables the stealth mode on the interface.
In this mode
.Nm
never sends packets with its virtual MAC address as source.
This is necessary to receive incoming traffic on all hosts in switched networks.
Stealth mode prevents a switch from learning the virtual MAC
address, so that it has to flood the traffic to all its ports.
The
.Cm link1
flag can be avoided
only if using a hub or if the switch ports that are connected
to the cluster nodes can be configured into some sort of monitoring mode.
Please note that activating stealth mode on a
.Nm
interface that has already been running might not work instantly.
As a workaround the VHID can be changed to a previously unused
one, or just wait until the MAC table entry in the switch times out.
.Pp
Some Layer-3 switches do port learning based on ARP packets.
Therefore the stealth mode cannot hide the virtual MAC address
from these kind of devices.
In such cases,
.Nm
can be told to use a multicast MAC address by additionally enabling the
.Cm link2
flag.
.Pp
If IP balancing is being used on a firewall, it is recommended to
configure the load balancing group in a symmetrical manner.
This is achieved by prioritizing the interfaces in the same order
(ascending by VHID) on both sides of the firewall.
This ensures that packets of one connection will pass in and out
on the same host and are not routed asymmetrically.
.Sh EXAMPLES
For firewalls and routers with multiple interfaces, it is desirable to
failover all of the
.Nm
interfaces together, when one of the physical interfaces goes down.
This is achieved by the preempt option.
Enable it on both host A and B:
.Pp
.Dl # sysctl net.inet.carp.preempt=1
.Pp
Assume that host A is the preferred master and 192.168.1.x/24 is
configured on one physical interface and 192.168.2.y/24 on another.
This is the setup for host A:
.Bd -literal -offset indent
# ifconfig carp0 192.168.1.1 vhid 1
# ifconfig carp1 192.168.2.1 vhid 2
.Ed
.Pp
The setup for host B is identical, but it has a higher
.Cm advskew :
.Bd -literal -offset indent
# ifconfig carp0 192.168.1.1 vhid 1 advskew 100
# ifconfig carp1 192.168.2.1 vhid 2 advskew 100
.Ed
.Pp
Because of the preempt option, when one of the physical interfaces of
host A fails,
.Cm advskew
is adjusted to 240 on all its
.Nm
interfaces.
This will cause host B to preempt on both interfaces instead of
just the failed one.
.Ss LOAD BALANCING
In order to set up an load balanced virtual host, it is necessary to configure
one virtual host for each physical host.
In the following example, two virtual hosts are configured on two hosts to
provide balancing and failover for the IP address 192.168.1.10.
.Pp
First the
.Nm
interfaces on Host A are configured.
The
.Cm advskew
of 100 on the second virtual host means that its advertisements will be sent
out slightly less frequently and will therefore become the designated backup.
.Bd -literal -offset indent
# ifconfig carp0 192.168.1.10 vhid 1
# ifconfig carp1 192.168.1.10 vhid 2 advskew 100
.Ed
.Pp
The configuration for host B is identical, except the skew is on
virtual host 1 rather than virtual host 2.
.Bd -literal -offset indent
# ifconfig carp0 192.168.1.10 vhid 1 advskew 100
# ifconfig carp1 192.168.1.10 vhid 2
.Ed
.Pp
If ARP balancing is being used, it must be enabled on both hosts:
.Pp
.Dl # sysctl net.inet.carp.arpbalance=1
.Pp
If IP balancing is being used, instead enable the
.Cm link0
and
.Cm link1
flags on the first interface of the load balancing group on both hosts:
.Bd -literal -offset indent
A# ifconfig carp0 192.168.1.10 vhid 1 link0 link1
A# ifconfig carp1 192.168.1.10 vhid 2 advskew 100
.Pp
B# ifconfig carp0 192.168.1.10 vhid 1 advskew 100 link0 link1
B# ifconfig carp1 192.168.1.10 vhid 2
.Ed
.Sh SEE ALSO
.Xr sysctl 3 ,
.Xr inet 4 ,
.Xr pfsync 4 ,
.Xr hostname.if 5 ,
.Xr ifconfig 8 ,
.Xr ifstated 8 ,
.Xr netstart 8 ,
.Xr sysctl 8
.Sh HISTORY
The
.Nm
device first appeared in
.Ox 3.5 .