A major overhaul of the CARP implementation. The ip_carp.c was started

from scratch, copying needed functionality from the old implemenation
on demand, with a thorough review of all code. The main change is that
interface layer has been removed from the CARP. Now redundant addresses
are configured exactly on the interfaces, they run on.

The CARP configuration itself is, as before, configured and read via
SIOCSVH/SIOCGVH ioctls. A new prefix created with SIOCAIFADDR or
SIOCAIFADDR_IN6 may now be configured to a particular virtual host id,
which makes the prefix redundant.

ifconfig(8) semantics has been changed too: now one doesn't need
to clone carpXX interface, he/she should directly configure a vhid
on a Ethernet interface.

To supply vhid data from the kernel to an application the getifaddrs(8)
function had been changed to pass ifam_data with each address. [1]

The new implementation definitely closes all PRs related to carp(4)
being an interface, and may close several others. It also allows
to run a single redundant IP per interface.

Big thanks to Bjoern Zeeb for his help with inet6 part of patch, for
idea on using ifam_data and for several rounds of reviewing!

PR:		kern/117000, kern/126945, kern/126714, kern/120130, kern/117448
Reviewed by:	bz
Submitted by:	bz [1]

1 files changed, 115 insertions, 131 deletions

diff --git a/share/man/man4/carp.4 b/share/man/man4/carp.4
index 4064c6c..7c214ff 100644
--- a/share/man/man4/carp.4
+++ b/share/man/man4/carp.4
@@ -1,6 +1,7 @@
 .\"	$OpenBSD: carp.4,v 1.16 2004/12/07 23:41:35 jmc Exp $
 .\"
 .\" Copyright (c) 2003, Ryan McBride.  All rights reserved.
+.\" Copyright (c) 2011, Gleb Smirnoff <glebius@FreeBSD.org>
 .\"
 .\" Redistribution and use in source and binary forms, with or without
 .\" modification, are permitted provided that the following conditions
@@ -25,7 +26,7 @@
 .\"
 .\" $FreeBSD$
 .\"
-.Dd August 15, 2011
+.Dd December 16, 2011
 .Dt CARP 4
 .Os
 .Sh NAME
@@ -34,33 +35,17 @@
 .Sh SYNOPSIS
 .Cd "device carp"
 .Sh DESCRIPTION
-The
-.Nm
-interface is a pseudo-device that implements and controls the
-CARP protocol.
-CARP allows multiple hosts on the same local network to share a set of IP addresses.
+The CARP allows multiple hosts on the same local network to share a set of
+IPv4 and/or IPv6 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
-.Nm ifconfig Li carp Ns Ar N Cm create
-command or by configuring
-it via
-.Va cloned_interfaces
-in the
-.Pa /etc/rc.conf
-file.
+addresses are always available.
 .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:
+the administrator needs to configure at minimum a common virtual host ID
+(vhid) and attach at least one IP address to this vhid on each machine which
+is to take part in the virtual group.
+Additional parameters can also be set on a per-vhid basis:
 .Cm advbase
 and
 .Cm advskew ,
@@ -93,9 +78,20 @@ or through the
 .Dv SIOCSVH
 .Xr ioctl 2 .
 .Pp
+CARP virtual hosts can be configured on multicast capable interfaces: Ethernet,
+layer 2 VLAN, FDDI and Token Ring.
+An arbitrary number of virtual host IDs can be configured on an interface.
+An arbitrary number of IPv4 or IPv6 addresses can be attached to a particular
+vhid.
+It is important that all hosts participating in a vhid have the same list
+of prefixes configured on the vhid, since all prefixes are included in the
+cryptographic checksum supplied in each advertisement.
+Multiple vhids running on one interface participate in master/backup
+elections independently.
+.Pp
 Additionally, there are a number of global parameters which can be set using
 .Xr sysctl 8 :
-.Bl -tag -width ".Va net.inet.carp.arpbalance"
+.Bl -tag -width ".Va net.inet.carp.preempt"
 .It Va net.inet.carp.allow
 Accept incoming
 .Nm
@@ -125,9 +121,6 @@ Values above 1 enable logging of bad
 .Nm
 packets.
 Default value is 1.
-.It Va net.inet.carp.arpbalance
-Balance local traffic using ARP (see below).
-Disabled by default.
 .It Va net.inet.carp.suppress_preempt
 A read only value showing the status of preemption suppression.
 Preemption can be suppressed if link on an interface is down
@@ -138,36 +131,36 @@ Value of 0 means that preemption is not suppressed, since no
 problems are detected.
 Every problem increments suppression counter.
 .El
-.Sh ARP level load balancing
-The
-.Nm
-has limited abilities for load balancing the incoming connections
-between hosts in Ethernet network.
-For load balancing operation, one needs several CARP interfaces that
-are configured to the same IP address, but to a different VHIDs.
-Once an ARP request is received, the CARP protocol will use a hashing
-function against the source IP address in the ARP request to determine
-which VHID should this request belong to.
-If the corresponding CARP interface is in master state, the ARP request
-will be replied, otherwise it will be ignored.
-See the
-.Sx EXAMPLES
-section for a practical example of load balancing.
-.Pp
-The ARP load balancing has some limitations.
-First, 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.
-Second, ARP load balancing can lead to asymmetric routing
-of incoming and outgoing traffic, and thus combining it with
-.Xr pfsync 4
-is dangerous, because this creates a race condition between
-balanced routers and a host they are serving.
-Imagine an incoming packet creating state on the first router, being
-forwarded to its destination, and destination replying faster
-than the state information is packed and synced with the second router.
-If the reply would be load balanced to second router, it will be
-dropped due to no state.
+.\".Sh ARP level load balancing
+.\"The
+.\".Nm
+.\"has limited abilities for load balancing the incoming connections
+.\"between hosts in Ethernet network.
+.\"For load balancing operation, one needs several CARP interfaces that
+.\"are configured to the same IP address, but to a different vhids.
+.\"Once an ARP request is received, the CARP protocol will use a hashing
+.\"function against the source IP address in the ARP request to determine
+.\"which vhid should this request belong to.
+.\"If the corresponding CARP interface is in master state, the ARP request
+.\"will be replied, otherwise it will be ignored.
+.\"See the
+.\".Sx EXAMPLES
+.\"section for a practical example of load balancing.
+.\".Pp
+.\"The ARP load balancing has some limitations.
+.\"First, 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.
+.\"Second, ARP load balancing can lead to asymmetric routing
+.\"of incoming and outgoing traffic, and thus combining it with
+.\".Xr pfsync 4
+.\"is dangerous, because this creates a race condition between
+.\"balanced routers and a host they are serving.
+.\"Imagine an incoming packet creating state on the first router, being
+.\"forwarded to its destination, and destination replying faster
+.\"than the state information is packed and synced with the second router.
+.\"If the reply would be load balanced to second router, it will be
+.\"dropped due to no state.
 .Sh STATE CHANGE NOTIFICATIONS
 Sometimes it is useful to get notified about
 .Nm
@@ -175,13 +168,10 @@ status change events.
 This can be accomplished by using
 .Xr devd 8
 hooks.
-Master/slave events are signalled as
-.Nm
-interface
-.Dv LINK_UP
-or
-.Dv LINK_DOWN
-event.
+Master/slave events are signalled under system
+.Dv CARP .
+Subsystem specifies vhid and name of interface, where event occured.
+Type of the message displays new state of vhid.
 Please see
 .Xr devd.conf 5
 and 
@@ -197,23 +187,19 @@ 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.
+Assume that host A is the preferred master and we are running the
+192.168.1.0/24 prefix on em0 and 192.168.2.0/24 on em1.
 This is the setup for host A:
 .Bd -literal -offset indent
-ifconfig carp0 create
-ifconfig carp0 vhid 1 pass mekmitasdigoat 192.168.1.1/24
-ifconfig carp1 create
-ifconfig carp1 vhid 2 pass mekmitasdigoat 192.168.2.1/24
+ifconfig em0 vhid 1 pass mekmitasdigoat 192.168.1.1/24
+ifconfig em1 vhid 2 pass mekmitasdigoat 192.168.2.1/24
 .Ed
 .Pp
 The setup for host B is identical, but it has a higher
 .Cm advskew :
 .Bd -literal -offset indent
-ifconfig carp0 create
-ifconfig carp0 vhid 1 advskew 100 pass mekmitasdigoat 192.168.1.1/24
-ifconfig carp1 create
-ifconfig carp1 vhid 2 advskew 100 pass mekmitasdigoat 192.168.2.1/24
+ifconfig em0 vhid 1 advskew 100 pass mekmitasdigoat 192.168.1.1/24
+ifconfig em1 vhid 2 advskew 100 pass mekmitasdigoat 192.168.2.1/24
 .Ed
 .Pp
 Because of the preempt option, when one of the physical interfaces of
@@ -224,67 +210,60 @@ is adjusted to 240 on all its
 interfaces.
 This will cause host B to preempt on both interfaces instead of
 just the failed one.
-.Pp
-In order to set up an ARP balanced virtual host, it is necessary to configure
-one virtual host for each physical host which would respond to ARP requests
-and thus handle the traffic.
-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.
-.Bd -literal -offset indent
-ifconfig carp0 create
-ifconfig carp0 vhid 1 pass mekmitasdigoat 192.168.1.10/24
-ifconfig carp1 create
-ifconfig carp1 vhid 2 advskew 100 pass mekmitasdigoat 192.168.1.10/24
-.Ed
-.Pp
-The configuration for host B is identical, except the
-.Cm advskew
-is on virtual host 1 rather than virtual host 2.
-.Bd -literal -offset indent
-ifconfig carp0 create
-ifconfig carp0 vhid 1 advskew 100 pass mekmitasdigoat 192.168.1.10/24
-ifconfig carp1 create
-ifconfig carp1 vhid 2 pass mekmitasdigoat 192.168.1.10/24
-.Ed
-.Pp
-Finally, the ARP balancing feature must be enabled on both hosts:
-.Pp
-.Dl sysctl net.inet.carp.arpbalance=1
-.Pp
-When the hosts receive an ARP request for 192.168.1.10, the source IP address
-of the request is used to compute which virtual host should answer the request.
-The host which is master of the selected virtual host will reply to the
-request, the other(s) will ignore it.
-.Pp
-This way, locally connected systems will receive different ARP replies and
-subsequent IP traffic will be balanced among the hosts.
-If one of the hosts fails, the other will take over the virtual MAC address,
-and begin answering ARP requests on its behalf.
+.\".Pp
+.\"In order to set up an ARP balanced virtual host, it is necessary to configure
+.\"one virtual host for each physical host which would respond to ARP requests
+.\"and thus handle the traffic.
+.\"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.
+.\".Bd -literal -offset indent
+.\"ifconfig carp0 create
+.\"ifconfig carp0 vhid 1 pass mekmitasdigoat 192.168.1.10/24
+.\"ifconfig carp1 create
+.\"ifconfig carp1 vhid 2 advskew 100 pass mekmitasdigoat 192.168.1.10/24
+.\".Ed
+.\".Pp
+.\"The configuration for host B is identical, except the
+.\".Cm advskew
+.\"is on virtual host 1 rather than virtual host 2.
+.\".Bd -literal -offset indent
+.\"ifconfig carp0 create
+.\"ifconfig carp0 vhid 1 advskew 100 pass mekmitasdigoat 192.168.1.10/24
+.\"ifconfig carp1 create
+.\"ifconfig carp1 vhid 2 pass mekmitasdigoat 192.168.1.10/24
+.\".Ed
+.\".Pp
+.\"Finally, the ARP balancing feature must be enabled on both hosts:
+.\".Pp
+.\".Dl sysctl net.inet.carp.arpbalance=1
+.\".Pp
+.\"When the hosts receive an ARP request for 192.168.1.10, the source IP address
+.\"of the request is used to compute which virtual host should answer the request.
+.\"The host which is master of the selected virtual host will reply to the
+.\"request, the other(s) will ignore it.
+.\".Pp
+.\"This way, locally connected systems will receive different ARP replies and
+.\"subsequent IP traffic will be balanced among the hosts.
+.\"If one of the hosts fails, the other will take over the virtual MAC address,
+.\"and begin answering ARP requests on its behalf.
 .Pp
 Processing of
 .Nm
-status change events can be set up by using the following devd.conf rules:
+status change events can be set up by using the following devd.conf rule:
 .Bd -literal -offset indent
 notify 0 {
-	match "system"          "IFNET";
-	match "type"            "LINK_UP";
-	match "subsystem"       "carp*";
-	action "/root/carpcontrol.sh $type $subsystem";
-};
-
-notify 0 {
-	match "system"          "IFNET";
-	match "type"            "LINK_DOWN";
-	match "subsystem"       "carp*";
-	action "/root/carpcontrol.sh $type $subsystem";
+	match "system"          "CARP";
+	match "subsystem"       "[0-9]+@";
+	match "type"            "(MASTER|BACKUP)";
+	action "/root/carpcontrol.sh $subsystem $type";
 };
 .Ed
 .Sh SEE ALSO
@@ -303,3 +282,8 @@ The
 .Nm
 device was imported into
 .Fx 5.4 .
+In
+.Fx 10
+the
+.Nm
+was significantly rewritten, and is no longer a pseudo-interface.