2    Basic Network Connections

This chapter describes:

Note

This chapter discusses the configuration of network interfaces in an Internet Protocol Version 4 (IPv4) environment. All references to the Internet Protocol (IP) and the Transmission Control Protocol/Internet Protocol (TCP/IP) are IPv4-specific. For information about configuring IPv6 in a network environment, see Chapter 3.

For information about ATM and point-to-point connections, see Chapter 4 and Chapter 6, respectively.

For troubleshooting information, see Section 8.3.

2.1    Network Environment

When you install a system in a network, you need to know how to configure your network interface card (NIC) and how to route messages from your system to other systems. This section helps answer both of these questions.

2.1.1    Network Interface

Your system is connected to the network through a NIC (which is also called a network interface or network adapter). End systems or hosts can have the following interface options:

Routers typically have multiple interfaces, with each connected to a different subnet. Figure 2-1 shows a network with two hosts, Host A and Host B, each with a single network interface in a subnet.

Figure 2-1:  Sample Single Interface Configuration

If you need one of the multiple interface options, Table 2-1 summarizes the characteristics of each multiple interface option to help you choose the option that is right for you.

Table 2-1:  Comparison of Multiple Interface Configurations

Configuration Characteristics
Multiple interfaces in a subnet Higher throughput, load sharing across interfaces based on connections (outbound traffic only)
NetRAIN Reliability and availability
Link aggregation or trunking Higher throughput, load sharing across interfaces (inbound and outbound traffic), and availability

The following sections describe each option in more detail.

2.1.1.1    Multiple Interfaces in a Subnet

You can configure multiple active network interfaces in one system, even if they operate on the same subnetwork. For example, you can configure a tu0 interface at 16.1.1.1 and a tu1 interface at 16.1.1.2, both with the same netmask, as shown for Host A in Figure 2-2.

Figure 2-2:  Sample Multiple Interfaces in a Subnet

When you establish a connection, the kernel routes the connection through the interface that has the fewest number of connections. This connection-balancing effect can lead to greater throughput than on a system with just one network adapter per subnetwork.

This feature differs from NetRAIN because it does not give you increased reliability or failover, it simply gives a system multiple paths to access the network.

Network administrators might choose to configure a system with multiple interfaces in the same subnetwork for various reasons. For example:

For the system to function properly when configured in this manner, it must meet all of these conditions:

This feature might affect the operation of network software or commands that rely on the network interface staying constant for the life of a connection. For example:

No special settings are required to use this feature. Configure the network interfaces as directed in Section 2.3.1 and assign the interfaces IP addresses in the same subnet.

By default, configuring an interface adds interface route into the routing table. If you want to add routes using the route command or the /etc/routes file, see route(8) for details on adding routes on multiple interfaces. For example, you might want to add a default route on multiple interfaces. See netstat(1) for information on how to view the kernel routing table.

2.1.1.2    NetRAIN

The Redundant Array of Independent Network Adaptors (NetRAIN) interface provides a mechanism to protect against certain kinds of network connectivity failures.

NetRAIN integrates multiple network interfaces on the same local area network (LAN) segment into a single virtual interface called a NetRAIN set. One network interface in the set is always active while the others remain idle. If the active interface fails, one of the idle set members comes online with the same IP address within an adjustable failover time period. Figure 2-3 shows Host A with three interfaces that are part of a NetRAIN set. The NetRAIN virtual interface is assigned the address 16.1.1.1.

Figure 2-3:  Sample NetRAIN Configuration

See Section 2.4.1 for information on configuring NetRAIN.

NetRAIN monitors the status of its network interfaces with the Network Interface Failure Finder (NIFF), a tool used to detect and report possible network failures. This tool can be used independently of NetRAIN. For more information about NIFF, see niff(7).

NetRAIN and MAC Address Licensing Schemes

Licensing schemes that use a network adapter's Media Access Control (MAC) address to uniquely identify a machine can be affected by how NetRAIN changes the MAC address.

All network drivers support the SIOCRPHYSADDR ioctl that fetches MAC addresses from the interface. This ioctl returns two addresses in an array:

Licensing schemes based on MAC addresses must use the default hardware address returned by the SIOCRPHYSADDR ioctl; do not use the current physical address because NetRAIN modifies this address for its own use. See the reference page for your network adapter (for example ln(7) and tu(7)) for a sample program that uses the SIOCRPHYSADDR ioctl. For more information about ioctls, see ioctl(2).

2.1.1.3    Link Aggregation

Link aggregation, or trunking, enables administrators to combine one or more physical Ethernet NICs and create a single logical link. (Upper-layer software sees this link aggregation group as a single logical interface.) The single logical link can carry traffic at higher data rates than a single interface because the traffic is distributed across all of the physical ports that make up the link aggregation group.

Using link aggregation provides the following capabilities:

You can use a link aggregation group virtual interface for the following point-to-point connections: server-to-server and server-to-switch. Figure 2-4 shows Server A and Server B, each with two interfaces in a link aggregation group, attached to a switch. A single IP address is assigned to each link aggregation virtual interface.

Figure 2-4:  Sample Link Aggregation Configuration

See Section 2.4.3 for information on configuring link aggregation.

2.1.2    Routing

All systems (hosts and routers) connected to a network must be configured to support network routing in order to communicate with other systems on other networks. A route is the path a packet takes through a network from one system to another. As such it enables you to communicate with other systems on other networks. Routes are stored on each system in the routing tables or routing database. Each route entry consists of the following:

When you configure your system you automatically get a route for your loopback interface (lo0). In addition, you get a route for each interface that you configure by using the SysMan Configure Interfaces application. If you want additional routes, you can do one of the following:

In addition to either of the previous choices, additional routes might be added to your routing tables based on Internet Control Message Protocol (ICMP) redirect messages. These are messages from routers to hosts that tell the host to forward traffic to another router on the local network. Section 2.2 presents the routing choices and information to help you make the correct choice.

2.2    Preparing for the Configuration

You configure the network components by using the Network Configuration application. The following sections contain worksheets that you can use to record the information required to configure the network components.

2.2.1    Information for Interfaces and Daemons

Figure 2-5 shows the Interface and Daemon Worksheet. The following sections explain the information you need to record on this worksheet. If you are viewing this manual online, you can use the print feature to print a copy of the worksheet.

Figure 2-5:  Interface and Daemon Worksheet

2.2.1.1    All Network Interfaces

Adapter name

The device names of the network interfaces. The following table contains a list of selected network interfaces that the operating system supports:

Interface Device Name

Ethernet

ee

le

ln

tu

xna

Fiber Distributed Data Interface (FDDI)

faa

fta

fza

Gigabit Ethernet

alt

Token Ring

tra

Note that if you configuring a NetRAIN interface, as described in Section 2.4.1, the adapter name is the virtual device name of your NetRAIN set (nr). If you are configuring a link aggregation group, as described in Section 2.4.3, the adapter name is the virtual device name of your group (lag).

Host name

The fully qualified host name assigned to your system. A fully qualified host name contains the host name and the domain name, with host name and each level of the domain name separated by a period (.). Ask the network administrator for a unique host name.

Internet address source

The source of your system's network address for Ethernet, FDDI, and NetRAIN interfaces only. If your network uses a Dynamic Host Configuration Protocol (DHCP) server to assign IP addresses to systems at boot time, check the DHCP server box. If you plan to assign an IP address and network mask as part of system configuration, check the User supplied box.

Internet address

The IP address of your system. If you are going to supply your own IP address, write it in this space. If you will be using DHCP to assign IP addresses on a temporary basis, leave this space blank.

If you do not have a designated IP address for your network, you need to obtain one from one of the following services. Then, after you receive your network's address, assign a unique IP address and host name to each system on your network.

To obtain an Internet address for your network, contact:

American Registry for Internet Numbers
4506 Daly Drive, Suite 200
Chantilly, VA  20151
 
Voice: (703) 227-0660
FAX: (703) 227-0676
E-mail: reg-services@arin.net (for general information)
               hostmaster@arin.net (for IP address registrations)
WWW: http://www.arin.net

In Europe, you can contact:

RIPE Network Coordination Center
Singel 258
1016 AB Amsterdam
The Netherlands
 
Voice: +31 20 535 4444
FAX: +31 20 535 4445
 
E-mail: ncc@ripe.net (for general information)
                   hostmaster@ripe.net (for IP address registrations)
WWW: http://www.ripe.net

In Asia and the Pacific region, you can contact:

Asia Pacific Network Information Center
Level 1, 33 Park Road
P.O. Box 2131
Milton, QLD 4064
Australia
 
Voice: +61 7 3367 0490
FAX: +61 7 3367 0482
 
E-mail: info@apnic.net (for general information)
                   hostmaster@apnic.net (for IP address registrations)
WWW: http://www.apnic.net

Note

It is a good idea to register your network even if you do not intend to connect to the Internet network. Then, if you decide to connect to the Internet network later, you will not have to change all the host addresses on your network.

Network mask

Your network's subnet mask. Subnetworks allow the systems on a LAN be known by one address to the Internet network, while being known locally by a set of addresses. Subnetworks can represent logical groupings of hosts, or different physical networks. If your network uses subnetwork routing, each system on the network must have the same subnet mask defined. Use the following table to help identify your subnet mask. If you are not using subnetworks, the n is zero (0); otherwise, the n is greater than zero and less than or equal to 255.

Class IP Address Range Subnet Mask
A 0.0.0.0 to 127.0.0.0 255.n.n.n
B 128.0.0.0 to 191.0.0.0 255.255.n.n
C 192.0.0.0 to 223.0.0.0 255.255.255.n

If you are connecting your system to an existing network that is using subnetwork routing, ask the network administrator for the correct subnet mask.

2.2.1.2    Token Ring Interface

Adapter speed

If your system supports token ring, the speed of your system's token ring adapter. Two speeds are supported: 4Mb/s and 16Mb/s. The default speed is 16Mb/s.

2.2.1.3    NetRAIN Interface

NetRAIN interfaces provide higher availability on systems that contain multiple network adapters. See Section 2.1.1.2 for more information.

Set members

The device names of the network interfaces that are part of the NetRAIN set. When one interface in the set ceases to function, NetRAIN will fail over to another interface on this list.

2.2.1.4    LAG Interface

Link aggregation interfaces provide higher availability, fault tolerance, and load sharing on systems that contain multiple network adapters. See Section 2.1.1.3 for more information.

Ports

The device names of the network interfaces that are ports in a link aggregation group. When one interface in the group ceases to function, traffic is rerouted to the other available port or ports.

2.2.1.5    rwhod Daemon

The rwhod daemon maintains the database that is used by the rwho and ruptime programs. These programs provide basic information about the system and its current users to users on remote systems.

rwhod

If you want to run the rwhod daemon, check Yes; otherwise, check No.

Running the rwhod daemon allows you to use the rwho and ruptime commands.

Flags

If the rwhod daemon is to send rwho packets and ignore incoming packets, check Broadcast Only. If the daemon is to collect incoming packets, but not broadcast rwho packets, check Listen Only. If the daemon is to do both, check Both.

See rwhod(8) for additional information.

2.2.1.6    routed Daemon

The routed daemon allows your system's internal routing tables for the Routing Information Protocol (RIP) to be updated automatically.

routed

If you want to run the routed daemon, check Yes; otherwise, check No. Use the routed daemon to manage your routes dynamically only if your network and system requirements match the criteria in the following table:

Criterion Type or Value
Size of network Medium to large LAN or WAN, with multiple subnets
Network Topology Variable
Number of routes required Loopback, network interface route, and many others
Routers advertising routes Yes
Configuration complexity Low
System overhead Low

You can choose to run the routed daemon or gated daemon, but not both. For more information about these daemons and static routing, see the Best Practice for Network Routing on the Tru64 UNIX Publications Home Page at the following URL:

http://www.tru64unix.compaq.com/docs/

Flags

Specifies how you want the routed daemon to run. You can run the routed daemon on a gateway host, write all packets to standard output, or log debugging information. Check the options you want. See routed(8) for more information.

RIP data

If the routed daemon is to supply RIP information, check Supply; otherwise, check Run Quietly.

2.2.1.7    Gateways File

The gateways file contains Internet routing information for the routed daemon. Specify the following parameters for the file:

Destination Type

If the route is to a network, check Net. If the route is to a specific host, check Host.

Destination

The destination name or IP address (in dotted-decimal format).

Gateway

The name or IP address of the gateway host to which messages will be forwarded.

Hop count

The hop count, or number of gateways, from the local network to the destination network.

Route type

If the gateway is expected to exchange RIP routing information, check Active. If the gateway is not expected to exchange routing information, check Passive. If the gateway is to notify routed that another routing process will install the route (it is not advertised through RIP), check External.

See gateways(4) for additional information.

2.2.1.8    gated Daemon

The gated daemon allows your system's internal routing tables for various routing protocols to be updated automatically.

gated

If you want to run the gated daemon, check Yes; otherwise, check No. Use the gated daemon to manage your routes dynamically only if your network and system requirements match the criteria in the following table:

Criterion Type or Value
Size of network Medium to large, with multiple subnets
Network Topology Variable
Number of routes required Loopback, network interface route, and many others
Routers advertising routes Yes
Configuration complexity Moderate to high
System overhead Low
System role Host, router, or cluster member

You can choose to run the gated daemon or routed daemon, but not both. For more information about these daemons and static routing, see the Best Practice for Network Routing on the Tru64 UNIX Publications Home Page at the following URL:

http://www.tru64unix.compaq.com/docs/

Configuration file

The name of an alternate configuration file. By default, the gated daemon uses the /etc/gated.conf file.

2.2.1.9    IP Router

An IP router is a gateway host connected to more than one TCP/IP network that receives and forwards packets between the networks.

You can configure your system as an IP router if you have more than one network interface installed and configured. In addition, you must have configured either the routed or the gated daemon.

IP router

If you want the system to run as an IP router, check Yes; otherwise, check No.

2.2.2    Information for Network Files

Figure 2-6 shows the Network Files Worksheet. The following sections explain the information you need to record on this worksheet. If you are viewing this manual online, you can use the print feature to print a copy of the worksheet.

Figure 2-6:  Network Files Worksheet

2.2.2.1    Static Routes File (/etc/routes)

The routes file specifies static routes that will be added to your system's internal routing tables when the system boots.

Use static routes only if your network and system requirements match the criteria in the following table:

Criterion Type or Value
Size of network Small LAN (hosts and one gateway/router)
Network Topology Stable
Number of routes required Loopback, network interface route, and a few others
Routers advertising routes No
Configuration complexity Low
System overhead None

For more information about static routing, as well as the gated and routed daemons, see the Best Practice for Network Routing on the Tru64 UNIX Publications Home Page at the following URL:

http://www.tru64unix.compaq.com/docs/

If you choose to use static routes, specify the following parameters for the routes file:

Destination type

The specific path, as stored in the /etc/routes file, from your system to another host or network. A static route is not updated by network software. If you want to route to a default gateway, check Default Gateway; to a host, check Host; or to a network, check Network.

Destination

The name or IP address of the route destination. For default gateway, the default destination is default.

Route via

If you are routing through a gateway, check Gateway. If you are routing through an interface, check Interface.

Gateway

The name or IP address of the gateway or interface.

See routes(4) for additional information.

2.2.2.2    Hosts File (/etc/hosts)

The hosts file contains critical address information for the known hosts on the network. Specify the following parameters for the file:

Host name

The names of other hosts on the network to be added to the /etc/hosts file.

If your network is running a distributed database lookup service (DNS/BIND or NIS), you do not need to list each host on your network in your /etc/hosts file. However, it is a good idea to list four or five systems on the network designated as DNS/BIND or NIS servers in your /etc/hosts file.

Internet address

The IP addresses of other hosts on the network to be added to the /etc/hosts file.

Alias

The aliases, if any, of other hosts on the network to be added to the /etc/hosts file.

See hosts(4) for additional information.

2.2.2.3    Hosts Equivalencies File (/etc/hosts.equiv)

The hosts.equiv file contains the names of remote systems and users that can execute commands on the local system. Specify the following parameters for the file:

Host name

The name of the trusted hosts to be put in the /etc/hosts.equiv file. Systems listed in the /etc/hosts.equiv file are logically equivalent to, and therefore treated exactly the same as, the local system.

Setting up an /etc/hosts.equiv file is optional but, if you choose to have one on your system, you need to create it and add the names of any trusted hosts.

User name

The name of a user on a trusted host.

See hosts.equiv(4) for additional information.

2.2.2.4    Networks File (/etc/networks)

The networks file contains information about the known networks that your system needs to access. Specify the following parameters for the file:

Network name

The official Internet name of the network.

Network address

The IP address of the network.

Alias

The unofficial names used for the network to be added to the /etc/networks file.

See networks(4) for additional information.

2.3    Configuring the Network Components

Use the SysMan Menu application of the Common Desktop Environment (CDE) Application Manager to configure the following network components on your system:

To invoke the SysMan Menu application, follow the instructions in Section 1.2.1. See the same section for information about time-saving alternatives for configuration tasks.

2.3.1    Configuring Network Interfaces

Use the following procedure to configure Ethernet, FDDI, or Token Ring network interfaces. For information about how to configure NetRAIN, see Section 2.4.1. For information about how to configure a link aggregation group, see Section 2.4.3.

Note

If you are configuring a system that is new to this environment, verify that the network adapter mode is set correctly at the console level before continuing. For example, if you have a 10base2 Ethernet network and your system is configured to use 10baseT Ethernet, your system fails to see the network until you set the appropriate console variable. See the prerequisite tasks for a full installation in the Installation Guide for more information.

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up Network Interface Card(s) to display the Network Interface Card (NIC) dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman interface

    All network adapters that are installed on the system are listed in the dialog box.

  2. Select the network adapter that you want to configure. The dialog box for the selected interface is displayed.

  3. Enter the name for the interface in the Host Name field.

  4. To configure an Ethernet interface, do the following:

    1. To obtain the IP address data from the DHCP server, select the Use DHCP radio button. Otherwise, select the User Supplied Value radio button and enter the IP address and network mask data in the appropriate fields.

    2. Select the Additional Flags button to display the Additional Flags dialog box, which shows advanced configuration parameters for the selected interface.

    3. Select the check boxes and radio buttons for the other interface options that you want to enable and enter values where necessary for optional ifconfig arguments.

    4. Go to step 7.

  5. To configure an FDDI interface, do the following:

    1. If you are to obtain the IP address data from the DHCP server, select the Use DHCP radio button. Otherwise, select the User Supplied Value radio button and enter the IP address and network mask data in the appropriate fields.

    2. Select the Additional Flags button to display the Additional Flags dialog box, which shows advanced configuration parameters for the selected interface.

    3. Select the check boxes and radio buttons for the interface options that you want to enable and enter values where necessary for optional ifconfig arguments.

    4. Go to step 7.

  6. To configure a Token Ring interface, do the following:

    1. Enter the IP address for the host device in the IP Address field.

    2. Enter the mask variable for the interface in the Network Mask field.

    3. Select the Additional Flags button to display the Additional Flags dialog box, which shows advanced configuration parameters for the selected interface.

    4. Select the check boxes and radio buttons for the interface options that you want to enable and enter values where necessary for optional ifconfig arguments. Select the appropriate adapter speed: 4 or 16.

    5. Go to step 7.

  7. Select OK to validate the parameters you entered and to close the Additional Flags dialog box. The dialog box for the adapter you are configuring is displayed.

  8. Select OK to validate the configuration for network interface and close the dialog box for the adapter. The NIC dialog box is displayed.

  9. Repeat steps 2 through 8, if necessary, to configure additional adapters; otherwise, select OK start network services and apply your changes now. The system applies the changes and closes the NIC dialog box.

You can also use the NIC dialog box to modify and deconfigure network interfaces. See the online help for more information.

Note

After you have configured a system to use the network for the first time, CDE becomes network-dependent, and it might function inconsistently if network services become unavailable. Therefore, if you modify or deconfigure the network interface on a system with only one interface, your system might be left in a unpredictable state. For this reason, it is best to reboot immediately after modifying the network interface to prevent problems. Furthermore, if you deconfigure the network interface, you must configure a new network interface to replace it before rebooting.

For information about monitoring and testing the connectivity of the network interfaces that you have configured, see Chapter 9.

2.3.2    Configuring the rwhod Daemon

To configure the rwhod daemon, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up remote who services (rwhod) to display the Remote Who dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman rwhod

    The utility asks if you want to run the remote who service on your system.

  2. Select the Yes radio button to enable the remote who service.

  3. Select the appropriate rwhod flag radio button.

  4. Select OK to save the changes. The utility notifies you that the changes are saved and asks if you want to apply the changes now.

  5. Select Yes to apply your changes now, or select No to close the Routing Services dialog box and apply the changes the next time you reboot your system.

  6. Select OK to dismiss the informational message and to close the Remote Who dialog box.

You can also use the Remote Who dialog box to disable the rwhod daemon. See the online help for more information.

2.3.3    Configuring the routed Daemon

To configure the routed daemon, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up routing services (gated, routed, IP Router) to display the Routing Services dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman routing

    The utility displays a list of options you can use to configure the gated and routed daemons and to set up your system as an IP router.

  2. Select Yes (use routed) radio button to enable the routed daemon.

  3. Select the appropriate checkbox if you want to run your system as an IP router.

  4. Select the appropriate check box if you want to run the routed daemon on a gateway.

  5. Select the Supply RIP Data radio button if you want the routed daemon to run on a gateway host and supply Routing Information Protocol (RIP) data. Select the Run Quietly radio button if you do not want the routed daemon to supply RIP information.

  6. Select the Configure Gateways button to display the Gateways dialog box. Do the following:

    1. Select Add to add a new gateway. The Add/Modify dialog box is displayed.

    2. In the Destination Type field, select the Network radio button if the destination is a network. Select the Specific Host radio button if the destination is a host.

    3. Enter the destination name, IP address, or "default" in the Destination field.

    4. Enter the name or IP address of the gateway host in the Gateways field.

    5. Enter the hop count in the Hop Count field.

    6. Select one of the Gateway Type radio buttons.

    7. Select OK to validate the information you entered and close the Add/Modify dialog box. Repeat steps a through g for additional gateways.

    8. Select OK to save the changes and close the Gateways dialog box.

  7. Select OK in the Routing Services dialog box to save the changes. The utility displays a dialog box to confirm the changes and to ask if you want to start the daemon now.

  8. Select Yes to start the daemon and apply your changes now, or select No to close the Routing Services dialog box and apply the changes the next time you reboot your system.

    If you choose Yes, you are informed that the daemon is running. Select OK to dismiss the message and to close the Routing Services dialog box.

You can also use the Routing Services dialog box to disable the routed daemon. See the online help for more information.

See routed(8) and gateways(4) for more information about the routed daemon and the gateways file.

2.3.4    Configuring the gated Daemon

To configure the gated daemon, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up routing services (gated, routed, IP Router) to display the Routing Services dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman routing

    The utility displays a list of options you can use to configure the gated and routed daemons and to set up your system as an IP router.

  2. Select the Yes (use gated) radio button to enable the gated daemon.

  3. Select the appropriate check box if you want to run your system as an IP router.

  4. Enter the file name of the gated configuration file in the Configuration File field.

    Note

    To configure the gated daemon, you must set up the /etc/gated.conf file in the format specified in gated.conf(4). A default /etc/gated.conf file is provided when you install the software.

  5. Select OK in the Routing Services dialog box to save the changes. A dialog box is displayed to confirm the changes and to ask if you want to start the daemon now.

  6. Select Yes to start the daemon and apply your changes now, or select No to close the Routing Services dialog box and apply the changes the next time you reboot your system.

    If you choose Yes, you are informed that the daemon is running. Select OK to dismiss the message and to close the Routing Services dialog box.

You can also use the Routing Services dialog box to disable the gated daemon. See the online help for more information.

See gated(8) and gated.conf(4) for more information about the gated daemon and the gated.conf file.

2.3.5    Configuring the System as an IP Router

In order to function as an IP router, your system must have two network interfaces installed and configured and must have the routed or gated daemon configured. To configure the system as an IP router, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up routing services (gated, routed, IP Router) to display the Routing Services dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman routing

    The utility displays a list of options you can use to configure the gated and routed daemons and to set up your system as an IP router.

  2. Select the appropriate check box to run your system as an IP router.

  3. Select OK to save the changes. A dialog box is displayed to confirm the changes and to ask if you want to start or restart the routed or gated daemon.

  4. Select Yes to start the daemon and apply your changes now, or select No to close the Routing Services dialog box and apply the changes the next time you reboot your system.

    If you choose Yes, you are informed that the daemon is running. Select OK to dismiss the message and to close the Routing Services dialog box.

You can also use the Routing Services dialog box to deconfigure the system as an IP router. See the online help for more information.

2.3.6    Configuring the Static Routes File

To configure the routes file, you add entries (static routes) to the routes file. Do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up static routes (/etc/routes) to display the Static Routes dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman route

  2. Select Add to add a static route. The Add/Modify dialog box is displayed.

  3. Select one of the Destination Type radio buttons.

  4. For host and net destinations:

    1. Enter the full name or IP address of the destination network or host in the Destination field.

    2. Select one of the Route Via radio buttons. Select the Gateway button if the route is through a gateway. Select the Interface button and skip to step 6 if the route is through an interface.

  5. For a gateway, enter the full name or IP address of the gateway host to which messages will be forwarded in the Gateway field.

  6. Select OK to validate the entry and add it to the list. Repeat steps 2 through 6 for additional static routes.

  7. Select OK to save the current changes. A dialog box is displayed to confirm the changes and to ask if you want to start the static routes service.

  8. Select Yes to start the service and apply your changes now. Or, select No to close the Static Routes dialog box and apply the changes the next time you reboot your system.

    If you choose Yes, select OK to close the Static Routes dialog box.

You can also use the Static Routes dialog box to modify and delete entries in the routes file. See the online help for more information.

See routes(4) for more information about the routes file.

2.3.7    Configuring the hosts File

To configure the hosts file, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up hosts file (/etc/hosts) to display the Hosts dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman host

  2. Select Add to add a host. The Add/Modify dialog box is displayed.

  3. Enter an official host name in the Host Name field.

  4. Enter the IP address of the new host in the Host Address field.

  5. Optionally, enter any unofficial name or names for this host in the Aliases field. Also, provide pertinent information, for example, the location of the host, in the Comment field.

  6. Select OK to validate the entry and add it to the list. Repeat steps 2 through 6 for additional hosts.

  7. Select OK to update the /etc/hosts file and to close the Hosts dialog box.

You can also use the Hosts dialog box to modify and delete entries in the hosts file. See the online help for more information.

See hosts(4) for more information about the hosts file.

2.3.8    Configuring the hosts.equiv File

To configure the hosts.equiv file, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up host equivalency file (/etc/hosts.equiv) to display the Hosts Equivalency dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman hosteq

  2. Select Add to add a host. The Add/Modify dialog box is displayed.

  3. Enter the remote host name in the Host field.

    Note

    If the host is not on the network, you cannot add the host.

  4. Enter the name of a user on the remote host in the User field.

  5. Select OK to validate the entry and add it to the list. Repeat steps 2 through 5 for additional remote hosts.

  6. Select OK to update the /etc/hosts.equiv file and to close the Hosts Equivalency dialog box.

The Hosts Equivalency dialog box also enables you to modify and delete entries in the hosts.equiv file. See the online help for additional information.

See hosts.equiv(4) for more information about the hosts.equiv file.

2.3.9    Configuring the networks File

To configure the networks file, do the following:

  1. From the SysMan Menu, select Networking-->Basic Network Services-->Set up the networks file (/etc/networks) to display the Networks dialog box.

    Alternatively, enter the following command on a command line: 

    # /usr/bin/sysman networks

  2. Select Add to add a network. The Add/Modify dialog box is displayed.

  3. Enter the official network name in the Network Name field.

  4. Enter the IP address of the network in the Network Address field.

  5. If an unofficial name (alias) is assigned to the new network, enter the aliases in the Aliases field.

  6. Select OK to validate the entry and add it to the list. Repeat steps 2 through 6 for additional networks.

  7. Select OK to update the /etc/networks file and to close the Networks dialog box.

You can also use the Networks dialog box to modify and delete entries in the networks file. See the online help for more information.

See networks(4) for more information about the networks file.

2.3.10    Configuring IP Aliases

An IP alias is an additional network address for an interface. The alias is usually an address in the same subnet as the primary IP address on the interface.

To configure an IP alias, you need the following information:

To configure an IP alias, do the following:

  1. Add the IP address and host name to the /etc/hosts file (see Section 2.3.7).

  2. Edit the /etc/inet.local file and add the command to configure the alias. Use the following syntax: 

    
ifconfig interface alias IP_alias_address netmask IP_alias_netmask

    For example: 

    ifconfig tu0 alias 18.54.76.129 netmask 255.255.255.0

    See ifconfig(8) for more information on ifconfig parameters.

  3. Restart network services by entering the following command: 

    # rcinet restart

2.4    Managing Multiple Network Interfaces

This section describes how to perform the following tasks on systems that contain multiple network interfaces:

2.4.1    Configuring NetRAIN

Before you set up the NetRAIN virtual interface, note the following hardware restrictions and configuration tips:

NetRAIN configuration parameters are stored in the /etc/rc.config file along with the parameters for other network interfaces. Use the rcmgr utility to change the values of the variables. For more information about the rcmgr utility, see rcmgr(8).

Note

The NetRAIN parameters in the following steps are case sensitive and must be typed in uppercase as shown.

To configure NetRAIN, do the following:

  1. Log in as root.

  2. Construct the NetRAIN set or sets, as follows:

    1. Set the NetRAIN interface name or names: 

      # rcmgr set NRDEV_n netrain-interface-id

      The netrain-interface-id must have the form nrn.

      Specify the same integer n for the NRDEV_n variable and the nrn interface. For example, if no NetRAIN interfaces are configured on your system, you can specify NRDEV_0 and nr0, respectively.

    2. Indicate which network interfaces will be part of the NetRAIN set or sets and, if necessary, provide failover timeout values: 

      # rcmgr set NRCONFIG_n interface-id,interface-id [nrtimers integer,integer]

      Note

      When specifying the interfaces, do not leave any spaces between the interface-id parameters and the commas. For example, for two Ethernet interfaces, you can specify tu0,tu1 but not tu0, tu1.

      The nrtimers values dictate how long the system is to wait before switching between interfaces. For more information about nrtimers values, see ifconfig(8).

    3. Indicate to the system that you have configured a NetRAIN set: 

      # rcmgr set NR_DEVICES integer

      Increment integer by the number of NetRAIN sets you have created. For example, if you create one NetRAIN set, integer is 1.

  3. Configure the network parameters for the NetRAIN set or sets that you created, as follows:

    1. Set the interface name: 

      # rcmgr set NETDEV_n netrain-interface-id

      For netrain-interface-id, use the same nrn ID you specified in step 2a.

      If you configured other network interfaces in the rc.config file, you need to find and use the next available NETDEV_n variable. For example, if you used NETDEV_0 to configure an Ethernet card that is not part of the NetRAIN set, the next available variable is NETDEV_1.

    2. Set the ifconfig parameters that will be used to initialize the NetRAIN interface: 

      # rcmgr set IFCONFIG_n IP-address netmask network-mask

      As in step 3a, if you configured other network interfaces in the rc.config file, you need to use the next available IFCONFIG_n variable.

    3. Indicate to the system that you have configured an additional network interface: 

      # rcmgr set NUM_NETCONFIG integer

      Increment integer by the number of NetRAIN interfaces you have created. If you configured other network interfaces in the rc.config file, you need to add the number of NetRAIN interfaces to the current NUM_NETCONFIG value from that file.

  4. Restart network services to apply the changes.

After you configure a NetRAIN set, the NetRAIN interface is available each time you restart your system.

Optionally, you can configure NetRAIN interfaces from the command line by using the ifconfig command, but the changes are not preserved when you reboot. For more information, see ifconfig(8).

Example 2-1 and Example 2-2 show the commands you would enter to establish two different NetRAIN configurations.

To create one NetRAIN set with two Ethernet interfaces, tu0 and tu1, on a system where no other network interfaces have been configured, you would enter the commands in Example 2-1.

Example 2-1:  Creating One NetRAIN Set

# rcmgr set NRDEV_0 nr0 [1]
# rcmgr set NRCONFIG_0 tu0,tu1 [2]
# rcmgr set NR_DEVICES 1 [3]
# rcmgr set NETDEV_0 nr0 [4]
# rcmgr set IFCONFIG_0 18.240.32.40 netmask 255.255.255.0 [5]
# rcmgr set NUM_NETCONFIG 1 [6]

  1. Creates a NetRAIN set called nr0. [Return to example]

  2. Indicates that the nr0 set consists of the tu0 and tu1 interfaces. Both interfaces must be marked "down" prior to this command. [Return to example]

  3. Indicates to the system that there is one NetRAIN set. [Return to example]

  4. Creates a network interface called nr0 for the NetRAIN virtual interface. [Return to example]

  5. Defines the IP address and network mask for the NetRAIN virtual interface. [Return to example]

  6. Indicates to the system that there is one network interface. [Return to example]

To create two NetRAIN sets, one with two FDDI interfaces called fta0 and fta1 and the other with two ATM LANE interfaces called elan0 and elan 1, on a system where one other network interface has been configured (suppose NETDEV_0 is tu0), you would enter the commands in Example 2-2.

Example 2-2:  Creating Two NetRAIN Sets

# rcmgr set NRDEV_0 nr0 [1]
# rcmgr set NRDEV_1 nr1
# rcmgr set NRCONFIG_0 fta0,fta1 [2]
# rcmgr set NRCONFIG_1 elan0,elan1 nrtimers 4,16 [3]
# rcmgr set NR_DEVICES 2 [4]
# rcmgr set NETDEV_1 nr1 [5]
# rcmgr set NETDEV_2 nr2
# rcmgr set IFCONFIG_1 18.240.31.40 netmask 255.255.255.0 [6]
# rcmgr set IFCONFIG_2 18.240.31.42 netmask 255.255.255.0
# rcmgr set NUM_NETCONFIG 3 [7]

  1. Creates two NetRAIN sets called nr0 and nr1. [Return to example]

  2. Indicates that the nr0 set consists of the tu0 and tu1 interfaces. Both interfaces must be marked "down" prior to issuing this command. [Return to example]

  3. Indicates that the nr1 set consists of the elan0 and elan1 interfaces. Both interfaces are currently idle. Also provides nrtimers failover values for the set. The values in this example are suggested starting values for ATM LANE. They might not work for your configuration, as described at the beginning of this section. For more information about nrtimers values, see ifconfig(8). [Return to example]

  4. Indicates to the system that there are two NetRAIN sets. [Return to example]

  5. Creates network interfaces called nr0 and nr1 for the two NetRAIN virtual interfaces. [Return to example]

  6. Defines the IP address and network mask for each NetRAIN virtual interface. [Return to example]

  7. Indicates to the system that there are three network interfaces, the two NetRAIN virtual interfaces and the preexisting Ethernet interface. [Return to example]

2.4.2    Monitoring NetRAIN Activity

To check which member of a NetRAIN set is the active interface, use the ifconfig command. For example: 

# ifconfig nr0
nr0: flags=8c63     NetRAIN Attached Interfaces: ( fta0 fta1 ) Active Interface:
 ( fta0 )inet 18.240.32.40 netmask ffffff00 broadcast 18.240.32.255 ipmtu 4352

This example shows that:

To see the status of all set members while the NetRAIN interface is running, use the niffconfig command. For example: 

# niffconfig -u
Interface:   tu1, state: DEAD, t1: 4, dt: 2, t2: 10, time to dead: 0,
current_interval: 2, next time: 2
Interface:   nr0, state: GREEN, t1: 4, dt: 2, t2: 10, time to dead: 0,
current_interval: 4, next time: 4
Interface:   tu0, state: GREEN, t1: 4, dt: 2, t2: 10, time to dead: 0,
current_interval: 4, next time: 4

In this example, you can see that the virtual interface nr0 is running and NetRAIN is using tu0 for communication. This example also shows the nrtimers values for each member of the set. See ifconfig(8) for more information on these values.

For more information about monitoring the connectivity of network interfaces, see Section 9.1.

2.4.3    Configuring a Link Aggregation Group

Before configuring a link aggregation group, verify that the link aggregation kernel subsystem (lag.mod) is configured in the kernel, by issuing the following command: 

# sysconfig -q lag

If the lag: subsystem attributes are not displayed, do the following:

  1. Edit the system configuration file and add the following entry to it: 

    options LAG

    The default configuration file is /sys/conf/SYSTEM_NAME, where SYSTEM_NAME is the name of your host processor, in uppercase letters.

  2. Build a new kernel by issuing the doconfig -c command. If you are unfamiliar with rebuilding the kernel, see System Administration.

  3. Reboot the system. Make sure that there are no other users on the system. Use a command similar to the following: 

    # shutdown -r +5 "Adding Link Aggregation software option ..."

You are now ready to configure a link aggregation group. Before you set up the link aggregation virtual interface, note the following hardware restrictions and configuration tips:

To configure a link aggregation group, do the following:

  1. Log in as root.

  2. Edit the /etc/inet.local file.

  3. Enter a lagconfig -c statement to create a link aggregation group.

  4. Enter a lagconfig -p statement to enable one port (physical interface) for link aggregation. To enable additional ports, enter additional lagconfig -p statements.

  5. Enter an ifconfig statement to assign an IP address to the link aggregation group virtual interface and enable it.

  6. Save the changes and close the file.

  7. Restart network services by entering the following command: 

    # rcinet restart

After you configure a link aggregation group, it is available each time you restart your system.

Optionally, you can configure a link aggregation group from the command line by using the lagconfig and ifconfig commands. However, the changes do not persist across system reboots. For more information, see lagconfig(8) and ifconfig(8).

Example 2-3 shows the statements you would add to the /etc/inet.local file to create a link aggregation group made up of three ports or interfaces.

Example 2-3:  Sample Link Aggregation Statements

# lagconfig -c  [1]
# lagconfig -p tu0 key=1  [2]
# lagconfig -p tu1 key=1  [3]
# lagconfig -p tu2 key=1  [4]
# ifconfig lag0 16.1.2.3 netmask 255.255.255.0 up  [5]

  1. Creates a link aggregation group with a default key value and the next available interface number. Since no link aggegation group is configured on the system, this creates a group with a key value of 1 and an interface number of 0 (lag0). [Return to example]

  2. Enables tu0 for link aggregation. The interface must be marked "down" prior to issuing this command. [Return to example]

  3. Enables tu1 for link aggregation. The interface must be marked "down" prior to issuing this command. [Return to example]

  4. Enables tu2 for link aggregation. The interface must be marked "down" prior to issuing this command. [Return to example]

  5. Sets the IP address of the link aggregation virtual interface to 16.1.2.3. The enabled ports then attach to the link aggregation group that has the same key assigned to it, and are available to carry traffic. [Return to example]

2.5    Enabling Access Filtering on an Interface

Interface access filtering helps you detect and prevent IP spoofing attacks. To enable interface access filtering on an interface, do the following:

  1. Create an /etc/ifaccess.conf file and add entries against which the source address of input packets are checked.

  2. Use the ifconfig command with the +filter parameter to enable access filtering on the network interface.

See ifaccess.conf(4) and ifconfig(8) for more information.

2.6    Displaying and Modifying the FDDI Parameters

You use the fddi_config command to display and modify the FDDI adapter parameters.

To display the FDDI adapter parameters, use the fddi_config command with the following syntax:

fddi_config -i interface_name -d

To modify the FDDI adapter parameters, log in as root and use the fddi_config command with one or more of the options in Table 2-2.

Table 2-2:  Options to the fddi_config Command

Option Function
-i interface_name Changes or displays the FDDI characteristics for interface_name. You must provide the interface name.
-c counter_update_interval Determines how often the driver counters are updated by the DEFTA adapter. The default is 1 second. Setting the interval time to zero (0) disables counter updates. (For the DEFTA (fta) FDDI interface only.)
-d Displays the FDDI interface parameters you can set.
-l lem_threshold Sets the error rate threshold of Link Error Monitor (LEM). The LEM error rate threshold is 1×10-n, where n ranges from 5 to 8, inclusively. The default LEM threshold is 1×10-8.
-p [1|0] Sets the ring purger state for the specified FDDI interface. A value of 1 enables the ring purger ability; a value of 0 disables it.
-r restricted_token_timeout Sets the Restricted Token Timeout parameter, defining how long a single restricted mode dialog can last before being terminated. The range for this parameter is from 0 to 10000 milliseconds. The default value is 1000 milliseconds.
-t token_request_time Sets the Request Token Rotation Time (T_req) for interface_name. T_req is used during the ring initialization process to negotiate a Target Token Rotation Time (TTRT) for the ring. The range for this parameter is from 4.0 milliseconds to 167.77208 milliseconds. The default value is 8.0 milliseconds.
-v valid_transmit_time Sets the Valid Transmission Time (TVX) timer for a specific FDDI interface. The range for the TVX timer is from 2.35 milliseconds to 5.2224 milliseconds. The default is 2.6214 milliseconds.
-x [1|0] Enables (1) or disables (0) full-duplex operation for the interface. If the full-duplex operation is enabled, the interface is in one of the following states: Idle, Request, Confirm, or Operational. (For the DEFTA (fta) FDDI interface only.)

See fddi_config(8) for more information on this command and its options.

The following example shows how to display the FDDI interface parameters you can set: 

% /usr/sbin/fddi_config -i fza0 -d
fza0 ANSI FDDI settable parameters
 
Token Request Time:                0.0000 ms
Valid Transmission Time:           0.0000 ms
LEM Threshold:                     0
Restricted Token Timeout:          15.8314 ms
Ring Purger State:                 (null)
 
fza0 Full Duplex Mode: Disabled
 
fza0 Counter Update Interval: 10 sec

The following example shows how to change the Token Request Time (TRT) value for the fza0 interface to 10.2: 

# fddi_config -t10.2 -i fza0

The following example shows how to turn the ring purger off: 

# fddi_config -p 0 -i mfa0

2.7    Managing Token Ring Source Routing

Source routing is a bridging mechanism that systems on a token ring LAN use to send messages to a system on another interconnected token ring LAN. Under this mechanism, the system that is the source of a message uses a route discovery process to determine the optimum route over token ring LANs and bridges to a destination system. The source system stores the optimum routes in its source routing table.

When the system is booted with the DETRA adapter installed and configured, token ring source routing is initialized by default. To manage token ring source routing, use the srconfig command.

Table 2-3 shows the srconfig command options. All srconfig command options are case insensitive; type them in uppercase, lowercase, or mixed case. The short form for each flag is indicated by uppercase letters.

Table 2-3:  Options to the srconfig Command

Option Function
-DElentry mac_address [Footnote 1] Deletes a source routing table entry.
-DISEntry mac_address [Footnote 1] Disables a source routing table entry. This marks the entry as Stale.
-RAttr Displays the source routing attributes.
-RCounter Displays the source routing counters.
-REntry mac_address Displays a specific source routing table entry.
-RTable Displays the source routing table.
-SETAgetimer timer [Footnote 1] Sets the value of the Source Routing Aging Timer, specifying the length of time a source routing table entry remains valid until being marked as invalid or Stale. If not set, the system default is 120 seconds.
-SETDsctimer timer [Footnote 1] Sets the Source Routing Discovery Timer, specifying the amount of time a route discovery process can take before it terminates. If not set, the system default is 5 seconds.
-SETMaxentry value [Footnote 1] Sets the maximum number of entries allowed in the source routing table. The range for this entry is a multiple of 256 from 1024 to 2048. This parameter can be increased, but not decreased. If not set, the system default is 1024.
-u Specifies that the MAC addresses are in uncanonical form. This option can be used with the -DElEntry mac_address, -DISEntry mac_address, and -RTable options only.
-Zcounter Sets the source routing counters to zero.

See srconfig(8) for more information on this command and its options.

The following example increases the number of routing table entries from 1024 to 1280 by using the shortened form of the -SetMaxEntry option: 

# srconfig -setm 1280
Current SR Table size is : 1024
New SR Table size is : 1280

The following example displays the source routing attributes by using the shortened form of the -RAttr option: 

# srconfig -ra
Source Routing is enabled
Current SR Aging Timer     : 120
Current SR Discovery Timer : 10
Current SR Table size is   : 1024

The following example displays the source routing counters by using the shortened form of the -RCounter option: 

# srconfig -rc
ARE Frames Sent          : 00000001
ARE Frames received      : 00000000
Route Discovery Failures : 00000001

The following example displays all entries, with MAC addresses in canonical form, in the source routing table, by using the shortened form of the -RTable option. The backslash (\) character indicates line continuation and does not appear in the actual output. 

# srconfig -rt
Target Node MAC Address 00-00-0C-01-08-E9 (ip = 130.180.4.3) \
Have Route  [1]
Routing Information: SRF, length 8, direction 0,largest frame \
4472 octets  [2]
Route Descriptors: 021C 7FFC 0220 0000 0000 0000 0000 0000    [3]
 
Target Node MAC Address 00-00-C9-10-1B-F5 On Ring    [4]
 
Target Node MAC Address 08-00-2B-2C-F1-F9 (ip = 130.180.4.2)  \
Stale (Have Route)   [5]
Routing Information: SRF, length 8, direction 0,largest frame 4472 octets
Route Descriptors: 021C 7FFC 0220 0000 0000 0000 0000 0000
 
Target Node MAC Address 00-00-C9-0B-33-80 Stale (On Ring)

  1. Have Route indicates the source system has a valid path to the destination system. [Return to example]

  2. Information returned by the destination system in response to the route discovery process. [Return to example]

  3. The LAN segments and bridges that constitute the path to the destination system. [Return to example]

  4. On Ring indicates the destination system is on the same ring as the source system and does not need source routing. [Return to example]

  5. Stale indicates the entry is invalid and needs to be updated by the route discovery process. [Return to example]

The following example shows all entries, with MAC addresses in noncanonical form, in the source routing table by using the shortened form of the -RTable option. The backslash (\) character indicates line continuation and does not appear in the actual output. 

# srconfig -rt -u
Target Node MAC Address 00:00:30:80:10:97 (ip = 130.180.4.3) Have Route
Routing Information: SRF, length 8, direction 0,largest frame 4472 octets
Route Descriptors: 021C 7FFC 0220 0000 0000 0000 0000 0000
 
Target Node MAC Address 00:00:93:08:D8:AF On Ring
 
Target Node MAC Address 10:00:D4:34:8F:9F (ip = 130.180.4.2) Stale \
 (Have Route)
Routing Information: SRF, length 8, direction 0,largest frame 4472 octets
Route Descriptors: 021C 7FFC 0220 0000 0000 0000 0000 0000
 
Target Node MAC Address 00:00:93:D0:CC:01 Stale (On Ring)

2.8    Displaying and Modifying the Token Ring IP MTU Size

By default, the DETRA adapter uses an IP maximum transfer unit (MTU) size of 4092 bytes. In a multivendor environment with different adapters using different IP MTU sizes, the bridges connecting different networks can be set up to forward smaller packet sizes. As a result, bridges might drop packets or remote hosts might reject packets. If either occurs on your network, reduce the IP MTU size for all hosts on the network and ensure that all hosts use the same size.

The following command displays the DETRA interface IP MTU size as 4092 bytes: 

% ifconfig tra0
tra0: flags=9863<UP,BROADCAST,NOTRAILERS,RUNNING>
     inet 16.141.208.3 netmask ffffff00 broadcast 16.141.208.255 ipmtu 4092
 

The following example sets the IP MTU size of DETRA interface to 2044 bytes: 

% ifconfig tra0 ipmtu 2044

2.9    Managing Network Quality of Service

As applications place increasing demands for bandwidth on the Internet network, increasing the network bandwidth is only a temporary solution. Newer real-time applications demand both increased bandwidth and low latency. Clearly, the importance of bandwidth management is increasing.

An IP network with its Best Effort delivery service performs a form of passive bandwidth management. If an outgoing queue is full, indicating high network traffic and congestion, the packets are quietly dropped. Some upper-level protocols can detect data loss, others cannot.

Quality of service (QoS) is the phrase commonly associated with the concept of actively managing network bandwidth. In this scenario, all network elements (for example, hosts, applications, and routers) and all network protocol layers cooperate to ensure consistent traffic and service end-to-end in a network. Network bandwidth for real-time applications is reserved, while sufficient bandwidth remains for best-effort traffic.

The major network QoS components in this operating system are as follows:

2.9.1    Managing the Traffic Control Subsystem

The Traffic Control subsystem performs the following tasks:

See iftcntl(8) for more information.

The rsvpd daemon requires that traffic control be enabled on the local system in order to install and modify flows and filters for a specific network interface. To enable traffic control on your local system, check that the ether_cl_scheduler system attribute is enabled (set to 1). If it is not enabled, enable it by using the sysconfig command or dxkerneltuner. Then, reboot the system.

2.9.2    Managing RSVP

RSVP assigns QoS to specific IP data flows or sessions, which can be either multipoint-to-multipoint or point-to-point. In order to receive data packets for a particular multicast session, a host must have joined the corresponding IP multicast group. A given session may have multiple senders and if the destination is a multicast address, multiple receivers.

The rsvpd daemon performs the following functions:

See rsvpd(8) for more information.

2.9.2.1    Starting and Stopping rsvpd

To start the rsvpd daemon, enter the following command: 

# /usr/sbin/rsvpd

If you want to start the daemon automatically at system boot time, include the command in the /etc/inet.local file. See rsvpd(8) for more information on the daemon and its options.

To stop the rsvpd daemon, enter the following command: 

# kill -9 `cat /var/run/rsvpd.pid`

The rsvpd daemon does not start or stop any applications during its startup or shutdown procedures. It also does not maintain any on-disk configuration information about applications. Whenever the rsvpd daemon starts, it has no information about previous reservations.

Typically all daemons on the operating system are started or stopped together, as the system changes run levels. But applications must correctly handle situations where they start before the rsvpd daemon, or are running while the rsvpd daemon is restarted. In these situations, local applications need to reinitiate communications with the rsvpd daemon.

2.9.2.2    Adding and Deleting Network Interfaces

When you add or delete a network interface on your system, you must stop and restart the rsvpd daemon in order to for it to update its table of available interfaces. Enter the following commands: 

# kill -9 `cat /var/run/rsvpd.pid`
# /usr/sbin/rsvpd

2.9.2.3    Displaying RSVP Session Information

You can display RSVP session information on routing systems or end systems to determine if RSVP is working correctly on your system. RSVP session information will show you if connections are are being set up and if reservations are being honored.

To monitor active RSVP sessions on the local system, enter the following command: 

# /usr/sbin/rsvpstat

By default, the rsvpstat command displays a list of all RSVP sessions, sender and receiver, active on this system. Information includes the session number, destination address, IP protocol, port number, and the number of PATH and RESV states for the session.

To display sender information, including the contents of the actual PATH message from the sender, enter the following command: 

# /usr/sbin/rsvpstat -Sv

To display receiver information, including the contents of the actual RESV message from the receiver, enter the following command: 

# /usr/sbin/rsvpstat -Rv

See rsvpstat(8) for more information.