User's Guide

Introduction

This chapter briefly describes the functions, capabilities, and benefits of the 8275 Model 416 High Performance Ethernet Workgroup Switch. This information helps you to plan for and use the switch in your network.

Product overview

Fast Ethernet switching continues to evolve from high-end backbone applications to desktop-switching applications. The switch provides a low-cost and powerful Layer 2 switch solution. It is an attractive base switch offering with the following key functions:

• High-performance, Layer 2, managed switch
• 16 base ports (10/100BASE-TX), expandable from 18 to 32 ports, depending on the combination of the following optional feature modules:
  • 8-Port 10/100BASE-TX
  • 8-Port 100BASE-FX
  • 4-Port 100BASE-FX
  • 2-Port 1000BASE-SX
• Robust management support; VT100 terminal interface, Web interface, SNMP
• Backplane performance 10 gigabits per second Ethernet switching
• Desktop and segment switching infrastructure
• Affordable migration to higher-performance networks

As a network administrator, you have a choice of three easy-to-use management methods: VT100 terminal interface, Web-based, and Simple Network Management Protocol (SNMP). These management methods enable you to configure, manage, and control the switch locally or from anywhere on the network.

The Spanning Tree Protocol (STP) provides fault tolerance on the network.

Switch functions

This section describes the functional support included in the switch:

• Layer 2 switching
• Virtual local area networks (VLANs)
• Management and user interface
• Self-learning IP
• Link aggregation (trunking)
• Fast spanning tree mode
• Security
• Reliability and serviceability
• Performance
• Flow Control
• Year 2000 (Y2K) compliance

Layer 2 switching

The 8275-416 is a Layer 2 Ethernet switch in which frame forwarding is based on MAC addresses and VLAN membership. The switch supports the IEEE 802.1D (1998) and 802.1Q standards.

802.3x flow control

The switch supports 802.3x flow control, which, when enabled, allows the transmission of data frames to be inhibited for a specified period of time. The default for 802.3x flow control is Disabled. 802.3x flow control is valid only when the port is in full-duplex mode.

Broadcast storm recovery

The switch detects broadcast storms and automatically blocks broadcast traffic to minimize the impact of the broadcast storm on the rest of the network. You can enable or disable this function at a switch level. For all broadcast frames received by the switch, the broadcast storm recovery operation depends on port speed and is described as follows:

1. If the Broadcast Storm Recovery Mode is Enable, and if the broadcast traffic on a 10 Mbps Ethernet port exceeds 20% of the link speed, then the switch blocks the broadcast traffic on the port until the broadcast traffic is returned to 10% or below.
2. If the Broadcast Storm Recovery Mode is Enable, and if the broadcast traffic on a 100 Mbps Ethernet port exceeds 5% of the link speed, then the switch blocks the broadcast traffic on the port until the broadcast traffic is returned to 2.5% or below.
3. Broadcast Storm Recovery is not supported on 1000 Mbps Ethernet ports.
4. If Broadcast Storm Recovery Mode is Disable, the switch does not block the broadcast traffic on the Ethernet port.
5. The switch issues a trap message when traffic exceeds a port's broadcast threshold and when it returns to or below the port's recovery threshold.

Forwarding database

The switch port MAC addresses are stored in the forwarding database. An address learned by the switch is removed from the forwarding database after a period of time if no frames have been received from that address. The default value for the aging period is 300 seconds (5 minutes), but it can be configured by the user. The time values range from 10 seconds to 600 seconds.

The switch forwarding database stores 12 000 entries. When the database is full, no new entries are learned until an existing entry ages out. All frames with unknown destination addresses are multicast to all ports in the appropriate VLAN.

Virtual local area networks (VLANs)

The switch supports VLANs. Appendix E, Introduction to Virtual LANs (VLANs) provides an introduction to VLANs. It describes concepts and terminology, as well as, the benefits of using VLANs. The switch is manageable only through the ports which are members of the Default VLAN (VLAN 1).

Figure 36 and Figure 37 show examples of the panels and descriptions of the parameters used to configure VLANs.

Self Learning IP

Self Learning IP is a configurable function of the switch that learns where IP addresses are in the network so that packets normally sent from one host to another through a router can bypass the router and be sent directly to the destination host address. Self Learning IP is most effective when the switch is used to "front" the router (the switch is positioned logically between the router and the networks to which it belongs). Because of this strategic vantage point in the network (Figure 1), the switch has visibility to all packets flowing to and from the router, as well as between any two switch ports. The switch monitors the traffic to determine if a Layer 2 shortcut can be used instead of subjecting the packet to Layer 3 router processing which can be relatively lengthy.

Figure 1. Self-Learning IP in the network

(artname: fgl2a002.EPS)

The Self Learning IP function essentially:

• Learns the network structure, classifying attached devices as routers or hosts.
• Maintains knowledge of network structure, aging out unused devices over time.
• Expedites packet flow through the network by circumventing the router whenever possible.

Device learning and classification is accomplished by watching ARP replies that flow naturally through the switch with the addition of active probing to determining whether a device is a router or a host. When an IP packet enters the switch, the Self Learning IP function compares the destination MAC address against the list of known routers, then checks if the destination IP address is a known host. If both tests pass, the packet is automatically re-addressed to the destination IP host device and is sent out the appropriate switch port.

A Router Table and Host Table are used to manage information learned about router and host devices, respectively. To keep this network information current, the Router Table entries are refreshed every 5 minutes while Host Table entries are checked every 3 minutes. Devices which are no longer active are dropped from their table; devices may be relearned at a later time as conditions change.

To use the Self Learning IP function, the switch must be configured with the following:

• IP information (see Network Connectivity Configuration Menu in Chapter 4).
• Enable the Self Learning IP function (see the Switch Configuration Menu in Chapter 4).

Once Self Learning IP is enabled, the following information is available:

• Self Learning IP statistics for the switch (see Switch Detailed Statistics Menu in Chapter 4).
• IP and MAC addresses of routers learned (see Self Learning IP Router Table Menu in Chapter 4).
• Host IP statistics (see Self Learning IP Host Address Menu in Chapter 4).

Note that the Packets Switched count included in the Switch Detailed Statistics Menu may not reflect the absolute latest value. The information used for updating this count is obtained as individual Host Table entries are refreshed, so while this value can change over time, it does not necessarily update at the same frequency as other statistics on the menu.

While intended as an autonomous feature, Self Learning IP is affected by certain changes in switch configuration. If Self Learning IP is enabled and the switch IP address is reset, the Self Learning IP function is automatically disabled. The Self Learning IP Router and Host Tables are cleared whenever there is a link aggregation configuration change (see Link aggregation (trunking) in this chapter), forcing router and host devices to be relearned.

Link aggregation (trunking)

Link aggregation, also called trunking allows 802.3 MAC interfaces to be grouped together logically to appear as one super-link. The super-link or Link Aggregation Group (LAG) has access to the combined bandwidth of all links.

The Sun Trunking(TM) 1.0 specification is supported for the 10/100BASE-TX and 100BASE-FX ports. All members of the trunk must support Sun Trunking(TM) 1.0. For information about configuring trunks, see Trunk management menu. Up to 8 trunks can be configured.

Advantages of trunking are:

• Fault tolerance: Failure of one or more of the links in the LAG are handled gracefully. If a link of the LAG fails, the flows mapped to that link are dynamically reassigned to the remaining links of the LAG.
• Redundancy: Link aggregation also provides automatic, point-to-point redundancy between two devices (switch-to-switch).

Fast spanning tree mode

The IEEE 802.1D spanning tree protocol (STP) is designed to prevent loops in Ethernet networks. To achieve this objective the STP does not allow switches to forward data frames on a link immediately after the link is activated. The STP first listens for spanning tree BPDUs from other switches, then determines whether to put the link into forwarding state. When a default IEEE spanning tree timer value of 15 seconds is used for the forward delay timer, a link can start forwarding traffic 30 seconds after it becomes active on the network.

In networks with shared media hubs, there is a trend to attach network stations (or hosts) directly to multi-port bridges (otherwise known as switches). Unfortunately, the 802.1D spanning tree protocol has not been changed to accommodate this trend. So, when a network station is ready to send data, the switch does not allow the network station to communicate on the network until STP puts the port in forwarding state. The 30 second delay forces the network station users to wait longer before accessing the network. Even worse, some higher level protocols running on the network station may time out, generate error messages or not work at all.

The 802.1D standard specifies that when a link comes up on the network, the spanning tree state is set to "Listening". After the forward delay timer expires, the spanning tree state is set to "Learning", and after another forward delay timer interval the state is set to "Forwarding". The forward delay timer is set for the entire network by the root bridge. The default value for this timer is 15 seconds.

With the Fast Spanning Tree function, the transition from "Listening" to "Learning" and the transition from "Learning" to "Forwarding" takes approximately 6 to 8 seconds. The forward delay timer behavior reverts to the 802.1D standard (15 seconds) after a port goes into forwarding state or blocking state.

Fast Spanning Tree mode is configurable by port. The default is the standard 802.1D protocol. When a port is configured in Fast Spanning Tree mode, it takes approximately 6 to 8 seconds before traffic can be forwarded on the link. Spanning tree can also be disabled on the port. For details about configuring ports for Fast Spanning Tree mode, see the Port Configuration Menu in Chapter 4.

Management and user interfaces

You have a choice of these easy-to-use management methods:

• A VT100 terminal interface allows you to fully manage the switch using a standard terminal or terminal emulator connected over the network using Telnet or connected to the switch's serial port (EIA 232).

  Accessing the switch describes how to access the switch using this interface and Using the Terminal Interface instructs you about using this interface.

• A Web-based interface enables you to manage the switch through standard Web browsers. There must be a physical path between the Web browser and the switch over the Ethernet network to use this method of connectivity.

  Accessing the switch describes how to access the switch using this interface and Using the Web Interface instructs you about using this interface.

• The switch has a Simple Network Management Protocol (SNMP) agent that the network administrator can access with a standard network manager. The following MIBs (Management Information Base) are supported:
  • MIB II (RFC 1213)
  • 8275-416 Enterprise MIB
  • RMON MIB (RFC 1757)
  • Bridge MIB (RFC 1493)
  • IEEE 802.3 Ethernet (RFC 1643)
• The switch interoperates with the following SNMP Managers:
  • Any standard MIB browser (SNMPv1)
  • IBM Nways(R) Manager for Windows NT(R) (V2.0 or later)
  • IBM Nways Manager for HP-UX (V2.0 or later)
  • IBM Nways Manager for AIX(R) (V2.0 or later)

Security

User access security can be implemented using the following functions of the 8275-416:

• User Accounts: The switch supports up to six accounts (one user with read/write status and five with read-only status) for terminal interface and Web access. Access to the switch configuration panels is password protected. Only one user name with read/write status is allowed to be configured, which prevents potential conflicts in configuration changes. The default Read/Write user name is: admin, and the default password consists of blanks (no password). If you lose the password, contact your IBM service representative.
• SNMP read/write protection based on community name.

Reliability and serviceability

The switch:

• Provides a comprehensive power-on self-test (POST) that ensures that all of its components are functioning correctly.
• Controls a seven-segment LED that allows you to follow the boot sequence.
• Allows you to download software upgrades using any of the management methods.
• Allows you to implement parallel paths for network traffic through the use of spanning tree protocol (STP), which provides a level of fault tolerance and ensures that:
  • Redundant paths are disabled when the main paths are operational
  • Redundant paths are enabled if the main paths fail
• Allows you to configure a port to "see" traffic going into and out of another port on the switch (port monitoring).
• Provides statistics for all ports.

Performance

High performance, Layer 2 switching for the switch consists of:

• Switching for up to 32 ports
• Supporting up to 12 000 end stations
• Processing 64-byte packets at the following rates:
  • 14 880 packets per second to 10-Mbps ports.
  • 148 800 packets per second to 100-Mbps ports
• Detecting broadcast storms and preventing them from impacting the network (Broadcast Storm Recovery).

Year 2000 (Y2K) Compliance

The 8275-416 is Y2K compliant.

When used in accordance with its associated documentation, it is capable of correctly processing and/or receiving date data within and between the 20th and 21st centuries providing all other products (for example, hardware, software, and firmware) used with the switch properly exchange accurate date data.

For additional information about Year 2000 related topics, visit:

http://www.ibm.com/year2000

Hardware

Cabling requirements

Ethernet cables are not provided and must be separately purchased. You can order them through your IBM representative.

Table 1 shows cable type and length requirements. Cable requirements depend on the speed of the network. Cables and connecting hardware must meet the standards specified in the ANSI/TIA/EIA 856-A or CSA T529 standards.

Table 1. Ethernet cable requirements

Ethernet Type	Cable Requirements	Max. Cable Length
10BASE-T	Category 3, 4 or 5 100-ohm STP/UTP cable	100 m (328 ft)
100BASE-TX	Category 5, 100-ohm STP or UTP cable and connecting hardware	100 m (328 ft)
100BASE-FX	62.5-micron multimode fiber (MMF) cabling	2 km (6561 ft) at full-duplex; 412 m (1352 ft) at half-duplex
1000BASE-SX	50/125-micron or 62.5-micron multimode fiber (MMF) cabling	550 m (1804 ft) at full-duplex.

10/100BASE-TX

10BASE-T connections are MDX ports and operate correctly with standard Category 3, 4, or 5 100-ohm UTP or STP cable and connecting hardware, as specified in the ANSI/TIA/EIA 856-A or CSA T529 standards when connected to MDI ports. When connecting to other MDX ports, such as ports of other 8275-416 switches, you must use crossover cables.

Do not use telephone extension cables in 10/100BASE-TX networks. The wire pairs in those cables are not twisted and the cables do not meet other requirements for use in a 10BASE-T network.

For connections to 10/100BASE-TX networks, you can use only Category 5 STP or UTP cables.

100BASE-FX

For connection to 100BASE-FX networks, you can use only 62.5/125 MMF cabling with MTRJ connectors.

1000BASE-SX

For connection to 1000BASE-SX networks, you can use 62.5/125 &mu.m or 50/125 &mu.m multi-mode fiber (MMF) cabling with SC fiber optic connectors.

Front panel

Figure 2. Front panel of the switch.

(artname: fgb0a001.EPS)

Switch LEDs

Switch LEDs are located at the lower left corner of the front panel (left of the single-digit display) and are identified with a vertical bar (I), OK, and Fault. The LED identified with the vertical bar and the OK LED are Green; the Fault LED is amber. The states of the LEDs are on, off, or blinking. They are explained later in this chapter.

Single-Digit Display

The single-digit display is located at the lower left corner of the front panel as shown in Figure 3. During diagnostics, the character displayed indicates the diagnostic test being executed. Once the switch is operational, the character displayed is its unit ID (Table 3).

Serial Port (EIA 232)

The serial port is a standard DB-9 male connector that provides an EIA 232 serial interface (sometimes referred to as the out-of-band management port). Use a null-modem serial cable when connecting to a workstation (Appendix C, Cable Pinout Diagrams). Use a VT100 terminal emulator program to configure your terminal's attached COM port as follows:

• 19200 baud
• 8 data bits
• 1 stop bit
• No parity
• Hardware flow control OFF

See Accessing the switch for more information about connectivity.

Ethernet 10/100BASE-TX Ports

The switch has 16 Ethernet 10/100BASE-TX ports. Each port has two LEDs located at the lower right and left of the connector. Status indications of the Port LEDs are explained later in this chapter.

Feature Module Slots 1 and 2

These feature modules are available to expand port connections for your switch:

• 8-Port 10/100BASE-TX Ethernet Feature Module, P/N 30L6661
• 8-Port 100BASE-FX Ethernet Feature Module, P/N 30L6662
• 4-Port 100BASE-FX Ethernet Feature Module, P/N 31L4054
• 2-Port 1000BASE-SX Ethernet Feature Module, P/N 30L6663

Switch LED status

Switch LEDs are shown in Figure 3 and LED status is explained in the table that follows:

Figure 3. LEDs for the switch.

(artname: fgb0a018.EPS)

Table 2. LED status for the switch.

LEDs			Explanation
I (Green)	OK (Green)	Fault (Yellow)
Off	Off	Off	No power is present, or there is a power supply failure. The switch is not operational.
On	On	Off	The switch is operational.
On	Blinking	Off	Configuration file or Operational Code file transfer is in process. Do not power-off or reset the switch.
On	Off	On	There is a hardware fault. The switch is not operational.
On	Off	Blinking	Diagnostics are in process. The switch is not yet operational.

Note:

Any other state of the LEDs indicates an LED failure.

Single-digit display

The single-digit display (shown in Figure 3) displays characters while diagnostics are running after power is applied to the switch. At the successful completion of diagnostics, the unit number appears in the display (for example, "1" indicates Unit Number 1). Table 3 gives the meaning of other digits that can be displayed and the corrective actions required.

Table 3. Problem indications on the single-digit display when the Fault LED is ON.

Character	Problem	Corrective Action
d	Board RAM problem	Replace the switch.
3	Detected an unsupported feature module.	Remove the feature module and update the operational code, or the feature module is not fully seated in its connector.
4	PIF fault on the feature module or base board.	If feature module Fault LED is On, remove the feature module. If no feature module Fault LED is On, replace the switch.
5 or 6	Non-volatile memory problem.	Replace the switch.
7	Switch memory problem.	Replace the switch.
8	Base board loopback problem.	Replace the switch.
9 or a	Feature module loopback problem.	9 = Feature module in Slot 1 has the fault; remove this feature module. a = Feature module in Slot 2 has the fault; remove this feature module.

Base ports LEDs

The switch has 16 base 10/100BASE-TX ports. LED status for these 16 base ports are shown in Figure 4 and they are explained in Table 4.

Figure 4. LEDs for the base 10/100BASE-TX ports on the switch

(artname: fgb0a021.EPS)

Table 4. Status of LEDs for 16 base 10/100BASE-TX ports

LED	Color	State	Explanation
Right Ethernet Port LED	Green	ON	Indicates a 100-Mbps port.
		OFF	Indicates a 10-Mbps port.
Left Ethernet Port LED	Green	ON	The link is up.
		OFF	The link is down.
		Blinking	Transmitting (Tx) and Receiving (Rx) traffic.

Feature module LEDs

Each feature module has an OK and a Fault LED located at the left side of the faceplate. The OK LED is green and the Fault LED is yellow. LED locations are shown in Figure 5, Figure 6, Figure 7, and Figure 8; LED status of the feature modules are explained in Table 5, Table 6, Table 7, and Table 8.

Status LEDs for the 8-port 10/100BASE-TX Ethernet feature module

Figure 5. LEDs for the 8-port 10/100BASE-TX feature module.

(artname: fgb0a019.EPS)

Table 5. Status of LEDs for 8-port 10/100BASE-TX feature module

LED	Color	State	Explanation
OK	Green	ON	There is power to feature module.
		OFF	There is no power to feature module, no power to the switch, or the module has failed.
Fault	Yellow	ON	There is a module fault.
		OFF	There is no module fault.
Right Ethernet Port LED	Green	ON	Indicates a 100-Mbps port.
		OFF	Indicates a 10-Mbps port.
Left Ethernet Port LED	Green	ON	The link is up.
		OFF	The link is down.
		Blinking	Transmitting (Tx) and Receiving (Rx) traffic.

Status LEDs for the 8-port 100BASE-FX Ethernet feature module

Figure 6. LEDs for the 8-port 100BASE-FX feature module.

(artname: fgb0a020.EPS)

Table 6. Status of LEDs for 8-port 100BASE-FX feature module

LED	Color	State	Explanation
OK	Green	ON	There is power to the feature module.
		OFF	There is no power to the feature module, no power to the switch, or the module has failed.
Fault	Yellow	ON	There is a module fault.
		OFF	There is no module fault.
Port LED	Green	ON	Link is up.
		OFF	Link is down.
		Blinking	Transmitting (Tx) and receiving (Rx) traffic.

Status LEDs for the 4-port 100BASE-FX Ethernet feature module

Figure 7. LEDs for the 4-port 100BASE-FX feature module.

(artname: fgb1a001.EPS)

Table 7. Status of LEDs for 4-port 100BASE-FX feature module

LED	Color	State	Explanation
OK	Green	ON	There is power to the feature module.
		OFF	There is no power to the feature module, no power to the switch,or the module has failed.
Fault	Yellow	ON	There is a module fault.
		OFF	There is no module fault.
Port LED	Green	ON	Link is up.
		OFF	Link is down.
		Blinking	Transmitting (Tx) and receiving (Rx) traffic.

Status LEDs for the 2-port 1000BASE-SX Ethernet feature module

Figure 8. LEDs for the 2-port 1000BASE-SX feature module.

(artname: fgl2a001.EPS)

Table 8. Status of LEDs for 2-port 1000BASE-SX feature module

LED	Color	State	Explanation
OK	Green	ON	There is power to the feature module.
		OFF	There is no power to the feature module, no power to the switch,or the module has failed.
Fault	Yellow	ON	There is a module fault.
		OFF	There is no module fault.
Port LED	Green	ON	Link is up.
		OFF	Link is down.
		Blinking	Transmitting (Tx) and receiving (Rx) traffic.

Physical characteristics

Table 9 summarizes the physical characteristics for the switch:

Table 9. Summary of physical characteristics for the switch

Characteristic	Specification
Physical Dimensions	Height 63.0 mm (2.48 in.) 1.5 EIA rack units Width 440.0 mm (17.16 in.) Depth 355.6 mm (14 in.)
Weight (estimate)	6.0 kg (13 lb)
Minimum Service Clearance	Front 15.3 mm (6 in.) for cooling, cables, and to view LEDs Sides 50 mm (2 in.) for cooling Rear 15.3 mm (6 in.) for cooling and power cord
Environment	Operating Temperature 10° - 40° C (50° - 104° F) Operating Humidity 8% - 80% Storage Temperature 1° - 60° C (33.8° - 140° F) Storage Humidity 8% - 80% Shipment Temperature -40°C - 60°C (-40°F - 140°F) Shipment Humidity 5% - 100%