dhcpd.conf.5

.\"	dhcpd.conf.5
.\"
.\" Copyright (c) 2004-2017 by Internet Systems Consortium, Inc. ("ISC")
.\" Copyright (c) 1996-2003 by Internet Software Consortium
.\"
.\" This Source Code Form is subject to the terms of the Mozilla Public
.\" License, v. 2.0. If a copy of the MPL was not distributed with this
.\" file, You can obtain one at http://mozilla.org/MPL/2.0/.
.\"
.\" THE SOFTWARE IS PROVIDED "AS IS" AND ISC DISCLAIMS ALL WARRANTIES
.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
.\" MERCHANTABILITY AND FITNESS.  IN NO EVENT SHALL ISC BE LIABLE FOR
.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT
.\" OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
.\"
.\"   Internet Systems Consortium, Inc.
.\"   950 Charter Street
.\"   Redwood City, CA 94063
.\"   <info@isc.org>
.\"   https://www.isc.org/
.\"
.\" This software has been written for Internet Systems Consortium
.\" by Ted Lemon in cooperation with Vixie Enterprises and Nominum, Inc.
.\"
.\" Support and other services are available for ISC products - see
.\" https://www.isc.org for more information or to learn more about ISC.
.\"
.\" $Id: dhcpd.conf.5,v 1.114 2012/04/02 22:47:35 sar Exp $
.\"
.TH dhcpd.conf 5
.SH NAME
dhcpd.conf - dhcpd configuration file
.SH DESCRIPTION
The dhcpd.conf file contains configuration information for
.IR dhcpd,
the Internet Systems Consortium DHCP Server.
.PP
The dhcpd.conf file is a free-form ASCII text file.  It is parsed by
the recursive-descent parser built into dhcpd.  The file may contain
extra tabs and newlines for formatting purposes.  Keywords in the file
are case-insensitive.  Comments may be placed anywhere within the
file (except within quotes).  Comments begin with the # character and
end at the end of the line.
.PP
The file essentially consists of a list of statements.  Statements
fall into two broad categories - parameters and declarations.
.PP
Parameter statements either say how to do something (e.g., how long a
lease to offer), whether to do something (e.g., should dhcpd provide
addresses to unknown clients), or what parameters to provide to the
client (e.g., use gateway 220.177.244.7).
.PP
Declarations are used to describe the topology of the
network, to describe clients on the network, to provide addresses that
can be assigned to clients, or to apply a group of parameters to a
group of declarations.  In any group of parameters and declarations,
all parameters must be specified before any declarations which depend
on those parameters may be specified.
.PP
Declarations about network topology include the \fIshared-network\fR
and the \fIsubnet\fR declarations.  If clients on a subnet are to be
assigned addresses
dynamically, a \fIrange\fR declaration must appear within the
\fIsubnet\fR declaration.  For clients with statically assigned
addresses, or for installations where only known clients will be
served, each such client must have a \fIhost\fR declaration.  If
parameters are to be applied to a group of declarations which are not
related strictly on a per-subnet basis, the \fIgroup\fR declaration
can be used.
.PP
For every subnet which will be served, and for every subnet
to which the dhcp server is connected, there must be one \fIsubnet\fR
declaration, which tells dhcpd how to recognize that an address is on
that subnet.  A \fIsubnet\fR declaration is required for each subnet
even if no addresses will be dynamically allocated on that subnet.
.PP
Some installations have physical networks on which more than one IP
subnet operates.  For example, if there is a site-wide requirement
that 8-bit subnet masks be used, but a department with a single
physical ethernet network expands to the point where it has more than
254 nodes, it may be necessary to run two 8-bit subnets on the same
ethernet until such time as a new physical network can be added.  In
this case, the \fIsubnet\fR declarations for these two networks must be
enclosed in a \fIshared-network\fR declaration.
.PP
Note that even when the \fIshared-network\fR declaration is absent, an
empty one is created by the server to contain the \fIsubnet\fR (and any scoped
parameters included in the \fIsubnet\fR).  For practical purposes, this means
that "stateless" DHCP clients, which are not tied to addresses (and therefore
subnets) will receive the same configuration as stateful ones.
.PP
Some sites may have departments which have clients on more than one
subnet, but it may be desirable to offer those clients a uniform set
of parameters which are different than what would be offered to
clients from other departments on the same subnet.  For clients which
will be declared explicitly with \fIhost\fR declarations, these
declarations can be enclosed in a \fIgroup\fR declaration along with
the parameters which are common to that department.  For clients
whose addresses will be dynamically assigned, class declarations and
conditional declarations may be used to group parameter assignments
based on information the client sends.
.PP
When a client is to be booted, its boot parameters are determined by
consulting that client's \fIhost\fR declaration (if any), and then
consulting any \fIclass\fR declarations matching the client,
followed by the \fIpool\fR, \fIsubnet\fR and \fIshared-network\fR
declarations for the IP address assigned to the client.  Each of
these declarations itself appears within a lexical scope, and all
declarations at less specific lexical scopes are also consulted for
client option declarations.  Scopes are never considered
twice, and if parameters are declared in more than one scope, the
parameter declared in the most specific scope is the one that is
used.
.PP
When dhcpd tries to find a \fIhost\fR declaration for a client, it
first looks for a \fIhost\fR declaration which has a
\fIfixed-address\fR declaration that lists an IP address that is valid
for the subnet or shared network on which the client is booting.  If
it doesn't find any such entry, it tries to find an entry which has
no \fIfixed-address\fR declaration.
.SH EXAMPLES
.PP
A typical dhcpd.conf file will look something like this:
.nf

.I global parameters...

subnet 204.254.239.0 netmask 255.255.255.224 {
  \fIsubnet-specific parameters...\fR
  range 204.254.239.10 204.254.239.30;
}

subnet 204.254.239.32 netmask 255.255.255.224 {
  \fIsubnet-specific parameters...\fR
  range 204.254.239.42 204.254.239.62;
}

subnet 204.254.239.64 netmask 255.255.255.224 {
  \fIsubnet-specific parameters...\fR
  range 204.254.239.74 204.254.239.94;
}

group {
  \fIgroup-specific parameters...\fR
  host zappo.test.isc.org {
    \fIhost-specific parameters...\fR
  }
  host beppo.test.isc.org {
    \fIhost-specific parameters...\fR
  }
  host harpo.test.isc.org {
    \fIhost-specific parameters...\fR
  }
}

.ce 1
Figure 1

.fi
.PP
Notice that at the beginning of the file, there's a place
for global parameters.  These might be things like the organization's
domain name, the addresses of the name servers (if they are common to
the entire organization), and so on.  So, for example:
.nf

	option domain-name "isc.org";
	option domain-name-servers ns1.isc.org, ns2.isc.org;

.ce 1
Figure 2
.fi
.PP
As you can see in Figure 2, you can specify host addresses in
parameters using their domain names rather than their numeric IP
addresses.  If a given hostname resolves to more than one IP address
(for example, if that host has two ethernet interfaces), then where
possible, both addresses are supplied to the client.
.PP
The most obvious reason for having subnet-specific parameters as
shown in Figure 1 is that each subnet, of necessity, has its own
router.  So for the first subnet, for example, there should be
something like:
.nf

	option routers 204.254.239.1;
.fi
.PP
Note that the address here is specified numerically.  This is not
required - if you have a different domain name for each interface on
your router, it's perfectly legitimate to use the domain name for that
interface instead of the numeric address.  However, in many cases
there may be only one domain name for all of a router's IP addresses, and
it would not be appropriate to use that name here.
.PP
In Figure 1 there is also a \fIgroup\fR statement, which provides
common parameters for a set of three hosts - zappo, beppo and harpo.
As you can see, these hosts are all in the test.isc.org domain, so it
might make sense for a group-specific parameter to override the domain
name supplied to these hosts:
.nf

	option domain-name "test.isc.org";
.fi
.PP
Also, given the domain they're in, these are probably test machines.
If we wanted to test the DHCP leasing mechanism, we might set the
lease timeout somewhat shorter than the default:

.nf
	max-lease-time 120;
	default-lease-time 120;
.fi
.PP
You may have noticed that while some parameters start with the
\fIoption\fR keyword, some do not.  Parameters starting with the
\fIoption\fR keyword correspond to actual DHCP options, while
parameters that do not start with the option keyword either control
the behavior of the DHCP server (e.g., how long a lease dhcpd will
give out), or specify client parameters that are not optional in the
DHCP protocol (for example, server-name and filename).
.PP
In Figure 1, each host had \fIhost-specific parameters\fR.  These
could include such things as the \fIhostname\fR option, the name of a
file to upload (the \fIfilename\fR parameter) and the address of the
server from which to upload the file (the \fInext-server\fR
parameter).  In general, any parameter can appear anywhere that
parameters are allowed, and will be applied according to the scope in
which the parameter appears.
.PP
Imagine that you have a site with a lot of NCD X-Terminals.  These
terminals come in a variety of models, and you want to specify the
boot files for each model.  One way to do this would be to have host
declarations for each server and group them by model:
.nf

group {
  filename "Xncd19r";
  next-server ncd-booter;

  host ncd1 { hardware ethernet 0:c0:c3:49:2b:57; }
  host ncd4 { hardware ethernet 0:c0:c3:80:fc:32; }
  host ncd8 { hardware ethernet 0:c0:c3:22:46:81; }
}

group {
  filename "Xncd19c";
  next-server ncd-booter;

  host ncd2 { hardware ethernet 0:c0:c3:88:2d:81; }
  host ncd3 { hardware ethernet 0:c0:c3:00:14:11; }
}

group {
  filename "XncdHMX";
  next-server ncd-booter;

  host ncd1 { hardware ethernet 0:c0:c3:11:90:23; }
  host ncd4 { hardware ethernet 0:c0:c3:91:a7:8; }
  host ncd8 { hardware ethernet 0:c0:c3:cc:a:8f; }
}
.fi
.SH ADDRESS POOLS
.PP
The
\fBpool\fR and \fBpool6\fR
declarations can be used to specify a pool of addresses that will be
treated differently than another pool of addresses, even on the same
network segment or subnet.  For example, you may want to provide a
large set of addresses that can be assigned to DHCP clients that are
registered to your DHCP server, while providing a smaller set of
addresses, possibly with short lease times, that are available for
unknown clients.  If you have a firewall, you may be able to arrange
for addresses from one pool to be allowed access to the Internet,
while addresses in another pool are not, thus encouraging users to
register their DHCP clients.  To do this, you would set up a pair of
pool declarations:
.PP
.nf
subnet 10.0.0.0 netmask 255.255.255.0 {
  option routers 10.0.0.254;

  # Unknown clients get this pool.
  pool {
    option domain-name-servers bogus.example.com;
    max-lease-time 300;
    range 10.0.0.200 10.0.0.253;
    allow unknown-clients;
  }

  # Known clients get this pool.
  pool {
    option domain-name-servers ns1.example.com, ns2.example.com;
    max-lease-time 28800;
    range 10.0.0.5 10.0.0.199;
    deny unknown-clients;
  }
}
.fi
.PP
It is also possible to set up entirely different subnets for known and
unknown clients - address pools exist at the level of shared networks,
so address ranges within pool declarations can be on different
subnets.
.PP
As you can see in the preceding example, pools can have permit lists
that control which clients are allowed access to the pool and which
aren't.  Each entry in a pool's permit list is introduced with the
.I allow
or \fIdeny\fR keyword.  If a pool has a permit list, then only those
clients that match specific entries on the permit list will be
eligible to be assigned addresses from the pool.  If a pool has a
deny list, then only those clients that do not match any entries on
the deny list will be eligible.   If both permit and deny lists exist
for a pool, then only clients that match the permit list and do not
match the deny list will be allowed access.
.PP
The \fBpool6\fR declaration is similar to the \fBpool\fR declaration.
Currently it is only allowed within a \fBsubnet6\fR declaration, and
may not be included directly in a shared network declaration.
In addition to the \fBrange6\fR statement it allows the \fBprefix6\fR
statement to be included.  You may include \fBrange6\fR statements
for both NA and TA and \fBprefixy6\fR statements in a single
\fBpool6\fR statement.
.SH DYNAMIC ADDRESS ALLOCATION
Address allocation is actually only done when a client is in the INIT
state and has sent a DHCPDISCOVER message.  If the client thinks it
has a valid lease and sends a DHCPREQUEST to initiate or renew that
lease, the server has only three choices - it can ignore the
DHCPREQUEST, send a DHCPNAK to tell the client it should stop using
the address, or send a DHCPACK, telling the client to go ahead and use
the address for a while.
.PP
If the server finds the address the client is requesting, and that
address is available to the client, the server will send a DHCPACK.
If the address is no longer available, or the client isn't permitted
to have it, the server will send a DHCPNAK.  If the server knows
nothing about the address, it will remain silent, unless the address
is incorrect for the network segment to which the client has been
attached and the server is authoritative for that network segment, in
which case the server will send a DHCPNAK even though it doesn't know
about the address.
.PP
There may be a host declaration matching the client's identification.
If that host declaration contains a fixed-address declaration that
lists an IP address that is valid for the network segment to which the
client is connected, the DHCP server will never do dynamic address allocation.
In this case, the client is \fIrequired\fR to take the address specified
in the host declaration.  If the client sends a DHCPREQUEST for some other
address, the server will respond with a DHCPNAK.
.PP
When the DHCP server allocates a new address for a client (remember,
this only happens if the client has sent a DHCPDISCOVER), it first
looks to see if the client already has a valid lease on an IP address,
or if there is an old IP address the client had before that hasn't yet
been reassigned.  In that case, the server will take that address and
check it to see if the client is still permitted to use it.  If the
client is no longer permitted to use it, the lease is freed if the
server thought it was still in use - the fact that the client has sent
a DHCPDISCOVER proves to the server that the client is no longer using
the lease.
.PP
If no existing lease is found, or if the client is forbidden to
receive the existing lease, then the server will look in the list of
address pools for the network segment to which the client is attached
for a lease that is not in use and that the client is permitted to
have.  It looks through each pool declaration in sequence (all
.I range
declarations that appear outside of pool declarations are grouped into
a single pool with no permit list).  If the permit list for the pool
allows the client to be allocated an address from that pool, the pool
is examined to see if there is an address available.  If so, then the
client is tentatively assigned that address.  Otherwise, the next
pool is tested.  If no addresses are found that can be assigned to
the client, no response is sent to the client.
.PP
If an address is found that the client is permitted to have, and that
has never been assigned to any client before, the address is
immediately allocated to the client.  If the address is available for
allocation but has been previously assigned to a different client, the
server will keep looking in hopes of finding an address that has never
before been assigned to a client.
.PP
The DHCP server generates the list of available IP addresses from a
hash table.  This means that the addresses are not sorted in any
particular order, and so it is not possible to predict the order in
which the DHCP server will allocate IP addresses.  Users of previous
versions of the ISC DHCP server may have become accustomed to the DHCP
server allocating IP addresses in ascending order, but this is no
longer possible, and there is no way to configure this behavior with
version 3 of the ISC DHCP server.
.SH IP ADDRESS CONFLICT PREVENTION
The DHCP server checks IP addresses to see if they are in use before
allocating them to clients.  It does this by sending an ICMP Echo
request message to the IP address being allocated.  If no ICMP Echo
reply is received within a second, the address is assumed to be free.
This is only done for leases that have been specified in range
statements, and only when the lease is thought by the DHCP server to
be free - i.e., the DHCP server or its failover peer has not listed
the lease as in use.
.PP
If a response is received to an ICMP Echo request, the DHCP server
assumes that there is a configuration error - the IP address is in use
by some host on the network that is not a DHCP client.  It marks the
address as abandoned, and will not assign it to clients. The lease will
remain abandoned for a minimum of abandon-lease-time seconds.
.PP
If a DHCP client tries to get an IP address, but none are available,
but there are abandoned IP addresses, then the DHCP server will
attempt to reclaim an abandoned IP address.  It marks one IP address
as free, and then does the same ICMP Echo request check described
previously.  If there is no answer to the ICMP Echo request, the
address is assigned to the client.
.PP
The DHCP server does not cycle through abandoned IP addresses if the
first IP address it tries to reclaim is free.  Rather, when the next
DHCPDISCOVER comes in from the client, it will attempt a new
allocation using the same method described here, and will typically
try a new IP address.
.SH DHCP FAILOVER
This version of the ISC DHCP server supports the DHCP failover
protocol as documented in draft-ietf-dhc-failover-12.txt.  This is
not a final protocol document, and we have not done interoperability
testing with other vendors' implementations of this protocol, so you
must not assume that this implementation conforms to the standard.
If you wish to use the failover protocol, make sure that both failover
peers are running the same version of the ISC DHCP server.
.PP
The failover protocol allows two DHCP servers (and no more than two)
to share a common address pool.  Each server will have about half of
the available IP addresses in the pool at any given time for
allocation.  If one server fails, the other server will continue to
renew leases out of the pool, and will allocate new addresses out of
the roughly half of available addresses that it had when
communications with the other server were lost.
.PP
It is possible during a prolonged failure to tell the remaining server
that the other server is down, in which case the remaining server will
(over time) reclaim all the addresses the other server had available
for allocation, and begin to reuse them.  This is called putting the
server into the PARTNER-DOWN state.
.PP
You can put the server into the PARTNER-DOWN state either by using the
.B omshell (1)
command or by stopping the server, editing the last failover state
declaration in the lease file, and restarting the server.  If you use
this last method, change the "my state" line to:
.PP
.nf
.B failover peer "\fIname\fB" state {
.B   my   state partner-down;.
.B   peer state \fIstate\fB at \fIdate\fB;
.B }
.fi
.PP
It is only required to change "my state" as shown above.
.PP
When the other server comes back online, it should automatically
detect that it has been offline and request a complete update from the
server that was running in the PARTNER-DOWN state, and then both
servers will resume processing together.
.PP
It is possible to get into a dangerous situation: if you put one
server into the PARTNER-DOWN state, and then *that* server goes down,
and the other server comes back up, the other server will not know
that the first server was in the PARTNER-DOWN state, and may issue
addresses previously issued by the other server to different clients,
resulting in IP address conflicts.  Before putting a server into
PARTNER-DOWN state, therefore, make
.I sure
that the other server will not restart automatically.
.PP
The failover protocol defines a primary server role and a secondary
server role.  There are some differences in how primaries and
secondaries act, but most of the differences simply have to do with
providing a way for each peer to behave in the opposite way from the
other.  So one server must be configured as primary, and the other
must be configured as secondary, and it doesn't matter too much which
one is which.
.SH FAILOVER STARTUP
When a server starts that has not previously communicated with its
failover peer, it must establish communications with its failover peer
and synchronize with it before it can serve clients.  This can happen
either because you have just configured your DHCP servers to perform
failover for the first time, or because one of your failover servers
has failed catastrophically and lost its database.
.PP
The initial recovery process is designed to ensure that when one
failover peer loses its database and then resynchronizes, any leases
that the failed server gave out before it failed will be honored.
When the failed server starts up, it notices that it has no saved
failover state, and attempts to contact its peer.
.PP
When it has established contact, it asks the peer for a complete copy
its peer's lease database.  The peer then sends its complete database,
and sends a message indicating that it is done.  The failed server
then waits until MCLT has passed, and once MCLT has passed both
servers make the transition back into normal operation.  This waiting
period ensures that any leases the failed server may have given out
while out of contact with its partner will have expired.
.PP
While the failed server is recovering, its partner remains in the
partner-down state, which means that it is serving all clients.  The
failed server provides no service at all to DHCP clients until it has
made the transition into normal operation.
.PP
In the case where both servers detect that they have never before
communicated with their partner, they both come up in this recovery
state and follow the procedure we have just described.  In this case,
no service will be provided to DHCP clients until MCLT has expired.
.SH CONFIGURING FAILOVER
In order to configure failover, you need to write a peer declaration
that configures the failover protocol, and you need to write peer
references in each pool declaration for which you want to do
failover.  You do not have to do failover for all pools on a given
network segment.   You must not tell one server it's doing failover
on a particular address pool and tell the other it is not.  You must
not have any common address pools on which you are not doing
failover.  A pool declaration that utilizes failover would look like this:
.PP
.nf
pool {
	failover peer "foo";
	\fIpool specific parameters\fR
};
.fi
.PP
The  server currently  does very  little  sanity checking,  so if  you
configure it wrong, it will just  fail in odd ways.  I would recommend
therefore that you either do  failover or don't do failover, but don't
do any mixed pools.  Also,  use the same master configuration file for
both  servers,  and  have  a  separate file  that  contains  the  peer
declaration and includes the master file.  This will help you to avoid
configuration  mismatches.  As our  implementation evolves,  this will
become  less of  a  problem.  A  basic  sample dhcpd.conf  file for  a
primary server might look like this:
.PP
.nf
failover peer "foo" {
  primary;
  address anthrax.rc.example.com;
  port 519;
  peer address trantor.rc.example.com;
  peer port 520;
  max-response-delay 60;
  max-unacked-updates 10;
  mclt 3600;
  split 128;
  load balance max seconds 3;
}

include "/etc/dhcpd.master";
.fi
.PP
The statements in the peer declaration are as follows:
.PP
The
.I primary
and
.I secondary
statements
.RS 0.25i
.PP
[ \fBprimary\fR | \fBsecondary\fR ]\fB;\fR
.PP
This determines whether the server is primary or secondary, as
described earlier under DHCP FAILOVER.
.RE
.PP
The
.I address
statement
.RS 0.25i
.PP
.B address \fIaddress\fR\fB;\fR
.PP
The \fBaddress\fR statement declares the IP address or DNS name on which the
server should listen for connections from its failover peer, and also the
value to use for the DHCP Failover Protocol server identifier.  Because this
value is used as an identifier, it may not be omitted.
.RE
.PP
The
.I peer address
statement
.RS 0.25i
.PP
.B peer address \fIaddress\fR\fB;\fR
.PP
The \fBpeer address\fR statement declares the IP address or DNS name to
which the server should connect to reach its failover peer for failover
messages.
.RE
.PP
The
.I port
statement
.RS 0.25i
.PP
.B port \fIport-number\fR\fB;\fR
.PP
The \fBport\fR statement declares the TCP port on which the server
should listen for connections from its failover peer.  This statement
may be omitted, in which case the IANA assigned port number 647 will be
used by default.
.RE
.PP
The
.I peer port
statement
.RS 0.25i
.PP
.B peer port \fIport-number\fR\fB;\fR
.PP
The \fBpeer port\fR statement declares the TCP port to which the
server should connect to reach its failover peer for failover
messages.  This statement may be omitted, in which case the IANA
assigned port number 647 will be used by default.
.RE
.PP
The
.I max-response-delay
statement
.RS 0.25i
.PP
.B max-response-delay \fIseconds\fR\fB;\fR
.PP
The \fBmax-response-delay\fR statement tells the DHCP server how
many seconds may pass without receiving a message from its failover
peer before it assumes that connection has failed.  This number
should be small enough that a transient network failure that breaks
the connection will not result in the servers being out of
communication for a long time, but large enough that the server isn't
constantly making and breaking connections.  This parameter must be
specified.
.RE
.PP
The
.I max-unacked-updates
statement
.RS 0.25i
.PP
.B max-unacked-updates \fIcount\fR\fB;\fR
.PP
The \fBmax-unacked-updates\fR statement tells the remote DHCP server how
many BNDUPD messages it can send before it receives a BNDACK
from the local system.  We don't have enough operational experience
to say what a good value for this is, but 10 seems to work.  This
parameter must be specified.
.RE
.PP
The
.I mclt
statement
.RS 0.25i
.PP
.B mclt \fIseconds\fR\fB;\fR
.PP
The \fBmclt\fR statement defines the Maximum Client Lead Time.  It
must be specified on the primary, and may not be specified on the
secondary.  This is the length of time for which a lease may be
renewed by either failover peer without contacting the other.  The
longer you set this, the longer it will take for the running server to
recover IP addresses after moving into PARTNER-DOWN state.  The
shorter you set it, the more load your servers will experience when
they are not communicating.  A value of something like 3600 is
probably reasonable, but again bear in mind that we have no real
operational experience with this.
.RE
.PP
The
.I split
statement
.RS 0.25i
.PP
.B split \fIbits\fR\fB;\fR
.PP
The split statement specifies the split between the primary and
secondary for the purposes of load balancing.  Whenever a client
makes a DHCP request, the DHCP server runs a hash on the client
identification, resulting in value from 0 to 255.  This is used as
an index into a 256 bit field.  If the bit at that index is set,
the primary is responsible.  If the bit at that index is not set,
the secondary is responsible.  The \fBsplit\fR value determines
how many of the leading bits are set to one.  So, in practice, higher
split values will cause the primary to serve more clients than the
secondary.  Lower split values, the converse.  Legal values are between
0 and 256 inclusive, of which the most reasonable is 128.  Note that
a value of 0 makes the secondary responsible for all clients and a value
of 256 makes the primary responsible for all clients.
.RE
.PP
The
.I hba
statement
.RS 0.25i
.PP
.B hba \fIcolon-separated-hex-list\fB;\fR
.PP
The hba statement specifies the split between the primary and
secondary as a bitmap rather than a cutoff, which theoretically allows
for finer-grained control.  In practice, there is probably no need
for such fine-grained control, however.  An example hba statement:
.PP
.nf
  hba ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:ff:
      00:00:00:00:00:00:00:00:00:00:00:00:00:00:00:00;
.fi
.PP
This is equivalent to a \fBsplit 128;\fR statement, and identical.  The
following two examples are also equivalent to a \fBsplit\fR of 128, but
are not identical:
.PP
.nf
  hba aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:
      aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa:aa;

  hba 55:55:55:55:55:55:55:55:55:55:55:55:55:55:55:55:
      55:55:55:55:55:55:55:55:55:55:55:55:55:55:55:55;
.fi
.PP
They are equivalent, because half the bits are set to 0, half are set to
1 (0xa and 0x5 are 1010 and 0101 binary respectively) and consequently this
would roughly divide the clients equally between the servers.  They are not
identical, because the actual peers this would load balance to each server
are different for each example.
.PP
You must only have \fBsplit\fR or \fBhba\fR defined, never both.  For most
cases, the fine-grained control that \fBhba\fR offers isn't necessary, and
\fBsplit\fR should be used.
.RE
.PP
The
.I load balance max seconds
statement
.RS 0.25i
.PP
.B load balance max seconds \fIseconds\fR\fB;\fR
.PP
This statement allows you to configure a cutoff after which load
balancing is disabled.  The cutoff is based on the number of seconds
since the client sent its first DHCPDISCOVER or DHCPREQUEST message,
and only works with clients that correctly implement the \fIsecs\fR
field - fortunately most clients do.  We recommend setting this to
something like 3 or 5.  The effect of this is that if one of the
failover peers gets into a state where it is responding to failover
messages but not responding to some client requests, the other
failover peer will take over its client load automatically as the
clients retry.
.PP
It is possible to disable load balancing between peers by setting this
value to 0 on both peers.  Bear in mind that this means both peers will
respond to all DHCPDISCOVERs or DHCPREQUESTs.
.RE
.PP
The
.I auto-partner-down
statement
.RS 0.25i
.PP
.B auto-partner-down \fIseconds\fR\fB;\fR
.PP
This statement instructs the server to initiate a timed delay upon entering
the communications-interrupted state (any situation of being out-of-contact
with the remote failover peer).  At the conclusion of the timer, the server
will automatically enter the partner-down state.  This permits the server
to allocate leases from the partner's free lease pool after an STOS+MCLT
timer expires, which can be dangerous if the partner is in fact operating
at the time (the two servers will give conflicting bindings).
.PP
Think very carefully before enabling this feature.  The partner-down and
communications-interrupted states are intentionally segregated because
there do exist situations where a failover server can fail to communicate
with its peer, but still has the ability to receive and reply to requests
from DHCP clients.  In general, this feature should only be used in those
deployments where the failover servers are directly connected to one
another, such as by a dedicated hardwired link ("a heartbeat cable").
.PP
A zero value disables the auto-partner-down feature (also the default), and
any positive value indicates the time in seconds to wait before automatically
entering partner-down.
.RE
.PP
The Failover pool balance statements.
.RS 0.25i
.PP
 \fBmax-lease-misbalance \fIpercentage\fR\fB;\fR
 \fBmax-lease-ownership \fIpercentage\fR\fB;\fR
 \fBmin-balance \fIseconds\fR\fB;\fR
 \fBmax-balance \fIseconds\fR\fB;\fR
.PP
This version of the DHCP Server evaluates pool balance on a schedule,
rather than on demand as leases are allocated.  The latter approach
proved to be slightly klunky when pool misbalanced reach total
saturation \(em when any server ran out of leases to assign, it also lost
its ability to notice it had run dry.
.PP
In order to understand pool balance, some elements of its operation
first need to be defined.  First, there are \'free\' and \'backup\' leases.
Both of these are referred to as \'free state leases\'.  \'free\' and
\'backup\'
are \'the free states\' for the purpose of this document.  The difference
is that only the primary may allocate from \'free\' leases unless under
special circumstances, and only the secondary may allocate \'backup\' leases.
.PP
When pool balance is performed, the only plausible expectation is to
provide a 50/50 split of the free state leases between the two servers.
This is because no one can predict which server will fail, regardless
of the relative load placed upon the two servers, so giving each server
half the leases gives both servers the same amount of \'failure endurance\'.
Therefore, there is no way to configure any different behaviour, outside of
some very small windows we will describe shortly.
.PP
The first thing calculated on any pool balance run is a value referred to
as \'lts\', or "Leases To Send".  This, simply, is the difference in the
count of free and backup leases, divided by two.  For the secondary,
it is the difference in the backup and free leases, divided by two.
The resulting value is signed: if it is positive, the local server is
expected to hand out leases to retain a 50/50 balance.  If it is negative,
the remote server would need to send leases to balance the pool.  Once
the lts value reaches zero, the pool is perfectly balanced (give or take
one lease in the case of an odd number of total free state leases).
.PP
The current approach is still something of a hybrid of the old approach,
marked by the presence of the \fBmax-lease-misbalance\fR statement.  This
parameter configures what used to be a 10% fixed value in previous versions:
if lts is less than free+backup * \fBmax-lease-misbalance\fR percent, then
the server will skip balancing a given pool (it won't bother moving any
leases, even if some leases "should" be moved).  The meaning of this value
is also somewhat overloaded, however, in that it also governs the estimation
of when to attempt to balance the pool (which may then also be skipped over).
The oldest leases in the free and backup states are examined.  The time
they have resided in their respective queues is used as an estimate to
indicate how much time it is probable it would take before the leases at
the top of the list would be consumed (and thus, how long it would take
to use all leases in that state).  This percentage is directly multiplied
by this time, and fit into the schedule if it falls within
the \fBmin-balance\fR and \fBmax-balance\fR configured values.  The
scheduled pool check time is only moved in a downwards direction, it is
never increased.  Lastly, if the lts is more than double this number in
the negative direction, the local server will \'panic\' and transmit a
Failover protocol POOLREQ message, in the hopes that the remote system
will be woken up into action.
.PP
Once the lts value exceeds the \fBmax-lease-misbalance\fR percentage of
total free state leases as described above, leases are moved to the remote
server.  This is done in two passes.
.PP
In the first pass, only leases whose most recent bound client would have
been served by the remote server - according to the Load Balance Algorithm
(see above \fBsplit\fR and \fBhba\fR configuration statements) - are given
away to the peer.  This first pass will happily continue to give away leases,
decrementing the lts value by one for each, until the lts value has reached
the negative of the total number of leases multiplied by
the \fBmax-lease-ownership\fR percentage.  So it is through this value that
you can permit a small misbalance of the lease pools - for the purpose of
giving the peer more than a 50/50 share of leases in the hopes that their
clients might some day return and be allocated by the peer (operating
normally).  This process is referred to as \'MAC Address Affinity\', but this
is somewhat misnamed: it applies equally to DHCP Client Identifier options.
Note also that affinity is applied to leases when they enter the state
\'free\' from \'expired\' or \'released\'.  In this case also, leases will not
be moved from free to backup if the secondary already has more than its
share.
.PP
The second pass is only entered into if the first pass fails to reduce
the lts underneath the total number of free state leases multiplied by
the \fBmax-lease-ownership\fR percentage.  In this pass, the oldest
leases are given over to the peer without second thought about the Load
Balance Algorithm, and this continues until the lts falls under this
value.  In this way, the local server will also happily keep a small
percentage of the leases that would normally load balance to itself.
.PP
So, the \fBmax-lease-misbalance\fR value acts as a behavioural gate.
Smaller values will cause more leases to transition states to balance
the pools over time, higher values will decrease the amount of change
(but may lead to pool starvation if there's a run on leases).
.PP
The \fBmax-lease-ownership\fR value permits a small (percentage) skew
in the lease balance of a percentage of the total number of free state
leases.
.PP
Finally, the \fBmin-balance\fR and \fBmax-balance\fR make certain that a
scheduled rebalance event happens within a reasonable timeframe (not
to be thrown off by, for example, a 7 year old free lease).
.PP
Plausible values for the percentages lie between 0 and 100, inclusive, but
values over 50 are indistinguishable from one another (once lts exceeds
50% of the free state leases, one server must therefore have 100% of the
leases in its respective free state).  It is recommended to select
a \fBmax-lease-ownership\fR value that is lower than the value selected
for the \fBmax-lease-misbalance\fR value.  \fBmax-lease-ownership\fR
defaults to 10, and \fBmax-lease-misbalance\fR defaults to 15.
.PP
Plausible values for the \fBmin-balance\fR and \fBmax-balance\fR times also
range from 0 to (2^32)-1 (or the limit of your local time_t value), but
default to values 60 and 3600 respectively (to place balance events between
1 minute and 1 hour).
.RE
.SH CLIENT CLASSING
Clients can be separated into classes, and treated differently
depending on what class they are in.  This separation can be done
either with a conditional statement, or with a match statement within
the class declaration.  It is possible to specify a limit on the
total number of clients within a particular class or subclass that may
hold leases at one time, and it is possible to specify automatic
subclassing based on the contents of the client packet.
.PP
Classing support for DHCPv6 clients was added in 4.3.0.  It follows
the same rules as for DHCPv4 except that support for billing classes
has not been added yet.
.PP
To add clients to classes based on conditional evaluation, you can
specify a matching expression in the class statement:
.PP
.nf
class "ras-clients" {
  match if substring (option dhcp-client-identifier, 1, 3) = "RAS";
}
.fi
.PP
Note that whether you use matching expressions or add statements (or
both) to classify clients, you must always write a class declaration
for any class that you use.  If there will be no match statement and
no in-scope statements for a class, the declaration should look like
this:
.PP
.nf
class "ras-clients" {
}
.fi
.SH SUBCLASSES
.PP
In addition to classes, it is possible to declare subclasses.  A
subclass is a class with the same name as a regular class, but with a
specific submatch expression which is hashed for quick matching.
This is essentially a speed hack - the main difference between five
classes with match expressions and one class with five subclasses is
that it will be quicker to find the subclasses.  Subclasses work as
follows:
.PP
.nf
class "allocation-class-1" {
  match pick-first-value (option dhcp-client-identifier, hardware);
}

class "allocation-class-2" {
  match pick-first-value (option dhcp-client-identifier, hardware);
}

subclass "allocation-class-1" 1:8:0:2b:4c:39:ad;
subclass "allocation-class-2" 1:8:0:2b:a9:cc:e3;
subclass "allocation-class-1" 1:0:0:c4:aa:29:44;

subnet 10.0.0.0 netmask 255.255.255.0 {
  pool {
    allow members of "allocation-class-1";
    range 10.0.0.11 10.0.0.50;
  }
  pool {
    allow members of "allocation-class-2";
    range 10.0.0.51 10.0.0.100;
  }
}
.fi
.PP
The data following the class name in the subclass declaration is a
constant value to use in matching the match expression for the class.
When class matching is done, the server will evaluate the match
expression and then look the result up in the hash table.  If it
finds a match, the client is considered a member of both the class and
the subclass.
.PP
Subclasses can be declared with or without scope.  In the above
example, the sole purpose of the subclass is to allow some clients
access to one address pool, while other clients are given access to
the other pool, so these subclasses are declared without scopes.  If
part of the purpose of the subclass were to define different parameter
values for some clients, you might want to declare some subclasses
with scopes.
.PP
In the above example, if you had a single client that needed some
configuration parameters, while most didn't, you might write the
following subclass declaration for that client:
.PP
.nf
subclass "allocation-class-2" 1:08:00:2b:a1:11:31 {
  option root-path "samsara:/var/diskless/alphapc";
  filename "/tftpboot/netbsd.alphapc-diskless";
}
.fi
.PP
In this example, we've used subclassing as a way to control address
allocation on a per-client basis.  However, it's also possible to use
subclassing in ways that are not specific to clients - for example, to
use the value of the vendor-class-identifier option to determine what
values to send in the vendor-encapsulated-options option.  An example
of this is shown under the VENDOR ENCAPSULATED OPTIONS head in the
.B dhcp-options(5)
manual page.
.SH PER-CLASS LIMITS ON DYNAMIC ADDRESS ALLOCATION
.PP
You may specify a limit to the number of clients in a class that can
be assigned leases.  The effect of this will be to make it difficult
for a new client in a class to get an address.  Once a class with
such a limit has reached its limit, the only way a new client in that
class can get a lease is for an existing client to relinquish its
lease, either by letting it expire, or by sending a DHCPRELEASE
packet.  Classes with lease limits are specified as follows:
.PP
.nf
class "limited-1" {
  lease limit 4;
}
.fi
.PP
This will produce a class in which a maximum of four members may hold
a lease at one time.
.SH SPAWNING CLASSES
.PP
It is possible to declare a
.I spawning class\fR.
A spawning class is a class that automatically produces subclasses
based on what the client sends.  The reason that spawning classes
were created was to make it possible to create lease-limited classes
on the fly.  The envisioned application is a cable-modem environment
where the ISP wishes to provide clients at a particular site with more
than one IP address, but does not wish to provide such clients with
their own subnet, nor give them an unlimited number of IP addresses
from the network segment to which they are connected.
.PP
Many cable modem head-end systems can be configured to add a Relay
Agent Information option to DHCP packets when relaying them to the
DHCP server.  These systems typically add a circuit ID or remote ID
option that uniquely identifies the customer site.  To take advantage
of this, you can write a class declaration as follows:
.PP
.nf
class "customer" {
  spawn with option agent.circuit-id;
  lease limit 4;
}
.fi
.PP
Now whenever a request comes in from a customer site, the circuit ID
option will be checked against the class\'s hash table.  If a subclass
is found that matches the circuit ID, the client will be classified in
that subclass and treated accordingly.  If no subclass is found
matching the circuit ID, a new one will be created and logged in the
.B dhcpd.leases
file, and the client will be classified in this new class.  Once the
client has been classified, it will be treated according to the rules
of the class, including, in this case, being subject to the per-site
limit of four leases.
.PP
The use of the subclass spawning mechanism is not restricted to relay
agent options - this particular example is given only because it is a
fairly straightforward one.
.SH COMBINING MATCH, MATCH IF AND SPAWN WITH
.PP
In some cases, it may be useful to use one expression to assign a
client to a particular class, and a second expression to put it into a
subclass of that class.  This can be done by combining the \fBmatch
if\fR and \fBspawn with\fR statements, or the \fBmatch if\fR and
\fBmatch\fR statements.  For example:
.PP
.nf
class "jr-cable-modems" {
  match if option dhcp-vendor-identifier = "jrcm";
  spawn with option agent.circuit-id;
  lease limit 4;
}

class "dv-dsl-modems" {
  match if option dhcp-vendor-identifier = "dvdsl";
  spawn with option agent.circuit-id;
  lease limit 16;
}
.fi
.PP
This allows you to have two classes that both have the same \fBspawn
with\fR expression without getting the clients in the two classes
confused with each other.
.SH DYNAMIC DNS UPDATES
.PP
The DHCP server has the ability to dynamically update the Domain Name
System.  Within the configuration files, you can define how you want
the Domain Name System to be updated.  These updates are RFC 2136
compliant so any DNS server supporting RFC 2136 should be able to
accept updates from the DHCP server.
.PP
There are two DNS schemes implemented.  The interim option is
based on draft revisions of the DDNS documents while the standard
option is based on the RFCs for DHCP-DNS interaction and DHCIDs.
A third option, ad-hoc, was deprecated and has now been removed
from the code base.  The DHCP server must be configured to use
one of the two currently-supported methods, or not to do DNS updates.
.PP
New installations should use the standard option. Older
installations may want to continue using the interim option for
backwards compatibility with the DNS database until the database
can be updated.  This can be done with the
.I ddns-update-style
configuration parameter.
.SH THE DNS UPDATE SCHEME
the interim and standard DNS update schemes operate mostly according
to work from the IETF.  The interim version was based on the drafts
in progress at the time while the standard is based on the completed
RFCs.  The standard RFCs are:
.PP
.nf
.ce 3
RFC 4701 (updated by RF5494)
RFC 4702
RFC 4703
.fi
.PP
And the corresponding drafts were:
.PP
.nf
.ce 3
draft-ietf-dnsext-dhcid-rr-??.txt
draft-ietf-dhc-fqdn-option-??.txt
draft-ietf-dhc-ddns-resolution-??.txt
.fi
.PP
The basic framework for the two schemes is similar with the main
material difference being that a DHCID RR is used in the standard
version while the interim versions uses a TXT RR.  The format
of the TXT record bears a resemblance to the DHCID RR but it is not
equivalent (MD5 vs SHA2, field length differences etc).
.PP
In these two schemes the DHCP server does not necessarily
always update both the A and the PTR records.  The FQDN option
includes a flag which, when sent by the client, indicates that the
client wishes to update its own A record.  In that case, the server
can be configured either to honor the client\'s intentions or ignore
them.  This is done with the statement \fIallow client-updates;\fR or
the statement \fIignore client-updates;\fR.  By default, client
updates are allowed.
.PP
If the server is configured to allow client updates, then if the
client sends a fully-qualified domain name in the FQDN option, the
server will use that name the client sent in the FQDN option to update
the PTR record.  For example, let us say that the client is a visitor
from the "radish.org" domain, whose hostname is "jschmoe".  The
server is for the "example.org" domain.  The DHCP client indicates in
the FQDN option that its FQDN is "jschmoe.radish.org.".  It also
indicates that it wants to update its own A record.  The DHCP server
therefore does not attempt to set up an A record for the client, but
does set up a PTR record for the IP address that it assigns the
client, pointing at jschmoe.radish.org.  Once the DHCP client has an
IP address, it can update its own A record, assuming that the
"radish.org" DNS server will allow it to do so.
.PP
If the server is configured not to allow client updates, or if the
client doesn\'t want to do its own update, the server will simply
choose a name for the client. By default, the server will choose
from the following three values:
.PP
     1. \fBfqdn\fR option (if present)
     2. hostname option (if present)
     3. Configured hostname option (if defined).
.PP
If these defaults for choosing the host name are not appropriate
you can write your own statement to set the ddns-hostname variable
as you wish.  If none of the above are found the server will use
the host declaration name (if one) and use-host-decl-names is on.
.PP
It will use its own domain name for the client.  It will then update
both the A and PTR record, using the name that it chose for the client.
If the client sends a fully-qualified domain name in the \fBfqdn\fR option,
the server uses only the leftmost part of the domain name - in the example
above, "jschmoe" instead of "jschmoe.radish.org".
.PP
Further, if the \fIignore client-updates;\fR directive is used, then
the server will in addition send a response in the DHCP packet, using
the FQDN Option, that implies to the client that it should perform its
own updates if it chooses to do so.  With \fIdeny client-updates;\fR, a
response is sent which indicates the client may not perform updates.
.PP
Both the standard and interim options also include a method to
allow more than one DHCP server to update the DNS database without
accidentally deleting A records that shouldn\'t be deleted nor failing
to add A records that should be added.  For the standard option the
method works as follows:
.PP
When the DHCP server issues a client a new lease, it creates a text
string that is an SHA hash over the DHCP client\'s identification (see
RFCs 4701 & 4702 for details).  The update attempts to add an A
record with the name the server chose and a DHCID record containing the
hashed identifier string (hashid).  If this update succeeds, the
server is done.
.PP
If the update fails because the A record already exists, then the DHCP
server attempts to add the A record with the prerequisite that there
must be a DHCID record in the same name as the new A record, and that
DHCID record\'s contents must be equal to hashid.  If this update
succeeds, then the client has its A record and PTR record.  If it