Appendix A: Configuration parameters

Table of contents

Appendix A

Configuration parameters

For information on hardware and software configuration parameters required for basic system and network configuration, see:

General configuration parameters

General configuration parameters include:

DMA channel

Some devices require that you specify the Direct Memory Access (DMA) channel that you want the adapter to use. Direct memory access permits data to be transferred between memory and a device without the intervention of the system's central processing unit (CPU). Each DMA chip on the system motherboard has eight circuits (or ``channels'') for conveying data. DMA channel 4 is reserved for use by the system. During configuration, you may choose to use any (or none) of the channels.

Interrupt vectors

Generally, each driver on your system, including those for network adapters and SLIP lines, must have its own interrupt vector (or ``IRQ''). An interrupt halts processing momentarily so that input/output or other operations can occur. Processing resumes after the specific operation takes place. Consequently, it is important that each device installed in your system be provided with an interrupt setting that does not conflict with the settings used by the hardware and other peripherals (unless the device in question supports sharing of interrupts).

Unless a device supports sharing of interrupts, its interrupt vector must not be used by any other device on the system. Refer to your networking hardware documentation to determine which vectors the hardware supports. The hwconfig(C) and vectorsinuse(ADM) commands list the hardware already installed on your system and what vectors are already in use.



Table A-1 Typical interrupt vectors

 ----------------------------------------------------------------
 Hardware                    Interrupt Vector (IRQ)
 ----------------------------------------------------------------
 ISA, EISA, or MC machine
   clock                               0
   console (keyboard)                  1
   floppy disk controller              6
 Hard disk controller                varies
 Serial ports
   COM1                                4
   COM2                                3
 Parallel ports
   lpt0, lpt1                          7
   lpt2                                5
To install a networking adapter and driver software, you must choose an interrupt vector (IRQ) setting for the adapter. Consult the adapter's documentation for acceptable IRQ settings. 

After you determine your hardware's IRQ settings, choose settings for each networking adapter that you plan to install, making sure that the settings do not conflict with each other. The documentation for each networking adapter should indicate whether you need to configure the adapter physically to use the chosen IRQ setting. The operating system reserves interrupt vectors 4 and 7 for COM1 and lpt0, respectively. If you choose any setting that is either a setting reserved for another use or is in use by another device, a conflict occurs. Some SCOadmin managers may be able to detect conflicts.

If yours is an ISA system, your networking hardware might be preconfigured to use a particular vector. If you want to change this vector setting, you might also need to change the physical jumper settings on the adapter or run a setup program provided with the adapter.

NOTE: A number of networking adapters are preconfigured to use interrupt vector 3. Your operating system has reserved IRQ3 for the sio (serial input-output) device.

I/O base address

Each hardware driver on your system that performs I/O (input/output) must have a unique memory base address so that the system can locate it. This memory address is a three- or four-digit hexadecimal number, must match the settings on the adapter, and must not conflict with any other hardware on your system. Valid base addresses are displayed when you configure your adapter.

The I/O base address is the initial address for a unique area of memory allocated for input/output data control to a specific hardware adapter. For example, 0x300, or 300, is the default I/O base address for the 3Com 3c501 network adapter. The I/O base address must match the hardware configuration on the adapter, and other devices must not use this I/O base address.

WARNING: If you enter an I/O base address that is already in use by another device, you might delete that device.

Network driver configuration parameters

In addition to the general configuration parameters, you can set these configuration parameters for network drivers:

10BaseT mode enforced

Under normal circumstances, the adapter detects the cable type connected to it. This option turns off auto-detection and enforces a 10BaseT connection. Use this option only if the 10BaseT connector is not detected.

NOTE: If you want to use the AUI port, do not plug a cable into the 10BaseT port.

See also:

Cable (media) type

When you configure certain adapters (for example, the 3Com 503 driver), you need to specify whether the adapter connects to ``thick,'' ``thin'', or ``twisted-pair'' Ethernet cable. Thick and thin are both coaxial cable. Thick cable is approximately 1/2-inch in diameter, and thin is approximately 1/4-inch in diameter. Twisted pair resembles slightly thicker phone cable. 

Data rate (Token-Ring)

Some Token-Ring drivers are able to automatically detect the data transmission rate on the ring and configure the adapter accordingly. However, if the adapter is the first active station on the ring, auto-detection will not work and the data rate must be set manually, usually to 4Mbps or 16Mbps. We recommend that you configure servers for a specific ring speed.

NOTE: Although it is desirable to set the ring to the highest rate possible, all adapters on the ring should operate at the same rate. Some older Token-Ring adapters can only be set to 4Mbps.

Frame size

Frame size is the maximum amount of the data that can be sent out on the network in one packet. The maximum frame size for Ethernet is fixed at approximately 1500 bytes of data. On networks such as Token-Ring, different frame sizes can be used. Larger frame sizes generally increase network performance, although not all systems support larger frame sizes. If you are using a network adapter that supports configurable frame sizes, you may want to experiment with larger sizes in your network.

Full-duplex mode enabled

Full-duplex mode can improve performance for 10BaseT (twisted-pair) connections when using a switched Ethernet hub. It is disabled by default.

NOTE: You must be connected to a switched Ethernet network to use full-duplex mode.

See also:

Hardware (MAC) address

The hardware address -- often referred to as the MAC (Media Access Control) address -- is assigned by the adapter manufacturer and, under normal circumstances, is not configurable. By default, the Network Configuration Manager associates the adapter's hard-coded MAC address with the adapter driver. However, you can supply an alternate MAC address if you want to:

Configurable hardware addresses do not change the MAC address on the adapter, but the Network Configuration Manager can map the physical MAC address to the desired alternate.

A MAC address consists of 6 hexidecimal numbers separated by colons. For example: 

   00:00:c0:34:f1:52

See also:

Interrupt service routine assignment

By default, on multi-processor systems, multi-threaded network drivers are assigned interrupts dynamically. In other words, when the interrupt is received, the system assigns it to a CPU that is available to service the interrupt. To bind network driver interrupt processing to a specific CPU, use this advanced option in the Network Configuration Manager.

Media speed

Some Ethernet adapter drivers detect the media speed automatically, although you can manually set it to 10Mbps or 100Mbps. Set the speed manually of hardware incompatibilities cause autodetection to fail.

NOTE: You must be connected to a 100Mbps Ethernet network to use the 100Mbps option.

PCI bus, device, and function numbers

PCI machines use unique bus, device, and function numbers to identify each installed device. Bus numbers range from 0-255. Device numbers range from 0-31. Function numbers range from 0-7. PCI bus, device, and function numbers can be displayed using your PCI setup program.

Primary/Alternate adapter

This parameter applies to these adapters:

You can install up to two Token-Ring adapters in your system. If you are using two adapters, make sure to use the Primary-Alternate switches on the adapters to set up one as the primary adapter and the other as the alternate adapter. You must also ensure that the adapters are designated as ``Primary'' or ``Alternate'' during configuration with the Network Configuration Manager. Obtain switch information for these adapters from the documentation supplied by the adapter manufacturer.

ROM base address

Some adapters (for example, IBM Token-Ring adapters) require that you specify a ROM base address. Like the RAM base address, the ROM base address specifies a specific part of ROM memory that is set aside for networking use. If you have more than one adapter requiring that you specify a ROM address, make sure that you do not assign addresses that conflict with another ROM region.

Shared RAM address

The Shared RAM address is the address of a unique area of memory allocated to the host machine and network or graphics adapter for read/write operations; this area is sometimes called the ``RAM buffer''. Other devices must not use any address in the range from the RAM base address to the end of the allocated memory area (that is, the RAM base address + the RAM buffer size).

The Network Configuration Manager notes any conflicting Shared RAM addresses with an asterisk (*).

Slot number

Some adapters (for example, the HP EtherTwist EISA Adapter Card/32) are meant to be used on machines using a EISA, MCA, or PCMCIA bus. If you install one of these adapters and the adapter is not detected in the machine, you may be required to provide its ``slot number'' during the configuration process. This number refers to the slot in which the adapter is inserted.

Source routing

IBM Token-Ring networking allows you to establish connections from your machine to other machines in these ways:

on a local ring
A local ring is the Token-Ring physically attached to your machine.

to other rings using gateways
A TCP/IP gateway is created when you configure TCP/IP over more than one Token-Ring adapter on the same machine. Those adapters must be connected to different rings and have different IP subnet addresses. TCP/IP gateways connected to Token-Ring networks in this manner will route TCP/IP traffic to the rings without using Token-Ring source routing. Similar gateways can be set up using the OSI and IPX/SPX protocols.

to other networks using a Token-Ring bridge
A Token-Ring bridge is a dedicated piece of computer hardware connected to several Token-Rings. The bridge routes frames between the rings. All Token-Rings connected via bridges will appear as a single ring to each station on the network. Token-Ring source routing allows your adapter to route network traffic across Token-Ring bridges regardless of the protocol stack used by any of the connected networks.

If you intend to connect your machine to a network that includes a bridge, and if you intend to send information from your machine across the bridge, you must use automatic Token-Ring source routing.

The Network Configuration Manager offers you two Token-Ring source routing options:

none
Source routing is not enabled; frames are not routed beyond the local ring.

auto
Source routing is enabled; frames include source routing information and the DLPI module performs source routing on behalf of the protocol stack. This is the default setting.
These options are set for individual adapters; they are not global to all Token-Ring adapters configured in your system.

These options take effect for all protocol stacks using the specified adapter. It is possible for stacks to override default source routing without affecting the source routing mode used by other stacks. In such cases, the source routing is said to be in ``stack'' mode for the specific protocol stack. For example, SCO TCP/IP and IPX/SPX can be configured to use automatic source routing of a Token-Ring adapter, while a third-party SNA product can provide stack mode routing when using the same adapter.

Protocol stacks provide stack mode routing if the characteristics of the protocol prevent it from working with the general purpose SCO source routing facility, or if a more specialized source routing that is designed to work optimally for a particular protocol is desired. Although no SCO protocol stacks currently provide stack mode source routing, third-party networking products might contain such functionality.

Source route optimization

The Network Configuration Manager allows you to find the optimal route between machines on different Token-Ring rings:

yes
Use All Routes Explorers (AREs) to discover the optimal route. This is the default.

no
Do not try to find the optimal route. Do not use AREs.
It is usually desirable to use ARE explorer frames to discover routes because they find the fastest route between two points and store the information in the Token-Ring source routing table for future transmissions. However, in large networks, using AREs can cause additional network traffic and slow down the network. If minimizing frame traffic is desirable, you may not want to optimize routing.

See also:

tx/rx (transmit/receive) buffers

Transmit and receive buffers are used to regulate the flow of data frames between adapters and protocol stacks. Although the default settings are usually acceptable, increasing the number may improve performance if network traffic is heavy, but it will also use system memory.

Serial connection configuration parameters

The SCO PPP and SLIP drivers take a variety of configuration parameters, most of which have different possible values and implications dependent upon the type of link configuration you are specifying.

Because of the complexity of these configurations, these parameters are defined by configuration type in the Networking Guide. If you are connecting to a service provider, or to another system that is already configured, ask the system administrator of the remote site for the correct values to enter at the configuration prompts. If you are configuring both sides of the connection, refer to this documentation:

TCP/IP configuration parameters

TCP/IP configuration parameters include these basic options:

TCP/IP configuration advanced options include:

IP address

The ``IP address'' identifies and differentiates a given machine from all others on the network. It consists of a 32-bit binary number that is usually displayed as four octets expressed in decimal and separated by periods. You must have a unique IP address for each machine on the network. In addition, if your machine serves as a router to another network (it contains two or more network adapters and belongs to two or more networks), you must assign each adapter a unique IP address on the appropriate network. 

NOTE: The IP address differs from a MAC (Media Access Control) address in that it is configurable. A MAC address is a 6-byte address that is unique to each physical network adapter. This non-configurable address is assigned by the adapter manufacturer.

The IP address consists of two parts: a network address that identifies the network and a host address that identifies the particular host, or node. 

Table A-2 IP address derivation

 ---------------------------------------------------------------------------------------

 binary (32-bit)   1 0 0 0 0 1 0 0 1 0 0 0 1 1 1 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0
 binary (octets)   1 0 0 0 0 1 0 0   1 0 0 0 1 1 1 1   0 0 0 0 0 0 1 0   0 0 0 0 0 0 1 0
 decimal octets    10                0                 2                 2
 ---------------------------------------------------------------------------------------
                      IP address (in standard notation) = 10.0.2.2
Several classes of TCP/IP networks are available, each based on the number of hosts a network needs. Network classes supported by SCO are Class A, B, and C. Use the smallest network class that can accommodate all of your network's hosts. Most TCP/IP installations use Class C, but some larger installations might need to use Class B.



Table A-3 Internet address classes

 -----------------------------------------------------------------------------
 Class      Available Hosts   Valid Address
            per Network       Ranges
 -----------------------------------------------------------------------------
 A          16777216          1.0.0.1 through 126.255.255.254
 B          65534             128.0.0.1 through 191.255.255.254
 C          254               192.0.0.1 through 222.255.255.254
 Reserved                     224.0.0.0 through 255.255.255.254
If you are connecting your machine to a pre-existing network, the network address (for Class A, the first octet; for Class B, the first two octets; and for Class C, the first three octets) is the same as those of other machines on the network. In this case, you only need to create a unique host address.

If you are creating an entirely new network and you want to connect to the Internet, you need to contact the Network Information Center (NIC) to have a network address assigned; see ``Domain name'' for the Network Information Center address. If you do not want to connect to an outside network, you can choose any network address that conforms to the syntax shown previously. In either case, once you determine the network address, you can then create the unique host address.

When you determine the IP address, remember:

Netmask setting

The ``netmask'' strips the network ID from the IP address, leaving only the host ID. Each netmask consists of binary ones (decimal 255) to mask the network ID and binary zeroes (decimal 0) to retain the host ID of the IP address. For example, the default netmask setting for a Class B address is 255.255.0.0.

NOTE: Always use the default netmask that the installation program prompts you for unless you are creating a subnet (a logical division of a physical network). If you create a subnet, also mask the portion of the address that indicates the subnet. For example, the netmask for a machine on a Class B subnet is 255.255.255.0. For more information, see ``Creating subnets'' in the Networking Guide.

Broadcast address parameters

All datagrams sent by TCP/IP move through all machines in the network path. However, each host adapter ignores any packet that does not include that particular computer's IP address in the datagram header. Occasionally, you might want to send a message to all machines on a particular network. To do so, select a ``broadcast address'' for your machine. A broadcast address is one in which the host portion of the IP address consists either of all 0's or all 255's. The configuration procedure prompts you to choose between these address schemes: 

Table A-4 Broadcast address schemes

 ---------------------------------------------------------------------
 Scheme                Example        Purpose
 ---------------------------------------------------------------------
 all zeroes (decimal   10.0.0.0       provides compatibility with
 0)                                   4.2BSD systems
 all ones (decimal     10.0.255.255   UNIX Operating System Standard
 255)                                 (RFC-919)
The addresses shown in the previous table are for a class B network, and are shown as examples only; your values will be different. If you are on a network that does not contain any machines running 4.2BSD systems or earlier BSD versions, choose all ones. If such machines exist on your network, choose all zeroes.

System name

The ``system name'' (``host name'') should be unique on the network. It can consist of lowercase letters and numbers, must begin with a letter, and should be no longer than eight characters. mail and other programs use the system name to identify the correct data destination. Some sample valid machine names are scosysv, tcpdev, and account1.

Domain name

The MMDF mail router uses the ``domain name'' to route messages, such as mail, from machine to machine. The domain name allows your network to fit into a hierarchical network structure. Some common domains at the top of this hierarchy include: 

 .com   commercial organizations
 .edu   educational institutions
 .gov   government institutions
 .mil   military institutions
 .org   miscellaneous organizations

Sample domain names are sco.com (the domain name used by SCO) and berkeley.edu (the domain name used by the University of California at Berkeley).

Base your domain name choice on:

TCP/IP connections

This TCP/IP configuration parameter sets the maximum number of ``TCP/IP connections''.

Gateway status

A machine that has interfaces (adapters or serial lines) to more than one network may operate as a ``gateway'' between networks by forwarding and redirecting packets from one network to another.

When you configure a second adapter under TCP/IP, you are prompted to turn on this gateway behavior or leave your machine in the default, non-gateway mode. If you do not make your machine into a gateway, it will continue to receive packets on each network at the specified IP addresses, but will not forward packets between networks.

See also:


Administrator for this system

This optional item lists the name of the network administrator. The information is used by those who need to know who to contact in case of SNMP problems and is stored in the /etc/snmpd.conf file.

Location of this system

This optional item lists the system location. The information is used by those who need to know who to contact in case of SNMP network problems and is stored in the /etc/snmpd.conf file.

Token-Ring Frame format

If your network adapter is a token ring card, then the default frame format is 802.n; otherwise it is Ethernet-II. During card reconfiguration, this setting defaults to the previously configured value. The choices for any card are either 802.n framing or Ethernet-II framing.

If you are connecting to any network running LLI Release 3.0 or older, then only Ethernet-II framing can be used. 802.n framing is only allowed for SCO OpenServer network adapters. For more information on frame formats, see ``Framing type''.

IPX/SPX configuration parameters

IPX/SPX configuration parameters include these basic options:

IPX/SPX configuration advanced options include:

NVT logins

IPX/SPX allows you to set a maximum number of simultaneous NVT connections. You may want to adjust this parameter if there is an increase or decrease in the number of clients that try to access this server.

The default value is 16 and is reflected by the parameter nvt_max_logins in the file /etc/ipx.d/NPSConfig.

The maximum value is hard-coded at the driver level and can be changed by manually editing the NVT_MAX_LOGINS parameter in the file /etc/conf/pack.d/nvt/nvt_tune.h and relinking the kernel. See ``Relinking the kernel'' in the SCO OpenServer Handbook.

The nvt_max_logins parameter must be set to a value less than or equal to NVT_MAX_LOGINS. For more information, see NPSConfig(SFF) and nvt_tune.h(SFF).

Internal network number

The ``internal network number'' must be unique for every machine on all network segments; it also must not conflict with any network segment's network number. This number is entered in hexadecimal format. You may want to change this parameter if conflicts develop as new servers are added to the network or numbering conventions change.

Network number

The ``network number'', entered in hexadecimal format, is a unique number identifying a single LAN segment; every node on the same LAN segment must have the same network number. If the SCO system is acting as a router, that is, it is connected to more than one network segment, each connected network segment must be assigned the correct unique network number. You may want to change this parameter if there is a conflict between network numbers or if the numbering conventions change.

Framing type

IPX/SPX allows you to set the type of ``framing'' performed by the network adapter driver. You may want to change this parameter if you reconfigure a network segment to use a different framing type.

The framing types supported by IPX/SPX are described in:



802.3 with 802.2 headers

The Ethernet 802.3 with 802.2 headers framing type conforms to the IEEE 802.3 and IEEE 802.2 standards. This framing type is the default Ethernet framing type in most Novell® NetWare networks.

The main feature of this framing type is the SAP (Service Advertising Protocol) fields, which indicate the protocol type. For Novell networks, these fields are set to 0xe0, which indicates that the upper layer protocol is IPX. Ethernet 802.3 with 802.2 header frames have this format: 

Table A-5 802.3 with 802.2 headers

 ---------------------------------------
 Header   Field                 Size
 ---------------------------------------
 802.3    destination address   6 octets
 802.3    source address        6 octets
 802.3    length                2 octets
 802.2    destination SAP       1 octet
 802.2    source SAP            1 octet
 802.2    control               1 octet
          data                  ...
 ---------------------------------------


802.3

The Ethernet 802.3 framing type is commonly used in Novell networks, particularly with native NetWare 2.x and 3.x servers. This framing type was developed by Novell before the IEEE 802.2 standard was complete and is sometimes referred to as ``802.3 raw''.

The main feature of the Ethernet 802.3 framing type is the use of the hardcoded 0xffff value in the frame header. Ethernet 802.3 frames have this format: 

Table A-6 802.3

 ----------------------------------------------------
 Header   Field                              Size
 ----------------------------------------------------
 802.3    destination address                6 octets
 802.3    source address          6 octets
 802.3    length                             2 octets
 802.3    0xffff                             2 octets
          data                               ...
 ----------------------------------------------------


Ethernet II

The Ethernet II framing type is sometimes referred to as the ``XEROX PARC version of Ethernet''.

The main feature of this framing type is its simple frame structure. Ethernet II frames have this format: 

Table A-7 Ethernet II

 ---------------------------------------
 Header        Field            Size
 ---------------------------------------
 Ethernet II   address          6 octets
 Ethernet II   source address   6 octets
 Ethernet II   type             2 octets
               data             ...
 ---------------------------------------


Ethernet SNAP

Ethernet SNAP framing allows network protocol stacks to use Ethernet II frames on IEEE style networks without modification. SNAP framing is most commonly used for Token-Ring networks.

The main feature of this framing type is the use of three protocol headers: IEEE 802.3, IEEE 802.2, and SNAP. Ethernet SNAP frames have this format: 

Table A-8 Ethernet SNAP

 ---------------------------------------
 Header   Field                 Size
 ---------------------------------------
 802.3    destination address   6 octets
 802.3    source address        6 octets
 802.3    length                2 octets
 802.2    0xaa                  1 octet
 802.2    0xaa                  1 octet
 802.2    UI                    1 octet
 SNAP     protocol ID           1 octet
 SNAP     type                  1 octet
          data                  ...
 ---------------------------------------


Token-Ring

The Token-Ring framing type conforms to the IEEE 802.5 and IEEE 802.2 standards.

The main feature of this framing type is the SAP (Service Advertising Protocol) fields, which indicate the protocol type. For Novell networks, these fields are set to 0xe0, which indicates that the upper layer protocol is IPX. Token-Ring frames have this format: 

Table A-9 Token-Ring

 ------------------------------------------
 Header   Field                 Size
 ------------------------------------------
 802.5    AC                    1 octet
 802.5    FC                    1 octet
 802.5    destination address   6 octets
 802.5    source address        6 octets
 802.5    routing information   0-18 octets
 802.2    destination SAP       1 octet
 802.2    source SAP            1 octet
 802.2    control               1 octet
          data                  ...
 ------------------------------------------


Token-Ring SNAP

Token-Ring SNAP allows network protocol stacks to use Ethernet II frames.

The main feature of this framing type is the use of three protocol headers: IEEE 802.5, IEEE 802.2, and SNAP. Token-Ring SNAP frames have this format: 

Table A-10 Token-Ring SNAP

 ------------------------------------------
 Header   Field                 Size
 ------------------------------------------
 802.5    AC                    1 octet
 802.5    FC                    1 octet
 802.5    destination address   6 octets
 802.5    source address        6 octets
 802.5    routing information   0-18 octets
 802.2    0xaa                  1 octet
 802.2    0xaa                  1 octet
 802.2    UI                    1 octet
 SNAP     protocol ID           1 octet
 SNAP     type                  1 octet
          data                  ...
 ------------------------------------------

NVT login banner file

IPX/SPX allows you to specify a file from which the NVT daemon prints user login banners. If no banner is desired, simply remove or move the file, or set this field to reference a non-existent file. Do not set this field to null ("").

SAP reply delay

IPX/SPX allows you to set the delay period that SAPD (Service Advertising Protocol Daemon) should wait before responding to a Nearest Server Request. This allows some control over the order in which servers will respond to Nearest Server Requests. Servers that should be used first for efficiency reasons should be set low. For example, on a network with native NetWare servers, it might be desirable to set the SAP reply delay high on all SCO servers to ensure that the native servers always reply first when there are duplicate services offered on both the SCO host and the native NetWare server. (If there are no duplicate services, a low value should be used on all servers to ensure rapid replies.)

The default value is 1, indicating a minimal delay. The maximum value for this parameter is 300. For more information, see NPSConfig(SFF).

Source routing (Token-Ring framing only)

If you selected Token-Ring as your framing type, you can enable IPX/SPX to pass source routing information to your adapter driver. See ``Source routing'' concerning source routing options for your adapter. The Network Configuration Manager enables Token-Ring source routing by default.

NetBIOS configuration parameters

TPI (Transport Provider Interface) NetBIOS for TCP/IP configuration is required when running LAN Manager over the TCP/IP protocol stack.

Basic NetBIOS configuration parameters include:

NetBIOS host name
The system name (host name) should be unique on the network. It can consist of lowercase letters and numbers, must begin with a letter, and should be no longer than eight characters. It must be same as the TCP/IP system name. Here are some sample valid system names: scosysv, srv555b, and account1. The ``values'' of the other two NetBIOS basic parameters are dependent on the ``value'' of this parameter. The default is the output of uname -n.

NetBIOS IP address
The NetBIOS IP address identifies and differentiates your machine from all others on the network. It consists of a 32-bit binary number that is usually displayed as four octets expressed in decimal and separated by periods (for example, 10.0.65.75). Each machine on your network must have a unique NetBIOS IP address that is the same as the IP address assigned for TCP/IP.

NetBIOS broadcast address
All datagrams sent by NetBIOS move through all machines in the network path. However, each host adapter ignores any packet that does not include that particular computer's IP address in the datagram header. Occasionally, you might want to send a message to all machines on a particular network. To do so, select a ``broadcast address'' either of all 0's or all 255's. It must be same as the TCP/IP Broadcast address.

NetBIOS scope identifier
The NetBIOS scope identifier sets the NetBIOS domain. Only users with equivalent scope identifiers can communicate with each other. It is similar to the TCP/IP domain name (for example sco.com or lachman.com).
Advanced NetBIOS configuration parameters are LAN Manager related ``per-user sessions'' parameters. If any of the LAN Manager parameters are increased, the corresponding TCP/IP NetBIOS related parameter should also be increased.

Default pending commands per user
The default number of maximum active commands per user.

Default sessions per user
The default number of NetBIOS sessions per user.

Maximum names per user
The maximum number of NetBIOS names (local/remote) per user. A user cannot register more than the maximum number of NetBIOS names on the machine.

Maximum pending commands per user
The maximum number of commands a user can have pending. This is used by nbstatus to display the maximum number of active/pending commands a user can set.

Maximum sessions per user
The maximum number of NetBIOS sessions a user may set. This is used by nbstatus to display the maximum number of configured sessions a user can set.

Maximum IP datagram size
The maximum IP datagram size before fragmentation can take place. If any datagram is larger than the size set, the packet is fragmented.

LAN Manager Client configuration parameters

The LAN Manager Client configuration parameter is:

Number of transport connections
This sets the maximum number of concurrent sessions that the transport provider will support (default value is 64). The value you enter can be obtained from the documentation for the transport provider.