keepalived.conf man page

keepalived.conf — configuration file for Keepalived

Description

keepalived.conf is the configuration file which describes all the Keepalived keywords. Keywords are placed in hierarchies of blocks and subblocks, each layer being delimited by '{' and '}' pairs.

Comments start with '#' or '!' to the end of the line and can start anywhere in a line.

The keyword 'include' allows inclusion of other configuration files from within the main configuration file, or from subsequently included files.

The format of the include directive is:

include FILENAME

FILENAME can be a fully qualified or relative pathname, and can include wildcards, including csh style brace expressions such as "{foo/{,cat,dog},bar}" if glob() supports them.

After opening an included file, the current directory is set to the directory of the file itself, so any relative paths included from a file are relative to the directory of the including file itself.

Parameter Syntax

<BOOL> is one of on|off|true|false|yes|no

Conditional configuration and configuration id

The config-id defaults to the first part of the node name as returned by uname, and can be overridden with the -i or --config-id command line option.

Any configuration line starting with '@' is a conditional configuration line. The word immediately following (i.e. without any space) the '@' character is compared against the config-id, and if they don't match, the configuration line is ignored.

Alternatively, '@^' is a negative comparison, so if the word immediately following does NOT match the config-id, the configuration line IS included.

The purpose of this is to allow a single configuration file to be used for multiple systems, where the only differences are likely to be the router_id, vrrp instance priorities, and possibly interface names and unicast addresses.

For example:

global_defs
{
@main   router_id main_router
@backup router_id backup_router
}
...
vrrp_instance VRRP1 {
...
@main    unicast_src_ip 1.2.3.4
@backup  unicast_src_ip 1.2.3.5
@backup2 unicast_src_ip 1.2.3.6
         unicast_peer {
@^main         1.2.3.4
@^backup       1.2.3.5
@^backup2      1.2.3.6
         }
}

If keepalived is invoked with -i main, then the router_id will be set to main_router, if invoked with -i backup, then backup_router, if not invoked with -i, or with -i anything else, then the router_id will not be set. The unicast peers for main will be 1.2.3.5 and 1.2.3.6.

Top Hierachy

Global Configuration

VRRPD Configuration

LVS Configuration

Global Configuration

contains subblocks of Global definitions, Static routes, and Static rules

Global definitions

global_defs           # Block id
{
notification_email    # To:
       {
       admin@example1.com
       ...
       }
# From: from address that will be in the header (default keepalived@<local host name>)
notification_email_from admin@example.com
smtp_server 127.0.0.1 [<PORT>]
                             # IP address or domain name
                             #  with optional port number (default 25)
smtp_helo_name <HOST_NAME>   # name to use in HELO messages
                             #  defaults to local host name
smtp_connect_timeout 30      # integer, seconds
router_id my_hostname        # string identifying the machine,
                             # (doesn't have to be hostname).
                             # default: local host name
vrrp_mcast_group4 224.0.0.18 # optional, default 224.0.0.18
vrrp_mcast_group6 ff02::12   # optional, default ff02::12
default_interface p33p1.3    # sets the default interface for static addresses, default eth0

lvs_sync_daemon <INTERFACE> <VRRP_INSTANCE> [id <SYNC_ID>] [maxlen <LEN>] [port <PORT>] [ttl <TTL>] [group <IP ADDR>]
                             # Binding interface, vrrp instance and optional
                             #  syncid for lvs syncd
                             #  syncid (0 to 255) for lvs syncd
                             #  maxlen (1..65507) maximum packet length
                             #  port (1..65535) UDP port number to use
                             #  ttl (1..255)
                             #  group - multicast group address (IPv4 or IPv6)
                             # NOTE: maxlen, port, ttl and group are only available on Linux 4.3 or later.
lvs_flush                    # flush any existing LVS configuration at startup

# delay for second set of gratuitous ARPs after transition to MASTER
vrrp_garp_master_delay 10    # seconds, default 5, 0 for no second set

# number of gratuitous ARP messages to send at a time after transition to MASTER
vrrp_garp_master_repeat 1    # default 5

# delay for second set of gratuitous ARPs after lower priority advert received when MASTER
vrrp_garp_lower_prio_delay 10

# number of gratuitous ARP messages to send at a time after lower priority advert received when MASTER
vrrp_garp_lower_prio_repeat 1

# minimum time interval for refreshing gratuitous ARPs while MASTER
vrrp_garp_master_refresh 60  # secs, default 0 (no refreshing)

# number of gratuitous ARP messages to send at a time while MASTER
vrrp_garp_master_refresh_repeat 2 # default 1

# Delay in ms between gratuitous ARP messages sent on an interface
vrrp_garp_interval 0.001          # decimal, seconds (resolution usecs). Default 0.

# Delay in ms between unsolicited NA messages sent on an interface
vrrp_gna_interval 0.000001        # decimal, seconds (resolution usecs). Default 0.

# If a lower priority advert is received, don't send another advert. This causes
# adherence to the RFCs. Defaults to false, unless strict_mode is set.
vrrp_lower_prio_no_advert [<BOOL>]

# If we are master and receive a higher priority advert, send an advert (which will be
# lower priority than the other master), before we transition to backup. This means
# that if the other master has garp_lower_priority_repeat set, it will resend garp messages.
# This is to get around the problem of their having been two simultaneous masters, and the
# last GARP messages seen were from us.
vrrp_higher_prio_send_advert [<BOOL>]

# Set the default VRRP version to use
vrrp_version <2 or 3>        # default version 2

# Specify the iptables chain for ensuring a version 3 instance
# doesn't respond on addresses that it doesn't own.
# Note: it is necessary for the specified chain to exist in
# the iptables and/or ip6tables configuration, and for the chain
# to be called from an appropriate point in the iptables configuration.
# It will probably be necessary to have this filtering after accepting
# any ESTABLISHED,RELATED packets, because IPv4 might select the VIP as
# the source address for outgoing connections.
vrrp_iptables keepalived     # default INPUT

# or for outbound filtering as well
# Note, outbound filtering won't work with IPv4, since the VIP can be selected as the source address
# for an outgoing connection. With IPv6 this is unlikely since the addresses are deprecated.
vrrp_iptables keepalived_in keepalived_out

# or to not add any iptables rules:
vrrp_iptables

# Keepalived may have the option to use ipsets in conjunction with iptables.
# If so, then the ipset names can be specified, defaults as below.
# If no names are specified, ipsets will not be used, otherwise any omitted
# names will be constructed by adding "_if" and/or "6" to previously specified
# names.
vrrp_ipsets [keepalived [keepalived6 [keepalived_if6]]]

# The following enables checking that when in unicast mode, the source
# address of a VRRP packet is one of our unicast peers.
vrrp_check_unicast_src

# Checking all the addresses in a received VRRP advert can be time consuming.
# Setting this flag means the check won't be carried out if the advert is
# from the same master router as the previous advert received.
vrrp_skip_check_adv_addr     # Default - don't skip

# Enforce strict VRRP protocol compliance. This will prohibit:
#   0 VIPs
#   unicast peers
#   IPv6 addresses in VRRP version 2
vrrp_strict

# The following 4 options can be used if vrrp or checker processes
#   are timing out. This can be seen by a backup vrrp instance becoming
#   master even when the master is still running because the master or
#   backup system is too busy to process vrrp packets.
vrrp_priority <-20 to 19>    # Set the vrrp child process priority
                             #   Negative values increase priority.
checker_priority <-20 to 19> # Set the checker child process priority
vrrp_no_swap                 # Set the vrrp child process non swappable
checker_no_swap              # Set the checker child process non swappable

# If Keepalived has been build with SNMP support, the following keywords are available
# Note: Keepalived, checker and RFC support can be individually enabled/disabled
snmp_socket udp:1.2.3.4:705  # specify socket to use for connecting to SNMP master agent (default unix:/var/agentx/master)
                             # unless using a network namespace, when the default is udp:localhost:705
enable_snmp_keepalived       # enable SNMP handling of vrrp element of KEEPALIVED MIB
enable_snmp_checker          # enable SNMP handling of checker element of KEEPALIVED MIB
enable_snmp_rfc              # enable SNMP handling of RFC2787 and RFC6527 VRRP MIBs
enable_snmp_rfcv2            # enable SNMP handling of RFC2787 VRRP MIB
enable_snmp_rfcv3            # enable SNMP handling of RFC6527 VRRP MIB
enable_traps                 # enable SNMP traps

# If Keepalived has been build with DBus support, the following keywords are available
enable_dbus                       # enable the DBus interface
dbus_service_name SERVICE_NAME    # Name of DBus service (default org.keepalived.Vrrp1)
                                  # Useful if you want to run multiple keepalived processes with DBus enabled

# Specify the default username/groupname to run scripts under.
# If this option is not specified, the user defaults to keepalived_script
#   if that user exists, otherwise root.
script_user username [groupname] # If groupname is not specified, it defaults to the user's group
enable_script_security       # Don't run scripts configured to be run as root if any part of the path
                             #   is writable by a non-root user.

# Rather than using notify scripts, specifying a fifo allows more efficient processing of notify events, and guarantees that they will be delivered in the correct sequence.
# NOTE: the FIFO names must all be different
notify_fifo FIFO_NAME        # FIFO to write notify events to
                             # See vrrp_notify_fifo and lvs_notify_fifo for format of output
                             # For further details, see the description under vrrp_sync_group see
                             # doc/samples/sample_notify_fifo.sh for sample usage.
notify_fifo_script STRING [username [groupname]]
                             # script to be run by keepalived to process notify events
                             # The FIFO name will be passed to the script as the last parameter
vrrp_notify_fifo FIFO_NAME   # FIFO to write vrrp notify events to
                             # The string written will be a line of the form: INSTANCE "VI_1" MASTER 100
                             # and will be terminated with a new line character.
                             # For further details of the output, see the description under vrrp_sync_group
                             # and doc/samples/sample_notify_fifo.sh for sample usage.
vrrp_notify_fifo_script STRING [username [groupname]]
                             # script to be run by keepalived to process vrrp notify events
                             # The FIFO name will be passed to the script as the last parameter
lvs_notify_fifo FIFO_NAME    # FIFO to write notify healthchecker events to
                             # The string written will be a line of the form:
                             #   VS [192.168.201.15]:tcp:80 {UP|DOWN}
                             #   RS [1.2.3.4]:tcp:80 [192.168.201.15]:tcp:80 {UP|DOWN}
                             # and will be terminated with a new line character.
lvs_notify_fifo_script STRING [username [groupname]]
                             # script to be run by keepalived to process healthchecher notify events
                             # The FIFO name will be passed to the script as the last parameter
}

# For running keepalived in a separate network namespace
net_namespace NAME           # Set the network namespace to run in
                             # The directory /var/run/keepalived will be created as an unshared mount point,
                             #   for example for pid files.
                             # syslog entries will have _NAME appended to the ident.
                             # Note: the namespace cannot be changed on a configuration reload
namespace_with_ipsets        # ipsets wasn't network namespace aware until Linux 3.13, and so if running with
                             # an earlier version of the kernel, by default use of ipsets is disabled if using
                             # a namespace and vrrp_ipsets has not been specified. This options overrides the
                             # default and allows ipsets to be used with a namespace on kernels prior to 3.13.

instance NAME                # If multiple instances of keepalived are run in the same namespace, this will
                             #   create pid files with NAME as part of the file names, in /var/run/keepalived.
                             # Note: the instance name cannot be changed on a configuration reload

use_pid_dir                  # Create pid files in /var/run/keepalived

linkbeat_use_polling         # Poll to detect media link failure otherwise attempt to use ETHTOOL or MII interface

Static routes/addresses/rules

Keepalived can configure static addresses, routes, and rules. These addresses are NOT moved by vrrpd, they stay on the machine. If you already have IPs and routes on your machines and your machines can ping each other, you don't need this section. The syntax for rules and routes is that same as for ip rule add/ip route add.

The syntax is the same for virtual addresses and virtual routes. If no dev element is specified, it defaults to default_interface (default eth0). Note: the broadcast address may be specified as '-' or '+' to clear or set the host bits of the address.

static_ipaddress
{
<IPADDR>[/<MASK>] [brd <IPADDR>] [dev <STRING>] [scope <SCOPE>] [label <LABEL>] [home] [-nodad] [mngtmpaddr] [noprefixroute] [autojoin]
192.168.1.1/24 dev eth0 scope global
...
}

static_routes
{
192.168.2.0/24 via 192.168.1.100 dev eth0
192.168.100.0/24 table 6909 nexthop via 192.168.101.1 dev wlan0 onlink weight 1 nexthop via 192.168.101.2 dev wlan0 onlink weight 2
192.168.200.0/24 dev p33p1.2 table 6909 tos 0x04 protocol bird scope link priority 12 mtu 1000 hoplimit 100 advmss 101 rtt 102 rttvar 103 reordering 104 window 105 cwnd 106 ssthresh lock 107 realms PQA/0x14 rto_min 108 initcwnd 109 initrwnd 110 features ecn
2001:470:69e9:1:2::4 dev p33p1.2 table 6909 tos 0x04 protocol bird scope link priority 12 mtu 1000 hoplimit 100 advmss 101 rtt 102 rttvar 103 reordering 104 window 105 cwnd 106 ssthresh lock 107 rto_min 108 initcwnd 109 initrwnd 110 features ecn
...
}

static_rules
{
from 192.168.2.0/24 table 1
to 192.168.2.0/24 table 1
from 192.168.28.0/24 to 192.168.29.0/26 table small iif p33p1 oif wlan0 tos 22 fwmark 24/12 preference 39 realms 30/20 goto 40
to 1:2:3:4:5:6:7:0/112 from 7:6:5:4:3:2::/96 table 6908 uidrange 10000-19999
...
}

VRRPD Configuration

contains subblocks of VRRP script(s), VRRP synchronization group(s), VRRP gratuitous ARP and unsolicited neighbour advert delay group(s) and VRRP instance(s)

VRRP Script(s)

# Adds a script to be executed periodically. Its exit code will be
# recorded for all VRRP instances which are monitoring it.
vrrp_script <SCRIPT_NAME> {
   script <STRING>|<QUOTED-STRING> # path of the script to execute
   interval <INTEGER>  # seconds between script invocations, default 1 second
   timeout <INTEGER>   # seconds after which script is considered to have failed
   weight <INTEGER:-254..254>  # adjust priority by this weight, default 0
   rise <INTEGER>              # required number of successes for OK transition
   fall <INTEGER>              # required number of successes for KO transition
   user USERNAME [GROUPNAME]   # user/group names to run script under
                               #   group default to group of user
   init_fail                   # assume script initially is in failed state
}

VRRP Synchronization Group(s)

#string, name of group of IPs that failover together
vrrp_sync_group VG_1 {
   group {
     inside_network   # name of the vrrp_instance (see below)
     outside_network  # One for each movable IP
     ...
   }

   # notify scripts and alerts are optional
   #
   # filenames of scripts to run on transitions
   # can be unquoted (if just filename)
   # or quoted (if it has parameters)
   # The username and groupname specify the user and group
   # under which the scripts should be run. If username is
   # specified, the group defaults to the group of the user.
   # If username is not specified, they default to the
   # global script_user and script_group
   # to MASTER transition
   notify_master /path/to_master.sh [username [groupname]]
   # to BACKUP transition
   notify_backup /path/to_backup.sh [username [groupname]]
   # FAULT transition
   notify_fault "/path/fault.sh VG_1" [username [groupname]]

   # for ANY state transition.
   # "notify" script is called AFTER the
   # notify_* script(s) and is executed
   # with 4 arguments provided by Keepalived
   # (so don't include parameters in the notify line).
   # arguments
   # $1 = "GROUP"|"INSTANCE"
   # $2 = name of the group or instance
   # $3 = target state of transition (stop only applies to instances)
   #     ("MASTER"|"BACKUP"|"FAULT"|"STOP")
   # $4 = priority value
   notify /path/notify.sh [username [groupname]]

   # Send email notification during state transition,
   # using addresses in global_defs above.
   smtp_alert

   global_tracking     # All VRRP share the same tracking config
}

VRRP Gratuitous ARP and Unsolicited Neighbour Advert Delay Group(s)

specifies the setting of delays between sending gratuitous ARPs and unsolicited neighbour advertisements. This is intended for when an upstream switch is unable to handle being flooded with ARPs/NAs.

Use interface when the limits apply on the single physical interface. Use interfaces when a group of interfaces are linked to the same switch and the limits apply to the switch as a whole.

If the global vrrp_garp_interval and/or vrrp_gna_interval are set, any interfaces that aren't specified in a garp_group will inherit the global settings.

garp_group {
   # Sets the interval between Gratuitous ARP (in seconds, resolution microseconds)
   garp_interval <DECIMAL>
   # Sets the default interval between unsolicited NA (in seconds, resolution microseconds)
   gna_interval <DECIMAL>
   # The physical interface to which the intervals apply
   interface <STRING>
   # A list of interfaces accross which the delays are aggregated.
   interfaces {
       <STRING>
       <STRING>
       ...
       }
   }

VRRP Instance(s)

describes the movable IP for each instance of a group in vrrp_sync_group. Here are described two IPs (on inside_network and on outside_network), on machine "my_hostname", which belong to the group VG_1 and which will transition together on any state change.

#You will need to write another block for outside_network.
vrrp_instance inside_network {
   # Initial state, MASTER|BACKUP
   # As soon as the other machine(s) come up,
   # an election will be held and the machine
   # with the highest priority will become MASTER.
   # So the entry here doesn't matter a whole lot.
   state MASTER

   # interface for inside_network, bound by vrrp
   interface eth0

   # Use VRRP Virtual MAC.
   # NOTE: If sysctl net.ipv4.conf.all.rp_filter is set,
   # and this vrrp_instance is an IPv4 instance, using
   # this option will cause the individual interfaces to be
   # updated to the greater of their current setting, and
   # all.rp_filter, as will default.rp_filter, and all.rp_filter
   # will be set to 0.
   # The original settings are restored on termination.
   use_vmac [<VMAC_INTERFACE>]

   # Send/Recv VRRP messages from base interface instead of
   # VMAC interface
   vmac_xmit_base

   native_ipv6         # force instance to use IPv6 (this option is deprecated since the virtual ip addresses determine whether IPv4 or IPv6 is used).

   # Ignore VRRP interface faults (default unset)
   dont_track_primary

   # optional, monitor these as well.
   # go to FAULT state if any of these go down.
   track_interface {
     eth0
     eth1
     eth2 weight <-254..254>
     ...
   }

   # add a tracking script to the interface (<SCRIPT_NAME> is the name of the vrrp_script entry)
   track_script {
       <SCRIPT_NAME>
       <SCRIPT_NAME> weight <-254..254>
   }

   # default IP for binding vrrpd is the primary IP
   # on interface. If you want to hide the location of vrrpd,
   # use this IP as src_addr for multicast or unicast vrrp
   # packets. (since it's multicast, vrrpd will get the reply
   # packet no matter what src_addr is used).
   # optional
   mcast_src_ip <IPADDR>
   unicast_src_ip <IPADDR>

   version <2 or 3>            # VRRP version to run on interface
                               #  default is global parameter vrrp_version.

   # Do not send VRRP adverts over a VRRP multicast group.
   # Instead it sends adverts to the following list of
   # ip addresses using unicast. It can be cool to use
   # the VRRP FSM and features in a networking
   # environment where multicast is not supported!
   # IP addresses specified can be IPv4 as well as IPv6.
   unicast_peer {
     <IPADDR>
     ...
   }
   # The checksum calculation when using VRRPv3 changed after v1.3.6.
   #  Setting this flag forces the old checksum algorithm to be used
   #  to maintain backward compatibility, although keepalived will
   #  attempt to maintain compatibility anyway if it sees an old
   #  version checksum. Sepcifying never will turn off auto detection
   #  of old checksums. [This option may not be enabled - check output
   #  of `keepalived -v` for OLD_CHKSUM_COMPAT.]
   old_unicast_checksum [never]

   # interface specific settings, same as global parameters; default to global parameters
   garp_master_delay 10
   garp_master_repeat 1
   garp_lower_prio_delay 10
   garp_lower_prio_repeat 1
   garp_master_refresh 60
   garp_master_refresh_repeat 2
   garp_interval 100
   gna_interval 100

   lower_prio_no_advert [<BOOL>]
   higher_prio_send_advert [<BOOL>]

   # arbitrary unique number from 0 to 255
   # used to differentiate multiple instances of vrrpd
   # running on the same NIC (and hence same socket).
   virtual_router_id 51

   # for electing MASTER, highest priority wins.
   # to be MASTER, make this 50 more than on other machines.
   priority 100

   # VRRP Advert interval in seconds (e.g. 0.92) (use default)
   advert_int 1

   # Note: authentication was removed from the VRRPv2 specification by RFC3768 in 2004.
   #   Use of this option is non-compliant and can cause problems; avoid using if possible,
   #   except when using unicast, where it can be helpful.
   authentication {     # Authentication block
       # PASS||AH
       # PASS - Simple password (suggested)
       # AH - IPSEC (not recommended))
       auth_type PASS
       # Password for accessing vrrpd.
       # should be the same on all machines.
       # Only the first eight (8) characters are used.
       auth_pass 1234
   }

   #addresses add|del on change to MASTER, to BACKUP.
   #With the same entries on other machines,
   #the opposite transition will be occurring.
   virtual_ipaddress {
       <IPADDR>[/<MASK>] [brd <IPADDR>] [dev <STRING>] [scope <SCOPE>] [label <LABEL>] [home] [-nodad] [mngtmpaddr] [noprefixroute] [autojoin]
       192.168.200.17/24 dev eth1
       192.168.200.18/24 dev eth2 label eth2:1
   }

   #VRRP IP excluded from VRRP
   #optional.
   #For cases with large numbers (eg 200) of IPs
   #on the same interface. To decrease the number
   #of packets sent in adverts, you can exclude
   #most IPs from adverts.
   #The IPs are add|del as for virtual_ipaddress.
   # Can also be used if you want to be able to add
   # a mixture of IPv4 and IPv6 addresses, since all
   # addresses in virtual_ipaddress must be of the
   # same family.
   virtual_ipaddress_excluded {
       <IPADDR>[/<MASK>] [brd <IPADDR>] [dev <STRING>] [scope <SCOPE>] [label <LABEL>] [home] [-nodad] [mngtmpaddr] [noprefixroute] [autojoin]
       <IPADDR>[/<MASK>] ...
       ...
   }

   # Set the promote_secondaries flag on the interface to stop other
   # addresses in the same CIDR being removed when 1 of them is removed
   # For example if 10.1.1.2/24 and 10.1.1.3/24 are both configured on an
   # interface, and one is removed, unless promote_secondaries is set on
   # the interface the other address will also be removed.
   prompte_secondaries

   # routes add|del when changing to MASTER, to BACKUP.
   # See static_routes for more details
   virtual_routes {
       # src <IPADDR> [to] <IPADDR>/<MASK> via|gw <IPADDR> [or <IPADDR>] dev <STRING> scope <SCOPE> table <TABLE>
       src 192.168.100.1 to 192.168.109.0/24 via 192.168.200.254 dev eth1
       192.168.110.0/24 via 192.168.200.254 dev eth1
       192.168.111.0/24 dev eth2
       192.168.112.0/24 via 192.168.100.254
       192.168.113.0/24 via 192.168.200.254 or 192.168.100.254 dev eth1
       blackhole 192.168.114.0/24
       0.0.0.0/0 gw 192.168.0.1 table 100  # To set a default gateway into table 100.
   }

   # rules add|del when changing to MASTER, to BACKUP
   # See static_rules for more details
   virtual_rules {
       from 192.168.2.0/24 table 1
       to 192.168.2.0/24 table 1
   }

   # VRRPv3 has an Accept Mode to allow the virtual router when not the address owner to
   # receive packets addressed to a VIP. This is the default setting unless strict mode is set.
   # As an extension, this also works for VRRPv2 (RFC 3768 doesn't define an accept mode).
   accept     # Accept packets to non address-owner
   no_accept  # Drop packets to non address-owner.

   # VRRP will normally preempt a lower priority
   # machine when a higher priority machine comes
   # online.  "nopreempt" allows the lower priority
   # machine to maintain the master role, even when
   # a higher priority machine comes back online.
   # NOTE: For this to work, the initial state of this
   # entry must be BACKUP.
   nopreempt
   preempt             # for backwards compatibility

   # See description of global vrrp_skip_check_adv_addr, which
   # sets the default value. Defaults to vrrp_skip_check_adv_addr
   skip_check_adv_addr [on|off|true|false|yes|no]

   # See description of global vrrp_strict
   # If vrrp_strict is not specified, it takes the value of vrrp_strict
   # If strict_mode without a parameter is specified, it defaults to on
   strict_mode [on|off|true|false|yes|no]

   # Seconds after startup or seeing a lower priority master until preemption
   # (if not disabled by "nopreempt").
   # Range: 0 (default) to 1000
   # NOTE: For this to work, the initial state of this
   # entry must be BACKUP.
   preempt_delay 300    # waits 5 minutes

   # Debug level, not implemented yet.
   debug <LEVEL>        # LEVEL is a number in the range 0 to 4

   # notify scripts, alert as above
   notify_master <STRING>|<QUOTED-STRING> [username [groupname]]
   notify_backup <STRING>|<QUOTED-STRING> [username [groupname]]
   notify_fault <STRING>|<QUOTED-STRING> [username [groupname]]
   notify_stop <STRING>|<QUOTED-STRING> [username [groupname]]      # executed when stopping vrrp
   notify <STRING>|<QUOTED-STRING> [username [groupname]]
   smtp_alert
}

# Parameters used for SSL_GET check.
# If none of the parameters are specified, the SSL context will be auto generated.
SSL {
   password <STRING>   # password
   ca <STRING>         # ca file
   certificate <STRING>  # certificate file
   key <STRING>        # key file
}

LVS Configuration

contains subblocks of Virtual server group(s) and Virtual server(s)

The subblocks contain arguments for ipvsadm(8). Knowledge of ipvsadm(8) will be helpful here.

Virtual Server Group(s)

# optional
# this groups allows a service on a real_server
# to belong to multiple virtual services
# and to only be health checked once.
# Only for very large LVSs.
virtual_server_group <STRING> {
       #VIP port
       <IPADDR> <PORT>
       <IPADDR> <PORT>
       ...
       #
       # <IPADDR RANGE> has the form
       # XXX.YYY.ZZZ.WWW-VVV eg 192.168.200.1-10
       # range includes both .1 and .10 address
       <IPADDR RANGE> <PORT># VIP range VPORT
       <IPADDR RANGE> <PORT>
       ...
       fwmark <INT>  # fwmark
       fwmark <INT>
       ... }

Virtual Server(s)

A virtual_server can be a declaration of one of

vip vport (IPADDR PORT pair)
fwmark <INT>
(virtual server) group <STRING>

   #setup service
   virtual_server IP port |
   virtual_server fwmark int |
   virtual_server group string
   {
   # delay timer for checker polling
   delay_loop <INT>

   # LVS scheduler
   lvs_sched rr|wrr|lc|wlc|lblc|sh|dh|fo|ovf|lblcr|sed|nq

   # Enable hashed entry
   hashed
   # Enable flag-1 for scheduler (-b flag-1 in ipvsadm)
   flag-1
   # Enable flag-2 for scheduler (-b flag-2 in ipvsadm)
   flag-2
   # Enable flag-3 for scheduler (-b flag-3 in ipvsadm)
   flag-3
   # Enable sh-port for sh scheduler (-b sh-port in ipvsadm)
   sh-port
   # Enable sh-fallback for sh scheduler  (-b sh-fallback in ipvsadm)
   sh-fallback
   # Enable One-Packet-Scheduling for UDP (-O in ipvsadm)
   ops
   # Default LVS forwarding method
   lvs_method NAT|DR|TUN
   # LVS persistence engine name
   persistence_engine <STRING>
   # LVS persistence timeout in seconds, default 6 minutes
   persistence_timeout [<INT>]
   # LVS granularity mask (-M in ipvsadm)
   persistence_granularity <NETMASK>
   # L4 protocol
   protocol TCP|UDP|SCTP
   # If VS IP address is not set,
   # suspend healthchecker's activity
   ha_suspend

   lb_algo    # deprecated synonym for lvs_sched
   lb_kind    # deprecated synonym for lvs_method

   # Default VirtualHost string for HTTP_GET or SSL_GET
   # eg virtualhost www.firewall.loc
   # Overridden by virtualhost config of real server or checker
   virtualhost <STRING>

   # On daemon startup assume that all RSs are down
   # and healthchecks failed. This helps to prevent
   # false positives on startup. Alpha mode is
   # disabled by default.
   alpha

   # On daemon shutdown consider quorum and RS
   # down notifiers for execution, where appropriate.
   # Omega mode is disabled by default.
   omega

   # Minimum total weight of all live servers in
   # the pool necessary to operate VS with no
   # quality regression. Defaults to 1.
   quorum <INT>

   # Tolerate this much weight units compared to the
   # nominal quorum, when considering quorum gain
   # or loss. A flap dampener. Defaults to 0.
   hysteresis <INT>

   # Script to execute when quorum is gained.
   quorum_up <STRING>|<QUOTED-STRING> [username [groupname]]

   # Script to execute when quorum is lost.
   quorum_down <STRING>|<QUOTED-STRING> [username [groupname]]

   # IP family for a fwmark service (optional)
   ip_family inet|inet6

   # setup realserver(s)

   # RS to add to LVS topology when the quorum isn't achieved.
   #  If a sorry server is configured, all real servers will
   #  be brought down when the quorum is not achieved.
   sorry_server <IPADDR> <PORT>
   # applies inhibit_on_failure behaviour to the sorry_server
   sorry_server_inhibit
   # Sorry server LVS forwarding method
   sorry_server_lvs_method NAT|DR|TUN

   # Retry count to make additional checks if check
   # of an alive server fails. Default: 1 unless specified below
   retry <INT>

   # delay before retry
   delay_before_retry <INT>

   # Optional random delay to start the initial check
   # for maximum N seconds.
   # Useful to scatter multiple simultaneous
   # checks to the same RS. Enabled by default, with
   # the maximum at delay_loop. Specify 0 to disable
   warmup <INT>

   # delay timer for checker polling
   delay_loop <INT>

   # Set weight to 0 when healthchecker detects failure
   inhibit_on_failure

   # one entry for each realserver
   real_server <IPADDR> <PORT>
      {
          # relative weight to use, default: 1
          weight <INT>
          # LVS forwarding method
          lvs_method NAT|DR|TUN

          # Script to execute when healthchecker
          # considers service as up.
          notify_up <STRING>|<QUOTED-STRING> [username [groupname]]
          # Script to execute when healthchecker
          # considers service as down.
          notify_down <STRING>|<QUOTED-STRING> [username [groupname]]

          uthreshold <INTEGER> # maximum number of connections to server
          lthreshold <INTEGER> # minimum number of connections to server

          # Default VirtualHost string for HTTP_GET or SSL_GET
          # eg virtualhost www.firewall.loc
          # Overridden by virtualhost config of a checker
          virtualhost <STRING>

          alpha <BOOL>                    # see above
          retry <INTEGER>                 # see above
          delay_before_retry <INTEGER>    # see above
          warmup <INTEGER>                # see above
          delay_loop <INTEGER>            # see above
          inhibit_on_failure <BOOL>       # see above

          # healthcheckers. Can be multiple of each type
          # HTTP_GET|SSL_GET|TCP_CHECK|SMTP_CHECK|DNS_CHECK|MISC_CHECK

          # All checkers have the following options, except MISC_CHECK
          # which only has options alpha onwards:
          CHECKER_TYPE {
              # ======== generic connection options
              # Optional IP address to connect to.
              # The default is the realserver IP
              connect_ip <IP ADDRESS>

              # Optional port to connect to
              # The default is the realserver port
              connect_port <PORT>

              # Optional address to use to
              # originate the connection
              bindto <IP ADDRESS>

              # Optional interface to use; needed if
              # the bindto address is IPv6 link local
              bind_if <IFNAME>

              # Optional source port to
              # originate the connection from
              bind_port <PORT>

              # Optional connection timeout in seconds.
              # The default is 5 seconds
              connect_timeout <INTEGER>

              # Optional fwmark to mark all outgoing
              # checker packets with
              fwmark <INTEGER>

              alpha <BOOL>                    # see above
              retry <INTEGER>                 # see above
              delay_before_retry <INTEGER>    # see above
              warmup <INTEGER>                # see above
              delay_loop <INTEGER>            # see above
              inhibit_on_failure <BOOL>       # see above
          }

          # The following options are additional checker specific

          # HTTP and SSL healthcheckers
          HTTP_GET|SSL_GET
          {
              # An url to test
              # can have multiple entries here
              url {
                #eg path / , or path /mrtg2/
                path <STRING>
                # healthcheck needs status_code
                # or status_code and digest
                # Digest computed with genhash
                # eg digest 9b3a0c85a887a256d6939da88aabd8cd
                digest <STRING>
                # status code returned in the HTTP header
                # eg status_code 200. Default is any 2xx value
                status_code <INT>
                # VirtualHost string. eg virtualhost www.firewall.loc
                # If not set, uses virtualhost from real or virtual server
                virtualhost <STRING>
              }
          }

          # TCP healthchecker
          TCP_CHECK
          {
              # No additional options
          }

          # SMTP healthchecker
          SMTP_CHECK
          {
              # Optional string to use for the SMTP HELO request
              helo_name <STRING>|<QUOTED-STRING>
          }

          # DNS healthchecker
          DNS_CHECK
          {
              # The retry default is 3.

              # DNS query type
              #   A|NS|CNAME|SOA|MX|TXT|AAAA
              # The default is SOA
              type <STRING>

              # Domain name to use for the DNS query
              # The default is . (dot)
              name <STRING>
          }

          # MISC healthchecker, run a program
          MISC_CHECK
          {
              # The retry default is 0.

              # External script or program
              misc_path <STRING>|<QUOTED-STRING>
              # Script execution timeout
              misc_timeout <INT>

              # If set, the exit code from healthchecker is used
              # to dynamically adjust the weight as follows:
              #   exit status 0: svc check success, weight
              #     unchanged.
              #   exit status 1: svc check failed.
              #   exit status 2-255: svc check success, weight
              #     changed to 2 less than exit status.
              #   (for example: exit status of 255 would set
              #     weight to 253)
              # NOTE: do not have more than one dynamic MISC_CHECK per real_server.
              misc_dynamic

              # Specify the username/groupname that the script should
              #   be run under.
              # If GROUPNAME is not specified, the group of the user
              #   is used
              user USERNAME [GROUPNAME]
          }
      } # realserver defn
   } # virtual service

Author

Joseph Mack.
Information derived from doc/keepalived.conf.SYNOPSIS, doc/samples/keepalived.conf.* and Changelog by Alexandre Cassen for keepalived-1.1.4, and from HOWTOs by Adam Fletcher and Vince Worthington.

See Also

ipvsadm(8), ip --help.

Referenced By

genhash(1), keepalived(8).

Apr 2016 V1.0