diff options
Diffstat (limited to 'share/man/man4/carp.4')
| -rw-r--r-- | share/man/man4/carp.4 | 283 |
1 files changed, 145 insertions, 138 deletions
diff --git a/share/man/man4/carp.4 b/share/man/man4/carp.4 index 4064c6c..1b59e72 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 20, 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 @@ -125,49 +121,69 @@ 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 -or when +.It Va net.inet.carp.demotion +This value shows current level of CARP demotion. +The value is added to the actual advskew sent in announcements for +all vhids. +At normal system operation the demotion factor is zero. +However, problematic conditions raise its level: when +.Nm +experiences problem with sending announcements, when an interface +running a vhid goes down, or while the .Xr pfsync 4 interface is not synchronized. -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 +The demotion value is writable, so that user may alter it +depending on some external conditions, for example on status of some +daemon utility. +However, altering the value should be performed with care, do +not conflict with subsystems that adjust demotion factor +automatically: .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. +and +.Xr pfsync 4 . +.It Va net.inet.carp.ifdown_demotion_factor +Value added to +.Va net.inet.carp.demotion +when interface running a vhid goes down. +Default value is 240 (maximum advskew value). +.It Va net.inet.carp.senderr_demotion_factor +Value added to +.Va net.inet.carp.demotion +when +.Nm +experiences errors sending its announcements. +Default value is 240 (maximum advskew value). +.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 STATE CHANGE NOTIFICATIONS Sometimes it is useful to get notified about .Nm @@ -175,13 +191,10 @@ 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 +210,19 @@ 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 +233,60 @@ 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 +305,8 @@ The .Nm device was imported into .Fx 5.4 . +In +.Fx 10 +the +.Nm +was significantly rewritten, and is no longer a pseudo-interface. |
