IBM Front cover Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter Full Layer 2 switching and Layer 3 routing Six external multimode fiber or copper GbE interfaces Hot pluggable switch modules Rufus Credle Stephan Fleck Scott Lorditch Jeremy Oliver ibm.
International Technical Support Organization Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter September 2005
Note: Before using this information and the product it supports, read the information in “Notices” on page vii. First Edition (September 2005) This edition applies to Nortel Networks Layer 2/3 Copper and Fiber GbE Switch Modules for IBM Eserver BladeCenter. © Copyright International Business Machines Corporation 2005. All rights reserved. Note to U.S. Government Users Restricted Rights -- Use, duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp.
Contents Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Trademarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix The team that wrote this Redpaper . . . . . . . . . . . . . . . . . . . . . . . .
.2.3 Broadcom Advanced Control Suite installation . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 6.3 Firmware and device drivers used in this example . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Chapter 7. Nortel Networks L2/3 GbESM configuration and network integration . . . 57 7.1 Standards and technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 7.1.1 VLAN tagging - 802.1Q . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9.3.2 Data collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.3 Data analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.4 Action plan creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.5 Action plan implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3.6 Observation of results . . .
vi Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
Notices This information was developed for products and services offered in the U.S.A. IBM may not offer the products, services, or features discussed in this document in other countries. Consult your local IBM representative for information on the products and services currently available in your area. Any reference to an IBM product, program, or service is not intended to state or imply that only that IBM product, program, or service may be used.
Trademarks The following terms are trademarks of the International Business Machines Corporation in the United States, other countries, or both: Eserver® Eserver® Redbooks (logo) ™ Redbooks (logo)™ eServer™ xSeries® AIX® BladeCenter® Domino® Electronic Service Agent™ Enterprise Storage Server® HelpCenter® HelpWare® IntelliStation® IBM® PartnerLink® Redbooks™ ServerGuide™ Summit® Tivoli® TotalStorage® WebSphere® The following terms are trademarks of other companies: Java, Sun, and all Java-based trademark
Preface This IBM® Redpaper positions the Nortel Networks Layer 2/3 Fiber and Copper GbE Switch Modules for IBM Eserver BladeCenter and describes how its integrated switch options enable the consolidation of full Layer 2-3 LAN switching and routing capabilities. The Nortel Networks switch modules also provide an upgrade path to full Layer 4-7 services by including 4-7 switch intelligence.
Senior Network Architect for a multi-national soft drink company, and as Product Manager for managed hosting services for a large telecommunications provider. He holds a BS in Operations Research with specialization in Computer Science from Cornell University. Jeremy Oliver is a Staff Engineer with the System Validation and Storage group of xSeries Development. He has worked at IBM in Research Triangle Park, North Carolina for seven years.
Comments welcome Your comments are important to us! We want our papers to be as helpful as possible. Send us your comments about this Redpaper or other Redbooks in one of the following ways: Use the online Contact us review redbook form found at: ibm.com/redbooks Send your comments in an email to: redbook@us.ibm.com Mail your comments to: IBM Corporation, International Technical Support Organization Dept. HQ7 Building 662 P.O.
xii Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
1 Chapter 1. Executive summary IBM and Nortel Networks are committed to collaborating on the design and development of server and networking technology to address customer requirements by establishing a joint development center. The Nortel Networks Layer 2/3 Copper and Fiber GbE Switch Modules for IBM Eserver BladeCenter (Nortel Networks L2/3 GbESM and Nortel GbESM) represents a new height in this alliance.
The Nortel and IBM agreement to form a joint development center equips Nortel as it becomes an on demand company that can generate customized products for its network equipment marketplace. This ensures that your needs of high availability, scalability, security, and manageability are addressed. Combined with the integration of IBM Tivoli®, Nortel, and Cisco management products, these architectures bring higher value solutions with lower operational expense.
2 Chapter 2. IBM Eserver BladeCenter overview IBM designed the IBM Eserver BladeCenter innovative modular technology, leadership density, and availability to help solve a multitude of real-world issues. For organizations seeking server consolidation, the IBM Eserver BladeCenter centralizes servers for increased flexibility, ease of maintenance, reduced cost, and streamlined human resources.
2.
IBM Eserver BladeCenter IBM Eserver LS20 IBM Eserver HS20 IBM Eserver HS40 IBM Eserver JS20 Figure 2-1 IBM Eserver BladeCenter and blade modules Blade development is ongoing for the BladeCenter platform. Therefore, we suggest that you regularly visit the following Web site for the latest information about IBM Eserver BladeCenter: http://www.ibm.com/servers/eserver/bladecenter/index.html 2.1.
2.1.2 IBM Eserver BladeCenter system management To get the most value from your IBM Eserver BladeCenter investment throughout its life cycle, you need smart, effective systems management which will keep your availability high and costs low.
that is connected to the lower section of the midplane. All other components plug into one section only (upper or lower). However, there is another matching component that can plug into the other midplane section for redundancy.
2.2.2 Management Module Ethernet Figure 2-4 illustrates the Management Module Ethernet interface. The switch modules are configured by the active Management Module through the use of a 100 Mb Ethernet interface. Each Management Module has four 100 Mb Ethernet interfaces, one for each switch module. Each switch module has two 100 Mb Ethernet interfaces, one for each Management Module. Note: On the Nortel Networks L2/3 GbESM, the management Ethernet ports on the switch are referred to as MGT1 and MGT2.
2.2.3 Gigabit Ethernet path Figure 2-5 on page 10 illustrates the Gigabit Ethernet path. Each processor blade has a minimum of two and a maximum of four EtherLAN interfaces. In particular, the BladeCenter HS20 processor blade has two serializer/deserializer SERDES-based Gb Ethernet interfaces, one for each midplane connector. With a daughter card installed, two more network interfaces can be added. Each switch module (SW Module) receives one LAN input from each processor blade, for a total of 14 inputs.
IBM ^ BladeCenter™ Gigabit Ethernet path SERDES Ethernet LAN 1 SW Module 1 SW Module 3 1………..14 1………..14 Midplane (Upper Section) LAN 3 Network Interface Daughter Card SERDES Ethernet LAN 4 LAN 2 Processor blade #1 Midplane (Lower Section) 1………..14 1………..14 SW Module 2 SW Module 4 Figure 2-5 Gigabit Ethernet path 2.3 IBM Eserver HS20 architecture In this section, we discuss the architectural design of the IBM Eserver BladeCenter HS20.
The BladeCenter HS20 uses the Intel® Lindenhurst chipset (see the HS20 architecture in Figure 2-6). Servicing the IBM ^ HS20 (M/T 8843) and Blade Storage Expansion-II Option 8843 HS20 Block Diagram Nocona XEON CPU To SP I2C bus ICH-S PCI bus 0 LPC LSI 1020 SCSI PCIExpress x8 PCIX 66 PCI 32/33 PCI Express x4 PXH Pri Sec To SP I2C bus Data A MCH Hublink 1.
PCI Express features include: PCI software compatibility Chip-to-chip, board-to-board implementations Support for end-to-end data integrity Advanced error reporting and handling for fault isolation and system recovery Low-overhead, low-latency data transfers and maximized interconnect efficiency High-bandwidth, low pin-count implementations for optimized performance 2.
Management LAN Internal LAN connection Higher security Production LAN BladeCenter™ Drawer Management Module MM external Ethernet port Telnet to MM port to manage switch DHCP lease or 192.168.70.125 Browser Any four Anyofofthe the six external ports external ports Can be disabled Command Line (Telnet) Web interface Switch Module 1 I2C bus Default is 192.168.70.127* Web interface Always static, default is 192.168.70.
14 Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
3 Chapter 3. Nortel Networks Layer 2/3 GbE Switch Modules This chapter discusses the Nortel Networks Layer 2/3 Copper and Fiber GbE Switch Modules for IBM Eserver BladeCenter and its set of features and services. © Copyright IBM Corp. 2005. All rights reserved.
3.1 Product description The new Nortel Networks Layer 2/3 Copper and Fiber GbE Switch Modules for IBM Eserver BladeCenter serve as a switching and routing fabric for the BladeCenter server chassis. In addition to the Layer 2 switching capabilities, these switches introduce the expanded capabilities of Layer 3 routing. Up to four copper or fiber Gb Ethernet modules can reside in the I/O module bays of the BladeCenter chassis.
level of support for a number of users. The IEEE 802.1D Spanning Tree Protocol (STP) support can be enabled or disabled on a per-port basis. Multiple instances of STP are supported (that is, 16 STP groups). Virtual Local Area Network (VLAN) support includes 802.1Q tagged VLANs and support for IEEE 802.3 support on six external ports for up to three static trunk groups. Dynamic trunking using LACP as well as static trunking is supported.
At Layer 3, special configurations of Virtual Router Redundancy Protocol (VRRP) allow all switches in the VRRP group to concurrently process traffic by using multiple instances of VRRP. Such configurations enable maximum switch performance while also ensuring seamless failover in the unlikely event of a failure. VRRP Hot Standby is also supported to enable effective use of NIC Teaming in Layer 3 network topologies much as Trunk Failover facilitates HA designs with NIC Teaming at Layer 2. 3.
3.3 Supported hardware Table 3-1 lists the following IBM hardware platforms which support Nortel Networks Layer 2/3 Copper Gigabit Ethernet Switch Module for IBM Eserver BladeCenter (26K6530) and Nortel Networks Layer 2/3 Fiber Gigabit Ethernet Switch Module for IBM Eserver BladeCenter (26K6531).
20 Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
4 Chapter 4. Nortel Networks Layer 2/3 GbE Switch Module architecture This chapter provides a system overview of the Nortel Networks Layer 2/3 GbE Switch Module. © Copyright IBM Corp. 2005. All rights reserved.
4.1 Nortel GbESM architecture overview The Nortel Networks Layer 2/3 GbE Switch Module is a fully functional Layer 2 and 3 switch that includes Layer 4 awareness and capability. Figure 4-1 shows the architecture overview of the Nortel Networks Layer 2/3 GbE Switch Module. The Nortel GbESM has 14 internal 1 Gbps links to connect to blade servers and six external Gigabit ports to connect to upstream switches. The switch module has two 100 Mbps connections to the Management Modules.
Uplinks 123456 GbESM1 123456 GbESM2 M M 1 M M M M 22 1 2 Blade Server1 1 2 Blade Server2 1 2 Blade Server14 Management Module Blade Servers Figure 4-2 BladeCenter Ethernet connectivity Internal Layer 2 traffic flow in the Nortel Networks L2/3 GbESM Figure 4-3 shows the internal Layer 2 traffic flow in the Nortel GbESM. The hard coded filter in the Nortel GbESM blocks all traffic between the external ports and the Management Module ports.
Figure 4-3 on page 23 also indicates the following: Two Nortel GbESMs in the same BladeCenter chassis can ping or telnet to each other without connecting external ports. They cannot pass user data to each other via this path, which passes traffic through the Management Module. The internal blade ports cannot be on the same VLAN as the Management Module ports.
4.2.1 Nortel Networks L2/3 GbESM ports specific roles Figure 4-5, Figure 4-6, and Figure 4-7 on page 26 show different examples of the port connections to the Nortel GbESM(s) within the IBM Eserver BladeCenter. We then discuss the specific roles and restrictions for the various ports.
EXT1-6 Top GbESM (Bay 1) 1 2 3 4 5 INT1 – 14 6 7 8 9 10 11 12 13 14 1 1 1 1 1 1 1 1 1 1 1 1 1 1 B S S 1 B S S 2 B S S 3 B S S 4 B S S 5 B S S 6 B S S 7 B S S 8 B S S 9 B S S 1 0 B S S 1 1 B S S 1 2 B S S 1 3 B S S 1 4 2 2 2 2 2 2 2 2 2 2 2 2 2 2 MGT2 16 MGT1 15 1 2 3 4 5 6 GbESM External uplinks Eth1 MM1 Uplink Eth0 MM1 MM2 Eth0 MM2 Uplink Eth1 2 3 4 5 6 7 8 9 10 11 12 13 14 INT1 - 14 MGT1 15 MGT2 16 EXT1-6 1 Bottom GbESM (Bay 2) Bla
Spanning Tree (STP) is disabled by default for all internal ports. Preset default values for ports going to the Management Modules (includes ports MGT1 and MGT2): Speed is hard-coded at 100 full and cannot be changed. Ports MGT1 and MGT2 cannot be disabled. – This is by design to ensure that the links to the BladeCenter Management Modules are not inadvertently or intentionally brought down by the administrator.
Also, if you change this setting to disabled, it is assumed that you plan on managing the Nortel Networks L2/3 GbESM via its own uplinks. The default Nortel GbESM IP addressing that is provided by the Management Module for a new IBM Eserver BladeCenter is as follows: Switch bay 1: 192.168.70.127/24 Switch bay 2: 192.168.70.128/24 Switch bay 3: 192.168.70.129/24 Switch bay 4: 192.168.70.
5 Chapter 5. Nortel Networks L2/3 GbESM management and administration In this chapter, we discuss tools, techniques, and applications that help with the management and deployment of the Nortel GbESM in an IBM Eserver BladeCenter. We also discuss the management paths and rules for connecting to and accessing the Nortel GbESM. Note: As noted elsewhere in this document, the information herein applies to the 6-port Nortel Networks Layer 2/3 Copper and Fiber GbE Switch Modules for IBM Eserver BladeCenter.
5.1 Nortel Networks L2/3 GbESM management connectivity In this section, we look at the basic management connectivity and management pathways to the Nortel GbESM., as shown in Figure 5-1. Important: Properly managing the Nortel GbESM in the IBM Eserver BladeCenter actually requires proper management of the Management Module within the BladeCenter chassis.
5.1.1 Out-of-band management It is common to provide a (physically) separate management interface for all of the devices and to carry only management traffic. This is referred to as out-of-band management and is sometimes a separate Ethernet connection (path 1) or a whole different physical connection such as the console port (path 2). Management Module (Path 1) The IBM Eserver BladeCenter comes with at least one Management Module.
Further configuration of the Nortel Networks L2/3 GbESM module is performed by using Telnet (for the Command Line Interface) or a Web browser (for the Browser Based Interface) and accessing the address of the MGT1 or 2 ports. Note: It is recommended (and easier) to use a server or mobile computer that is external to the IBM Eserver BladeCenter chassis to perform initial configuration of the Nortel Networks L2/3 GbESM module.
Figure 5-3 Enabling management over all ports using the Management Module Web interface Internal Ethernet ports (Path 3B) The internal ports can be used to provide management access to the switch from the server blades in the same chassis. In-band management considerations In order to use in-band management paths, you must configure at least one additional IP address on the Nortel Networks L2/3 GbESM beyond the address that is provided through the Management Module and attached to VLAN 4095.
5.2 Nortel Networks L2/3 GbESM user interface This section discusses the management interface of the switch module and what each task represents. To configure and manage the switch module, you can use the following interfaces: IBM Eserver BladeCenter Management Module and I2C Management functions that are necessary for initial setup are provided through the Management Module Web interface. I2C is the communication that is used between the Management Module and Ethernet switch.
The maint menu contains all of the commands for maintenance of the switch. The commands to manipulate the ARP cache and forwarding database are here, as well as the commands to obtain dumps of the current state of the switch for technical support. Figure 5-4 CLI Main Menu Global commands The remainder of the options on the Main Menu — diff, apply, save, revert, and exit — are all global commands that work anywhere on the switch. Figure 5-4 shows what each of the commands does.
Navigation commands There are several commands that are useful in moving from one part of the menu tree to another. The commands are similar to those used in a UNIX® shell: cd This command moves you to a given spot in the menu tree. Entering cd / always takes you back to the main menu. pwd This command displays the current menu path where you are in the menu tree. up This takes you back to the last menu that you touched. .. or cd .. Both of these commands take you up one level in the menu tree.
Upgrading the firmware To upgrade the firmware on the Nortel Networks L2/3 GbESM, you must use Trivial File Transfer Protocol (TFTP) or File Transfer Protocol (FTP). It is not possible to use the Management Module menu item for upgrading firmware at this time. However, this is a planned feature for a future software release. Important: Before updating the firmware, save any configuration changes to the Nortel Networks L2/3 GbESM. From the Telnet session, enter apply, then press Enter.
7. If the download location is the same as the location for the currently loaded OS image, the switch warns you that a failed download could result in an inoperative switch. If the download location is different from the location of the currently loaded OS image, the image file downloads. After the download is finished, the switch asks whether you want to use the old location or the new location. Figure 5-6 on page 37 shows a successful download of the OS image to image2. 8.
name "VLAN_Red" def INT1 INT2 EXT1 EXT2 /c/l2/vlan 99 ena name "MGMT" def INT4 EXT1 EXT2 /c/l2/stg 1/clear /c/l2/stg 1/add 1 10 20 99 /c/l2/lacp/port EXT1 mode active /c/l2/lacp/port EXT2 mode active adminkey 17 /c/l3/if 99 ena addr 10.99.0.243 mask 255.255.255.0 broad 10.99.0.255 vlan 99 /c/l3/gw 1 ena addr 10.99.0.245 /c/l3/gw 2 ena addr 10.99.0.246 / script end /**** DO NOT EDIT THIS LINE! Configuring user accounts This section describes the user accounts on the switch.
Local mode allows the definition of individual user IDs with associated authority levels and passwords. This is configured in the /cfg/sys/access/user menu. For example, Figure 5-7 shows the configuration that is necessary to create the IBM Eserver BladeCenter default USERID and PASSW0RD account as an administrator.
2. Click the folder icon next to System in the left-hand frame. 3. Click CONFIGURE at the top of the page. 4. Click the icon next to General in the drop-down list under System. On a window similar to Figure 5-8, you see options, such as IP Address and Network Mask fields, that can be configured on this page. Other options on this page include date and time settings, syslog settings (if you have a syslog server), and SNMP settings.
other network management products. The MIBs that are provided include Nortel proprietary extensions to the standard MIB1 and MIB2 objects. Both read and write access to these variables can be configured. 5.3 Multiple Nortel Networks L2/3 GbESMs in a BladeCenter If there are two (or more) switches in a single IBM Eserver BladeCenter chassis, the management (MGTx) interfaces of all of the switches are on VLAN 4095.
6 Chapter 6. IBM Eserver BladeCenter system initial setup This chapter discusses the network topology and the hardware that is configured to provide a tested and working configuration to help implement your Nortel Networks L2/3 GbESM for the IBM Eserver BladeCenter. © Copyright IBM Corp. 2005. All rights reserved.
6.1 IBM Eserver BladeCenter system In this section, we discuss the stages of preparing IBM Eserver BladeCenter for operation. 6.1.1 Management Module firmware After the required hardware has been installed in your IBM Eserver BladeCenter, you should update the Management Module using IBM Eserver® BladeCenter - Management Module Firmware Update Version 1.18 or later. To acquire the firmware, go to: http://www.ibm.com/pc/support/site.wss/document.
6.1.2 Management Module network interface In this section, we configure the Management Module external and internal network interfaces to exist upon the management subnet. The external network interface IP address is attached to the network outside of the IBM Eserver BladeCenter. This is the address used to contact the Management Module from an external device.
Configuring the Management Module network interfaces After you access the Management Module Web interface, you can configure the external and internal network interfaces. From the BladeCenter Management Module Web interface, click MM Control → Network Interfaces. Figure 6-2 Management Module External Network Interface window The BladeCenter Management Module defaults to the IP address 192.168.70.125.
Click Save at the bottom of the page. You must restart the Management Module to implement the changes. 6.1.3 I/O module management tasks In this section, we set up and configure the Nortel Networks Layer 2/3 Copper Gigabit Ethernet Switch Module for IBM Eserver BladeCenter. Nortel Networks L2/3 GbESM setup and configuration You can install the Nortel Networks L2/3 GbESM into any of the four BladeCenter switch bays in the rear of the chassis.
Enabling Nortel Networks L2/3 GbESM uplink ports through the Management Module To enable the Ethernet ports of the Nortel Networks Layer 2/3 GbE Switch Module from the BladeCenter Management Module: 1. In the I/O Module Tasks → Management (Bay 1 Ethernet SM) window shown in Figure 6-4 on page 47, click Advanced Management. 2. If necessary, scroll down to the Advanced Setup section. You must at least set the External ports to Enabled for data to be sent out through the switch (Figure 6-5). 3.
this switch. If this field is set to Disabled, the factory default IP configuration will become active when the switch factory defaults are restored or when a switch reset is initiated by a source other than the Management Module. In this case, any user-defined IP configuration for the Nortel Networks L2/3 GbESM stored on the Management Module will not be used.
To transfer the software image files from the TFTP server to the switch, you must establish a Telnet session through the Management Module. To make sure that you have a connection, ping the TFTP server. The Telnet session performs optimally if all three network entities (TFTP server, Management Module, and switch IP addresses) are on the same subnet. Otherwise, you must use a router.
Always update your system in the following order: 1. Update the device drivers. (Start from the hard disk and access the IBM UpdateXpress CD.) 2. Update the firmware. (Start from the UpdateXpress CD.) Before the firmware update, make sure that your server can successfully restart. Note: In our example, we dealt with pristine HS20 systems. Therefore, we uploaded our firmware to the HS20 servers first.
Figure 6-6 BladeCenter Firmware VPD window 6.2.2 Operating systems In this section, we prepare the use of our operating systems for the BladeCenter HS20s. Note you can use IBM Director and Remote Deployment Manager (RDM) to customize and deploy your network operating systems to the HS20s. However, if you are building your network operating system manually, after Windows 2003 Server has been successfully installed, go the next section 6.2.3, “Broadcom Advanced Control Suite installation” on page 53.
IBM Eserver 325, 326 IBM Eserver MXE-460 IBM Eserver xSeries 205, 225 (Type 8647), 226, 235, 236, 255, 305, 335, 336, 346, 365, 440, 445, 450, 455, 460 IBM IntelliStation A Pro (Type 6224) IBM IntelliStation E Pro (Type 6216, 6226) IBM IntelliStation Z Pro (Type 6221, 6223, 6227) IBM IntelliStation M Pro (Type 6219, 6225, 6228) Red Hat Linux Broadcom driver installation To perform a driver installation for Red Hat Linux®, use the example Red Hat Linux AS 2.
Figure 6-7 Broadcom selection window 2. Click MANAGEMENT PROGRAMS, and a window similar to Figure 6-8 opens. Figure 6-8 Select Features window 3. Select Control Suite and BASP. 4. Click Next to continue, and then click Finish.
6.3 Firmware and device drivers used in this example We applied the following firmware and drivers to our environment: IBM Eserver BladeCenter Management Module: – Management Module Firmware Update Version 1.10 BladeCenter HS20(8832) firmware: – BladeCenter HS20 (Type 8832) - Flash BIOS Update Version 1.09 – BladeCenter HS20 (Type 8678, 8832) - blade server integrated system management processor firmware update Version 1.09 – Broadcom NetXtreme firmware level 3.
56 Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
7 Chapter 7. Nortel Networks L2/3 GbESM configuration and network integration This chapter discusses the Best Practices for implementing and configuring Nortel Networks Layer 2/3 Fiber and Copper GbE Switch Modules in Nortel Networks, Cisco Systems, and Extreme Networks network environments. It provides several network topology examples to help you successfully implement the Nortel Networks Layer 2/3 GbE Switch Module. © Copyright IBM Corp. 2005. All rights reserved.
7.1 Standards and technologies This section provides a brief overview of the networking standards which are supported on the Nortel GbESM. Detailed examples of configurations which include these standards are included later in this chapter. The bulk of this chapter shows interconnection with Cisco Systems’ devices. In 7.9, “Configuration for Extreme switches” on page 143, we include configurations for some functions that were tested on a pair of Extreme switches.
proprietary Per VLAN Spanning Tree (PVST). An example of this is shown in 7.7, “Advanced Layer 2 topology sample configurations” on page 80. Two of the key shortcomings of the original STP standard is that it takes as much as 50 seconds to recover from the failure of a link or device, and that it does not deal well with multiple VLANs carried over the same physical link (typically by using the 802.1q standard described above.) These shortcomings are remedied by the 802.
7.1.6 Virtual Router Redundancy Protocol (VRRP) - RFC 3768 Virtual Router Redundancy Protocol is used by Layer 3 routers to enable one (or more) routers to back up a primary or master router seamlessly. Workstations and other devices are typically not aware when a backup router takes over for a primary router which has failed. VRRP recovery time can be as little as one second, or less.
section. You can find more information about this configuration in 7.6, “Basic Layer 2 entry topology” on page 69. 7.2.2 Advanced Layer 2 configurations These configurations add additional connections between the GbESM modules and the Core switches, using a mesh topology. This topology includes loops on most of the VLANs used, and therefore requires the use of Spanning Tree Protocol (STP). Testing was performed using the original STP (802.1D) as well as Rapid Spanning Tree and Multiple Spanning Tree (802.
existing network. You can find more information about this configuration at 7.8, “Layer 3 topology sample configurations” on page 108. 7.3 Introduction to High Availability This section provides an explanation of the trunk failover feature, the Broadcom Advanced Services Protocol driver, and VRRP and of how they work together to provide a highly available IBM Eserver BladeCenter environment. 7.3.
123456 123456 GbESM GbESM X XX If failure anywhere on the link toward the upstream switch, the NIC on Blade server does not know about the failure and may continue to send traffic toward the top switch, which will discard the traffic. - The Trunk Failover feature addresses this issue. X If the switch fails in such a way that the link toward the Blade server goes down, or NIC fails, Blade server can sense this and redirect traffic out the other NIC toward the bottom switch.
configuration of this capability is included in 7.5.3, “Base configuration common to all examples” on page 68. Notes: The BASP driver can be configured to use standards-based Port Aggregation (802.3-ad) teaming. This is useful on HS40 blades or HS20 blades with the SCSI sidecar, both of which have two ports connecting them to each switch module. Only ports connected to the same switch should be teamed in this way. The current production version of the GbESM software(1.0.1.
7.3.4 Some important rules for ensuring High Availability For High Availability (HA) to be truly effective it needs to be well thought out. A complete High Availability design should encompass servers, storage, and more of the network than just the portions connected to the BladeCenter chassis. The object is to ensure that there is no single point of failure which can cause the application(s) to become unavailable or unreachable.
Fiber connections The fiber connectors on the Nortel Networks Layer 2/3 Fiber Gigabit Ethernet Switch Module for IBM Eserver BladeCenter must be Multi-Mode Fiber (MMF). Either 50 or 62.5 micron fiber can be used. Single-Mode (9 micron) fiber is not supported. The fiber should be terminated with LC connectors on both ends. Speed or duplex selection The decision to allow a port to negotiate its speed and duplex automatically or to force it to a set value is a subject of frequent debate.
• • • • • BIOS build SBJT58AUS Diagnostic build SBY113AUS Integrated System Management Processor (ISMP) build BRMK27A Windows 2003 Standard Edition operating system Intel driver version 7.3.13.
7.5.3 Base configuration common to all examples This section lists some established configuration options that are common to all of the examples. These are only for demonstration purposes in the examples and might or might not be duplicated in your particular environment. Management Module settings for Nortel GbESMs Each Nortel GbESM is configured with an IP address for the MGT1 ports (see Figure 5-2 on page 31 for more detail): GbESM_1 is configured with 9.42.171.243 and mask 255.255.255.
For blade server 1, a team is made using the first Ethernet interface as the primary and the second as standby. The IP address for the new BASP interface is set to 10.20.0.1. This is called Active/Standby mode. Note: First Ethernet interface refers to the blade’s physical connection to the first Ethernet switch module in slot 1. For blade server 2, a team is made the same way as blade server 1. Here two VLANs are created (VL10 and VL20). Both of these must be set as tagged.
noted is that this offering is basic and has limited redundancy that relies on port aggregation and trunk failover. 7.6.1 Layer 2 configuration with 802.1Q tagging and trunk failover Figure 7-3 illustrates the first basic Layer 2 topology. 9.0.0.0 G0/24 Cisco 3560G G0/24 G0/23 G0/23 Cisco 3560G Core 1 G0/2 G0/1 Core 2 Port Channel G0/1 G0/2 Port Channel VLAN10, 20, 99 Ext1 VLAN10, 20, 99 Ext2 Trunk Ext1 GbESM_1 10.99.0. 243 Int1 Int2 Ext2 Trunk GbESM_2 Int1 Int4 Int3 10.99.0.
Summary of disconnect procedure to be performed for each example When performing initial configurations or making changes to existing configurations that might have an impact on Spanning Tree (such as changing link aggregation), it is recommended that you leave connections uncabled or shut down prior to making the configuration changes. This reduces the likelihood of any temporary Spanning Tree loops and possible network-down conditions that might result in the process of adding or changing configurations.
1003 token-ring-default 1004 fddinet-default 1005 trnet-default act/unsup act/unsup act/unsup VLAN ---1 5 10 20 Type ----enet enet enet enet SAID ---------100001 100005 100010 100020 MTU ----1500 1500 1500 1500 Parent ------ RingNo ------ BridgeNo -------- Stp ---- BrdgMode -------- Trans1 -----0 0 0 0 Trans2 -----0 0 0 0 VLAN ---99 1002 1003 1004 1005 Type ----enet fddi tr fdnet trnet SAID ---------100099 101002 101003 101004 101005 MTU ----1500 1500 1500 1500 1500 Parent ------ RingNo --
2. Enable 802.1Q tagging with VLAN 5 untagged and allow the VLANs which should be carried over the aggregated ports: switchport switchport switchport switchport trunk encapsulation dot1q trunk native vlan 5 trunk allowed vlan 5,10,20,99 mode trunk 3. Disable the Cisco proprietary dynamic trunk protocol DTP: switchport nonegotiate 4. Enable aggregation by choosing a channel-group number and mode on: channel-group 1 mode on This creates and enables a virtual interface called Port-channel1 (po1).
2. Create the VLANs, enable them, and add ports, as shown in Example 7-5 Create and enable VLANs /c/l2/vlan 5 ena name “Native” def EXT1 EXT2 /c/l2/vlan 10 ena name "VLAN_Green" def INT2 INT3 EXT1 EXT2 /c/l2/vlan 20 ena name "VLAN_Red" def INT1 INT2 EXT1 EXT2 /c/l2/vlan 99 ena name "MGMT" def INT4 EXT1 EXT2 /* def INT2 EXT1 EXT2 /* def INT1 INT2 INT3 EXT1 EXT2 3. In Example 7-6, INT ports must have PVIDs set. Blade servers on INT1, INT3, and INT4 as configured are untagged.
Example 7-7 shows what has been configured so far. Example 7-7 Verifying the VLAN setup of the Nortel GbESM >> Main# /i/l2/vlan VLAN Name Status Ports ---- -------------------------------- ------ ----------------------1 Default VLAN ena INT1-INT14 EXT3-EXT6 5 Native ena EXT1 EXT2 10 VLAN_Green ena INT2 EXT1 EXT2 20 VLAN_Red ena INT1-INT3 EXT1 EXT2 99 MGMT ena INT4 EXT1 EXT2 4095 Mgmt VLAN ena INT1-MGT2 Spanning tree can be disabled for this topology.
Re-enable the ports After verifying the correct cabling between all the devices, the ports can be re-enabled. Enabling the ports on GbESM_1 and GbESM_2 /oper/port EXT1/ena /oper/port EXT2/ena Enabling the ports on Core1 and Core2 conf t int range g0/1-2 no shut After the ports have been enabled, the aggregated links should be functioning correctly. Example 7-8 shows the trunk in forwarding state on the Nortel GbESM.
name "VLAN_Green" def INT2 INT3 EXT1 EXT2 /c/l2/vlan 20 ena name "VLAN_Red" def INT1 INT2 EXT1 EXT2 /c/l2/vlan 99 ena name "MGMT" def INT4 EXT1 EXT2 /c/l2/stg 1/off /c/l2/stg 1/clear /c/l2/stg 1/add 1 5 10 20 99 /c/l2/trunk 1 ena failovr ena add EXT1 add EXT2 /c/l3/if 99 ena addr 10.99.0.243 mask 255.255.255.0 broad 10.99.0.255 vlan 99 /c/l3/gw 1 ena addr 10.99.0.245 / script end /**** DO NOT EDIT THIS LINE! /* def INT2 EXT1 EXT2 /* def INT1 INT2 INT3 EXT1 EXT2 /* addr 10.99.0.244 /* addr 10.99.0.
spanning-tree extend system-id ! vlan internal allocation policy ascending ! ! interface Port-channel1 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate ! interface GigabitEthernet0/1 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate channel-group 1 mode on ! interface GigabitEthernet0/2 switchport trun
interface GigabitEthernet0/19 ! interface GigabitEthernet0/20 ! interface GigabitEthernet0/21 ! interface GigabitEthernet0/22 ! interface GigabitEthernet0/23 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk ! interface GigabitEthernet0/24 no switchport ip address 9.42.171.245 255.255.255.
7.6.2 Basic topology conclusions Testing for this configuration involved pulling cables from the g0/1 and g0/2 ports of Core1 while pings were running. Removing the cables one by one shows that first the trunk redundancy switches the link. After removing the second cable, the trunk failover brings down the INT ports of GbESM_1. If the Nortel switch is being monitored, Figure 7-4 illustrates what you see.
9.0.0.0 G0/24 G0/24 G0/23 G0/23 Cisco 3560G Cisco 3560G Core 1 PO1 Core 2 PO2 PO2 G0/1 G0/2 G0/12 G0/11 G0/11 VLAN 5, 10, 20, 99 EXT1 EXT6 EXT5 G0/2 EXT6 EXT1 EXT2 GbESM_1 10.99.0. 243 Int1 Int2 EXT2 GbESM_2 Int3 10.99.0. 244 Int1 Int4 Int2 Int3 Int4 M M 1 M M 2 Management Workstation G0/1 G0/12 , EXT5 Management Network PO1 1 2 1 2 1 2 1 2 Team Team Blade Server 1 Blade Server 2 Blade Server 3 Blade Server 4 10.20.0.1 10.10.0.2 10.20.0.2 10.10.0.
The Advanced Layer 2 topology offers a good compromise between performance and high availability. It is made up of dual GbESMs, each with two, 2-port aggregated links, going to a pair of Core switches, each Core switch joined to each other via a single link (simulating a Layer 2 network beyond the switches). The 2-port aggregation itself provides for higher performance and the second 2-port aggregation link provides for full redundancy on a connection-loss or a switch-failure situation.
Configure the Nortel GbESMs for IEEE 802.1Q tagged LACP trunk Enable IEEE 802.
Configure the Cisco Core1 and Core2 for tagged LACP trunks In the configuration mode, access the interface level for G0/11 and G0/12: conf t interface range G0/11-12 Enable IEEE 802.
The advantages of this approach are: Traffic can be distributed across several ports by blocking different ports on different VLANs. This is done by setting different root bridge priorities or port costs for different VLANs Some known issues which arise when connecting PVST and plain Mono Spanning Tree can be avoided. For more information, ssee IBM Eserver BladeCenter Layer 2-7 Network Switching, REDP-3755, at: http://www.redbooks.ibm.com/redpapers/pdfs/redp3755.
Ensure that PVST is selected as Spanning Tree mode, enter: conf t spanning-tree mode pvst Core1 should become the Root Bridge for VLAN 5 and 10 while being backup Root for VLAN 20 and 99: spanning-tree vlan 1-10 root primary spanning-tree vlan 11-4094 root secondary end Vice versa for Core2, so that it will be elected as Root Bridge for VLAN 20 and 99, backing up the Root for VLAN 5 and 10: spanning-tree vlan 1-10 root secondary spanning-tree vlan 11-4094 root primary end Reenable the ports After verifyi
EXT3 EXT4 EXT5 EXT6 19 20 21 22 >> GbESM_1 Alias Port ----- ---INT1 1 INT2 2 INT3 3 INT4 4 INT5 5 INT6 6 INT7 7 INT8 8 INT9 9 INT10 10 INT11 11 INT12 12 INT13 13 INT14 14 MGT1 15 MGT2 16 EXT1 17 EXT2 18 EXT3 19 EXT4 20 EXT5 21 EXT6 22 any any 1000 1000 any any full full yes yes no no yes yes no no disabled down up up Information# /i/port Tag FAST PVID NAME VLAN(s) --- ---- ---- ------------- ----------------------y n 20 INT1 1 4095 20 y n 1 INT2 1 4095 10 20 y n 10 INT3 1 4095 10 y n 99 INT4 1 4095
INT14 0 0 DISABLED EXT3 0 0 DISABLED EXT4 0 0 DISABLED * = STP turned off for this port.
EXT6 128 4! FORWARDING ! = Automatic path cost. 6014-00:03:fd:6a:c9:80 8270 -----------------------------------------------------------------Spanning Tree Group 5: On (STP/PVST) VLANs: 99 Current Root: Path-Cost 6063 00:03:fd:6a:c9:80 4 Parameters: Priority 65535 Hello 2 Port Priority Cost ---- -------- ---INT4 128 4! EXT1 128 4! EXT2 128 4! EXT5 128 4! EXT6 128 4! ! = Automatic path cost.
Design or topology remarks Although this topology is very robust and offers high redundancy, newer topologies do exist. The drawback of this topology is the convergence time after a connection-loss or a switch-failure. Our tests showed that as long as only one link of an aggregated port group is going down, the switchover time is around 1 second, which is fairly good.
/c/port INT3 pvid 10 /* pvid 20 on GbESM_2 /c/port INT4 pvid 99 /c/port EXT1 tag ena pvid 5 /c/port EXT2 tag ena pvid 5 /c/port EXT5 tag ena pvid 5 /c/port EXT6 tag ena pvid 5 /c/l2/vlan 1 def INT1 INT2 INT3 INT4 INT5 INT6 INT7 INT8 INT9 INT10 INT11 INT12 INT13 INT14 EXT3 EXT4 /c/l2/vlan 5 ena name "Native" def EXT1 EXT2 EXT5 EXT6 /c/l2/vlan 10 ena name "VLAN_Green" def INT2 INT3 EXT1 EXT2 EXT5 EXT6 /* def INT2 EXT1 EXT2 EXT5 EXT6 on GbESM_2 /c/l2/vlan 20 ena name "VLAN_Red" def INT1 INT2 EXT1 EXT2 EXT5 EXT
mask 255.255.255.0 broad 10.99.0.255 vlan 99 /c/l3/gw 1 ena addr 10.99.0.245 /c/l3/gw 2 ena addr 10.99.0.246 /c/l3/frwd/off dirbr disabled / script end /**** DO NOT EDIT THIS LINE! Example 7-14 Advanced Layer 2 Topology with PVST configuration for the Cisco3560 Switches ! version 12.
switchport mode trunk switchport nonegotiate ! interface GigabitEthernet0/1 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate channel-group 1 mode active ! interface GigabitEthernet0/2 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate channel-group 1 mode active ! interface GigabitEthernet0/3 ! interfac
! interface GigabitEthernet0/18 ! interface GigabitEthernet0/19 ! interface GigabitEthernet0/20 ! interface GigabitEthernet0/21 ! interface GigabitEthernet0/22 ! interface GigabitEthernet0/23 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 2,10,20,99 switchport mode trunk ! interface GigabitEthernet0/24 no switchport ip address 9.42.171.245 255.255.255.0 ! ip address 9.42.171.246 255.255.255.
ip classless ip route 0.0.0.0 0.0.0.0 9.42.171.3 ip http server ip http secure-server ! control-plane ! line con 0 line vty 0 4 password cisco no login line vty 5 15 no login ! end 7.7.3 Rapid Spanning Tree IEEE 802.1w The GbESM supports the IEEE 802.1w RSTP (Rapid Spanning Tree) standard. When this protocol is used there can be only one Spanning Tree Group. When IEEE 802.1s Multiple Spanning Tree is used, multiple spanning tree instances can be configured.
Vice versa for Core2 to become the Root Bridge for VLANs 20 and 99: spanning-tree vlan 1-10 root secondary spanning-tree vlan 11-4094 root primary Finally, the port costs for the connections to the GbESMs are set to a higher value to make them relatively undesirable Spanning Tree Paths.
Full configuration snapshots Example 7-15 shows the output of /c/dump for the GbESM_1 while Example 7-16 on page 98 shows the output of show running-conf from the Core1 Cisco3560 Switch. See the differences to the configurations of GbESM_2 and Core2 as comments in the relevant lines respectively. The configurations are valid for copper and fiber infrastructure, it makes no difference to the GbESM.
name "VLAN_Green" def INT2 INT3 EXT1 EXT2 EXT5 EXT6 /* def INT2 EXT1 EXT2 EXT5 EXT6 on GbESM_2 /c/l2/vlan 20 ena name "VLAN_Red" def INT1 INT2 EXT1 EXT2 EXT5 EXT6 /* def INT1 INT2 INT3 EXT1 EXT2 EXT5 EXT6 on GbESM_2 /c/l2/vlan 99 ena name "MGMT" def INT4 EXT1 EXT2 EXT5 EXT6 /c/l2/stg 1/off /c/l2/stg 1/clear /c/l2/stg 1/add 1 5 10 20 99 /c/l2/lacp/port EXT1 mode active /c/l2/lacp/port EXT2 mode active adminkey 17 /c/l2/lacp/port EXT5 mode active /c/l2/lacp/port EXT6 mode active adminkey 21 /c/l3/if 99 ena ad
ip subnet-zero ip routing no ip domain-lookup ! ! no file verify auto ! spanning-tree mode rapid-pvst no spanning-tree optimize bpdu transmission spanning-tree extend system-id spanning-tree vlan 1-10 priority 24576 ! spanning-tree vlan 1-10 priority 28672 on Core2 spanning-tree vlan 11-4094 priority 28672 ! spanning-tree vlan 11-4094 priority 24576 on Core2 ! vlan internal allocation policy ascending ! ! interface Port-channel1 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport
interface GigabitEthernet0/8 ! interface GigabitEthernet0/9 ! interface GigabitEthernet0/10 ! interface GigabitEthernet0/11 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate channel-group 2 mode active ! interface GigabitEthernet0/12 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate channel-group 2 mode
interface GigabitEthernet0/26 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk switchport nonegotiate ! interface GigabitEthernet0/27 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport trunk allowed vlan 5,10,20,99 switchport mode trunk ! interface GigabitEthernet0/28 ! interface Vlan1 no ip address interface Vlan10 ip address 10.10.0.245 255.255.255.0 ! ip address 10.10.0.246 255.255.255.
spanning tree instance is associated with one or more VLANs, which must be explicitly identified as part of the configuration. It is good practice to associate VLANs with the same topology to a single instance of spanning tree. This reduces processor overhead on the GbESM and other switches in the network and it reduces the traffic generated by the BPDUs that the switches send to each other. In general, MSTP configuration mirrors the configuration of multiple instances of classic 802.
32768 300 Port Prio Cost State Role Designated Bridge Des Port ---- ---- --------- ----- ---- ---------------------- -------EXT1 128 20000 FWD ROOT 1000-00:12:7f:ea:76:00 8003 EXT2 128 20000 DSB -----------------------------------------------------------------Spanning Tree Group 2: On (MSTP) VLANs: 20 40 Current Root: 2000 00:12:7f:ea:76:00 Path-Cost 20000 Port EXT1 Aging 300 Parameters: Priority Aging 32768 300 Port Prio Cost State Role Designated Bridge Des Port ----- ---- --------- ----- ---- -------
On the Cisco 3750, the sh spanning-tree mst command provides similar data as shown below. It is possible to tell that the connected ports on the GbESM and 3750 are successfully interoperating using MST because neither of the ports is displayed as being a boundary port. Boundary ports on switches running MSTP are ports that connect to switches running other, earlier versions of spanning tree such as RSTP or original 802.1D spanning tree.
name "VLAN 10" def EXT1 EXT2 /c/l2/vlan 20 ena name "VLAN 20" def EXT1 EXT2 /c/l2/vlan 30 ena name "VLAN 30" def EXT1 EXT2 /c/l2/vlan 40 ena name "VLAN 40" def EXT1 EXT2 /c/l2/mrst/on /c/l2/mrst/mode mstp /c/l2/mrst/name INTEROP /c/l2/stg 1/clear /c/l2/stg 1/add 10 30 /c/l2/stg 2/clear /c/l2/stg 2/add 20 40 /c/l2/stg 16/clear /c/l3/if 1 ena addr 172.16.1.3 mask 255.255.255.0 broad 172.16.1.255 /c/l3/if 10 ena addr 172.16.10.3 mask 255.255.255.0 broad 172.16.10.255 vlan 10 /c/l3/if 20 ena addr 172.16.20.
Example 7-20 was tested by being interconnected with Example 7-18 on page 104. The same parameters for MSTP and VLANs are used. A Cisco 3750 switch is used to connect to the Nortel GbESM in the BladeCenter chassis. Example 7-20 Sample MSTP Configuration - Cisco version 12.
interface FastEthernet1/0/7 shutdown ! interface FastEthernet1/0/8 shutdown ! interface FastEthernet1/0/9 shutdown ! interface FastEthernet1/0/10 shutdown ! interface FastEthernet1/0/11 shutdown ! interface FastEthernet1/0/12 shutdown ! interface FastEthernet1/0/13 shutdown ! interface FastEthernet1/0/14 shutdown ! interface FastEthernet1/0/15 shutdown ! interface FastEthernet1/0/16 shutdown ! interface FastEthernet1/0/17 shutdown ! interface FastEthernet1/0/18 shutdown ! interface FastEthernet1/0/19 shutdo
! interface Vlan10 ip address 172.16.10.10 255.255.255.0 ! ip classless ip http server ip http secure-server ! control-plane ! line con 0 line vty 0 4 password c login line vty 5 15 no login ! end 7.8 Layer 3 topology sample configurations The configuration samples in this section use Layer 3 switching (routing) in addition to Layer 2 switching (bridging, forwarding). They are based on the simple configuration that was introduced in 7.6, “Basic Layer 2 entry topology” on page 69.
Reasons for choosing a Layer 3 design The reasons for choosing a Layer 3 design for blade server switching are outlined below. It is worth noting that in most discussions of network architecture, the blade servers would be connected to a so-called server access switch. This is a switch at the edge of the network which connects directly to servers as opposed to client (desktop, mobile computer) computers. Most discussions of network architecture recommend that server access switches be Layer 3 devices.
VRRP is configured on VLANs 10 and 20 which face towards the server blades and VLANs 35 and 46 which face the Core routers. VRRP uses an IP address which is shared between the two GbESM modules. At any moment, one of the GbESM modules is the VRRP master, and only the master responds to the shared address. In this sample, GbESM 1 is configured with a higher priority than GbESM 2 and it will therefore be the master if it is operational.
Summary of IP Addressing used in this sample Table 7-1 summarizes the IP addresses used in this sample. Figure 7-6 also illustrates these. Table 7-1 IP Addresses for Layer 3 Sample Configuration (Static routing, VRRP) Switch VLAN 35 VLAN 46 VLAN 10 VLAN 20 GbESM 1 10.35.0.243 10.46.0.243 10.10.0.243 10.20.0.243 GbESM 2 10.35.0.244 10.46.0.244 10.10.0.244 10.20.0.244 Core 1 10.35.0.245 none none none Core 2 none 10.46.0.246 none none VRRP - GbESMs 10.35.0.100 10.46.0.100 10.10.0.
Configure VLANs and subnets for the uplink ports The ports which connect the GbESMs to the Core switches are assigned to their own VLANs and associated subnets are configured. This process is done on the GbESM modules as well as the Core switches. Access to the GbESMs via the management module or the serial console port may be required to successfully achieve this while the external ports are disconnected.
/cfg/l2/vlan 46/def ext5 ext6 /* for GbESM 2 /* /cfg/l2/vlan 35/def ext5 ext6 /* /cfg/l2/vlan 46/def ext1 ext2 /* both GbESMs use the below... /cfg/l2/vlan 10/add ext3 /cfg/l2/vlan 20/add ext3 /* create L3 interfaces on VLANs /* these are for the subnets where the servers are /cfg/l3/if 10 ena vlan 10 addr 10.10.0.243 /* addr 10.10.0.244 on GbESM 2 mask 255.255.255.0 /cfg/l3/if 20 ena vlan 20 addr 10.20.0.243 /* addr 10.20.0.244 for GbESM 2 mask 255.255.255.
Example 7-22 Layer 3 VLANs and ports for Core switches Configuration for Core1 (address .
VRRP configuration The VRRP configuration establishes the .100 addresses on VLANs 10, 20, 35, and 46. The server blades and Core routers must also be configured to be aware of these addresses. Configuration for GbESMs The configuration in Example 7-23 enables VRRP globally and creates four instances of VRRP, one for each of the VLANs 10, 20, 35, and 46. VRRP priority is set to 101 instead of the default of 100 to ensure that GbESM 1 will be the VRRP master for all four VLANs.
Configuration of Hot-standby for VRRP The commands in Example 7-24 add the hot-standby feature to the VRRP configuration. Hot-standby is used in this sample with interface tracking. VRRP tracking allows a device to dynamically adjust its priority based on the availability of other resources, and allow a standby switch to become master if the current master loses access to those resources.
shared virtual address to the GbESMs. This virtual address is configured as the GbESM’s default gateway. The necessary static routes for the Core switch 1 are: ip route 10.10.0.0 255.255.255.0 10.35.0.100 ip route 10.20.0.0 255.255.255.0 10.35.0.100 Because Core switch 2 uses VLAN 46 rather than 35 to send traffic to the GbESMs, its static routes are different as shown below: ip route 10.10.0.0 255.255.255.0 10.46.0.100 ip route 10.20.0.0 255.255.255.0 10.46.0.
Example 7-25 Verifying operation of the Layer 3 configuration /i/port Alias Port ----- ---INT1 1 INT2 2 INT3 3 INT4 4 INT5 5 INT6 6 INT7 7 INT8 8 INT9 9 INT10 10 INT11 11 INT12 12 INT13 13 INT14 14 MGT1 15 MGT2 16 EXT1 17 EXT2 18 EXT3 19 EXT4 20 EXT5 21 EXT6 22 Tag --y y y y y y y y y y y y y y y y n n y n n n FAST ---n n n n n n n n n n n n n n n n n n n n n n PVID ---20 1 10 1 1 1 1 1 1 1 1 1 1 1 4095 4095 35 35 50 1 36 36 NAME VLAN(s) ------------- ----------------------INT1 1 4095 20 INT2 1 4095 10
Trunk group 24: Enabled port state: EXT5: STG 1 forwarding EXT6: STG 1 forwarding VRRP operation can be verified using the /i/l3/vrrp command. The /i/l3/ip command provides general information about TCP/IP. The four VRRP instances should be up and they should show as master on GbESM 1 and standby on GbESM 2. The default gateways shown in the /i/l3/ip command should be up.
Complete configuration snapshots Example 7-27 and Example 7-28 on page 122 show complete configuration files for the GbESM and upstream Core switches. One configuration for each type of switch is shown with notes where the second switch of the same type has a different configuration.
/* the below is used solely to provide an unused VLAN as PVID for the crossover /c/l2/vlan 50 ena def EXT3 /c/l2/stg 1/off /c/l2/stg 1/clear /c/l2/stg 1/add 1 10 20 35 46 50 /c/l2/trunk 1 ena failovr dis add EXT1 add EXT2 /c/l2/trunk 2 ena failovr dis add EXT5 add EXT6 /c/l3/if 10 ena addr 10.10.0.243 /* addr 10.10.0.244 on GbESM 2 mask 255.255.255.0 broad 10.10.0.255 vlan 10 /c/l3/if 20 ena addr 10.20.0.243 /* addr 10.20.0.244 on GbESM 2 mask 255.255.255.0 broad 10.20.0.255 vlan 20 /c/l3/if 35 ena addr 10.
addr 10.46.0.246 /c/l3/vrrp/on /c/l3/vrrp/vr 1 ena vrid 1 if 10 prio 101 /* the above is only on GbESM 1; defaults addr 10.10.0.100 /c/l3/vrrp/vr 2 ena vrid 2 if 20 prio 101 /* the above is only on GbESM 1; defaults addr 10.20.0.100 /c/l3/vrrp/vr 3 ena vrid 3 if 35 prio 101 /* the above is only on GbESM 1; defaults addr 10.35.0.100 /c/l3/vrrp/vr 4 ena vrid 4 if 46 prio 101 /* the above is only on GbESM 1; defaults addr 10.46.0.
no service password-encryption ! hostname Core1 ! enable password cisco ! no aaa new-model ip subnet-zero ip routing no ip domain-lookup ! no file verify auto ! spanning-tree mode pvst no spanning-tree optimize bpdu transmission spanning-tree extend system-id no spanning-tree vlan 1-4094 ! vlan internal allocation policy ascending ! ! interface Port-channel1 switchport access vlan 35 ! -- vlan 46 on core 2 switchport trunk encapsulation dot1q switchport mode access switchport nonegotiate ! interface Port-ch
switchport nonegotiate channel-group 2 mode active ! interface GigabitEthernet0/12 switchport access vlan 35 ! -- vlan 46 on core 2 switchport trunk encapsulation dot1q switchport mode access switchport nonegotiate channel-group 2 mode active ! interface GigabitEthernet0/13 switchport trunk encapsulation dot1q switchport trunk native vlan 5 switchport mode trunk ! interface GigabitEthernet0/23 no switchport ip address 10.56.0.245 255.255.255.0 -- .
line vty 5 15 no login ! ! end 7.8.2 Dynamic routing options OSPF/RIP This section discusses an enhancement to the Layer 3 configuration that is shown in 7.8.1, “Layer 3 sample configuration with static routing and VRRP” on page 109. Instead of using static routes to enable the switches to know how to reach the various subnets, standard dynamic routing protocols is used.
Summary of IP Addressing used in this sample Table 7-2 summarizes the IP addresses in use. Figure 7-7 also shows these. Table 7-2 IP Addresses for Layer 3 Sample Configuration (Dynamic routing) Switch VLAN 35 VLAN 36 VLAN 45 VLAN 46 VLAN 10 VLAN 20 GbESM 1 10.35.0.243 10.36.0.243 none none 10.10.0.243 10.20.0.243 GbESM 2 none none 10.45.0.244 10.46.0.244 10.10.0.244 10.20.0.244 Core 1 10.35.0.245 none 10.45.0.245 none none none Core 2 none 10.36.0.246 none 10.46.0.
Configure additional VLANs and subnets for the uplink ports Note: The configuration instructions that follow build in part on the configuration in section 7.8.1, “Layer 3 sample configuration with static routing and VRRP” on page 109 Example 7-29 and Example 7-30 on page 129 use a VLAN and associated subnet for each pair of switches which are connected. VLAN 35 connects GbESM1 (.243) with Core1 (.245), and similarly for VLANs 36, 45, and 46.
/cfg/l3/vrrp/vrid 4/del /cfg/l3/vrrp/hot dis /cfg/l3/vrrp/group/dis /* Configuration below for GbESM 2 (address .244) only /cfg/l3/frwd/on /cfg/l3/gw 1 addr 10.45.0.245 ena /cfg/l3/gw 2 addr 10.46.0.
Example 7-30 Additional VLANs and Subnets - Core switches ! Configuration for Core 1 (address .245) ! first of all turn layer 3 functions on ip routing ! ! explicitly create new VLAN & remove unused - this does not appear in sh run vlan 45 no vlan 46 ! interface range Po1, g0/1-2 switchport access vlan 35 switchport mode access interface range Po2, g0/11-12 switchport access vlan 45 switchport mode access interface vlan 35 ip address 10.35.0.245 255.255.255.0 interface vlan 45 ip address 10.45.0.245 255.
Configuring RIP Routing Information Protocol (RIP) is the oldest and simplest of the dynamic routing protocols but it is nonetheless adequate for use in many BladeCenter networking configurations. The configurations in Example 7-31 and Example 7-32 use RIP version 2, which is supported by all of the devices used in our testing. Example 7-31 shows the addition of RIP to the GbESM switches. Example 7-32 shows the addition of RIP to the Core switches.
/c/l3/ospf/if 35 /* -- if 45 on GbESM 2 aindex 0 ena /c/l3/ospf/if 36 /* -- if 46 on GbESM 2 aindex 0 ena /* also turn RIP (from previous sample) off /cfg/l3/rip/off Example 7-34 and Example 7-35 show the addition of OSPF to the Core switches. As in Example 7-33 on page 130, it is assumed that RIP is not present or has been removed. Example 7-34 Configuration for Core switch 1 router ospf 9 network 10.35.0.0 0.0.255.255 area 0.0.0.9 network 10.45.0.0 0.0.255.255 area 0.0.0.
* * * * * * * * * * * * * * 10.20.0.0 10.20.0.243 10.20.0.255 10.35.0.0 10.35.0.243 10.35.0.255 10.36.0.0 10.36.0.243 10.36.0.255 10.45.0.0 10.46.0.0 10.99.0.0 127.0.0.0 255.255.255.255 255.255.255.0 255.255.255.255 255.255.255.255 255.255.255.0 255.255.255.255 255.255.255.255 255.255.255.0 255.255.255.255 255.255.255.255 255.255.255.0 255.255.255.0 255.255.255.0 255.0.0.0 255.255.255.255 10.20.0.243 10.20.0.243 10.20.0.255 10.35.0.243 10.35.0.243 10.35.0.255 10.36.0.243 10.36.0.243 10.36.0.255 10.10.0.
Interface Send Recv Triggered RIP Vlan35 2 2 Vlan36 2 2 Vlan45 2 2 Vlan46 2 2 GigabitEthernet0/23 2 2 GigabitEthernet0/24 2 2 Automatic network summarization is in effect Maximum path: 4 Routing for Networks: 10.0.0.0 Routing Information Sources: Gateway Distance Last Update 10.46.0.244 120 00:00:00 10.36.0.243 120 00:00:02 Distance: (default is 120) Key-chain Core2#sh ip rip database 10.0.0.0/8 auto-summary 10.10.0.0/24 [1] via 10.46.0.244, 00:00:07, Vlan46 [1] via 10.36.0.243, 00:00:07, Vlan36 10.20.0.
* * * * * * * * * * * * * * * * 10.36.0.0 10.36.0.243 10.36.0.255 10.45.0.0 10.45.0.0 10.46.0.0 10.46.0.0 10.46.0.0 127.0.0.0 224.0.0.0 224.0.0.0 224.0.0.2 224.0.0.5 224.0.0.6 224.0.0.18 255.255.255.255 255.255.255.0 255.255.255.255 255.255.255.255 255.255.255.0 255.255.255.0 255.255.255.0 255.255.255.0 255.255.255.0 255.0.0.0 240.0.0.0 224.0.0.0 255.255.255.255 255.255.255.255 255.255.255.255 255.255.255.255 255.255.255.255 10.36.0.243 10.36.0.243 10.36.0.255 10.20.0.244 10.10.0.244 10.20.0.244 10.10.0.
Started at 9936 and the process uptime is 11418 Area Border Router: no, AS Boundary Router: no External LSA count 1 Number of interfaces in this router is 4 Number of virtual links in this router is 0 61 new lsa received and 21 lsa originated from this router Total number of entries in the LSDB 9 Total neighbors are 3, of which 3 are >=INIT state, 3 are >=EXCH state, 3 are =FULL state Number of areas is 1, of which 1-transit 0-nssa Area Id : 0.0.0.
Cisco commands to verify OSPF The following commands are in addition to the Cisco commands used for RIP (see Example 7-39): sh ip ospf neighbor Lists the OSPF neighbors of the current switch. sh ip ospf database Provides a dump of the OSPF topology database, which includes all switches in the same area as the one where the command is issued.
10.36.0.246 10.46.0.244 10.56.0.246 9.42.171.244 1532 1507 0x80000008 0x00BDD5 0x80000008 0x00E064 Type-5 AS External Link States Link ID 0.0.0.0 ADV Router 10.56.0.246 Age 517 Seq# Checksum Tag 0x80000007 0x00D991 9 Complete configuration snapshots Configurations shown in Example 7-40 and Example 7-41 on page 140 include details for OSPF and RIP. In general they should not be run at the same time.
/* above also INT3 on GbESM 2 /c/l2/vlan 35 /* use vlan 45 on GbESM 2 ena name "VLAN 35" def EXT1 EXT2 /c/l2/vlan 36 /* use vlan 46 on GbESM 2 ena name "VLAN 36" def EXT5 EXT6 /c/l2/stg 1/off /c/l2/stg 1/clear /c/l2/stg 1/add 1 10 20 35 36 50 /c/l2/trunk 1 ena add EXT1 add EXT2 /c/l2/trunk 2 ena add EXT5 add EXT6 /c/l3/if 10 ena addr 10.10.0.243 /* -- 10.10.0.244 on GbESM 2 mask 255.255.255.0 broad 10.10.0.255 vlan 10 /c/l3/if 20 ena addr 10.20.0.243 /* -- 10.20.0.244 on GbESM 2 mask 255.255.255.0 broad 10.
/c/l3/vrrp/vr 2 ena vrid 2 if 20 prio 101 /* -- above only on GbESM 1 addr 10.20.0.
trans 5 retra 5 /c/l3/ospf/if 36 /* use if 46 instead of 36 for GbESM 2 ena aindex 1 prio 1 cost 1 hello 10 dead 40 trans 5 retra 5 script end /**** DO NOT EDIT THIS LINE! Example 7-41 Layer 3 Core switch configuration with dynamic routing Core1#sh run ! note that VLANs must be created with the Vlan command in “config t” mode ! these statements are not displayed by the “sh run” command Building configuration... Current configuration : 2935 bytes ! version 12.
switchport access vlan 45 ! -- vlan 46 on core 2 switchport trunk encapsulation dot1q switchport mode access switchport nonegotiate ! interface GigabitEthernet0/1 switchport access vlan 35 ! -- 36 on core 2 switchport trunk encapsulation dot1q switchport mode access switchport nonegotiate channel-group 1 mode on ! interface GigabitEthernet0/2 switchport access vlan 35 ! -- 36 on core 2 switchport trunk encapsulation dot1q switchport mode access switchport nonegotiate channel-group 1 mode on ! interface Giga
interface Vlan35 ip address 10.35.0.245 255.255.255.0 ! interface Vlan45 ip address 10.45.0.245 255.255.255.0 ! interface Vlan99 ip address 10.99.0.245 255.255.255.0 ! ! the below is for RIP - remove it for OSPF router rip redistribute static network 10.0.0.0 ! the below is for OSPF - remove it for RIP ! core 2 uses 10.36.0.0 and 10.46.0.0 router ospf 9 network 10.35.0.0 0.0.255.255 area 0.0.0.9 network 10.45.0.0 0.0.255.255 area 0.0.0.9 ! the below is used for OSPF and RIP ip default-gateway 9.42.171.
7.9 Configuration for Extreme switches The configuration in Example 7-42 is from one of a pair of Extreme switches which we tested in place of the Cisco Core switches. The Extreme switches had a subset of the functionality of the GbESM switch modules, so our testing was limited. We were able to successfully test the following: VLAN tagging using 802.1Q Link aggregation (trunking) with static trunks; the Extreme switch did not support LACP.
configure vlan "VLAN10" ipaddress 10.10.0.247 255.255.255.0 configure vlan "VLAN10" add port 1 tagged configure vlan "VLAN10" add port 11 tagged configure vlan "VLAN10" add port 23 tagged # # Config information for VLAN VLAN20. configure vlan "VLAN20" tag 20 # VLAN-ID=0x14 Global Tag 5 configure stpd s0 add vlan "VLAN20" configure vlan "VLAN20" qosprofile "QP1" configure vlan "VLAN20" ipaddress 10.20.0.247 255.255.255.
# -- IP Interface[0] = "Default" enable ipforwarding vlan "Default" disable ipforwarding broadcast vlan "Default" disable ipforwarding ignore-broadcast vlan "Default" disable isq vlan "Default" disable irdp vlan "Default" enable icmp unreachable vlan "Default" enable icmp redirects vlan "Default" enable icmp port-unreachables vlan "Default" enable icmp time-exceeded vlan "Default" enable icmp parameter-problem vlan "Default" enable icmp timestamp vlan "Default" enable icmp address-mask vlan "Default" config
configure ip-mtu 1500 vlan "VLAN99" # Global IP settings. configure irdp 450 600 1800 0 configure irdp broadcast disable icmp useredirects disable iproute sharing configure ipfdb route-add clear-all disable bootprelay configure ip-down-vlan-action forward # # IP ARP Configuration configure iparp timeout 20 configure iparp max-entries 4096 configure iparp max-pending-entries 256 enable iparp checking enable iparp refresh # # IP Route Configuration configure iproute add default 9.42.171.
disable rip export ospf-inter disable rip export ospf-extern1 disable rip export ospf-extern2 disable rip export direct disable rip originate-default configure rip updatetime 30 configure rip routetimeout 180 configure rip garbagetime 120 # RIP Global enable/disable state enable rip # # PIM Router Configuration # disable pim # remaining details omitted # Ospf Area Configuration create ospf area 0.0.0.9 configure ospf area 0.0.0.9 interarea-filter "None" configure ospf area 0.0.0.
# SNTP client configuration # omitted # # # # # # # Mac Vlan Configurations Access-mask Configuration Access-list Configuration Rate-limit Configuration # # System Dump Configuration # ## SNMPV3 EngineID Configuration # ## SNMPV3 USM Users Configuration # # # SNMPV3 MIB Views Configuration # # # SNMPV3 VACM Access Configuration # # # SNMPV3 USM Groups Configuration # # # SNMPV3 Community Table Configuration # # # SNMPV3 Target Addr Configuration # # # SNMPV3 Target Params Configuration # # # SNMPV3 Notify
# Network Login Configuration configure netlogin base-url "network-access.net" configure netlogin redirect-page "http://www.extremenetworks.
150 Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
8 Chapter 8. Serial over LAN feature description and configuration This chapter provides a brief introduction to the Serial over LAN (SOL) feature for the IBM Eserver BladeCenter. It also discusses the configuration of the Nortel Networks L2/3 GbESM along with rules to follow for establishing an SOL connection. © Copyright IBM Corp. 2005. All rights reserved.
8.1 SOL overview SOL is accomplished in the following way: Serial data that flows to and from the blade server(s) COM port is routed through the network infrastructure of the BladeCenter chassis. This network infrastructure includes the BladeCenter Management Module, the Nortel Networks L2/3 GbESM, and the onboard network adapter of the blade server. In addition, the blade server’s integrated system management processor also assists in handling the serial data to and from the COM port.
8.
The operating system of choice for this paper was Windows 2003 Standard Edition. The blade server was located in slot 5 of the BladeCenter chassis. The slot number is used for when you are attempting to start a SOL session from the Management Module. To administer the Windows 2003 blade server through SOL: 1. Telnet to the Management Module and login with USERID and PASSW0RD. 2. At the prompt type console -T blade[5]. 3. At the prompt type ? for a help menu.
9 Chapter 9. Nortel Networks Layer 2/3 GbE Switch Module troubleshooting In this chapter, we discuss troubleshooting techniques that you can use in support of the Nortel Networks L2/3 GbESM. © Copyright IBM Corp. 2005. All rights reserved.
9.1 Basic rules and unique symptoms Before going into detail about troubleshooting, it is important to first discuss certain common rules and symptoms for this environment. Certain interactions within the IBM Eserver BladeCenter between the Nortel Networks L2/3 GbESM and the Management Module require that certain important rules are followed. Failure to follow these rules can produce unexpected results when deploying the IBM Eserver BladeCenter containing an Nortel Networks L2/3 GbESM.
9.2 Nortel Networks L2/3 GbESM troubleshooting methodology In this section, we discuss general troubleshooting techniques and offer options for getting started. 9.2.1 General comments on troubleshooting Because of the highly integrated nature of the Nortel Networks L2/3 GbESM within the IBM Eserver BladeCenter, it is usually necessary to engage several teams for anything beyond basic hardware troubleshooting.
9.3 Systematic approach Figure 9-1 provides a basic flow diagram on how to approach troubleshooting Nortel Networks L2/3 GbESM. Define Problem Collect Data Analyse Data Create Action Plan Implement Action Plan Observe Results Problem Resolved Figure 9-1 Basic flow diagram 9.3.
9.3.2 Data collection In this particular step, it might be necessary to collect additional information to help isolate problem cause as noted in the following sections.
Network logs Review the network logs to help you determine possible errors: Console log messages Last 10 messages only Syslog messages Highly recommended to use SNMP trap messages Highly recommended to use Server logs Packet traces Collect packet traces as needed: Client side Server side 9.3.3 Data analysis You should consider specific causes and narrow down the list to the most likely issue or problem. If necessary, gather more data based on your analysis. 9.3.
9.3.7 Problem resolution Document the problem and resolution for future reference. 9.4 Troubleshooting tools This section lists troubleshooting tools for your use. Ping This tool, also known as the Packet INternet Groper, allows you to check connectivity between network devices. This tool is included with Windows, all variants of UNIX, and is also available on many network device CLIs. Traceroute Traceroute records the route through the Internet between your computer and a specified destination computer.
Difference tool The Difference tool allows you to compare the content of two files. For example, if you have two configuration files for a pair of active-standby switches, it helps identify any configuration discrepancies. There are text-based and visual versions for Windows. A text-based version is included with all variants of UNIX. Tools for Windows – Examdiff – VDiff32 Note: ExamDiff is a freeware difference tool that you can download at: http://www.prestosoft.com/ps.
Web load generator A Web load generator allows you to test and analyze the performance characteristics and bottlenecks of your Web site under various load conditions. Tools available for Windows: – – – – – Microsoft Web Stress Tool WebBench LoadRunner Socrates WAPT Note: Microsoft Web Stress Tool is a free load generator. WebBench is a free load generator and performance benchmark tool. LoadRunner is a commercial load generator and performance testing tool that is available at: http://www-heva.
Network monitors or management systems Use network monitor or management systems to create traffic profiles and to form baseline statistics. Such a system provides notification of network device errors and failures. Brand examples of such systems include Nortel Networks Optivity NMS (network management server), HP OpenView, IBM Tivoli, Ciscoworks, and Concord Network Health.
10 Chapter 10. Service and support This chapter explains the support methods that you can use if you have issues with the Nortel Networks L2/3 GbESM. © Copyright IBM Corp. 2005. All rights reserved.
10.1 Placing a call to IBM For U.S., AP, CAN, and EMEA, use one of the following numbers when calling IBM for technical support: Within the United States, call the IBM Support Center at 1-800-IBM-SERV (426-7378). Within Canada: – For support, call HelpPC at 800-426-7378. – For more information or to place an order, call 800-465-7999. Outside the United States and Canada, contact your IBM HelpWare® number, your place of purchase, or your local IBM office.
10.4 Other support sites Listed here are other helpful Web sites: Nortel Networks Technical Support http://www130.nortelnetworks.com/cgi-bin/eserv/cs/main.jsp Nortel Networks Products http://products.nortel.com/go/product_index.jsp Chapter 10.
168 Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
Abbreviations and acronyms AIX IBM open operating system AMD Manufacturer of personal computing microprocessors HTTPS Hypertext Transfer Protocol over Secure Socket Layer I2C Inter-IC IBM International Business Machines Corporation ANSI American National Standards Institute ICMP Internet Control Message Protocol ARP Address Resolution Protocol IDE Integrated Drive Electronics ATI World's largest supplier of 3D graphics and multimedia technology IEEE Institute of Electrical and Electroni
PVID Port VLAN ID PVST Per VLAN Spanning Tree PXE Preboot Execution Environment QOS Quality of Service RADIUS Remote Authentication and Remote Authentication Dial-in User Service Protocol RIPv1 Routing Information Protocol version 1 RIPv2 Routing Information Protocol version 2 SDRAM synchronous dynamic RAM SERDES serializer/deserializer is an integrated circuit SLB Smart Load Balancing SMP symmetric multiprocessing SNMPv3 Simple Network Management Protocol version 3 SoL Serial over
Related publications The publications listed in this section are considered particularly suitable for a more detailed discussion of the topics covered in this Redpaper. IBM Redbooks For information about ordering these publications, see “How to get IBM Redbooks” on page 173. Note that some of the documents referenced here may be available in softcopy only.
BladeCenter Management Module firmware http://www.ibm.com/pc/support/site.wss/document.do?lndocid=MIGR-54939 IBM Eserver BladeCenter firmware http://www-1.ibm.com/servers/eserver/support/bladecenter/index.html IBM Support http://www.ibm.com/support/ IBM UpdateExpress http://www.ibm.com/pc/support/site.wss/document.do?lndocid=MIGR-53046 Support for IBM xSeries products http://www.ibm.com/servers/eserver/support/xseries/index.html Broadcom NetXtreme Gigabit Ethernet drivers http://www.ibm.
IBM Directory of worldwide contacts http://www.ibm.com/planetwide/ IBM Latin America http://www.ibm.com/pc/la Nortel Networks Technical Support http://www130.nortelnetworks.com/cgi-bin/eserv/cs/main.jsp Nortel Networks Products http://products.nortel.com/go/product_index.jsp How to get IBM Redbooks You can search for, view, or download Redbooks, Redpapers, Hints and Tips, draft publications and Additional materials, as well as order hardcopy Redbooks or CD-ROMs, at this Web site: ibm.
174 Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter
Back cover Nortel Networks L2/3 Ethernet Switch Module for IBM Eserver BladeCenter Full Layer 2 switching and Layer 3 routing Six external multimode fiber or copper GbE interfaces Hot pluggable switch modules ® Redpaper This IBM Redpaper positions the Nortel Networks Layer 2/3 Fiber and Copper GbE Switch Modules for IBM Eserver BladeCenter and describes how its integrated switch options enable the consolidation of full Layer 2-3 LAN switching and routing capabilities.
