dhcp-options(5)

📦
version 4.4.3.P1 · 6 versions
, in arch rolling, debian bookworm, debian trixie, ubuntu noble, ubuntu resolute

💖 dhcp-options(5) 💖 — like, the whole entire tea on DHCP options

okay bestie so this is the section 5 file formats manual page for dhcp-options, and it's a big one, so grab a snack 🍓

✨ NAME ✨

dhcp-options — Dynamic Host Configuration Protocol options 🦋

🌸 DESCRIPTION 🌸

so basically the Dynamic Host Configuration protocol lets the client receive options from the DHCP server that describe the network configuration and all the various services chilling on the network. and when you're setting up dhcpd(8) or dhclient(8), you're gonna have to declare options like, a lot. so this page is where you find the syntax for declaring options, plus the names and formats of every option you can declare. periodt 💅

🔥 REFERENCE: OPTION STATEMENTS 🔥

DHCP option statements always start with the option keyword, then an option name, then the option data. names and data formats are all described below. and ngl, you do not have to exhaustively specify every single DHCP option — only the ones your clients actually need, bestie. no cap. ✨

Option data comes in a bunch of formats, defined below:

  • The ip-address data type — you can enter this either as an explicit IP address (e.g., 239.254.197.10) or as a domain name (e.g., haagen.isc.org). When you're entering a domain name, make sure that domain name resolves to a single IP address, fr fr. Also please note: the software (dhcpd or dhclient) will only attempt to resolve the domain name the first time the option is needed. For example, if the next-server option is defined as a domain name, dhcpd will attempt to resolve it while responding to the first client query dhcpd receives after startup. Should the domain's address subsequently change, the software has to be restarted in order to pick up the change. 😭
  • The ip6-address data specifies an IPv6 address, like ::1 or 3ffe:bbbb:aaaa:aaaa::1. 🌈
  • The int32 data type specifies a signed 32-bit integer. The uint32 data type specifies an unsigned 32-bit integer. The int16 and uint16 data types specify signed and unsigned 16-bit integers. The int8 and uint8 data types specify signed and unsigned 8-bit integers. Unsigned 8-bit integers are also sometimes referred to as octets. 🔢
  • The text data type specifies an NVT ASCII string, which must be enclosed in double quotes — for example, to specify a root-path option, the syntax would be
option root-path "10.0.1.4:/var/tmp/rootfs";
  • The domain-name data type specifies a domain name, which must not be enclosed in double quotes. The domain name is stored just as if it were a text option.
  • The domain-list data type specifies a list of domain names, enclosed in double quotes and separated by commas ("example.com", "foo.example.com"). 🌐
  • The flag data type specifies a boolean value. Booleans can be either true or false (or on or off, if that makes more sense to you 💜).
  • The string data type specifies either an NVT ASCII string enclosed in double quotes, or a series of octets specified in hexadecimal, separated by colons. For example:
  option dhcp-client-identifier "CLIENT-FOO";

or

  option dhcp-client-identifier 43:4c:49:45:54:2d:46:4f:4f;

🧚 SETTING OPTION VALUES USING EXPRESSIONS 🧚

sometimes it's genuinely helpful to be able to set the value of a DHCP option based on some value that the client has sent. to do this, you can use expression evaluation. the dhcp-eval(5) manual page describes how to write expressions. to assign the result of an evaluation to an option, define the option like this:

  option my-option = expression ;

For example:

  option hostname = binary-to-ascii (16, 8, "-",
                                     substring (hardware, 1, 6));

slay. 🎀

💅 INCLUDING OPTION DEFINITIONS 💅

starting with 4.3.0, when ISC adds new option definitions, those definitions will be included in the code based on the definition of an argument for the RFC that defines the option in includes/site.h. this gives you a method for over-riding the ISC definitions if you need to — for example if you have previously defined the option with a different format using the mechanism from DEFINING NEW OPTIONS below. 🫶

By default all of the options are enabled. In order to disable an option you would edit the includes/site.h file and comment out the definition for the proper RFC.

👑 STANDARD DHCPV4 OPTIONS 👑

the documentation for the various options mentioned below is taken from the latest IETF draft document on DHCP options. options not listed below may not yet be implemented, but it is possible to use such options by defining them in the configuration file — please see the DEFINING NEW OPTIONS heading later in this document for more info. 🔍

Some of the options documented here are automatically generated by the DHCP server or by clients, and cannot be configured by the user. The value of such an option can be used in the configuration file of the receiving DHCP protocol agent (server or client), for example in conditional expressions. However, the value of the option cannot be used in the configuration file of the sending agent, because the value is determined only after the configuration file has been processed. In the following documentation, such options will be shown as "not user configurable".

The standard options are:

  • option all-subnets-local flag; — this option specifies whether or not the client may assume that all subnets of the IP network to which the client is connected use the same MTU as the subnet of that network to which the client is directly connected. A value of true indicates that all subnets share the same MTU. A value of false means that the client should assume that some subnets of the directly connected network may have smaller MTUs. 🤔
  • option arp-cache-timeout uint32; — this option specifies the timeout in seconds for ARP cache entries. ⏳
  • option associated-ip ip-address [, ip-address... ]; — this option is part of lease query. It is used to return all of the IP addresses associated with a given DHCP client.

This option is not user configurable.

  • option bcms-controller-address ip-address [, ip-address... ]; — this option configures a list of IPv4 addresses for use as Broadcast and Multicast Controller Servers ("BCMS"). 📡
  • option bcms-controller-names domain-list; — this option contains the domain names of local Broadcast and Multicast Controller Servers ("BCMS") controllers which the client may use.
  • option bootfile-name text; — this option is used to identify a bootstrap file. If supported by the client, it should have the same effect as the filename declaration. BOOTP clients are unlikely to support this option. Some DHCP clients will support it, and others actually require it. 😤
  • option boot-size uint16; — this option specifies the length in 512-octet blocks of the default boot image for the client. 📦
  • option broadcast-address ip-address; — this option specifies the broadcast address in use on the client's subnet. Legal values for broadcast addresses are specified in section 3.2.1.3 of STD 3 (RFC1122). 📢
  • option capwap-ac-v4 ip-address [, ip-address ... ] ; — a list of IPv4 addresses of CAPWAP ACs that the WTP may use. The addresses are listed in preference order.

This option is included based on RFC 5417.

  • option client-last-transaction-time uint32; — this option is part of lease query. It allows the receiver to determine the time of the most recent access by the client. The value is a duration in seconds from when the client last communicated with the DHCP server.

This option is not user configurable.

  • option cookie-servers ip-address [, ip-address... ]; — the cookie server option specifies a list of RFC 865 cookie servers available to the client. Servers should be listed in order of preference. 🍪
  • option default-ip-ttl uint8; — this option specifies the default time-to-live that the client should use on outgoing datagrams.
  • option default-tcp-ttl uint8; — this option specifies the default TTL that the client should use when sending TCP segments. The minimum value is 1. ⭐
  • option default-url string; — the format and meaning of this option is not described in any standards document, but is claimed to be in use by Apple Computer. It is not known what clients may reasonably do if supplied with this option. Use at your own risk. 💀
  • option dhcp-client-identifier string; — this option can be used to specify a DHCP client identifier in a host declaration, so that dhcpd can find the host record by matching against the client identifier.

Please be aware that some DHCP clients, when configured with client identifiers that are ASCII text, will prepend a zero to the ASCII text. So you may need to write:

       option dhcp-client-identifier "\0foo";

rather than:

       option dhcp-client-identifier "foo";
  • option dhcp-lease-time uint32; — this option is used in a client request (DHCPDISCOVER or DHCPREQUEST) to allow the client to request a lease time for the IP address. In a server reply (DHCPOFFER), a DHCP server uses this option to specify the lease time it is willing to offer.

This option is not directly user configurable in the server; refer to the max-lease-time and default-lease-time server options in dhcpd.conf(5). 🕐

  • option dhcp-max-message-size uint16; — this option, when sent by the client, specifies the maximum size of any response that the server sends to the client. When specified on the server, if the client did not send a dhcp-max-message-size option, the size specified on the server is used. This works for BOOTP as well as DHCP responses. 📏
  • option dhcp-message text; — this option is used by a DHCP server to provide an error message to a DHCP client in a DHCPNAK message in the event of a failure. A client may use this option in a DHCPDECLINE message to indicate why the client declined the offered parameters.

This option is not user configurable.

  • option dhcp-message-type uint8; — this option, sent by both client and server, specifies the type of DHCP message contained in the DHCP packet. Possible values (taken directly from RFC2132) are:
    1. DHCPDISCOVER
    2. DHCPOFFER
    3. DHCPREQUEST
    4. DHCPDECLINE
    5. DHCPACK
    6. DHCPNAK
    7. DHCPRELEASE
    8. DHCPINFORM

This option is not user configurable.

  • option dhcp-option-overload uint8; — this option is used to indicate that the DHCP ´sname´ or ´file´ fields are being overloaded by using them to carry DHCP options. A DHCP server inserts this option if the returned parameters will exceed the usual space allotted for options.

If this option is present, the client interprets the specified additional fields after it concludes interpretation of the standard option fields.

Legal values for this option are:

  1. 1 — the file field is used to hold options
  2. 2 — the sname field is used to hold options
  3. 3 — both fields are used to hold options

This option is not user configurable.

  • option dhcp-parameter-request-list uint8 [, uint8... ];

okay so this is the ✨wishlist✨ option. When the client sends it, it's the client literally going "hey server, here's which options I want you to return." Normally, in the ISC DHCP client, this is done using the request statement. If the client doesn't specify this option, the DHCP server will normally return every option that is valid in scope and that fits into the reply. And when it's specified on the server? The server returns the specified options. This can be used to force a client to take options that it hasn't requested, and it can also be used to tailor the response of the DHCP server for clients that may need a more limited set of options than those the server would normally return. bespoke service, honestly 💅

  • option dhcp-rebinding-time uint32;

This one sets the number of seconds from the time a client gets an address until the client transitions to the REBINDING state.

It's user configurable, but it will be ignored if the value is greater than or equal to the lease time. 🙈

To make DHCPv4+DHCPv6 migration easier in the future, any value configured in this option is also used as a DHCPv6 "T1" (renew) time.

  • option dhcp-renewal-time uint32;

Same energy, different state: the number of seconds from the time a client gets an address until the client transitions to the RENEWING state.

User configurable, but it will be ignored if the value is greater than or equal to the rebinding time, or lease time. 🫠

To make DHCPv4+DHCPv6 migration easier in the future, any value configured in this option is also used as a DHCPv6 "T2" (rebind) time.

  • option dhcp-requested-address ip-address;

The client uses this one inside a DHCPDISCOVER to request that a particular IP address be assigned. main character behavior, honestly.

This option is not user configurable.

  • option dhcp-server-identifier ip-address;

Used in DHCPOFFER and DHCPREQUEST messages, and it may optionally be included in the DHCPACK and DHCPNAK messages too. DHCP servers include this option in the DHCPOFFER so the client can distinguish between lease offers. DHCP clients use the contents of the server identifier field as the destination address for any DHCP messages unicast to the DHCP server. Clients also indicate which of several lease offers is being accepted by including this option in a DHCPREQUEST message.

The value of this option is the IP address of the server.

This option is not directly user configurable. See the server-identifier server option in dhcpd.conf(5).

  • option domain-name text;

Specifies the domain name that client should use when resolving hostnames via the Domain Name System. 🦋

  • option domain-name-servers ip-address [, ip-address... ];

The domain-name-servers option specifies a list of Domain Name System (STD 13, RFC 1035) name servers available to the client. Servers should be listed in order of preference. order matters, bestie.

  • option domain-search domain-list;

The domain-search option specifies a search list of Domain Names to be used by the client to locate not-fully-qualified domain names. The difference between this option and historic use of the domain-name option for the same ends is that this option is encoded in RFC1035 compressed labels on the wire. For example:

option domain-search "example.com", "sales.example.com",
                     "eng.example.com";
  • option extensions-path text;

This option specifies the name of a file containing additional options to be interpreted according to the DHCP option format as specified in RFC2132. 📄

  • option finger-server ip-address [, ip-address... ];

The Finger server option specifies a list of Finger servers available to the client. Servers should be listed in order of preference.

  • option font-servers ip-address [, ip-address... ];

This option specifies a list of X Window System Font servers available to the client. Servers should be listed in order of preference.

  • option geoconf-civic string;

A string to hold the geoconf civic structure. 🗺️

This option is included based on RFC 4776.

  • option host-name string;

This option specifies the name of the client. The name may or may not be qualified with the local domain name (it is preferable to use the domain-name option to specify the domain name). See RFC 1035 for character set restrictions. Heads up: this option is only honored by dhclient-script(8) if the hostname for the client machine is not set.

  • option ieee802-3-encapsulation flag;

This option specifies whether or not the client should use Ethernet Version 2 (RFC 894) or IEEE 802.3 (RFC 1042) encapsulation if the interface is an Ethernet. A value of false indicates that the client should use RFC 894 encapsulation. A value of true means that the client should use RFC 1042 encapsulation. 🎀

  • option ien116-name-servers ip-address [, ip-address... ];

The ien116-name-servers option specifies a list of IEN 116 name servers available to the client. Servers should be listed in order of preference.

  • option impress-servers ip-address [, ip-address... ];

The impress-server option specifies a list of Imagen Impress servers available to the client. Servers should be listed in order of preference.

  • option interface-mtu uint16;

This option specifies the MTU to use on this interface. The minimum legal value for the MTU is 68. don't go lower than 68, fr fr.

  • option ip-forwarding flag;

This option specifies whether the client should configure its IP layer for packet forwarding. A value of false means disable IP forwarding, and a value of true means enable IP forwarding.

  • option irc-server ip-address [, ip-address... ];

The IRC server option specifies a list of IRC servers available to the client. Servers should be listed in order of preference. 💬

  • option loader-configfile text

This option is used to specify a boot loading configuration file a PXE client should use.

This option is included based on RFC 5071.

  • option loader-pathprefix text

This option is used to specify a path prefix a PXE client should use in conjunction with the boot load configuration file.

This option is included based on RFC 5071.

  • option loader-reboottime uint32

This option is used to dictate the maximum amount of time a PXE client should allow itself to achieve configured network resources before rebooting. clock's ticking ⏳

This option is included based on RFC 5071.

  • option log-servers ip-address [, ip-address... ];

The log-server option specifies a list of MIT-LCS UDP log servers available to the client. Servers should be listed in order of preference.

  • option lpr-servers ip-address [, ip-address... ];

The LPR server option specifies a list of RFC 1179 line printer servers available to the client. Servers should be listed in order of preference. 🖨️

  • option mask-supplier flag;

This option specifies whether or not the client should respond to subnet mask requests using ICMP. A value of false indicates that the client should not respond. A value of true means that the client should respond.

  • option max-dgram-reassembly uint16;

This option specifies the maximum size datagram that the client should be prepared to reassemble. The minimum legal value is 576.

  • option merit-dump text;

This option specifies the path-name of a file to which the client's core image should be dumped in the event the client crashes. The path is formatted as a character string consisting of characters from the NVT ASCII character set. 💀 (crash dump address, iconic)

  • option mobile-ip-home-agent ip-address [, ip-address... ];

This option specifies a list of IP addresses indicating mobile IP home agents available to the client. Agents should be listed in order of preference, although normally there will be only one such agent.

  • option name-service-search uint16 [, uint6... ];

This option specifies a list of name services in the order the client should attempt to use them.

This option is included based on RFC 2937.

  • option nds-context string;

The nds-context option specifies the name of the initial Netware Directory Service for an NDS client.

  • option nds-servers ip-address [, ip-address... ];

The nds-servers option specifies a list of IP addresses of NDS servers.

  • option nds-tree-name string;

The nds-tree-name option specifies NDS tree name that the NDS client should use. 🌳

  • option netbios-dd-server ip-address [, ip-address... ];

The NetBIOS datagram distribution server (NBDD) option specifies a list of RFC 1001/1002 NBDD servers listed in order of preference.

  • option netbios-name-servers ip-address [, ip-address...];

The NetBIOS name server (NBNS) option specifies a list of RFC 1001/1002 NBNS name servers listed in order of preference. NetBIOS Name Service is currently more commonly referred to as WINS. WINS servers can be specified using the netbios-name-servers option. 👑

  • option netbios-node-type uint8;

The NetBIOS node type option allows NetBIOS over TCP/IP clients which are configurable to be configured as described in RFC 1001/1002. The value is specified as a single octet which identifies the client type.

Possible node types are:

  1. 1 — B-node: Broadcast - no WINS
  2. 2 — P-node: Peer - WINS only
  3. 4 — M-node: Mixed - broadcast, then WINS
  4. 8 — H-node: Hybrid - WINS, then broadcast
  • option netbios-scope string;

The NetBIOS scope option specifies the NetBIOS over TCP/IP scope parameter for the client as specified in RFC 1001/1002. See RFC1001, RFC1002, and RFC1035 for character-set restrictions.

  • option netinfo-server-address ip-address [, ip-address... ];
⚠️ The netinfo-server-address option has not been described in any RFC, but has been allocated (and is claimed to be in use) by Apple Computers. It's hard to say if the above is the correct format, or what clients might be expected to do if values were configured. Use at your own risk.
  • option netinfo-server-tag text;
⚠️ The netinfo-server-tag option has not been described in any RFC, but has been allocated (and is claimed to be in use) by Apple Computers. It's hard to say if the above is the correct format, or what clients might be expected to do if values were configured. Use at your own risk.
  • option nis-domain text;

(the glow-up explanation for this one is right around the corner 🫶)

  • option nis-domain text; — this one tells the client what its NIS (Sun Network Information Services) domain is. It's formatted as a character string, and every character in it comes straight from the NVT ASCII character set. clean, simple, iconic 💅
  • option nis-servers ip-address [, ip-address... ]; — a whole list of IP addresses showing which NIS servers the client can hit up. Bestie, order is not a vibe suggestion here: servers should be listed in order of preference.
  • option nisplus-domain text; — same energy as the NIS one but for the client's NIS+ domain. It's a character string made of characters from the NVT ASCII character set. NIS but make it plus ✨
  • option nisplus-servers ip-address [, ip-address... ]; — a list of IP addresses for the NIS+ servers available to the client, again in order of preference. Most preferred first, periodt.
  • option nntp-server ip-address [, ip-address... ]; — the NNTP server option drops a list of NNTP servers the client can use. And yep, you guessed it: servers should be listed in order of preference. 🗞️
  • option non-local-source-routing flag; — this one decides whether the client configures its IP layer to allow forwarding of datagrams with non-local source routes (peep Section 3.3.5 of [4] if you want the whole discussion). false = disallow forwarding of those datagrams. true = allow forwarding. No in-between, no maybe.
  • option ntp-servers ip-address [, ip-address... ]; — a list of IP addresses for NTP (RFC 5905) servers available to the client. List them in order of preference — your clocks will thank you ⏰
  • option nwip-domain string; — the name of the NetWare/IP domain that a NetWare/IP client should use. just the domain name, nothing fancy
  • option nwip-suboptions string; — a whole sequence of suboptions for NetWare/IP clients — RFC2242 has the details if you're nosy. Normally you don't set this raw string directly; you set specific NetWare/IP suboptions and it works out. Scroll to the NETWARE/IP SUBOPTIONS section for the tea.
  • option pxe-system-type uint16 [, uint16 ... ]; — a list of one or more 16-bit integers, letting a client specify its pre-boot architecture type(s). Included based on RFC 4578. 🥾
  • option pxe-interface-id uint8 uint8 uint8 — a three octet value that lets a client specify its network interface type. Included based on RFC 4578.
  • option pxe-client-id uint8 string — a single octet indicating the type, followed by a string that lets a client specify its PXE client identity. Included based on RFC 4578. PXE girlies rise 🙈
  • option option-6rd uint8 uint8 ip6-address ip-address [, ip-address ...]; — all the rapid deployment info lives here: 8 bits of ipv4 mask length, 8 bits of 6rd prefix length, an ipv6 prefix as an ipv6 address, and a list of one or more ipv4 addresses. Included based on RFC 5969. 6rd said let me cook 🔥
  • option pana-agent ip-address [, ip-address ... ] ; — a set of IPv4 addresses of a PAA for the client to use, listed in preferred order. Included based on RFC 5192.
  • option path-mtu-aging-timeout uint32; — the timeout (in seconds) to use when aging Path MTU values discovered by the mechanism from RFC 1191. it's a uint32, so go wild (responsibly)
  • option path-mtu-plateau-table uint16 [, uint16... ]; — a table of MTU sizes to use while doing Path MTU Discovery as defined in RFC 1191. It's formatted as a list of 16-bit unsigned integers, ordered from smallest to largest. Heads up: the minimum MTU value cannot be smaller than 68. That's the floor, bestie.
  • option pcode text; — a string suitable for the TZ variable. Included based on RFC 4833. 🌍
  • option perform-mask-discovery flag; — whether or not the client should do subnet mask discovery using ICMP. false = nope, no mask discovery. true = yes, go discover that mask.
  • option policy-filter ip-address ip-address [, ip-address ip-address...]; — policy filters for non-local source routing. The filters are a list of IP addresses and masks that specify destination/mask pairs for filtering incoming source routes. any source routed datagram whose next-hop address doesn't match one of the filters should be discarded by the client. See STD 3 (RFC1122) for more. 🚫
  • option pop-server ip-address [, ip-address... ]; — the POP3 server option lists POP3 servers available to the client, in order of preference. 📬
  • option rdnss-selection uint8 ip-address ip-address domain-name; — this one specifies an 8 bit flags field, a primary and secondary ip address for the name server, plus a domainlist of domains for which the RDNSS has special knowledge. Included based on RFC 6731. flags field, primary, secondary, domains — a whole package deal 🎀
  • option resource-location-servers ip-address [, ip-address...]; — a list of RFC 887 Resource Location servers available to the client, in order of preference.
  • option root-path text; — the path-name that contains the client's root disk. Formatted as a character string built from NVT ASCII characters. point them home 🏠
  • option router-discovery flag; — whether or not the client should solicit routers using the Router Discovery mechanism defined in RFC 1256. false = don't do router discovery. true = do it.
  • option router-solicitation-address ip-address; — the address the client should transmit router solicitation requests to. 📡
  • option routers ip-address [, ip-address... ]; — a list of IP addresses for routers on the client's subnet, listed in order of preference. The classic. The main character of DHCP options, honestly 👑
  • option slp-directory-agent boolean ip-address [, ip-address... ]; — this does two things at once: it gives the IP addresses of one or more Service Location Protocol Directory Agents, and it says whether using those addresses is mandatory. If the initial boolean is true, the SLP agent should just use the IP addresses given. If it's false, the SLP agent may additionally do active or passive multicast discovery of SLP agents (RFC2165 has the details).

Quick note, bestie: in this option and in slp-service-scope, "SLP Agent" means a Service Location Protocol agent running on a machine that's being configured using the DHCP protocol. Not a random agent, that specific one.

Also — some companies call SLP "NDS". So if you've got an NDS directory agent whose address you need to configure, the slp-directory-agent option should work. naming drama, classic 💀

  • option slp-service-scope boolean text; — the Service Location Protocol Service Scope Option specifies two things: a list of service scopes for SLP, and whether using that list is mandatory. If the initial boolean is true, the SLP agent should only use the list of scopes provided in this option. Otherwise, it may use its own static configuration in preference to the list provided here.

The text string should be a comma-separated list of scopes the SLP agent should use. It can be omitted, and in that case the SLP Agent will use the aggregated list of scopes of all directory agents it knows about. commas, bestie, commas 🦋

  • option smtp-server ip-address [, ip-address... ]; — a list of SMTP servers available to the client, in order of preference. 📧
  • option static-routes ip-address ip-address [, ip-address ip-address...]; — a list of static routes the client should install in its routing cache. If multiple routes to the same destination are specified, they're listed in descending order of priority. The routes are a list of IP address pairs: first address = the destination address, second address = the router for that destination.
Real talk: the default route (0.0.0.0) is an illegal destination for a static route. If you want the default route, use the routers option instead. Also please know this option is not intended for classless IP routing — it doesn't include a subnet mask. Since classless IP routing is now the most widely deployed routing standard, this option is virtually useless, and it isn't implemented by any of the popular DHCP clients, for example the Microsoft DHCP client. it's giving legacy 😭
  • option streettalk-directory-assistance-server ip-address [, ip-address...]; — the StreetTalk Directory Assistance (STDA) server option lists STDA servers available to the client, in order of preference.
  • option streettalk-server ip-address [, ip-address... ]; — the StreetTalk server option lists StreetTalk servers available to the client, in order of preference. Same rules, different name.
  • option subnet-mask ip-address; — specifies the client's subnet mask as per RFC 950. Now listen closely: if no subnet mask option is provided anywhere in scope, as a last resort dhcpd will use the subnet mask from the subnet declaration for the network on which an address is being assigned. However — any subnet-mask option declaration that's in scope for the address being assigned will override the subnet mask specified in the subnet declaration. Scope wins, every time. 💅
  • option subnet-selection ip-address; — sent by the client when an address is required in a subnet other than the one that would normally be selected (based on the relaying address of the connected subnet the request is obtained from). See RFC3011.
Heads up: the option number used by this server is 118. That hasn't always been the defined number, and some clients may use a different value. Use of this option should be regarded as slightly experimental! living on the edge, lowkey 💀

This option is not user configurable in the server.

  • option swap-server ip-address; — the IP address of the client's swap server. That's it. short king 👑
  • option tftp-server-address ip-address [, ip-address... ]; — configures a list of one or more IPv4 addresses of tftp servers a client may use. This option is included based on RFC 5859
  • option tcp-keepalive-garbage flag;

So this next bit is all about that keepalive garbage octet situation 💀 — it decides whether or not the client should send TCP keepalive messages with an octet of garbage, purely for compatibility with older implementations. A value of false means nah, no garbage octet. A value of true means yep, send the garbage octet. iconic behaviour honestly.

  • option tcp-keepalive-interval uint32; — this one sets the interval (in seconds) that the client TCP should wait before sending a keepalive message on a TCP connection. It's a 32-bit unsigned integer, no cap. And a value of zero means the client should not generate keepalive messages on connections unless an application specifically asks for them. periodt.
  • option tcode text; — specifies a name of a zone entry in the TZ database. 🕰️

This option is included based on RFC 4833.

  • option tftp-server-name text; — used to identify a TFTP server, and if the client supports it, it should have the same effect as the server-name declaration. Heads up though: BOOTP clients are unlikely to support this option. Some DHCP clients will support it, and others actually require it. 🙈
  • option time-offset int32; — the time-offset option specifies the offset of the client's subnet in seconds from Coordinated Universal Time (UTC). 🦋
  • option time-servers ip-address [, ip-address... ]; — specifies a list of RFC 868 time servers available to the client. Servers should be listed in order of preference, bestie, don't just throw them in there.
  • option trailer-encapsulation flag; — specifies whether or not the client should negotiate the use of trailers (RFC 893 [14]) when using the ARP protocol. A value of false means the client should not attempt to use trailers. A value of true means the client should attempt to use trailers. that's it that's the whole vibe.
  • option uap-servers text; — specifies a list of URLs, each one pointing to a user authentication service capable of processing authentication requests encapsulated in the User Authentication Protocol (UAP). UAP servers can accept either HTTP 1.1 or SSLv3 connections. If the list includes a URL that doesn't contain a port component, the normal default port is assumed (i.e., port 80 for http and port 443 for https). If the list includes a URL that doesn't contain a path component, the path /uap is assumed. And if more than one URL is specified in this list, the URLs are separated by spaces.
  • option user-class string; — some DHCP clients use this as a way for users to specify identifying information to the client. It's kinda like the vendor-class-identifier option, except the value here is specified by the user, not the vendor. Most recent DHCP clients have a way in the user interface to specify the value for this identifier, usually as a text string. 💅
  • option v4-access-domain domain-name; — the domain name associated with the access network, for use with LIS Discovery.

This option is included based on RFC 5986.

  • option v4-lost domain-name; — the domain name of the LoST server for the client to use.

This option is included based on RFC 5223.

  • option v6-only-preferred uint32; — the number of seconds the client should disable DHCPv4 for. no cap, it just goes v6-only for a bit.

This option and its use by the client are specified in RFC 8925.

  • option vendor-class-identifier string; — some DHCP clients use this to identify the vendor type and possibly the configuration of a DHCP client. The info is a string of bytes whose contents are specific to the vendor and are not specified in a standard. To see what vendor class identifier clients are sending, you can write the following in your DHCP server configuration file:
set vendor-string = option vendor-class-identifier;

This will result in all entries in the DHCP server lease database file for clients that sent vendor-class-identifier options having a set statement that looks something like this:

set vendor-string = "SUNW.Ultra-5_10";

The vendor-class-identifier option is normally used by the DHCP server to determine the options that are returned in the vendor-encapsulated-options option. Please see the VENDOR ENCAPSULATED OPTIONS section later in this manual page for further information. 👑

  • option vendor-encapsulated-options string; — this option can contain either a single vendor-specific value or one or more vendor-specific suboptions. It is not normally specified in the DHCP server configuration file — instead, a vendor class is defined for each vendor, vendor class suboptions are defined, values for those suboptions are defined, and the DHCP server makes up a response on that basis.

Some default behaviours for well-known DHCP client vendors (currently, the Microsoft Windows 2000 DHCP client) are configured automatically, but otherwise this must be configured manually — see the VENDOR ENCAPSULATED OPTIONS section later in this manual page for details.

  • option vivso string; — the vivso option can contain multiple separate options, one for each 32-bit Enterprise ID. Each Enterprise-ID discriminated option then contains additional options whose format is defined by the vendor who holds that ID. This option is usually not configured manually, but rather is configured via intervening option definitions. Please also see the VENDOR ENCAPSULATED OPTIONS section later in this manual page for details. 🧚
  • option www-server ip-address [, ip-address... ]; — specifies a list of WWW servers available to the client. Servers should be listed in order of preference. 🌈
  • option x-display-manager ip-address [, ip-address... ]; — specifies a list of systems that are running the X Window System Display Manager and are available to the client. Addresses should be listed in order of preference. main character energy for your display manager fr fr.

🌸 RELAY AGENT INFORMATION OPTION 🌸

okay so there's an IETF draft, draft-ietf-dhc-agent-options-11.txt, that defines a whole series of encapsulated options a relay agent can add to a DHCP packet when it's relaying that packet to the DHCP server. The server can then make address allocation decisions (or literally whatever other decisions it wants, it's her world) based on these options. And the server also returns these options in any replies it sends through the relay agent, so the relay agent can use the info in these options for delivery or accounting purposes. team work bestie 🫶

The current draft defines two options. To reference these options in the dhcp server, specify the option space name, "agent", followed by a period, followed by the option name. It is not normally useful to define values for these options in the server, although it is permissible. And heads up — these options are not supported in the client. 💀

  • option agent.circuit-id string; — the circuit-id suboption encodes an agent-local identifier of the circuit from which a DHCP client-to-server packet was received. It's intended for use by agents in relaying DHCP responses back to the proper circuit. The format of this option is currently defined to be vendor-dependent, and will probably remain that way, although the current draft allows for the possibility of standardizing the format in the future. so like, watch this space.
  • option agent.remote-id string; — the remote-id suboption encodes information about the remote host end of a circuit. Examples of what it might contain include caller ID information, username information, remote ATM address, cable modem ID, and similar things. In principal, the meaning is not well-specified, and it should generally be assumed to be an opaque object that is administratively guaranteed to be unique to a particular remote end of a circuit. 🔍
  • option agent.DOCSIS-device-class uint32; — the DOCSIS-device-class suboption is intended to convey information about the host endpoint, hardware, and software that either the host operating system or the DHCP server may not otherwise be aware of (but the relay is able to distinguish). This is implemented as a 32-bit field (4 octets), each bit representing a flag describing the host in one of these ways. So far, only bit zero (being the least significant bit) is defined in RFC3256. If this bit is set to one, the host is considered a CPE Controlled Cable Modem (CCCM). All other bits are reserved. ⭐
  • option agent.link-selection ip-address; — the link-selection suboption is provided by relay agents to inform servers what subnet the client is actually attached to. This is useful in those cases where the giaddr (where responses must be sent to the relay agent) is not on the same subnet as the client. When this option is present in a packet from a relay agent, the DHCP server will use its contents to find a subnet declared in configuration, and from here take one step further backwards to any shared-network the subnet may be defined within; the client may be given any address within that shared network, as normally appropriate. lowkey genius ngl 💜

🌸 THE CLIENT FQDN SUBOPTIONS 🌸

okay bestie, so the Client FQDN option lives in the Internet Draft draft-ietf-dhc-fqdn-option-00.txt, and it is not a standard yet 💀 but it's in sufficiently wide use already that we went ahead and implemented it anyway. And because the option format is genuinely complex, we implemented it as a suboption space rather than one single option. In general you should not be configuring this one yourself — instead it should be used as part of an automatic DNS update system. let the robots handle it, fr fr. 🤖

  • option fqdn.no-client-update flag; — when the client sends this and it's true, it means the client will not attempt to update its A record. When the server sends it to the client, it means the client should not update its own A record.
  • option fqdn.server-update flag; — when the client sends this to the server, it's requesting that the server update its A record. When the server sends it, it means the server has updated (or is about to update) the client's A record.
  • option fqdn.encoded flag; — if true, this says the domain name inside the option is encoded in DNS wire format, not as plain ASCII text. The client normally sets this to false if it doesn't support DNS wire format in the FQDN option. The server should always send back the same value the client sent. When this value is set on the configuration side, it controls the format in which the fqdn.fqdn suboption is encoded.
  • option fqdn.rcode1 flag;
  • option fqdn.rcode2 flag;

these two specify the result of the updates of the A and PTR records, respectively 🫶 and they are only sent by the DHCP server to the DHCP client. Their values are the ones defined in the DNS protocol specification.

  • option fqdn.fqdn text; — specifies the domain name the client wishes to use. This can be a fully-qualified domain name, or a single label. If there's no trailing . character in the name, it is not fully-qualified, and the server will generally update that name in some locally-defined domain.
  • option fqdn.hostname --never set--; — this one should never be set, but it can be read back using the option and config-option operators in an expression, in which case it returns the first label in the fqdn.fqdn suboption — for example, if the value of fqdn.fqdn is "foo.example.com.", then fqdn.hostname will be "foo".
  • option fqdn.domainname --never set--; — also should never be set, but it can be read back using the option and config-option operators in an expression, in which case it returns all labels after the first label in the fqdn.fqdn suboption — for example, if the value of fqdn.fqdn is "foo.example.com.", then fqdn.domainname will be "example.com.". If this suboption value is not set, it means an unqualified name was sent in the fqdn option, or that no fqdn option was sent at all.
If you wish to use any of these suboptions, we strongly recommend that you refer to the Client FQDN option draft (or standard, when it becomes a standard) — the documentation here is sketchy and incomplete in comparison, and is just intended for reference by people who already understand the Client FQDN option specification. 💅

🍓 THE NETWARE/IP SUBOPTIONS 🍓

RFC2242 defines a whole set of encapsulated options for Novell NetWare/IP clients. To use these options in the dhcp server, specify the option space name "nwip", followed by a period, followed by the option name. Here's the lineup:

  • option nwip.nsq-broadcast flag; — if true, the client should use the NetWare Nearest Server Query to locate a NetWare/IP server. The behaviour of the Novell client if this suboption is false, or is not present, is not specified 🙈 (so like, nobody knows, don't ask).
  • option nwip.preferred-dss ip-address [, ip-address... ]; — this suboption specifies a list of up to five IP addresses, each of which should be the IP address of a NetWare Domain SAP/RIP server (DSS).
  • option nwip.nearest-nwip-server ip-address [, ip-address...]; — this suboption specifies a list of up to five IP addresses, each of which should be the IP address of a Nearest NetWare IP server.
  • option nwip.autoretries uint8; — specifies the number of times that a NetWare/IP client should attempt to communicate with a given DSS server at startup.
  • option nwip.autoretry-secs uint8; — specifies the number of seconds that a Netware/IP client should wait between retries when attempting to establish communications with a DSS server at startup.
  • option nwip.nwip-1-1 uint8; — if true, the NetWare/IP client should support NetWare/IP version 1.1 compatibility. This is only needed if the client will be contacting Netware/IP version 1.1 servers.
  • option nwip.primary-dss ip-address; — specifies the IP address of the Primary Domain SAP/RIP Service server (DSS) for this NetWare/IP domain. The NetWare/IP administration utility uses this value as the Primary DSS server when configuring a secondary DSS server.

🦋 STANDARD DHCPV6 OPTIONS 🦋

DHCPv6 options differ from DHCPv4 options, partly because they use 16-bit code and length tags — but also, semantically, zero-length options are legal in DHCPv6, and multiple options are treated differently. Here's the vibe check: in DHCPv4, multiple options would be concatenated to form one option, whereas in DHCPv6 they're expected to be individual instantiations. Understandably, lots of options are not "allowed" to have multiple instances in a packet — normally these are options that get digested by the DHCP protocol software itself, not by users or applications. 👑

  • option dhcp6.client-id string; — specifies the client's DUID identifier. DUIDs are similar to, but different from, DHCPv4 client identifiers — and there are documented duid types:
    • duid-llt
    • duid-en
    • duid-ll

This value should not be configured — it's provided by clients and treated as an opaque identifier key blob by servers.

  • option dhcp6.server-id string; — specifies the server's DUID identifier. You may use this option to configure an opaque binary blob for your server's identifier.
  • option dhcp6.ia-na string; — the Identity Association for Non-temporary Addresses (ia-na) carries assigned addresses that are not temporary addresses, for use by the DHCPv6 client. This option is produced by the DHCPv6 server software, and should not be configured.
  • option dhcp6.ia-ta string; — the Identity Association for Temporary Addresses (ia-ta) carries temporary addresses, which may change upon every renewal. There is no support for this in the current DHCPv6 software.
  • option dhcp6.ia-addr string; — the Identity Association Address option is encapsulated inside ia-na or ia-ta options in order to represent addresses associated with those IA's. These options are manufactured by the software, so should not be configured.
  • option dhcp6.oro uint16 [ , uint16, ... ]; — the Option Request Option ("ORO") is the DHCPv6 equivalent of the parameter-request-list. Clients supply this option to ask servers to reply with options relevant to their needs and use. This option must not be directly configured — the request syntax in dhclient.conf(5) should be used instead.
  • option dhcp6.preference uint8; — the preference option informs a DHCPv6 client which server is 'preferred' for use on a given subnet. This preference is only applied during the initial stages of configuration — once a client is bound to an IA, it will remain bound to that IA until it is no longer valid or has expired. This value may be configured on the server, and is digested by the client software.
  • option dhcp6.elapsed-time uint16; — the elapsed-time option is constructed by the DHCPv6 client software, and is potentially consumed by intermediaries. This option should not be configured.
  • option dhcp6.relay-msg string; — the relay-msg option is constructed by intervening DHCPv6 relay agent software. This option is entirely used by protocol software, and is not meant for user configuration.
  • option dhcp6.unicast ip6-address; — the unicast option is provided by DHCPv6 servers which are willing (or prefer) to receive Request, Renew, Decline, and Release packets from their clients via unicast. Normally, DHCPv6 clients will multicast these messages. Per RFC 3315, the server will reject a unicast message received from a client unless it previously sent (or would have sent) the unicast option to that client. This option may be configured on the server at the global and shared network level. When a unicast message is received, the server will check for an applicable definition of the unicast option. If such an option is found, the message will be accepted; if not, it will be rejected. 🔥
  • option dhcp6.status-code status-code [ string ] ; — the status-code option is provided by DHCPv6 servers to inform clients of error conditions during protocol communication. This option is manufactured and digested by protocol software, and should not be configured.
  • option dhcp6.rapid-commit ; — the rapid-commit option is a zero-length option that clients use to indicate their desire to enter into rapid-commit with the server.
  • option dhcp6.vendor-opts string; — the vendor-opts option is actually an encapsulated sub-option space, in which each Vendor-specific Information Option (VSIO) is identified by a 32-bit Enterprise-ID number. The encapsulated option spaces within these options are defined by the vendors.

To make use of this option, the best way is to examine the section titled VENDOR ENCAPSULATED OPTIONS below — in particular, the bits about the "vsio" option space. 👀

  • option dhcp6.interface-id string; — the interface-id option is manufactured by relay agents, and may be used to guide configuration differentiating clients by the interface they are remotely attached to. It does not make sense to configure a value for this option, but it may make sense to inspect its contents.
  • option dhcp6.reconf-msg dhcpv6-message; — the reconf-msg option is manufactured by servers, and sent to clients in Reconfigure messages to inform them of what message the client should Reconfigure using. There is no support for DHCPv6 Reconfigure extensions, and this option is documented informationally only.
  • option dhcp6.reconf-accept ; — the reconf-accept option is included by DHCPv6 clients that support the Reconfigure extensions, advertising that they will respond if the server were to ask them to Reconfigure. There is no support for DHCPv6 Reconfigure extensions, and this option is documented informationally only.
  • option dhcp6.sip-servers-names domain-list;
  • sip-servers-names — lets SIP clients find a local SIP server to use for all their outbound SIP requests, aka a so-called "outbound proxy server." Wanna use manually entered IPv6 addresses instead? Go peep the sip-servers-addresses option below. 👇
  • option dhcp6.sip-servers-addresses ip6-address [, ip6-address ... ] ; — lets SIP clients find a local SIP server to use for all their outbound SIP requests, aka "outbound proxy servers" (plural, bestie). If you'd rather use domain names than IPv6 addresses, the sip-servers-names option above has you covered. ✨
  • option dhcp6.name-servers ip6-address [, ip6-address ... ] ; — tells clients about locally available recursive DNS servers. Easiest way to think about it: it's the nameserver line in /etc/resolv.conf, fr fr. 🧚
  • option dhcp6.domain-search domain-list; — sets the client's domain search path, applied to recursive DNS queries. Again, easiest framing ever: it's the search line in /etc/resolv.conf. 🌸
  • option dhcp6.ia-pd string; — manufactured by clients and servers to create a Prefix Delegation binding, i.e. to delegate an IPv6 prefix to the client. You don't directly edit this in dhcpd.conf(5) or dhclient.conf(5) — the software manufactures it and consumes it. 🙈
  • option dhcp6.ia-prefix string; — sits inside ia-pd options to identify which prefix(es) got allocated to the client. Also not directly edited in dhcpd.conf(5) or dhclient.conf(5) — manufactured and consumed by the software, periodt. 💅
  • option dhcp6.nis-servers ip6-address [, ip6-address ... ] ; — identifies, in order, the NIS servers available to the client.
  • option dhcp6.nisp-servers ip6-address [, ip6-address ... ] ; — identifies, in order, the NIS+ servers available to the client.
  • option nis-domain-name domain-list; — the NIS domain name the client is expected to use, and it's related to the nis-servers option.
  • option dhcp6.nis-domain-name domain-name; — the NIS domain name the client is expected to use, related to the dhcp6.nis-servers option.
  • option nisp-domain-name domain-list; — the NIS+ domain name the client is expected to use, related to the nisp-servers option.
  • option dhcp6.nisp-domain-name domain-name; — okay so the page says "The dhcp6.nis-domain-name option" right here (yes, really, that's literally what it says 😭) — it specifies the NIS+ domain name the client is expected to use, and it's related to the dhcp6.nisp-servers option.
  • option dhcp6.sntp-servers ip6-address [, ip6-address ... ] ; — a list of local SNTP servers the client can use to synchronize its clock. ⭐
  • option dhcp6.info-refresh-time uint32; — gives DHCPv6 clients using Information-request messages a hint about how long they should go between refreshing the info they were handed. Heads up tho: this option only gets delivered to the client — and is only likely to actually affect the client's behaviour — if the client requested the option. 👑
  • option dhcp6.bcms-server-d domain-list; — the domain names of local BCMS (Broadcast and Multicast Control Services) controllers the client may use.
  • option dhcp6.bcms-server-a ip6-address [, ip6-address ... ] ; — the IPv6 addresses of local BCMS (Broadcast and Multicast Control Services) controllers the client may use.
  • option dhcp6.geoconf-civic string; — a string to hold the geoconf civic structure. This option is included based on RFC 4776.
  • option dhcp6.remote-id string; — constructed by relay agents to inform the server about details pertaining to what the relay knows about the client (like what port it's attached to, and so forth). Its contents have some vendor-specific structure (similar to VSIO), but this implementation chose to treat the whole thing as an opaque field. 💀
  • option dhcp6.subscriber-id string; — an opaque field provided by the relay agent with additional info about the subscriber in question. What's actually inside depends on the vendor and/or the operator's configuration of the remote device, so yeah: opaque field. 🫶
  • option dhcp6.fqdn string; — normally constructed by the client or server; it negotiates the client's Fully Qualified Domain Name, plus which party is responsible for Dynamic DNS Updates. Full deets live in the section on the Client FQDN SubOptions — and note that the DHCPv4 and DHCPv6 FQDN options use the same fqdn. encapsulated space, so they're in all ways identical. 🦋
  • option dhcp6.pana-agent ip6-address [, ip6-address ... ] ; — a set of IPv6 addresses of a PAA for the client to use. The addresses are listed in preferred order. This option is included based on RFC 5192.
  • option dhcp6.new-posix-timezone text; — a string suitable for the TZ variable. This option is included based on RFC 4833.
  • option dhcp6.new-tzdb-timezone text; — the name of a zone entry in the TZ database. This option is included based on RFC 4833.
  • option dhcp6.ero uint16 [, uint16 ... ] ; — a list of the options requested by the relay agent. This option is included based on RFC 4994.
  • option dhcp6.lq-query string; — used internally for lease query.
  • option dhcp6.client-data string; — used internally for lease query.
  • option dhcp6.clt-time uint32; — used internally for lease query.
  • option dhcp6.lq-relay-data ip6-address string; — used internally for lease query.
  • option dhcp6.lq-client-link ip6-address [, ip6-address ... ] ; — used internally for lease query.
  • option dhcp6.v6-lost domain-name; — the domain name of the LoST server for the client to use. This option is included based on RFC 5223.
  • option dhcp6.capwap-ac-v6 ip6-address [, ip6-address ... ] ; — a list of IPv6 addresses of CAPWAP ACs that the WTP may use. The addresses are listed in preference order. This option is included based on RFC 5417.
  • option dhcp6.relay-id string; — the DUID for the relay agent. This option is included based on RFC 5460.
  • option dhcp6.v6-access-domain domain-name; — the domain name associated with the access network, for use with LIS Discovery. This option is included based on RFC5986.
  • option dhcp6.sip-ua-cs-list domain-list; — the list of domain names in the SIP User Agent Configuration Service Domains. This option is included based on RFC 6011.
  • option dhcp6.bootfile-url text; — the URL for a boot file. This option is included based on RFC 5970.
  • option dhcp6.bootfile-param string; — a string for the parameters to the bootfile. See RFC 5970 for more description of how the parameters are laid out inside the string. This option is included based on RFC 5970.
  • option dhcp6.client-arch-type uint16 [, uint16 ... ] ; — a list of one or more architecture types, described as 16 bit values. This option is included based on RFC 5970.
  • option dhcp6.nii uint8 uint8 uint8; — the client network interface identitier option (that spelling is straight off the page, keeping it 💀) supplies info about a client's level of UNDI support. The values are, in order, the type, the major value and the minor value. This option is included based on RFC5970.
  • option dhcp6.aftr-name domain-name; — a domain name of the AFTR tunnel endpoint. This option is included based on RFC 6334.
  • option dhcp6.erp-local-domain-name domain-name; — a domain name for the ERP domain. This option is included based on RFC 6440.
  • option dhcp6.rdnss-selection ip6-address uint8 domain-name; — RDNSS information = an IPv6 address of RDNSS, an 8 bit flags field, and a domain-list of domains for which the RDNSS has special knowledge. This option is included based on RFC 6731.
  • option dhcp6.client-linklayer-addr string; — a client link-layer address. The first two bytes must be the type of the link-layer, followed by the address itself. This option is included based on RFC 6939.
  • option dhcp6.link-address ip6-address; — an IPv6 address a relay agent uses to indicate to the server which link the client is located on. This option is included based on RFC 6977.
  • option dhcp6.solmax-rt uint32; — a value to override the default for SOL_MAX_RT. It's a 32 bit value. This option is included based on RFC 7083.
  • option dhcp6.inf-max-rt uint32; — a value to override the default for INF_MAX_RT. Also a 32 bit value. This option is included based on RFC 7083.

🌈 ACCESSING DHCPV6 RELAY OPTIONS 🌈

v6relay (relay-number, option)

Okay so this one lets you access an option that a relay agent added to a packet. relay-number selects which relay to examine, and option is the option you're hunting for. 🔥

Here's the vibe of how DHCPv6 actually works: each relay encapsulates the entire previous message into an option, adds its own options (if any), and sends the result onwards. The RFC specifies a limit of 32 hops. 🎀

So the numbering, bestie:

  • A relay-number of 0 is a no-op — it means don't look at the relays.
  • 1 is the relay closest to the client.
  • 2 would be the next one in from the client, and so on.
  • Any value greater than the max number of hops is which is closest to the server, independent of number.
Yes, that last sentence is exactly as confusing on the original page as it is here. Not me making it up, promise. 😭

Wanna use this in a class statement? It looks a lil something like this:

match if v6relay(1, option dhcp6.subscriber-id) = "client_1";

Slay. 💖

✨ DEFINING NEW OPTIONS ✨

Okay bestie, so the Internet Systems Consortium DHCP client and server literally let you define brand-new options. Every DHCP option has a name, a code, and a structure. The name is what you use to refer to it. The code is a number that the DHCP server and client use to refer to it. The structure describes what the option’s contents look like. Iconic trio, honestly.

To define a new option, pick a name that isn’t already taken by some other option. For example, you can’t use "host-name" because the DHCP protocol already defines a host-name option, and it’s documented earlier in this manual page. If an option name doesn’t appear in this manual page, you can use it, but it’s probably a good idea to put some kind of unique string at the beginning so you can be sure future options don’t snatch your name. For example, you might define an option, "local-host-name", feeling some confidence that no official DHCP option name will ever start with "local". Manifesting that for you. 🫶

Once you’ve chosen a name, you gotta choose a code. All codes between 224 and 254 are reserved as ´site-local´ DHCP options, so you can pick any one of these for your site (not for your product/application). In RFC3942, site-local space was moved from starting at 128 to starting at 224. In practice, some vendors have interpreted the protocol rather loosely and have used option code values greater than 128 themselves. There’s no real way to avoid this problem, and it was thought to be unlikely to cause too much trouble in practice. If you come across a vendor-documented option code in either the new or old site-local spaces, please contact your vendor and inform them about rfc3942. Be that helpful queen. 💅

The structure of an option is simply the format in which the option data appears. The ISC DHCP server currently supports a few simple types, like integers, booleans, strings and IP addresses, and it also supports the ability to define arrays of single types or arrays of fixed sequences of types. So many possibilities, so little time.

New options are declared as follows:

option new-name code new-code = definition ;

The values of new-name and new-code should be the name you have chosen for the new option and the code you have chosen. The definition should be the definition of the structure of the option. Slay.

The following simple option type definitions are supported:

🦋 BOOLEAN

option new-name code new-code = boolean ;

An option of type boolean is a flag with a value of either on or off (or true or false). So an example use of the boolean type would be:

option use-zephyr code 180 = boolean;
option use-zephyr on;

🔢 INTEGER

option new-name code new-code = sign integer width ;

The sign token should either be blank, unsigned or signed. The width can be either 8, 16 or 32, and refers to the number of bits in the integer. So for example, the following two lines show a definition of the sql-connection-max option and its use:

option sql-connection-max code 192 = unsigned integer 16;
option sql-connection-max 1536;

🌐 IP-ADDRESS

option new-name code new-code = ip-address ;

An option whose structure is an IP address can be expressed either as a domain name or as a dotted quad. So the following is an example use of the ip-address type:

option sql-server-address code 193 = ip-address;
option sql-server-address sql.example.com;

🌈 IP6-ADDRESS

option new-name code new-code = ip6-address ;

An option whose structure is an IPv6 address must be expressed as a valid IPv6 address. The following is an example use of the ip6-address type:

option dhcp6.some-server code 1234 = array of ip6-address;
option dhcp6.some-server 3ffe:bbbb:aaaa:aaaa::1, 3ffe:bbbb:aaaa:aaaa::2;

📝 TEXT

option new-name code new-code = text ;

An option whose type is text will encode an ASCII text string. For example:

option sql-default-connection-name code 194 = text;
option sql-default-connection-name "PRODZA";

🔤 DATA STRING

option new-name code new-code = string ;

An option whose type is a data string is essentially just a collection of bytes, and can be specified either as quoted text, like the text type, or as a list of hexadecimal contents separated by colons whose values must be between 0 and FF. For example:

option sql-identification-token code 195 = string;
option sql-identification-token 17:23:19:a6:42:ea:99:7c:22;

🌸 DOMAIN-LIST

option new-name code new-code = domain-list [compressed] ;

An option whose type is domain-list is an RFC1035 formatted (on the wire, "DNS Format") list of domain names, separated by root labels. The optional compressed keyword indicates if the option should be compressed relative to the start of the option contents (not the packet contents). The way that detail matters? Chef’s kiss. 👑

When in doubt, omit the compressed keyword. When the software receives an option that is compressed and the compressed keyword is omitted, it will still decompress the option (relative to the option contents field). The keyword only controls whether or not transmitted packets are compressed.

Note that when domain-list formatted options are output as environment variables to dhclient-script(8), the standard DNS -escape mechanism is used: they are decimal. This is appropriate for direct use in eg /etc/resolv.conf.

🎁 ENCAPSULATION

option new-name code new-code = encapsulate identifier ;

An option whose type is encapsulate will encapsulate the contents of the option space specified in identifier. Examples of encapsulated options in the DHCP protocol as it currently exists include the vendor-encapsulated-options option, the netware-suboptions option and the relay-agent-information option. Nested option energy, fr fr.

option space local;
option local.demo code 1 = text;
option local-encapsulation code 197 = encapsulate local;
option local.demo "demo";

🧚 ARRAYS

Options can contain arrays of any of the above types except for the text and data string types, which aren’t currently supported in arrays. An example of an array definition is as follows:

option kerberos-servers code 200 = array of ip-address;
option kerberos-servers 10.20.10.1, 10.20.11.1;

🍓 RECORDS

Options can also contain data structures consisting of a sequence of data types, which is sometimes called a record type. For example:

option contrived-001 code 201 = { boolean, integer 32, text };
option contrived-001 on 1772 "contrivance";

It’s also possible to have options that are arrays of records, for example:

option new-static-routes code 201 = array of {
     ip-address, ip-address, ip-address, integer 8 };
option static-routes
     10.0.0.0 255.255.255.0 net-0-rtr.example.com 1,
     10.0.1.0 255.255.255.0 net-1-rtr.example.com 1,
     10.2.0.0 255.255.224.0 net-2-0-rtr.example.com 3;

🌸 VENDOR ENCAPSULATED OPTIONS 🌸

okay bestie, this is the part where vendors get to do their own thing 💅 The DHCP protocol defines the vendor-encapsulated-options option, which is basically a lil VIP pass letting vendors define their own options that get sent encapsulated inside a standard DHCP option. It also defines the Vendor Identified Vendor Sub Options option (aka "VIVSO"), and the DHCPv6 protocol defines the Vendor-specific Information Option ("VSIO"). ✨

The format of all of these options is usually internally a string of options, similarly to other normal DHCP options. The VIVSO and VSIO options are different tho, because they contain options that correspond to vendor Enterprise-ID numbers (assigned by IANA 👑), which then contain options according to each Vendor's specifications. So yeah, you will need to refer to your vendor's documentation in order to form options to their specification. iykyk 📖

The value of these options can be set in one of two ways. The first way is to simply specify the data directly, using a text string or a colon-separated list of hexadecimal values. For help in forming these strings, please refer to RFC2132 for the DHCPv4 Vendor Specific Information Option, RFC3925 for the DHCPv4 Vendor Identified Vendor Sub Options, or RFC3315 for the DHCPv6 Vendor-specific Information Option. For example:

option vendor-encapsulated-options
    2:4:
     AC:11:41:1:
    3:12:
     73:75:6e:64:68:63:70:2d:73:65:72:76:65:72:31:37:2d:31:
    4:12:
     2f:65:78:70:6f:72:74:2f:72:6f:6f:74:2f:69:38:36:70:63;
option vivso
    00:00:09:bf:0E:
     01:0c:
         48:65:6c:6c:6f:20:77:6f:72:6c:64:21;
option dhcp6.vendor-opts
    00:00:09:bf:
     00:01:00:0c:
         48:65:6c:6c:6f:20:77:6f:72:6c:64:21;

The second way of setting the value of these options is to have the DHCP server generate a vendor-specific option buffer. 🛠️ To do this, you must do four things: define an option space, define some options in that option space, provide values for them, and specify that that option space should be used to generate the relevant option.

To define a new option space in which vendor options can be stored, use the option space statement:

option space name [ [ code width number ] [ length width number ] [ hash
size number ] ] ;

Where the numbers following code width, length width, and hash size respectively identify the number of bytes used to describe option codes, option lengths, and the size in buckets of the hash tables to hold options in this space (most DHCPv4 option spaces use 1 byte codes and lengths, which is the default, whereas most DHCPv6 option spaces use 2 byte codes and lengths). 🧮

The code and length widths are used in DHCP protocol — you must configure these numbers to match the applicable option space you are configuring. They each default to 1. Valid values for code widths are 1, 2 or 4. Valid values for length widths are 0, 1 or 2. Most DHCPv4 option spaces use 1 byte codes and lengths, which is the default, whereas most DHCPv6 option spaces use 2 byte codes and lengths. A zero-byte length produces options similar to the DHCPv6 Vendor-specific Information Option — but not their contents! 😭

The hash size defaults depend upon the code width selected, and may be 254 or 1009. Valid values range between 1 and 65535. Note that the higher you configure this value, the more memory will be used. 💀 It is considered good practice to configure a value that is slightly larger than the estimated number of options you plan to configure within the space. Previous versions of ISC DHCP (up to and including DHCP 3.0.*), this value was fixed at 9973.

The name can then be used in option definitions, as described earlier in this document. For example:

option space SUNW code width 1 length width 1 hash size 3;
option SUNW.server-address code 2 = ip-address;
option SUNW.server-name code 3 = text;
option SUNW.root-path code 4 = text;

option space ISC code width 1 length width 1 hash size 3;
option ISC.sample code 1 = text;
option vendor.ISC code 2495 = encapsulate vivso-sample;
option vendor-class.ISC code 2495 = text;

option ISC.sample "configuration text here";
option vendor-class.ISC "vendor class here";

option space docsis code width 2 length width 2 hash size 17;
option docsis.tftp-servers code 32 = array of ip6-address;
option docsis.cablelabs-configuration-file code 33 = text;
option docsis.cablelabs-syslog-servers code 34 = array of ip6-address;
option docsis.device-id code 36 = string;
option docsis.time-servers code 37 = array of ip6-address;
option docsis.time-offset code 38 = signed integer 32;
option vsio.docsis code 4491 = encapsulate docsis;

Once you have defined an option space and the format of some options, you can set up scopes that define values for those options, and you can say when to use them. 👑 For example, suppose you want to handle two different classes of clients. Using the option space definition shown in the previous example, you can send different option values to different clients based on the vendor-class-identifier option that the clients send, as follows:

class "vendor-classes" {
  match option vendor-class-identifier;
}

subclass "vendor-classes" "SUNW.Ultra-5_10" {
  vendor-option-space SUNW;
  option SUNW.root-path "/export/root/sparc";
}

subclass "vendor-classes" "SUNW.i86pc" {
  vendor-option-space SUNW;
  option SUNW.root-path "/export/root/i86pc";
}

option SUNW.server-address 172.17.65.1;
option SUNW.server-name "sundhcp-server17-1";

option vivso-sample.sample "Hello world!";

option docsis.tftp-servers ::1;

As you can see in the preceding example, regular scoping rules apply, so you can define values that are global in the global scope, and only define values that are specific to a particular class in the local scope. 🫶 The vendor-option-space declaration tells the DHCP server to use options in the SUNW option space to construct the DHCPv4 vendor-encapsulated-options option. This is a limitation of that option — the DHCPv4 VIVSO and the DHCPv6 VSIO options can have multiple vendor definitions all at once (even transmitted to the same client), so it is not necessary to configure this. ✨

✨ SEE ALSO ✨

like, go read these too bestie, they're the whole friend group 💖

💖 AUTHOR 💖

Information about Internet Systems Consortium can be found at https://www.isc.org.

and that's a wrap on dhcp-options(5), bestie 🎀 this whole glow-up came to you from the source/version tag dhcp-options(5) — periodt, go configure something iconic 💅✨

📦 where this page comes from