|
| dhcptab - DHCP configuration parameter table |
|
The dhcptab configuration table allows network administrators to organize groups of configuration parameters as macro definitions, which can then be referenced in the definition of other useful macros. These macros are then used by the DHCP server to return their
values to DHCP and BOOTP clients.
The preferred method of managing the dhcptab is through the use of the dhcpmgr(1M) or dhtadm(1M) utility. The description of dhcptab entries included in this manual page is intended for informational purposes only, and should not be used to manually edit entries.
You can view the contents of the dhcptab using the DHCP manager's tabs for Macros and Options, or using the dhtadm -P command.
Syntax of dhcptab Entries
|
The format of a dhcptab table depends on the data store used to maintain it. However, any dhcptab must contain the following fields in each record:
-
Name
- This field identifies the macro or symbol record and is used as a search key into the dhcptab table. The name of a macro or symbol must consist of ASCII characters, with the length limited to 128 characters. Names can include spaces, except at the end of the name. The name is not case-sensitive.
-
Type
- This field specifies the type of record and is used as a search key into the dhcptab. Currently, there are only two legal values for Type:
-
m
- This record is a DHCP macro definition.
-
s
- This record is a DHCP symbol definition. It is used to define vendor and site-specific options.
-
Value
- This field contains the value for the specified type of record. For the m type, the value will consist of a series of symbol=value pairs, separated by the colon (:) character. For the s type, the value will consist of a series of fields, separated by a comma (,), which define a symbol's characteristics. Once defined, a symbol can be used in macro definitions.
|
Symbol Characteristics
|
The Value field of a symbols definition contain the following fields describing the characteristics of a symbol:
-
Context
- This field defines the context in which the symbol definition is to be used. It can have one of the following values:
-
Site
- This symbol defines a site-specific option, codes 128-254.
-
Vendor=Client Class ...
- This symbol defines a vendor-specific option, codes 1-254. The Vendor context takes ASCII string arguments which identify the client class that this vendor option is associated with. Multiple
client class names can be specified, separated by white space. Only those clients whose client class matches one of these values will see this option. For Sun machines, the Vendor client class matches the value returned by the command uname -i on the client, with periods replacing
commas.
-
Code
- This field specifies the option code number associated with this symbol. Valid values are 128-254 for site-specific options, and 1-254 for vendor-specific options.
-
Type
- This field defines the type of data expected as a value for this symbol, and is not case-sensitive. Legal values are:
-
ASCII
-
NVT ASCII text. Value is enclosed in double-quotes ("). Granularity setting has no effect on symbols of this type,
since ASCII strings have a natural granularity of one (1).
-
BOOLEAN
- No value is associated with this data type. Presence of symbols of this type denote boolean TRUE, whereas absence denotes FALSE.
Granularity and Miximum values have no meaning for symbols of this type.
-
IP
- Dotted decimal form of an Internet address. Multi-IP address granularity is supported.
-
NUMBER
- An unsigned number with a supported granularity of 1, 2, 4, and 8 octets.
Valid NUMBER types are: UNUMBER8, SNUMBER8, UNUMBER16, SNUMBER16, UNUMBER32, SNUMBER32, UNUMBER64, and SNUMBER64. See dhcp_inittab(4) for details.
-
OCTET
- Uninterpreted ASCII representation of binary data. The client identifier is one example of an OCTET string. Valid characters are 0-9, [a-f] [A-F]. One ASCII character represents one nibble (4 bits), thus two ASCII characters are needed to represent an 8 bit quantity. The granularity setting has no effect on symbols of this type, since OCTET strings have a natural granularity of one (1).
-
Granularity
- This value specifies how many objects of Type define a single instance of the symbol value. For example, the static route option is defined to be a variable list of routes. Each route
consists of two IP addresses, so the Type is defined to be IP, and the data's granularity is defined to be 2 IP addresses. The granularity field affects the IP and NUMBER data types.
-
Maximum
- This value specifies the maximum items of Granularity which are permissible in a definition using this symbol. For example, there can only be one IP address specified for a subnet mask,
so the Maximum number of items in this case is one (1). A Maximum value of zero (0) means that a variable number of items is permitted.
The following example defines a site-specific option (symbol) called MystatRt, of code 130, type IP, and granularity 2, and a Maximum of 0. This
definition corresponds to the internal definition of the static route option (StaticRt).
|
MystatRt s Site,130,IP,2,0
|
|
Macro Definitions
|
The following example illustrates a macro defined using the MystatRt site option symbol just defined:
|
10netnis m :MystatRt=3.0.0.0 10.0.0.30:
|
Macros can be specified in the Macro field in DHCP network tables (see dhcp_network(4)), which will bind particular
macro definitions to specific IP addresses.
Up to four macro definitions are consulted by the DHCP server to determine the options that are returned to the requesting client.
These macros are processed in the following order:
-
Client Class
- A macro named using the ASCII representation of the client class (e.g. SUNW.Ultra-30) is searched for in the dhcptab.
If found, its symbol/value pairs will be selected for delivery to the client. This mechanism permits the network administrator to select configuration parameters to be returned to all clients of the same class.
-
Network
- A macro named by the dotted Internet form of the network address of the client's network (for example, 10.0.0.0) is searched for in the dhcptab. If found, its symbol/value pairs will be combined
with those of the Client Class macro. If a symbol exists in both macros, then the Network macro value overrides the value defined in the Client Class macro. This mechanism permits the network administrator to select configuration parameters to be
returned to all clients on the same network.
-
IP Address
- This macro may be named anything, but must be specified in the DHCP network table for the IP address record assigned to the requesting client. If this macro is found in the dhcptab, then its symbol/value pairs will be combined with those of the Client Class macro and the Network macro. This mechanism permits the network administrator to select configuration parameters to be returned to clients using a particular IP address. It can also be used to deliver a macro defined to include "server-specific" information by including this macro definition in all DHCP network table entries owned by a specific server.
-
Client Identifier
- A macro named by the ASCII representation of the client's unique identifier as shown in the DHCP network table (see dhcp_network(4)). If found, its symbol/value pairs are combined to the sum of the Client Class, Network, and IP Address macros. Any symbol collisions are replaced with those
specified in the client identifier macro. The client mechanism permits the network administrator to select configuration parameters to be returned to a particular client, regardless of what network that client is connected to.
Refer to System Administration Guide: IP Services for more information about macro processing.
Refer to the dhcp_inittab(4) man page for more information about symbols used in Solaris DHCP.
|
|
|
dhcpmgr(1M), dhtadm(1M), in.dhcpd(1M), dhcp_inittab(4), dhcp_network(4), dhcp(5)
System Administration Guide: IP Services
Alexander, S., and R. Droms, DHCP Options and BOOTP Vendor Extensions, RFC 2132, Silicon Graphics, Inc., Bucknell University, March 1997.
Droms, R., Interoperation Between DHCP and BOOTP, RFC 1534, Bucknell University, October 1993.
Droms, R., Dynamic Host Configuration Protocol, RFC 2131, Bucknell University, March 1997.
Wimer, W., Clarifications and Extensions for the Bootstrap Protocol, RFC 1542, Carnegie Mellon University, October 1993.
|
| |