This chapter introduces the Dynamic Host Configuration Protocol (DHCP), and explains the concepts that underlie the protocol. This chapter also describes the advantages of using DHCP in your network.
This chapter contains the following information:
The DHCP protocol enables host systems in a TCP/IP network to be configured automatically for the network as the systems boot. DHCP uses a client-server mechanism. Servers store and manage configuration information for clients and provide that information upon a client's request. The information includes the client's IP address and information about network services that are available to the client.
DHCP evolved from an earlier protocol, BOOTP, which was designed for booting over a TCP/IP network. DHCP uses the same format as BOOTP for messages between the client and server. However, unlike BOOTP messages, DHCP messages can include network configuration data for the client.
A primary benefit of DHCP is its ability to manage IP address assignments through leases. Leases allow IP addresses to be reclaimed when they are not in use. The reclaimed IP addresses can be reassigned to other clients. A site that uses DHCP can use a smaller pool of IP addresses than would be needed if all clients were assigned a permanent IP address.
DHCP relieves you of some of the time-consuming tasks involved in setting up a TCP/IP network and in the daily management of that network. Note that Solaris DHCP works only with IPv4.
Solaris DHCP offers the following advantages:
IP address management – A primary advantage of DHCP is easier management of IP addresses. In a network without DHCP, you must manually assign IP addresses. You must be careful to assign unique IP addresses to each client and to configure each client individually. If a client moves to a different network, you must make manual modifications for that client. When DHCP is enabled, the DHCP server manages and assigns IP addresses without administrator intervention. Clients can move to other subnets without manual reconfiguration because they obtain, from a DHCP server, new client information appropriate for the new network.
Centralized network client configuration – You can create a tailored configuration for certain clients, or for certain types of clients. The configuration information is stored in one place, in the DHCP data store. You do not need to log in to a client to change its configuration. You can make changes for multiple clients just by changing the information in the data store.
Support of BOOTP clients – Both BOOTP servers and DHCP servers listen and respond to broadcasts from clients. The DHCP server can respond to requests from BOOTP clients as well as DHCP clients. BOOTP clients receive an IP address and the information needed to boot from a server.
Support of local clients and remote clients – BOOTP provides for the relaying of messages from one network to another network. DHCP takes advantage of the BOOTP relay feature in several ways. Most network routers can be configured to act as BOOTP relay agents to pass BOOTP requests to servers that are not on the client's network. DHCP requests can be relayed in the same manner because, to the router, DHCP requests are indistinguishable from BOOTP requests. The Solaris DHCP server can also be configured to behave as a BOOTP relay agent, if a router that supports BOOTP relay is not available.
Network booting – Clients can use DHCP to obtain the information that is needed to boot from a server on the network, instead of using RARP (Reverse Address Resolution Protocol) and the bootparams file. The DHCP server can give a client all the information that the client needs to function, including IP address, boot server, and network configuration information. Because DHCP requests can be relayed across subnets, you can deploy fewer boot servers in your network when you use DHCP network booting. RARP booting requires that each subnet have a boot server.
Large network support – Networks with millions of DHCP clients can use Solaris DHCP. The DHCP server uses multithreading to process many client requests simultaneously. The server also supports data stores that are optimized to handle large amounts of data. Data store access is handled by separate processing modules. This data store approach enables you to add support for any database that you require.
You must first install and configure the DHCP server. During configuration, you specify information about the network that clients need to operate on the network. After this information is in place, clients are able to request and receive network information.
The sequence of events for DHCP service is shown in the following diagram. The numbers in circles correlate to the numbered items in the description following the diagram.
The preceding diagram shows the following steps:
The client discovers a DHCP server by broadcasting a discover message to the limited broadcast address (255.255.255.255) on the local subnet. If a router is present and configured to behave as a BOOTP relay agent, the request is passed to other DHCP servers on different subnets. The client's broadcast includes its unique ID, which, in the Solaris DHCP implementation, is derived from the client's Media Access Control (MAC) address. On an Ethernet network, the MAC address is the same as the Ethernet address.
DHCP servers that receive the discover message can determine the client's network by looking at the following information:
Which network interface did the request come in on? The server determines either that the client is on the network to which the interface is connected, or that the client is using a BOOTP relay agent connected to that network.
Does the request include the IP address of a BOOTP relay agent? When a request passes through a relay agent, the relay agent inserts its address in the request header. When the server detects a relay agent address, the server knows that the network portion of the address indicates the client's network address because the relay agent must be connected to the client's network.
Is the client's network subnetted? The server consults the netmasks table to find the subnet mask used on the network indicated by the relay agent's address or by the address of the network interface that received the request. Once the server knows the subnet mask used, it can determine which portion of the network address is the host portion, and then it can select an IP address appropriate for the client. See the netmasks(4) man page for information on netmasks.
After the DHCP servers determine the client's network, the servers select an appropriate IP address and verify that the address is not already in use. The DHCP servers then respond to the client by broadcasting an offer message. The offer message includes the selected IP address and information about services that can be configured for the client. Each server temporarily reserves the offered IP address until the client determines whether to use the IP address.
The client selects the best offer, based on the number and type of services offered. The client broadcasts a request that specifies the IP address of the server that made the best offer. The broadcast ensures that all the responding DHCP servers know that the client has chosen a server. The servers that are not chosen can cancel the reservations for the IP addresses that they had offered.
The selected server allocates the IP address for the client and stores the information in the DHCP data store. The server also sends an acknowledgement message (ACK) to the client. The acknowledgement message contains the network configuration parameters for the client. The client uses the ping utility to test the IP address to make sure no other system is using it. The client then continues booting to join the network.
The client monitors the lease time. When a set period of time has elapsed, the client sends a new message to the chosen server to increase the lease time.
The DHCP server that receives the request extends the lease time if the lease still adheres to the local lease policy set by the administrator. If the server does not respond within 20 seconds, the client broadcasts a request so that one of the other DHCP servers can extend the lease.
When the client no longer needs the IP address, the client notifies the server that the IP address is released. This notification can happen during an orderly shutdown and can also be done manually.
The Solaris DHCP server runs as a daemon in the Solaris Operating System (Solaris OS) on a host system. The server has two basic functions:
Managing IP addresses – The DHCP server controls a range of IP addresses and allocates them to clients, either permanently or for a defined period of time. The server uses a lease mechanism to determine how long a client can use a nonpermanent address. When the address is no longer in use, it is returned to the pool and can be reassigned. The server maintains information about the binding of IP addresses to clients in its DHCP network tables, ensuring that no address is used by more than one client.
Providing network configuration for clients – The server assigns an IP address and provides other information for network configuration, such as a host name, broadcast address, network subnet mask, default gateway, name service, and potentially much more information. The network configuration information is obtained from the server's dhcptab database.
The Solaris DHCP server can also be configured to perform the following additional functions:
Responding to BOOTP client requests – The server listens for broadcasts from BOOTP clients discovering a BOOTP server and provides them with an IP address and boot parameters. The information must have been configured statically by an administrator. The DHCP server can simultaneously perform as a BOOTP server and as a DHCP server.
Relaying requests – The server relays BOOTP and DHCP requests to appropriate servers on other subnets. The server cannot provide DHCP or BOOTP service when configured as a BOOTP relay agent.
Providing network booting support for DHCP clients – The server can provide DHCP clients with information needed to boot over the network: an IP address, boot parameters, and network configuration information. The server can also provide information that DHCP clients need to boot and install over a wide area network (WAN).
Updating DNS tables for clients that supply a host name – For clients that provide a Hostname option and value in their requests for DHCP service, the server can attempt DNS updates on their behalf.
As superuser, you can start, stop, and configure the DHCP server with DHCP Manager or with command-line utilities described in DHCP Command-Line Utilities. Generally, the DHCP server is configured to start automatically when the system boots, and to stop when the system is shut down. You should not need to start and stop the server manually under normal conditions.
All the data used by the Solaris DHCP server is maintained in a data store. The data store might consist of plain text files, NIS+ tables, or binary-format files. While configuring the DHCP service, you choose the type of data store to be used. The section Choosing the DHCP Data Store describes the differences between the types of data stores. You can convert a data store from one format to another by using DHCP Manager or the dhcpconfig command.
You can also move data from one DHCP server's data store to another server's data store. You can use export and import utilities that work with the data stores, even if the servers are using different data store formats. You can export and import the entire content of a data store, or just some of the data within it, using DHCP Manager or the dhcpconfig command.
Any database or file format can be used for DHCP data storage if you develop your own code module to provide an interface between Solaris DHCP (server and management tools) and the database. .
Within the Solaris DHCP data store are two types of tables. You can view and manage the contents if these tables by using either DHCP Manager or the command-line utilities. The data tables are as follows:
dhcptab table – Table of configuration information that can be passed to clients.
DHCP network tables – Tables containing information about the DHCP and BOOTP clients that reside on the network specified in the table name. For example, the network 192.168.32.0 would have a table whose name includes 192_168_32_0.
The dhcptab table contains all the information that clients can obtain from the DHCP server. The DHCP server scans the dhcptab table each time it starts. The file name of the dhcptab table varies according to the data store used. For example, the dhcptab table created by the NIS+ data store SUNWnisplus is SUNWnisplus1_dhcptab.
The DHCP protocol defines a number of standard items of information that can be passed to clients. These items are referred to as parameters, symbols, or options. Options are defined in the DHCP protocol by numeric codes and text labels, but without values. Some commonly used standard options are shown in the following table.
Table 11–1 Sample DHCP Standard Options
Code |
Label |
Description |
---|---|---|
1 |
Subnet |
Subnet mask IP address |
3 |
Router |
IP address for the router |
6 |
DNSserv |
IP address for the DNS server |
12 |
Hostname |
Text string for the client host name |
15 |
DNSdmain |
DNS domain name |
Some options are automatically assigned values when you provide information during server configuration. You can also explicitly assign values to other options at a later time. Options and their values are passed to the client to provide configuration information. For example, the option/value pair, DNSdmain=Georgia.Peach.COM, sets the client's DNS domain name to Georgia.Peach.COM.
Options can be grouped with other options in containers known as macros, which makes it easier to pass information to a client. Some macros are created automatically during server configuration and contain options that were assigned values during configuration. Macros can also contain other macros.
The format of the dhcptab table is described in the dhcptab(4) man page. In DHCP Manager, all the information shown in the Options and Macros tabs comes from the dhcptab table. See About DHCP Options for more information about options. See About DHCP Macros for more information about macros.
Note that the dhcptab table should not be edited manually. You should use either the dhtadm command or DHCP Manager to create, delete, or modify options and macros.
A DHCP network table maps client identifiers to IP addresses and the configuration parameters associated with each address. The format of the network tables is described in the dhcp_network(4) man page. In DHCP Manager, all the information shown in the Addresses tab comes from the network tables.
DHCP Manager is a graphical user interface (GUI) tool you can use to perform all management duties associated with the DHCP service. You can use it to manage the server as well as the data the server uses. You must be superuser when you run DHCP Manager.
You can use DHCP Manager with the server in the following ways:
Configuring and unconfiguring the DHCP server
Starting, stopping, and restarting the DHCP server
Disabling and enabling DHCP service
Customizing DHCP server settings
DHCP Manager enables you to manage the IP addresses, network configuration macros, and network configuration options in the following ways:
Adding and deleting networks under DHCP management
Viewing, adding, modifying, deleting, and releasing IP addresses under DHCP management
Viewing, adding, modifying, and deleting network configuration macros
Viewing, adding, modifying, and deleting nonstandard network configuration options
DHCP Manager allows you to manage the DHCP data stores in the following ways:
Convert data to a new data store format
Move DHCP data from one DHCP server to another by exporting it from the first server and importing it on the second server
DHCP Manager includes extensive online help for procedures you can perform with the tool. For more information, see About DHCP Manager.
All DHCP management functions can be performed by using command-line utilities. You can run the utilities if you are logged in as superuser or as a user assigned to the DHCP Management profile. See Setting Up User Access to DHCP Commands.
The following table lists the utilities and describes the purpose of each utility.
Table 11–2 DHCP Command-Line Utilities
Security for the dhcpconfig, dhtadm, and pntadm commands is determined by role-based access control (RBAC) settings. By default, the commands can be run only by superuser. If you want to use the commands under another user name, you must assign the user name to the DHCP Management profile as described in Setting Up User Access to DHCP Commands.
You configure the Solaris DHCP server the first time you run DHCP Manager on the system where you want to run the DHCP server.
DHCP Manager server configuration dialog boxes prompt you for essential information needed to enable and run the DHCP server on one network. Some default values are obtained from existing system files. If you have not configured the system for the network, there are no default values. DHCP Manager prompts for the following information:
Role of the server, either as the DHCP server or as the BOOTP relay agent
Data store type (files, binary files, NIS+, or something specific to your site)
Data store configuration parameters for the data store type you selected
Name service to use to update host records, if any (/etc/hosts, NIS+, or DNS)
Length of lease time and whether clients should be able to renew leases
DNS domain name and IP addresses of DNS servers
Network address and subnet mask for the first network you want to configure for DHCP service
Network type, either local area network (LAN) or point-to-point network
Router discovery or the IP address of a particular router
NIS domain name and IP address of NIS servers
NIS+ domain name and IP address of NIS+ servers
You can also configure the DHCP server using the dhcpconfig command. This utility automatically gathers information from existing system files to provide a useful initial configuration. Therefore, you must ensure that the files are correct before running dhcpconfig. See the dhcpconfig(1M) man page for information about the files that dhcpconfig uses to obtain information.
The Solaris DHCP server supports the following types of IP address allocation:
Manual allocation – The server provides a specific IP address that you choose for a specific DHCP client. The address cannot be reclaimed or assigned to another client.
Automatic, or permanent, allocation – The server provides an IP address that has no expiration time, making it permanently associated with the client until you change the assignment or the client releases the address.
Dynamic allocation – The server provides an IP address to a requesting client, with a lease for a specific period of time. When the lease expires, the address is taken back by the server and can be assigned to another client. The period of time is determined by the lease time configured for the server.
You determine what information to provide to DHCP clients. When you configure the DHCP server, you provide essential information about the network. Later, you can add more information that you want to provide to clients.
The DHCP server stores network configuration information in the dhcptab table, in the form of option/value pairs and macros. Options are keywords for network data that you want to supply to clients. Values are assigned to options and passed to clients in DHCP messages. For example, the NIS server address is passed by way of an option called NISservs. The NISservs option has a value that is equal to a list of IP addresses, which is assigned by the DHCP server. Macros provide a convenient way to group together any number of options that you want to supply to clients. You can use DHCP Manager to create macros to group options and to assign values to the options. If you prefer a command-line tool, you can use dhtadm, the DHCP configuration table management utility, to work with options and macros.
In Solaris DHCP, an option is a piece of network information to be passed to a client. The DHCP literature also refers to options as symbols or tags. An option is defined by a numeric code and a text label. An option receives a value when it is used in the DHCP service.
The DHCP protocol defines a large number of standard options for commonly specified network data: Subnet, Router, Broadcst, NIS+dom, Hostname, and LeaseTim are a few examples. A complete list of standard options is shown in the dhcp_inittab(4) man page. You cannot modify the standard option keywords in any way. However, you can assign values to the options that are relevant to your network when you include the options in macros.
You can create new options for data that is not represented by the standard options. Options you create must be classified in one of three categories:
Extended – Reserved for options that have become standard DHCP options but are not yet included in the DHCP server implementation. You might use an extended option if you know of a standard option that you want to use, but you do not want to upgrade your DHCP server.
Site – Reserved for options that are unique to your site. You create these options.
Vendor – Reserved for options that should apply only to clients of a particular class, such as a hardware or vendor platform. The Solaris DHCP implementation includes a number of vendor options for Solaris clients. For example, the option SrootIP4 is used to specify the IP address of a server that a client that boots from the network should use for its root (/) file system.
Chapter 14, Administering DHCP (Tasks) includes procedures for creating, modifying, and deleting DHCP options.
In the Solaris DHCP service, a macro is a collection of network configuration options and the values that you assign to them. Macros are created to group options together to be passed to specific clients or types of clients. For example, a macro intended for all clients of a particular subnet might contain option/value pairs for subnet mask, router IP address, broadcast address, NIS+ domain, and lease time.
When the DHCP server processes a macro, it places the network options and values defined in the macro in a DHCP message to a client. The server processes some macros automatically for clients of a particular type.
For the server to process a macro automatically, the name of the macro must comply with one of the categories shown in the following table.
Table 11–3 DHCP Macro Categories for Automatic Processing
A macro with a name that does not use one of the categories listed in Table 11–3 can be processed only if one of the following is true:
The macro is mapped to an IP address.
The macro is included in another macro that is processed automatically.
The macro is included in another macro that is mapped to an IP address.
When you configure a server, a macro that is named to match the server's name is created by default. This server macro is not processed automatically for any client because it is not named with one of the name types that cause automatic processing. When you later create IP addresses on the server, the IP addresses are mapped to use the server macro by default.
When a DHCP client requests DHCP services, the DHCP server determines which macros match the client. The server processes the macros, using the macro categories to determine the order of processing. The most general category is processed first, and the most specific category is processed last. The macros are processed in the following order:
Client class macros – The most general category
Network address macros – More specific than Client class
Macros mapped to IP addresses – More specific than Network address
Client ID macros – The most specific category, pertaining to one client
A macro that is included in another macro is processed as part of the container macro.
If the same option is included in more than one macro, the value for that option in the macro with the most specific category is used because it is processed last. For example, if a Network address macro contains the lease time option with a value of 24 hours, and a Client ID macro contains the lease time option with a value of 8 hours, the client receives a lease time of 8 hours.
The sum total of the values assigned to all the options in a macro must not exceed 255 bytes, including the option codes and length information. This limit is dictated by the DHCP protocol.
The macros that are most likely to be impacted by this limit are macros that are used to pass paths to files on Solaris installation servers. Generally, you should pass the minimum amount of vendor information needed. You should use short path names in options that require path names. If you create symbolic links to long paths, you can pass the shorter link names.
The term “client” is sometimes used to refer to a physical machine that is performing a client role on the network. However, the DHCP client described in this document is a software entity. The Solaris DHCP client is a daemon (dhcpagent) that runs in the Solaris OS on a system that is configured to receive its network configuration from a DHCP server. DHCP clients from other vendors can also use the services of the Solaris DHCP server. However, this document describes only the Solaris DHCP client.
See Chapter 15, Configuring and Administering the DHCP Client for detailed information about the Solaris DHCP client.