HP 600 Series Services And Applications Download | Manualshive

Page: 187 / 316

background image

HP’s Compression Algorithm

Beginning November 1, 1993, HP has been shipping a variation of the Lempel-
Ziv (LZ) lossless compression algorithm on all HP routers. The HP scheme
(called HP Packet-by-Packet Compression, or HP PPC) compresses each
packet independently using the packet-by-packet dictionary method in
which the dictionary is reset with each packet. Also, HP PPC ensures that
the compressed packet is never larger than the original, otherwise the
original (uncompressed) packet will be sent. Sending the uncompressed
packet in this case saves CPU bandwidth by eliminating the need to uncom-
press the packet at the other end of the link. Additionally, sending the
uncompressed packet saves WAN link bandwidth by assuring that a packet
is never expanded beyond its original size. Not all compression algorithms
guarantee smaller, or even equal size data!

Run Length Encoding

In addition to packet-by-packet compression, the HP PPC algorithm employs
an additional encoding scheme, called run length encoding, to further
compress data. Run length encoding is a method of replacing repeated
occurrences of a certain character with a single occurrence, followed by the
number of times it occurs (i.e., the run length).

Currently, HP PPC is supported only on HP point-to-point links. Later
releases of router software will support other, more standard link technolo-
gies, such as Point-to-Point Protocol (PPP), including PPP over ISDN (V.25
bis) and PPP over frame relay. HP PPC uses very little memory, and is
optimized for running on WAN links that use non-reliable, datagram-oriented
protocols, such as HP point-to-point links running the LLC1 datagram
service. You can also use the LLC2 (reliable) service, which provides error
detection as well as error recovery by retransmission. However, LLC2 uses
more link overhead in acknowledgments and retransmissions, and is not
required for HP PPC to run efficiently.

R

o

u

ti

n

g

S

e

r

v

ic

e

s

N

o

te

s

2

Data Compression for WAN Links

HP’s Compression Algorithm

2-135

«
...
185
186
187
188
189
...
»

Summary of Contents for 600 Series

Page 1: ...Hewlett Packard Series 200 400 and 600 Routers HP Routing Services and Applications ...

Page 2: ...rmation contained in this guide is entirely unwarranted This document contains proprietary information which is protected by copy right No part of this docu ment may be photocopied reproduced or translated into another language without the prior written consent of Hewlett Packard The information contained in this document is subject to change without notice Microsoft is a U S registered trademark ...

Page 3: ...vice 2 107 DECnet Routing Service 2 115 A Primer on HP Probe 2 125 Data Compression for WAN Links 3 133 3 Application Notes and Case Studies Improving Network Availability 3 3 ISDN Wide Area Network Design Dry Creek Joint Elem School District 3 31 Shining a Light on FDDI 3 41 Using Synchronous Pass Through to Consolidate Synchronous Traffic 3 57 Routing with OSPF 3 69 Linking Up with Frame Relay 3...

Page 4: ......

Page 5: ...1 Product Notes ...

Page 6: ...Features of HP Routers Architecture and Technology Branch Office Routing Product Notes 1 2 ...

Page 7: ...on specific routers see the HP Network Connectivity Product Catalog or the installation manual and release notes for your specific router product Routing Services and Protocols The standard configuration of HP routers supports concurrent operation of five popular routing services Internet Protocol the TCP IP protocol suite Novell IPX AppleTalk Phase 2 DECnet Phase IV and Xerox XNS For packets that...

Page 8: ...ning Tree Protocol STP is available for management of bridged networks with mesh topologies STP allows a bridged network to have redundant paths In the event of a primary link failure a backup link takes over thereby ensuring continued data transmission between all reachable network segments Translational bridging is used between FDDI and either Ethernet or IEEE 802 3 networks Source route transla...

Page 9: ... allows the creation of subnets Subnets usually reflect the organizational and geographic structure of the entity Communication between subnets can optionally be enabled or disabled based on subnet address This capability simplifies traffic control and enhances network security Within the entity s internal network IP routers exchange network informa tion using an interior gateway protocol IGP HP r...

Page 10: ...X and Sequenced Packet Exchange SPX protocols used by Novell s NetWare The IPX routing service integrates NetWare based PC networks into the enterprise LAN The IPX routing service on HP routers is compatible with network nodes that use Novell NetWare version 2 15 or later HP routers can interoperate with routers using Netware and with Novell routers by negotiating data link characteristics when es...

Page 11: ... is called level 2 routing HP routers support both level 1 and level 2 routing HP routers can interoperate over Ethernet with routers from DEC however direct interoperability over serial links WAN links is not supported For more information refer to the DECnet Routing Service note in this manual N o t e The DECnet routing service does not support routing on token ring IEEE 802 5 networks Xerox XNS...

Page 12: ... RIP DDN Defense Data Network address resolution algorithm Open Shortest Path First OSPF PDN public data network RFC 877 compliant address resolution Static routes Novell IPX routing Novell Internet work Packet Exchange IPX Routing Information Protocol RIP Not needed station address used as network layer address Static routes Xerox XNS routing Internet Datagram Protocol IDP Routing Information Pro...

Page 13: ...supply can be added for load sharing and redun dancy All interfaces the redundant power supply and the fan module can be replaced online without interrupting network connections and compo nents that are still operating hot swap An LED PCA provides multiple status indicators for each port on each card for overall chassis status and for the fan and redundant power supply For more information see the...

Page 14: ...r used to connect to any of the other LAN media listed above this item plus the following medium Thick coaxial cable IEEE 802 3 Type 10Base5 On an HP Router 650 the Ethernet 802 3 ports are supplied on replaceable recessed transceiver modules All of the Ethernet 802 3 ports on the series 200 and 400 routers include an AUI connector for attachment of an external transceiver and some also include a ...

Page 15: ... public switched telephone network PSTN or POTS switched 56 Kbit second switched fractional T1 and ISDN networks Support link terminating equipment including modems on the PSTN ISDN manual terminal adapters modems in DTR mode V 25 terminal adapters for ISDN and CSU DSUs for the switched digital network Support data compression over all point to point links Whether data compresssion is enabled is a...

Page 16: ... 37 pin male connector V 35 with a 34 pin male connector and X 21 with a 15 pin male connector X 25 WAN Links AN HP router can transfer data through a public or private X 25 packet switching network to another HP router In addition IP traffic from the router can be sent through an X 25 network directly to any computer system that supports the standard for IP encapsulation in X 25 RFC 877 HP router...

Page 17: ...ument Dial on demand Using ISDN switched circuits the router can be configured to have a V 25 bis terminal adapter automatically dial different phone numbers for primary and backup circuits or config ured to dial a different number for each destination IP address Some more information is found in the application note Improving Net work Availability in this manual Various features that help control...

Page 18: ...e is automatically detected from the type of cable attached no switches or configuration needed WAN link configuration of leased lines to another HP router or bridge is auto detected and negotiated no configuration for data link layer is needed SmartBoot Branch office routers acquire their configuration automatically when placed in service no expert personnel required at the branch site only at th...

Page 19: ...2 console connection using a modem for instance N o t e You don t need a specific router from HP for branch routing Any HP router except the HP Router PR can be the central router Any router except the HP Router 650 can be the remote router 2 A network administrator has stored a configuration file for the HP router on any Bootp server on the network HP provides applica tions that specify the most ...

Page 20: ...tribution of Software and Configurations The operating system software is preloaded and stored in flash EEPROM There are no disks to insert Software can be updated over the network TFTP the Trivial File Transfer Protocol can be used to push or pull either the operating system or the configuration file through the network This can be done from one centralized facility on the network Each router can...

Page 21: ... serial console port The following is a summary of the functions you can perform Set date and time and passwords Configure links and routing and bridging services View the current configuration Manually enable and disable configured links and routing and bridging services Review logged events to monitor the links and services being established View online network traffic statistics View routing an...

Page 22: ...figuration Router configuration in a multiprotocol net work can be a complex task HP routers simplify this task with a Quick Configuration utility Quick Configuration allows you to configure the minimum required parameters to enable links and routing bridging services in the most common configurations Parameters are entered or modified in a spreadsheet style configuration summary screen with the a...

Page 23: ...ntion For each remote router Quick Remote sets up the WAN link and the first LAN port Even if IP and IPX routing services are not needed Quick Remote aids in configuration of the other services that are needed For example if only AppleTalk routing is being used the central site administrator could minimally configure IP and Telnet services After the remote router begins routing then the administra...

Page 24: ...representation of network traffic can be sent using SNMP to a network server with less than 1 4 of 1 network overhead Then EASE applications can be used standalone or integrated into HP OpenView to analyze this information You can constantly monitor six network activities utilization frame rate broadcast rate multicast rate CRC and alignment errors in real time with customized threshold levels for...

Page 25: ...or two printed circuit assemblies PCAs one or two fans and a power supply in an enclosure The enclosure of each of the routers is illustrated in the Release Notes for your router The PCAs which use surface mounted components include the mother board containing most of the hardware components as described in the architecture section below Some of the routers contain a smaller daughter PCA as well T...

Page 26: ... system through the network or the console ports Self Test EPROM This stores the router s self test and console downloading code Station ID PROM This stores the station MAC address for each of the router s ports The processor accesses two types of random access memory RAM on the bus Data RAM Program variables and data critical to packet throughput such as routing and bridging tables are stored in ...

Page 27: ...ing such as frame delimiting with flags and FCS frame check sequence generation and detection Other information about LAN and WAN ports is found in the Features of HP Routers product note in this book and in the installation guides for your hardware Packet RAM Each LAN or WAN coprocessor stores packets received from the line the network medium into packet RAM and transmits packets from packet RAM ...

Page 28: ...errupting network connections and components that are still operating here called hot swap and replacements will be autodetected and autoconfigured The LEDs are moved out behind all the cabling into plain view An LED PCA above the slots provides multiple status indicators for each port on each card for overall chassis status and for the fan and redundant power supply Swap Factory Test 1 2 3 4 HP J...

Page 29: ... and compression as well Figure 3 Logical View of Router Series 600 Architecture The high speed multiported memory architecture allows the DLA modules to access packets at the same time as the routing engine does with no degradation in performance The backplane allows each DLA module to access the multiported packet memory The memory bus is dedicated solely to switching packets in a cell interleav...

Page 30: ...ogy requires routing between various media types the DLA modules hide the performance robbing idiosyncrasies of some network media Ethernet to Ethernet routing performance is unaffected by the presence of WAN and token ring source routed traffic All WAN specific processing such as compression X 25 and frame relay is handled by the WAN DLA The performance of the memory bus is more than one gigabit ...

Page 31: ... National s SONIC Systems Oriented Network Interface Controller is a second generation Ethernet controller that integrates a fully compatible 802 3 encoder decoder The token ring coprocessor is Texas Instruments TMS380C16 04 or TMS380C25 They support the 4 Mbit s and 16 Mbit s data rates as well as universal and local ring addressing A Motorola FDDI chip set is used for the dual attached stations ...

Page 32: ...ssed Cache RAM 128 kilobytes of high speed RAM are used for external processor code and data cache PCMCIA Flash EEPROM The flash memory on the removable PCMCIA card is used for nonvolatile storage of the operating system code and the configuration data The PCMCIA card can be replaced to update the operating system code or to provide more memory for updating through electronic media Self Test EPROM...

Page 33: ...re module of its own Features for the transparent or source routing bridge including the spanning tree protocol are also part of the bridging software module The redirector for bridging is here called the bridge redirector Protocols for the higher layers such as transmission TCP and application TFTP FTP layers are not used in routing decisions on the router but support of these protocols depends o...

Page 34: ...er throughput performance those directly involved with routing or bridging packets through the router This modular design allows for the addition of new protocols and links The four basic elements are driver data link service redirectors and circuit group manager Figure 6 Block Diagram of Software Data Flow Packets Architecture and Technology Software Data Flow Architecture 1 30 ...

Page 35: ...e for forwarding packets from one circuit group to another They receive packets from the data link service Table lookups result in subsequent modification of the packet header s address and control fields Redirectors then transmit the modified packets through the data link service to the specified driver and device s It should be noted that the spanning tree protocol routing protocols such as RIP ...

Page 36: ...es of routing protocol and can involve other features as well such as the import and export route filters for the IP routing service The most complex flow is for IP as shown in figure 7 Figure 8 shows the flow for IPX These diagrams of the flow of routing information complement the diagram of the flow of the actual data packets which are most of the packets transmitted in figure 6 Figure 7 Block D...

Page 37: ...Figure 8 Block Diagram of IPX Routing Information Flow Product Notes 1 Architecture and Technology Software Control Path Architecture 1 33 ...

Page 38: ...Architecture and Technology Software Control Path Architecture 1 34 ...

Page 39: ...ill be asked routinely to provide customers with multiprotocol support easy connectivity from home offices to branch office local area networks and uncomplicated management of the remote networks In addition managing the costs of installation use and maintenance of these internetworks will be a prime consideration This note covers the issues related to the implementation of branch office networks ...

Page 40: ...t cost effectively Compounding the problem is the fact that there are typically no MIS resources at the branch office Just as every corporation s networking requirements are different so are the solutions for each environment Connecting branch offices for example may be as simple as establishing a single Ethernet link per remote office back to a corporate or regional site using a leased line as sh...

Page 41: ...the business and that is as simple as possible to administer Branch Routing Requirements To develop cost effective solutions for branch office networks network designers must determine the basic communications requirements of the applications for which the network is being designed including the underlying protocols and the relative priorities of applications based on mission criticality Next but ...

Page 42: ...proving Performance HP has several new features that improve WAN performance add robust ness and get the most out of wide area communications facilities avoiding additional communications charges wherever possible These include HP packet by packet compression Prioritization bandwidth reservation Latency control Branch Office Branch Office Branch Office Branch Office Corporate Office HEWLETT PACKAR...

Page 43: ...on schemes based on the Lempel Ziv algorithm the actual compression ratio obtained depends on the type of data being trans mitted Generally the compression ratio obtained over a long period of time averages about 1 8 to 1 HP PPC is implemented in software and is currently supported over leased lines at link speeds up to 64 Kbit s Compression will be available on dial up circuits and frame relay ci...

Page 44: ...ncy Reducing Cost and Improving Robustness Once relegated to the communications backwaters circuit switching dial up technology is becoming an increasingly attractive proposition for branch office networking The near universal availability of switched digital communications Switched 56 and or ISDN as well as the new CCITT V fast recommendation clearly put circuit switching in the mainstream as far...

Page 45: ...s a unique backup solution that uses a small number of backup circuits and modems at a regional or corporate site to service a much larger number of remote sites with dial backup capability This allows a network designer to provide only as many backup ports as there are likely to be primary circuit failures at any one point in time Planning Controlling WAN Capacity Understanding wide area networki...

Page 46: ...ces by isolating them i e using traffic filters or with further network segmentation or through better scheduling of network activity thus avoiding the need to add costly WAN bandwidth Understand network traffic trends to anticipate future capacity requirements This allows the network administrator to carefully manage the implementation of additional capacity to ensure band width is added only whe...

Page 47: ...ecause of the cumula tive effect they have on reducing the MIS administrative burden Simplifying the administration of router based networks is an area in which HP is making significant contributions and will continue doing so Instant On Branch Office Router Installation Imagine the ideal solution for making router products installable in remote branch offices The solution would merely require the...

Page 48: ...e Installation Dynamic link configuration and SmartBoot help speed the installation of branch office routers and eliminate costly and time consuming configuration errors Dynamic Link Configuration In the fall of 1993 HP introduced dynamic link configuration for wide area interfaces using leased lines With dynamic link configuration leased line interfaces no longer require the configuration of data...

Page 49: ...the exception of the HP Router PR can store one remote configuration per WAN interface The configuration files created with Quick Remote include high level information such as addresses and which of the bridging and rout ing IP and or IPX modules are to be enabled Two interfaces one LAN and one WAN can be configured per remote router using Quick Remote SmartBoot is ideal for routing applications u...

Page 50: ...fective Solutions From the introduction of the Router ER HP s first router HP has led the initiative to reduce the cost of routing platforms Branch solutions such as the HP Router PR start at 2395 ranging to the newly announced high end HP Router 650 with an entry level configuration priced at 12 000 The HP Router 650 extends HP s family of routers to address the specific requirements of a headqua...

Page 51: ...ic link types including packet switched networks More traffic control and management features to improve application access to WAN resources New WAN traffic planning and analysis tools To help control and contain the administrative costs associated with implementing and managing branch networks HP is developing Instant On installation for routers in packet switched environments New management tool...

Page 52: ...Branch Office Routing Future Directions 1 48 ...

Page 53: ...2 Routing Services Notes ...

Page 54: ...ng Service Internet Protocol Routing Service Novell IPX Routing Service AppleTalk Phase 2 Routing Service DECnet Routing Service A Primer on HP Probe Data Compression for WAN Links Routing Services Notes 2 2 ...

Page 55: ... have redundant paths In the event of a primary link failure a backup link takes over thereby ensuring continued data transmission between all reachable network segments Translational bridging is used between FDDI and either Ethernet or IEEE 802 3 networks Source route translational bridging is used for bridging between token ring IEEE 802 5 and Ethernet or IEEE 802 3 networks These addressing sch...

Page 56: ...tion address also known as the MAC address Ethernet address and physical address The end node s station addresses is found in the source address of each frame received by the bridging service As the bridge receives frames it builds and updates a database that lists each source address This bridging database is known as a forwarding and filtering table or an address table it is comparable to the ro...

Page 57: ...me on the circuit group found in the address table With the spanning tree algorithm enabled the transparent bridge ensures a loop free topology by cooperating with other bridges in the extended network The algorithm provides a single path composed of bridges and intervening LANs between any two end nodes Spanning Tree Algorithm The IEEE 802 1 committee has issued a standard applicable to all MAC l...

Page 58: ... address table of neither bridge contains an entry for J or K Figure 1 Parallel Bridge Topology Each bridge updates its address table to indicate that J is in the direction of the red LAN After updating its address table each bridge floods the frame bridge 1 relays the frame over interface 1 and bridge 2 relays the frame over interface 2 Bridge 2 also relays the frame over interface 3 to simplify ...

Page 59: ...n parallel bridges The algorithm provides a single path composed of bridges and intervening LANs between any two nodes in such an extended network It also provides a high degree of fault tolerance by allowing for the automatic reconfigura tion of the spanning tree topology in the face of bridge or data path failure Five management assigned values are required for derivation of the spanning tree to...

Page 60: ... said to be in the forwarding state carries all extended network traffic to and from the LAN This process ensures that all redundant ports those providing parallel connections are removed from service placed in the blocking state In the event of a topological change or in the event of bridge or data path failure however the algorithm derives a new spanning tree that may move some such ports from t...

Page 61: ...pplications and explains concepts central to routing and bridging in a token ring environ ment A few token ring limitations are described at the end of the section The Token Ring Network Figure 3 Typical IBM Token Ring Network Figure 3 shows a typical IBM token ring network Second Floor Source Routing Bridge Token Ring First Floor Token Ring T o k e n R i n g Routing Services Notes 2 Bridging Serv...

Page 62: ...ifiers in the packet s Routing Information Field and then retransmit the packet on all network interfaces other than the one on which the discovery packet was received The destination system receiving the packet responds by returning the route information received in the discovery packet to the originating system Once the route has been discovered originating systems send data to the destination w...

Page 63: ...t of the data link layer there are no media dependent topological restrictions on the use of routers in large mixed media networks The network topology shown in figure 4 illustrates this point Each computer in the network is able to communicate with any other computer This assumes that all systems communicate using routable protocols supported on token rings IP IPX XNS and AppleTalk and that the r...

Page 64: ...red for source routing bridging can be used in any application that would otherwise be performed by an IBM Source Routing Bridge Source Routing and Transparent Bridging When bridging software is enabled transparent bridging is always enabled Optionally the bridge will also perform source routing bridging Bridges that perform both types of bridging concurrently are referred to as source routing tra...

Page 65: ...as does the individual group bit in the destination station address Source Routing Transparent Bridge Versus Source Routing Bridge A source routing only bridge provides inter ring communication only for systems that support source routing Systems that do not support source routing can communicate only with systems on the same token ring An SRT bridge in contrast provides inter ring communication f...

Page 66: ...s on Ethernet IEEE 802 3 LANs Bridging in Mixed Media Networks Differences Between Token Ring and Ethernet Addressing IEEE 802 5 token ring networks and FDDI differ from Ethernet networks in the transmission order of bits in each octet Ethernet nodes always transmit the least significant bit of a byte first Token ring nodes always transmit the most significant bit of a byte first see figure 6 The ...

Page 67: ...multicast packet Other systems on the Ethernet will interpret the packet transmitted by system A as a multicast packet The accidental multicast packet may be misinterpreted by systems as a valid multicast packet with unpredictable results Worse all bridges connected to the Ethernet will forward the packet In this case normal unicast packets are treated the same as broadcast packets with potentiall...

Page 68: ...e it was returned in the ARP reply by a system on a token ring This is again the first problem Figure 8 A dynamically determined address is misordered Tunneling Although the SRT bridge does not support bridging data between Ethernet systems and token ring systems for the reasons described above tunneling provides a mechanism to transmit Ethernet data over token ring backbones and to transmit token...

Page 69: ...g network the bridged packet is encapsulated for transmission to another HP router Thus tunneling requires one HP router to encapsulate the packet and another HP router to de encap sulate the packet Figure 10 shows the encapsulation performed on Ethernet packets to be transmitted over token ring backbones and on token ring packets to be transmitted over Ethernet backbones DECTerminalServer ROUTER ...

Page 70: ...ng Header Data Tunnel Header Destination Station Address 48 bits Source Station Address 48 bits DSAP 8 bits SSAP 8 bits Control 8 bits MSID 24 bits Type 16 bits Destination SRT Bridge Source SRT Bridge AA SNAP AA SNAP 03 000000 8101 Note If destination SRT bridge address in unknown address 0100A2FFFFFF is used Figure 10 Encapsulation Used to Tunnel Bridge Packets Limitation Note that when tunnelin...

Page 71: ... discovery The route discovery process is enabled by four types of routing directives listed below Each type is known by several names some of which are listed here All routes explorer ARE All paths explorer APE All routes broadcast ARB All paths broadcast Generates multiple frames that traverse all paths between source and destination stations Such frames are called all routes explorer ARE or all...

Page 72: ...eiving an SRE each bridge on the spanning tree forwards the frame onto all active non blocked ports except the port on which the frame was received With spanning tree broadcast routing one copy of the SRE appears on each LAN and the frame recipient receives only a single copy of the frame Some other terms referring to a looped topology are alternate routes parallel bridges and redundant bridges Sp...

Page 73: ...Every source routing bridge that the frame passes through inserts routing designators in the frame s MAC header Transparent bridges however do not write to the MAC header Each routing designator pairs a LAN segment number with a bridge ID number in order to identify a portion of the frame s path through the bridge Figure 11 SRT Bridge Routing Designators In figure 11 the router receives the source...

Page 74: ...osses the network every source routing bridge that it passes through inserts routing designators in the frame s MAC header If the frame crosses a transparent bridge the bridge simply forwards the frame on the basis of its destination station address 8 0 Outgoing Path 5 1 Incoming Path A 1 Internal Path 5 Identifies the incoming LAN segment 1 Identifies the source routing bridge A Identifies the in...

Page 75: ...tors 1 1 7 1 5 0 identify the frame s path through bridge A Note that in figure 14 bridge A inserts the incoming LAN ID bridge ID pairing 1 1 only because it is the first source routing bridge to receive the frame If A was not the first source routing bridge to receive this frame the incoming LAN ID bridge ID pairing would be taken from the outgoing LAN ID bridge ID pairing inserted by the previou...

Page 76: ...Routing Designator 2 1 1 1 identifies the incoming LAN segment token ring 1 1 identifies bridge A Incoming Path 7 1 7 identifies the internal virtual LAN segment 1 identifies bridge A Internal Path Bridging Service Source Routing Bridging 2 24 ...

Page 77: ...ed since LAN segment 3 does not provide a path to H2 1 1 7 1 5 1 8 1 6 0 The above ARE frame is transmitted onto LAN segment 6 this frame passes through transparent bridge T Because T in figure 13 is a transparent bridge it does not recognize the routing designators in the source routed frame T treats the ARE frame as if it was a transparent bridging frame and simply forwards the frame to router C...

Page 78: ...smits the ARE frame to the destination node H2 H2 inspects the frame s MAC header to learn the route that this particular frame traversed The route is mapped out by the sequence of routing designators 1 1 7 1 5 1 8 1 6 1 9 1 2 0 inserted by source routing bridges A B and C N o t e Other route discovery protocols are available and extensively used All such protocols however exchange TSFs ARE frames...

Page 79: ... is performing source routing bridging is referred to as a bridge in this section Internal LAN ID 100 Internal LAN ID 101 Internal LAN ID 102 Group LAN ID FFF Group LAN ID FFF Group LAN ID FFF Bridge ID A Bridge ID A Bridge ID A Figure 17 Sample HP Source Routing Bridging Network The following three subsections describe three tracks listed below taken by source routing frames in the sample network...

Page 80: ... frame toward ring 2 Figure 19 illustrates the RIF in the frame Adds the incoming LAN ID and its bridge ID Adds the internal LAN ID and its bridge ID Adds the outgoing LAN ID and bridge ID of 0 Other bridges receiving the explorer frame The explorer frame received by bridges B and C contains internal LAN IDs thus indicating that this frame has traversed at least one other bridge Each of these brid...

Page 81: ...sent by bridge A onto ring 2 Incoming LAN ID bridge ID Internal LAN ID bridge ID Outgoing LAN ID bridge ID of 0 H2 H1 8A30 001A 002A 101A 0030 DSAP SSAP Data Frame sent by bridge B onto ring 3 Incoming LAN ID bridge ID Internal LAN ID bridge ID Outgoing LAN ID bridge ID of 0 H2 H1 8C30 001A 002A 003A 102A 004A DSAP SSAP Data Frame sent by bridge C onto ring 4 Figure 19 Structure of an Explorer Fra...

Page 82: ...FFF Group LAN ID FFF Bridge ID A Bridge ID A Bridge ID A Figure 20 Tracking a Specifically Routed Frame Back to Node 1 First of several bridges receiving the SRF The frame received by bridge C from ring 4 has not traversed any other bridges yet But there are multiple bridge IDs in the RIF so it will This bridge does the follow ing to the RIF before transmitting the frame toward ring 3 Figure 21 il...

Page 83: ... end of the HP group address to the bridge ID of the next bridge in the RIF In this example all bridge IDs are the same so the frame is not modified Last of several bridges receiving the SRF Bridge A is the last of several bridges traversed by the frame This bridge does the following to the RIF of the frame it receives before transmitting the frame toward ring 1 See figure 21 Replaces the HP group...

Page 84: ... 0CB0001A FFFA 002A 003A 0040 DSAP SSAP Data H1 Frame sent by bridge C onto ring 3 HP group address Next bridge ID C000A2FFFFFA H2 0CB0001A FFFA 002A 003A 0040 DSAP SSAP Data H1 Frame sent by bridge B onto ring 2 Saved destination station address Next bridge ID H1 H2 0CB0 001A 100A 002A 003A 0040 DSAP SSAP Data Frame sent by bridge A onto ring 1 Figure 21 Structure of a Specifically Routed Frame B...

Page 85: ...100 Internal LAN ID 101 Internal LAN ID 102 Group LAN ID FFF Group LAN ID FFF Group LAN ID FFF Bridge ID A Bridge ID A Bridge ID A Figure 22 Tracking a Specifically Routed Frame from Node 1 to Node 2 First of several bridges receiving the SRF Bridge A is the first to receive the frame from ring 1 and does the following to the RIF before transmitting the frame toward ring 2 Figure 23 illustrates th...

Page 86: ... to the bridge ID of the next bridge in the RIF In this example all bridge IDs are the same so the frame is not modified Last of several bridges receiving the SRF The last bridge to receive the frame does the following to the RIF before transmitting the frame toward ring 1 See figure 23 Replaces the HP group address with the destination station address that was saved to the data field Replaces the...

Page 87: ...03A FFFA 0040 DSAP SSAP Data H2 Frame sent by bridge A onto ring 2 HP group address Next bridge ID C000A2FFFFFA H1 0C30 001A 002A 003A FFFA 0040 DSAP SSAP Data H2 Frame sent by bridge B onto ring 3 Saved dest station address Next bridge ID H2 H1 0C30 001A 002A 003A 102A 0040 DSAP SSAP Data Frame sent by bridge C onto ring 4 Figure 23 Structure of a Specifically Routed Frame from Node 1 to Node 2 R...

Page 88: ...ferent methodologies are provided concurrently The transparent bridge treats all frames as if they are transparent bridging frames In order to effect route discovery however the SRT bridge needs to separate frames which require source routing service from those frames which require transparent bridging service Figure 24 Sample SRT Topology In order to identify source routing frames the SRT bridge ...

Page 89: ...e Before reading on ensure that you understand the explanation of trans parent bridging TB source routing bridging SRB and global source routing earlier in this section Figures 25 through 28 below illustrate some common topologies that can use TRNSB Figure 25 Simple One Hop TRNSB Topology In figure 25 the PC running the PC3270 application is limited to either 802 3 or Ethernet frame format and tra...

Page 90: ...etBEUI SRB 802 5 frames to NetBEUI TB 802 3 frames Figure 27 Token Ring Backbone Topology Figure 27 shows an example of a 16 Mbit s token ring SRB backbone used to interconnect two distributed Ethernets This topology is often found in envi ronments where many departments have installed their own LANs while the site services organization often in coordination with the information tech nology organi...

Page 91: ...ning tree environment If the network in figure 28 had mostly token ring SRB networks the WAN circuit groups would more likely be configured as source routing links This would work also Please note that with TRNSB functionality enabled the WAN links must all be config ured as source routing or all not source routing but cannot be mixed Token Ring HP Router TR HP Router ER Transparent WAN Backbone H...

Page 92: ...onverted to and from 802 3 frames Since bit ordering differs between these types of networks the bits of the station addresses DA SA must be reversed The Routing Information Field RIF is extracted when moving from 802 5 to 802 3 and inserted when going in the other direction The TRNSB maintains a store of RIFs see below The 802 3 length field is also created or removed as necessary Figure 30 802 5...

Page 93: ...the frame is converted to Ethernet 802 3 format and forwarded to that circuit group If the destination is not known then the TRNSB converts the frame and floods it to all transparent circuit groups configured for TRNSB If the destination address in the frame indicates a functional address or the broadcast address the frame is always flooded In all of these cases the source of the frame is learned ...

Page 94: ...addresses that use the other format can be specified Most end nodes use the 802 3 format Limitations TRNSB does not support protocols that can be routed TRNSB does not operate with looped topologies that mix SRB and TB That is TRNSB operates in looped topologies only when the links in the loops are either all SRB or TB but not a combination of both TRNSB does not support either the hop count reduc...

Page 95: ...hops set Hop Count Reduction to Yes But if your network never needs more than seven hops it is recommended that you leave this parameter set to the default No Note that when used with certain network topologies the hop count reduction algorithm generates extra explorer frame traffic and multiple redundant route responses In certain topologies such redundant traffic may be insignificant In others e...

Page 96: ... node 1 The RIF on ring 2 of the frame head ing back to end node 1 is then 88F0 0011 0023 0030 End node 1 saves this RIF as the path to end node 2 Hop Count Reduction Algorithm Disabled In figure 31 above with Hop Count Reduction set to the default No end node 1 sends an explorer frame The RIF at ring 1 is 8270 the RIF at ring 2 is 8670 0011 0020 The bridging HP router receives the frame from ring...

Page 97: ... on bridge IDs see an earlier section of this note How Source Routing Works and the next section on configuration In a network with multiple bridges when Hop Count Reduction is off and each router s bridge ID is different there is a routing designator pair for every LAN or ring incoming and outgoing path and for every bridge an internal path identifying a virtual LAN and the bridge All the bridges...

Page 98: ...own as a virtual LAN ID Each internal LAN ID must be unique among all internal LAN IDs and LAN IDs throughout the network and be different from the group LAN ID Bridge ID The bridge ID for this router identifies a specific source routing bridge this router To keep the hop count partially reduced see the preceding section on hop count reduction it should generally match the bridge ID assigned to al...

Page 99: ...s set to 1 which specifies that an additional path exists through router B Hop Count Reduction Configure it to be No unless you want infinite hop capability No specifies a maximum of seven hops between the source routing source and destination Group LAN ID This is used by the bridging router when transmitting specifically routed frames SRFs between HP or Wellfleet bridging routers Together with th...

Page 100: ...uration parameter is used to enable TRNSB for each circuit group that will translate frames between source routing bridging circuit groups and transparent bridging circuit groups Another parameter screen for the bridging service on this router covering all circuit groups has some parameters specifically for translational bridging The aging timer can be left at the default The default frame convers...

Page 101: ...s of destination address information six octets of source address information and two octets of protocol type information to the frame It appends a four octet frame check sequence to the frame Figure 34 802 2 Encapsulation 802 2 encapsulation shown in figure 34 prefixes one octet of destination service access point identification one octet of source service access point identification and one octe...

Page 102: ...re 36 Novell Proprietary Encapsulation Novell proprietary encapsulation shown in figure 36 prefixes an eight octet preamble six octets of destination address information six octets of source address information and two octets of frame length information to the unchecksummed IPX frame indicated by a value of FFFF It appends a four octet frame check sequence to the frame DSAP 1 octet SSAP 1 octet Co...

Page 103: ...ter Fields Encapsulation Method Predefined Fields All Source station addresses Destination station addresses Ethernet Type 802 2 SSAP DSAP SNAP Organization Ethertype The bridge supplements basic filtering functionality by providing the ability to specify user defined fields within each of the supported encapsulation formats It also provides the ability to specify lists which contain a collection ...

Page 104: ...ce that identifies which action to take in the event of a frame that meets the conditions of more than one rule A filter rule consists of three entities A specified field or fields in the frame header A value or range of values associated with the field An operator which specifies the relationship between field and value A filter operator may take one of three values ignore match or don t match A ...

Page 105: ...rity This option gives you the following capabilities Configure bridged traffic to a lower priority than routed traffic which is always normal Configure bridged traffic to a higher priority than routed traffic which is always normal Configure different priority levels for different types of bridged packets within a circuit group This is done with traffic filters by setting the Action to high or lo...

Page 106: ...Bridging Service Traffic Prioritization 2 54 ...

Page 107: ...s Local area networks that use the Internet TCP IP protocol suite can communicate with each other through IP routers A TCP IP network or IP network is typically an entity wide geographically dispersed network consisting of several subnetworks connected by routers Example entities include companies academic institutions government agencies or subgroups of these IP subnetworks may consist of LAN or ...

Page 108: ...nic mail and news Simple Mail Transfer Protocol SMTP Privacy Enhanced Mail PEM Remote terminal emulation Telnet virtual terminal rlogin tn3270 access to IBM VM CMS computers Window systems X Window System Time synchronization Time protocol Network time protocol NTP Security systems Kerberos authentication and authorization system Network management Simple Network Management Protocol SNMP ISO CMIP ...

Page 109: ...gn and register your IP addresses Telephone in U S only worldwide 800 365 3642 703 802 4535 Mailing Address Government Systems Inc Attn Network Information Center 14200 Park Meadow Drive Suite 200 Chantilly VA 22021 E mail hostmaster nic ddn mil If your network is isolated and will not ever be connected to any other IP networks you can build your own IP addresses If you use your own addressing sch...

Page 110: ...twork host 32 bits or network subnet host 32 bits The address field length of an IP address depends on the address class The class A address assigns 8 bits to the network field and a total of 24 bits to the rest of the address This address class can address almost 17 million different nodes on an IP network The class A address is used for very large networks The class B address assigns 16 bits to ...

Page 111: ...l potentially be needed on the network and the total number of nodes in each subnetwork Make sure that the lengths of each field will accommodate these projected totals and leave room for expansion See Suggestion following the discussion of subnet masks below The actual subnet and host number values are assigned by the network administrator Neither the subnet or host number can be all 1s or all 0s...

Page 112: ...11 0011 0100 0001 0011 dotted decimal 132 7 52 19 address fields network subnet host 132 7 52 19 The address class can be determined by the first decimal number This is because the leftmost bits are assigned according to the address class For a class A address the first digits are in the range 1 through 126 leftmost bit is 0 For a class B address the first digits are in the range 128 through 191 l...

Page 113: ...ask for an IP address with field lengths of network 8 subnet 8 and host 16 is 255 255 0 0 IP Addresses binary 1111 1111 1111 1111 1111 1111 0000 0000 subnet mask 255 255 255 0 address fields network subnet host 16 bits 8 bits 8 bits Notice that the field division may not correspond neatly to the dotted decimal divisions as in the above examples For example the subnet mask for an IP address with fi...

Page 114: ...numbers at the least significant bit of the host field For example starting with a class B address and subnet mask 255 255 255 0 start assigning subnets 128 64 192 32 160 etc Within each subnet start assigning nodes 1 2 3 4 5 etc Figure 1 illustrates an example class B network First Subnets First Hosts Binary Decimal Binary Decimal 1000 0000 128 0000 0001 1 0100 0000 64 0000 0010 2 1100 0000 192 0...

Page 115: ... H respectively The New Orleans bridged LAN and WAN links form a single subnetwork A The link between Paris and Lyon through the Transpac X 25 packet switching network forms a single subnetwork J Figure 2 Example IP Network New York Network Address Subnet Mask HP 9000 HP OpenView Interconnect Manager Fractional T 1 224 Kbit s A E T 1 1 544 Mbit s B 128 1 0 0 255 255 255 0 C Paris 64 Kbit s J Trans...

Page 116: ...ntify individual hosts and router interfaces All addresses will be of the form 128 1 sss hhh where sss is a number 1 254 identifying the subnetwork and hhh is a number 1254 identifying a node host on that subnetwork Note that each port network interface on the router resides on a different subnetwork and requires a unique IP address For more information on IP addressing refer to the student workbo...

Page 117: ... route is no longer available then the route is marked invalid The route will remain in the table until the router is notified of a valid route or is rebooted Route Learned indicates whether the route was statically learned that is the route was specifically configured or learned dynami cally through one of the routing protocols RIP OSPF or EGP Age indicates the number of seconds since the route w...

Page 118: ...erfaces with lower speed networks For further explanation refer to the IP Network Interface Definition section in the IP chapter in the reference manual OSPF Open Shortest Path First OSPF is an interior gateway protocol IGP for exchanging network reachability and routing information either within or between autonomous systems Configuring OSPF is more involved than configuring RIP However it is muc...

Page 119: ...o mous systems When connecting an autonomous system to the Internet every autonomous system must use EGP to advertise network reachabil ity to the Internet s core gateway system The Internet is a globally administered confederation of autonomous systems networks Note that any or all of the supported routing protocols may be active on the router simultaneously Each of the routing protocols derives ...

Page 120: ...wn unique IP address Note that the nodes on the two subnetworks will always communicate with each other through the router even if they are on the same cable For more than one subnetwork on a port configure that circuit group in more than one network interface definition each with a different IP address Suppose for example that LAN interface B shown in figure 3 accesses networks Nx and Ny In this ...

Page 121: ...will not receive routing information about the other network See figure 4 for an example For a node on the Red net to reach a node on the Green net on this router configure a static route on network inter face 1 using the destination IP address of router X s interface B and using the IP address of router X s interface A as the next hop The cost is 1 Router X also must have a static route configure...

Page 122: ...ound to router 1 The Red and White networks attached to router 1 are included in routing tables on all the other routers so the Red and White networks are reachable by all Network interface 3 on router 1 is a default route to the other routers interface A on router 2 is the next hop RIP Supply will be configured but RIP Listen Default Route Supply and Default Route Listen will not be configured Co...

Page 123: ...route to router X interface A on router X is the next hop RIP Supply RIP Listen Default Route Supply and Default Route Listen will not be configured Cost is not applicable Also configure static routes on router X to the Red network and to router 2 Figure 6 Default Route Example 2 Static route example 2 To restrict packets for security reasons to paths you specifically configure all routes would be...

Page 124: ...be will be used for address resolution Normal ARP should be on Figure 7 Conditional Next Hop Routing 10 1 8 x 10 2 8 x 10 3 8 x 10 4 8 x 11 1 8 1 11 1 8 3 11 1 8 4 12 1 8 1 12 1 8 2 12 1 8 3 12 1 8 4 ISDN Network WAN Network Next Hop for Static Routes Next Hop for Conditional Static Routes Router B Router C Router A Router D 11 1 8 2 Internet Protocol Routing Service Static Routing 2 72 ...

Page 125: ...t hop for router D to network 10 1 8 x through the ISDN dial up static route is 12 1 8 1 The configuration in figure 7 solves any single point WAN failure to routers A B or C For example if the WAN link to router B breaks traffic between B and all other routers will be conducted through the ISDN dial up static routes to D D will advertise the new route to B over its frame relay connec tion If howe...

Page 126: ...2 B n a 5 10 3 8 x 12 1 8 3 C n a 3 Preference and the Routing Pool The IP router maintains a routing pool which contains information supplied by up to three routing protocols RIP EGP OSPF in addition to statically configured routes Consequently the routing pool may contain multiple routes to the same destination Each route carries an associated preference value that determines the best route wher...

Page 127: ... by a neighboring router You can configure up to four default routes By configuring multiple routes you ensure that a packet can be re routed if the interface associated with the default route is disabled You assign preference values between 1 and 16 with 16 being the highest to each default route to determine which default route gets the highest prior ity If the highest priority default route is ...

Page 128: ...twork Specify an adjacent host if you are setting up a network or if a particular local host or hosts don t respond to ARP requests The static routing type is configured to be adjacent host and one of three encapsulation methods is configured as follows Internet Protocol Routing Service Static Routing 2 76 ...

Page 129: ...sulation 802 2 can be used for hosts supporting IEEE 802 2 over IEEE 802 3 LAN interfaces The 802 2 structure is encapsulated as shown and is further encapsulated within a medium specific 802 x packet Figure 10 IEEE 802 2 Encapsulation Protocol Packet Data 46 1500 octets FCS2 2 octets Source 6 octets Destination 6 octets Preamble Type1 IP Header 1 0800h 2 octets 2 FCS frame check sequence IP Heade...

Page 130: ...AP structure is encapsulated within a medium specific 802 x packet Figure 11 SNAP Encapsulation DSAP 1 octet 170 AAh IP Header Protocol Packet Control 1 octet Ether Type 2 octets 0800h Data SSAP 1 octet 170 AAh Organization 3 octets 0h Internet Protocol Routing Service Static Routing 2 78 ...

Page 131: ...xample upon receiving a new routing update RIP consults its specific import rules to validate the information before inserting the update in the routing pool Import rules contain search information used to match fields in incom ing routing updates and action information used to specify the action to take with matched fields RIP export rules govern the propagation of routing information by RIP RIP ...

Page 132: ...e active in the routing pool Most rules include a network number and mask however the network number and mask are not always required Each import rule can contain up to three search fields however it need not contain any As a result Ignore RIP is a valid rule If such a rule were not specified any received RIP packets that did not match a rule would be accepted into the routing pool If a rule is cr...

Page 133: ...yer services such as Telnet FTP and TFTP The User Datagram Protocol UDP and Transmission Control Protocol TCP are internet trans port level protocols TCP provides a reliable connection mode packet serv ice while UDP provides connectionless datagram service UDP packets and TCP segments are originated by ports and addressed to ports Ports are logical abstractions used by transport level protocols to...

Page 134: ...E TCP 17 QUOTE TCP UDP 102 ISO TSAP TCP 19 CHARGEN TCP UDP 113 AUTH TCP 20 FTP DATA TCP 117 UUCP PATH TCP 21 FTP TCP 123 NTP TCP UDP 23 TELNET TCP 133 159 UNASSIGNED 25 SMTP TCP 160 223 UNASSIGNED 37 TIME TCP UDP 224 241 UNASSIGNED 39 RLP TCP UDP 247 255 UNASSIGNED Filter Rules Filtering decisions are based on user defined rules An IP filter rule consists of An IP UDP TCP field or fields A value o...

Page 135: ...re a name to identify the list and one or more pairs of numbers that specify a range When a filter specifies a list name packets are checked against the range s of values specified by the list Two types of lists are defined for the two types of packet filters IP address lists specify ranges of IP network addresses Specify the lower and upper boundary of each IP address range to be in the list or j...

Page 136: ...unction with IP address and or UDP TCP port filters You can specify the location of the bit pattern to filter by giving an offset position starting position offset from the beginning of the selected header type and a length ending position See figure 14 Figure 14 User Defined Field Positioning IP Header or Header of Upper Level Protocol Length Offset Low Value hex High Value hex Internet Protocol ...

Page 137: ...ocol allows an HP router to be a time server for other network devices or a client of a time server or both SNMP Agent Simple Network Management Protocol agent enables the router to be managed by network management applications across the network HP OpenView products can access the vendor proprietary objects on the HP routers and can create configuration files for the router Addition ally many of ...

Page 138: ...eded by an HP router to boot and route and it can supply the names of the Bootp file server and full boot file This Smart boot use of Bootp and IP is configured on the HP routers by default See the discussion of Instant On and Smartboot earlier in this chapter In the remainder of this discussion Bootp is used to refer to the combination of Bootp and the DHCP extension The most common networking ap...

Page 139: ...a relay agent If that adjacent relay agent has a configured set of Bootrequest destinations that are either specific server addresses or subnet addresses then it is the only router that needs the relay agent functionality See figure 15 If the adjacent relay agent does not have either server or subnet addresses configured then the Bootrequest is sent out through all of the adjacent router s interfa...

Page 140: ...Address is an IP address with a valid value for the network number and subnet number but either all zeroes or all ones for the host part For example the class A network number 15 with a subnet mask of 255 255 255 0 has the valid subnet address 15 1 1 0 15 1 1 255 The address 15 1 255 255 would not be a valid subnet address This would be treated as a broadcast to all subnets in network 15 If a subn...

Page 141: ...he subnet on which the packet is sent for example 15 1 1 255 Since the packets are addressed only to the adjacent subnet all routers in the path between the client and server must have Bootp relay agent functionality For the second case routers that are not relay agents drop the Bootrequest packets The other relay agents relay the Bootrequest packets in accordance with their configured destination...

Page 142: ...es on which you anticipate having device management IP traffic information The example network in figure 17 is configured to have IP traffic on all links except the LANs in Chicago and Washington D C Note that the IP addresses are the same for all IP interfaces on any single router The subnet mask is the same for all IP router interfaces and devices throughout the network The same IP address resol...

Page 143: ... mask 255 0 0 0 Figure 17 Virtual IP Host Configuration on a Novell IPX Internetwork Novell IPX Only Chicago Washington D C A 1 0 0 10 Novell IPX only A Novell IPX and IP Novell IPX and IP Novell IPX and IP 1 0 0 2 1 0 0 2 Key Interface with Novell IPX enabled Interface with Novell IPX bridging and IP host only enabled A IP network network management station N B I N N B I N B I N 1 0 0 3 Router 3 ...

Page 144: ... is because it does not meet the criterial for a router even though its ID is for a router Also because there is no IP routing information being kept the router cannot act as a default gateway for HP management products to provide routing information for autodiscovery In an IPX network the router is described by an SNMP IPX Openview DOS platform as a router However the router only supports SNMP IP...

Page 145: ...route end node support Source route end node support allows IP nodes on Ethernet IEEE 802 3 and token ring IEEE 802 5 LANs to communicate with each other You should enable source routing when you are configuring the IP routing services for each of your router s network interfaces IP Router Operations Use the Network Control Language Interpreter NCL to display IP events access IP statistics view AR...

Page 146: ... IP routing The IP routing service on HP routers is compatible with them General IP TCP UDP Subnetting Gateways IEEE 802 networks MTU discovery RFC 791 RFC 793 RFC 768 RFC 950 RFC 1009 RFC 1042 RFC 1063 Routing Protocols RIP OSPF EGP Static routes RFC 1058 RFC 1247 RFC 904 Address Resolution ARP HP Probe IP over X 25 RFC 826 HP proprietary RFC 877 Filtering Source address Destination address TCP U...

Page 147: ...is now used IW2 specifies initial connection setup methods for running IPX traffic over WAN media such as PPP X 25 switched virtual circuits X 25 permanent virtual circuits and frame relay HP routers will use the IPXWAN and IW2 protocol when it is configured for a WAN link to interoperate with a Novell router NetWare Services Novell NetWare provides services that vendors and end users can use to d...

Page 148: ...erations in the case of system failures General Addressing Considerations An IPX packet header contains a source network address field and a destination network address field The network address is composed of three fields as shown in figure 1 The network number uniquely identifies a network in the internetwork A host number uniquely identifies a single host The socket number uniquely identifies a...

Page 149: ...ess also called MAC or physical address of the device is used as the host address This ensures uniqueness and is one less parameter to configure For the HP routers you do not specify an IPX host address the station address of WAN port 1 will be used except for operating system versions earlier than 5 70 Control Destination Network Address Source Network Address NetWare Core Protocol packet 0 to 54...

Page 150: ...ost number The net work number can be an 8 digit hexadecimal number Subnetworks that are connected by bridges are assigned the same network number Subnetworks that are connected by routers must be assigned different network numbers That is an 8 digit hexadecimal number must be assigned to each router LAN Ethernet IEEE 802 3 and or token ring and each WAN port that will be routing IPX packets For e...

Page 151: ...mation six octets of source address information and two octets of packet length information to the IPX packet It appends a four octet frame check sequence to the packet Choose Novell encapsulation when using HP routers in Novell NetWare internetworks N o t e This encapsulation choice will be valid only for IPX routing services and will have no effect on the encapsulation used by other enabled brid...

Page 152: ...d by the router For each destination network number entry the table also contains the following information Next Hop If the destination network is directly connected to this router then next hop is the host number of this router station address of WAN port 1 If the destination network is a remote network the next hop is the host number of the next router along the path to the destination Hop the n...

Page 153: ...ly routes can be statically configured Routing Information Protocol The Routing Information Protocol RIP is the dynamic routing protocol used in Novell IPX internetworks for gathering and maintaining routing infor mation Novell RIP supports both request and response operations RIP requests are used by hosts to determine the network number of the network that they are connected to or to determine t...

Page 154: ...ervice Advertising Protocol SAP is used by NetWare servers to inform NetWare clients of their presence SAP is an ISO application layer protocol that runs directly over IPX an ISO network layer protocol The Service Advertising Protocol allows a file print server or program to register its name on the network The programs may be application programs written by third party developers or users A SAP r...

Page 155: ...logically partition IPX internetworks by controlling the advertisement of servers A NetWare client cannot access a server that is not advertised to it All servers in the bindery are advertised if no filters are configured SAP filters are configured for each interface Two levels of SAP filters are supported Network level filters permit or deny advertising of servers that match a pattern of network ...

Page 156: ... rebroadcasting of NetBIOS broadcast packets can be configured for each interface Therefore the router has the capability to restrict server or client access to remote connections There are two parameters used to configure Novell NetBIOS broadcast handling on an IPX router interface Set the Accept NetBIOS Bcasts from Net parameter to enable or disable client access on a given IPX network to NetBIO...

Page 157: ... in the IPX internet To configure a NetBIOS static route you must specify a NetBIOS resource name and the destination network where it resides If the NetBIOS resource name in a broadcast packet matches a static route table entry then the NetBIOS packet is routed to the destination network If no match is found then the packet is treated as specified by the Accept NetBIOS Bcasts from Net and Deliver...

Page 158: ...Novell IPX Routing Service NetBIOS Protocol Support 2 106 ...

Page 159: ...er An AppleTalk internet is a collection of one or more AppleTalk networks connected by AppleTalk routers An AppleTalk network is a LAN that contains end nodes and connects to at least one AppleTalk router The HP routers support CSMA CD LANs EtherTalk and token ring LANs TokenTalk There are three ways that routers are used to build an internetwork Local router A router used to connect LANs only is...

Page 160: ...re 2 Half Router Configuration Backbone router Routers each connected to one or more AppleTalk LANs connected together through either an Ethernet backbone or an X 25 packet switching backbone network are called backbone routers Backbone networks do not have any AppleTalk end nodes attached Figure 3 Backbone Router Configuration AppleTalk Phase 2 Routing Service 2 108 ...

Page 161: ...gure 4 AppleTalk Network Map Example Zone Advertising Network 7 8 Zone Marketing Network 4 5 R3 R1 R2 Zone R D Network 3 3 S S S Network 6 6 Zone Marketing Network 9 11 S S S S Zone R D Network 1 1 Network 2 2 S Seed Routing Services Notes 2 AppleTalk Phase 2 Routing Service 2 109 ...

Page 162: ...etwork number range must be unique and must not overlap This network number range is configured by the user on at least one of the routers residing on the network That router is called the seed router for that network The AppleTalk nodes dynamically determine the network number portion of their node address If the node is being restarted and there is a network number stored in non volatile memory ...

Page 163: ...ame searching can be done within one or more user specified zones Each node in an AppleTalk internet can select one of several possible zone names from the list of allowable names for its network The seed router on each AppleTalk LAN network is configured with a list of zone names that can be used by nodes on that LAN If a node on a network is assigned a zone name that is not on the configured lis...

Page 164: ...outers using RTMP routing protocol messages If you do not configure the network number on a router that router can learn the number from the seed router The seed router can optionally be configured with a list of zone names that will be valid on the network Any nodes on the network with a node name not on this list will be assigned the default zone name by the seed router Note that the zone name l...

Page 165: ... addresses Ensures the integrity of node addresses When a node address is assigned AARP confirms the uniqueness of that address When a node identifier is dynamically determined it must be automatically validated as a unique node identifier on a given AppleTalk LAN AARP generates a random node identifier or if the node is being restarted and a node identifier already exists in non volatile memory t...

Page 166: ...names Note that the ZIT is not displayed by the HP routers However the Local Zone Table a table of zone names configured on a particular router can be displayed using NCL s Rgetat command Name Binding Protocol NBP on a Router Maps network entity names with internet addresses It allows you to specify descriptive or symbolic names while other processes may refer to the same entity numerically for in...

Page 167: ...es System Communications Architecture SCA also known as VAX clusters uses System Communication Services protocol at the network layer This architecture allows multiple computers to share disk drives over Ethernet at 10 Mbit s or over the Computer Interconnect CI bus at 70 Mbit s DECnet OSI Architecture DEC s implementation of the Open Systems Interconnect OSI protocols for DECnet Phase V It uses I...

Page 168: ...ion Distributed File System DFS allows a remote file to appear to the user as though it were local Distributed Queuing Service provides access to remote printers Videotex is an interactive system that allows subscribers to down load several pages of information to a local terminal or node General Addressing Considerations DECnet Phase IV uses a hierarchical addressing scheme DECnet internet works ...

Page 169: ...Planning for a DECnet internetwork involves defining areas and assigning unique addresses to each router and end node in the internetwork When configuring DECnet nodes individual addresses must be configured on each router and end node Note that a single address is assigned to the router itself not to each port or interface on the router DECnet addresses consist of a 6 bit area number 1 63 a 10 bi...

Page 170: ...n address resolution for DECnet node 1 10 is shown in figure 2 The station address for node 1 10 will be AA0004000A04h Decimal Hexadecimal Binary DEC Vendor Code 24 bits AA0004 Area Number 6 bits 1 000001 Node Number 10 bits 10 0000001010 Station Address 48 bits 6 octets Conversion Pattern Binary Representation Hexadecimal Representation 7 0 7 0 Octet 5 DEC Vendor Code 3 octets 1 0 1 0 1 0 1 0 A A...

Page 171: ...e associated with it Ideally the cost value should reflect the speed of the link The router uses this information to select the fastest least cost path to a destination Therefore cost values have a significant effect on what routing decisions are made by the router Planning a DECnet internetwork involves determining the cost value for each link These values should be documented on the network map ...

Page 172: ...ters with equally high priority the router with the highest node number will be elected as the designated router Note that a router that is connected to multiple LANs may be the designated router on some LANs but not on others A designated router will be elected even if there are no end nodes currently on the Ethernet In figure 1 designated router ports for each Ethernet LAN are highlighted with a...

Page 173: ...on routes to destination nodes within the area The level 2 routing table contains information on routes to destination areas Figure 3 shows a two area DECnet internetwork Each area has two routers When end node 1 10 sends a packet to end node 2 10 it is routed through all four routers in the internetwork When router 1 1 receives the packet des tined for node 2 10 the router must refer to its level...

Page 174: ...st assignments for a DECnet internetwork and configure them on HP routers see the DECnet Link Cost section above The path cost is the cumulative cost value of the links along the path to the destination node in the level 1 routing table or area in the level 2 routing table For a given destination node or area DECnet routing tables reflect the least cost path that the router has learned from period...

Page 175: ...t may or may not update its routing table based on whether the router learns of a new destination node or area or a lower cost path On HP routers routing messages are multicast every 3 minutes by default However a period as short as every 15 seconds can be configured by modify ing the Bcast Routing Timer parameter using the Configuration Editor Connecting to Routers from DEC HP routers use the HDL...

Page 176: ...se MIB variables For detailed infor mation on NCL and the various DECnet routing service operations refer to the user s guide and reference manual Specification DECnet routing service on HP routers is compatible with the following specification DECnet Digital Network Architecture Phase IV Routing Layer Functional Specification December 1983 DECnet Routing Service DECnet Router Operations 2 124 ...

Page 177: ...rmat node domain organization for its node names If it does not already know the address for the destination node then one or more Probe requests are generated The destination of the request is a target node or a Probe name server called a proxy server which will respond with a Probe reply that contains the requested path or address information A Probe name server is a machine that contains a mapp...

Page 178: ...re the well known NS services it supports the protocols it supports its name and machine addresses A node uses the Unsolicited Reply packet to announce its presence on the network or to update information about itself that was transmitted earlier and has changed HP routers send an Unsolicited Reply upon booting but do not listen for or receive Unsolicited Replies sent by other nodes Probe Name Req...

Page 179: ... router will generate a VNA when it needs to forward IP packets to an NS node and there is no entry for it in its ARP cache The router will have the destination IP address and requires the station address in order to forward the packet The target node responds to the request with a path report that contains its station address The router then forwards the IP packet onto the destination NS node The...

Page 180: ...outer will bridge these multicast requests to their target nodes and back to the requester If bridging is not enabled then other provisions must be made to resolve the name to IP mapping by using proxy servers or statically building the Nodal Registry in each node Connection Scenarios The following are two different TCP connection scenarios between node A and node B over a router network The first...

Page 181: ...ction to node B which is on the other side of the network Probe is enabled on the routers Figure 2 Connection Establishment Steps for Scenario 1 The steps required to establish the connection are shown in figure 2 above and are described below 1 5 7 19 20 21 13 11 8 3 10 12 16 14 18 15 6 4 2 9 17 Routing Services Notes 2 A Primer on HP Probe Connection Scenarios 2 129 ...

Page 182: ... TCP sync packet to the destination network 9 The remote router receives the TCP synchronous packet for node B If node B is not in the ARP cache it sends a VNA request and ARP if it is enabled for node B s station address 10 Node B replies to the router with a VNA with its station address 11 The remote router sends the TCP sync packet to node B s station address 12 When node B wants to reply to th...

Page 183: ... A proxy server is used in this scenario Each router is configured for IP routing only Node A wants to set up a TCP connection to node B which is on the other side of the network Probe is enabled on the routers There is a Probe proxy server on each subnet The steps required to establish the connection are as follows 1 Node A wants to connect to node B checks the Nodal Registry and does not have an...

Page 184: ...e mapping is performed using Probe name request reply packets If bridging is not enabled on the network then proxy servers need to be on each subnet and the mapping is performed by Probe proxy request reply packets As an alternative the Nodal Registry on each node communicating over the network could be configured with the entries required for nodes on remote subnets A Primer on HP Probe Summary 2...

Page 185: ...n the original data uses At the other end of the link the data is then uncompressed to its original size In this way one can gain improved WAN throughput without having to upgrade to higher speed lines For example using the HP AdvanceStack Router 650 running compression on 64 Kbit s links can yield throughputs of up to 96 Kbit s See the Design Guidelines section for more details Data compression i...

Page 186: ... may arrive at the receiving device out of order or corrupted In other words running dictionaries require that packets be reliably delivered across the link with no data loss or corruption Additionally if the WAN link uses multiple virtual circuits then a separate dictionary is used for each virtual circuit thus requiring more memory Packet by Packet Dictionary To overcome the memory and link qual...

Page 187: ... packet by packet compression the HP PPC algorithm employs an additional encoding scheme called run length encoding to further compress data Run length encoding is a method of replacing repeated occurrences of a certain character with a single occurrence followed by the number of times it occurs i e the run length Currently HP PPC is supported only on HP point to point links Later releases of rout...

Page 188: ...aper3 Witten In Search of Autonomy 46526 paper4 Cleary Programming by Example Revised 13286 paper5 Cleary A Logical Implementation of Arithmetic 11954 paper6 Cleary Compact Hash Tables Using Bidirectional Linear Probing 38105 pic Picture number 5 from the CCITT facsimile test files test srawings 513216 progc C source code compress version 4 0 39611 progl LISP source code system software 71646 prog...

Page 189: ...ng packet by packet compression techniques over non reliable wide area links 1 5 1 is a reasonable number Design Guidelines As a result of testing the following guidelines should be observed when im plementing compression on HP Point to Point links 1 HP Router 650 card cage model holding up to 4 interface cards or 16 ports Compression should not be run on WAN links with speeds greater than 256 Kbi...

Page 190: ...ll link speeds for a given card is less than 306 Kbit s per 4 port synchronous card then all links will be able to handle maximum load with no problems This 306 Kbit s value is very conservative as it takes into consideration the maximum amount of traffic going in both directions on the link simultaneously If the total of all link speeds for the card is between 306 Kbit s and 460 Kbit s then the c...

Page 191: ...higher speed links the throughput will likely not increase beyond that of a 64 Kbit s link due to the required CPU bandwidth to compress and decompress the data The 64 Kbit s limit is imposed on the these routers instead of the 256 Kbit s limit of the Router 650 because the architecture of fixed port routers is such that there is one processor for both routing and compression thus CPU demands are ...

Page 192: ...ze could result in many more retransmissions because if a packet is lost all packets in the window must be retransmitted not just the one packet that was lost The reason for the burst mode requirement in Novell networks is that the standard NCP NetWare Core Protocol uses a method that requires the acknowledgment of every packet that is transmitted Additionally standard Novell file servers acting a...

Page 193: ...at maximize available bandwidth Performance testing has shown that there are no performance issues with the HP AdvanceStack Router 650 running compression on all WAN links at 64 Kbit s or combinations of compression and non compression links up to 256 Kbit s With the fixed port routers there are certain limitations of the software based HP PPC algorithm that impose some design constraints with res...

Page 194: ...Data Compression for WAN Links Conclusion 2 142 ...

Page 195: ...3 Application Notes and Case Studies ...

Page 196: ...n Dry Creek Joint Elementary School District Shining a Light on FDDI Using Synchronous Pass Through to Consolidate Synchronous Traffic Routing with OSPF Linking Up withFrame Relay Frame Relay Network Design Fleet Call Inc Application Notes and Case Studies 3 2 ...

Page 197: ...ays to the strength of routers quickly detecting network failures and routing data around them This application note examines the different ways of providing alternate data paths to survive both data link and router failures Permanent Alternate Paths Consider the network in figure 1 There is only one path for data from any site to any other site Should link 1 fail for instance systems at site A wi...

Page 198: ...router operation the forwarding table indicates only the best lowest cost path to a destination network When link 1 fails router A and router B will exchange routing information with router C Router A will learn about its new path to the site B LAN through router C and router B will learn about its new path to the site A LAN again through router C Note that it takes a short period of time for the ...

Page 199: ...is a logical structure that contains one or more circuits Coterminous circuits are simply circuits that have common starting points and common end points Consider the network in figure 2 Two coterminous circuits are used to connect routers A and B Both circuits have been configured as members of a single circuit group on each router The forwarding process within each redirector directs traffic to ...

Page 200: ...e bridging redirector directs traffic to the single circuit group rather than the two circuits There are limitations on the use of coterminus circuits To understand those limitations you need to know how the circuit group manager works Circuit Group Manager After a routing or bridging forwarder has received a packet and has determined that the packet must be forwarded it hands the packet to the ci...

Page 201: ... that a computer at site A with IP address 15 8 128 210 sent a packet to a computer at site B with IP address 15 8 64 173 To determine which of the two circuits to use for sending the packet the circuit group manager adds the two IP addresses The result in hexadecimal is 1E10C17F This value is then MODed by 2 the number of available circuits in the circuit group The result is 1 the remainder when ...

Page 202: ...circuit Load Balancing Overrides The circuit assignment algorithms described above can be overridden for IP and Novell IPX traffic Answering Yes for the IP network interface defini tion s Load Balancing parameter changes load balancing for all IP packets on that particular circuit group to random circuit assignment algorithm 1 Random usually works quite well and is generally better than the indexe...

Page 203: ...measure of balanced circuit utilization As a final note it is generally a good practice when designing a network with coterminous circuits to purchase the circuits from different carriers to insulate your network from multiple simultaneous circuit failures Packet Switching Networks Packet switching networks PSNs based on technologies that include X 25 frame relay and Switched Multimegabit Data Ser...

Page 204: ...e failure if it does occur See the two routers in the upper right in figure 3 Finally when choosing a PSN to interconnect routers ask your network vendor for a description of the physical network to verify that the network has the redundancy you expect Circuit Switching Networks Circuit switching networks are those in which a physical circuit is established after a number is dialed Circuit switchi...

Page 205: ...etwork is usually referred to as a virtual circuit Users are typically charged for service on the basis of the number of packets transmitted Packet switched communications are thus multi point and may or may not be limited in duration Circuit Group Considerations The way in which a switched circuit operates is governed primarily by its type of circuit group membership Switched circuits can be conf...

Page 206: ...or packet switching networks When compared to leased line CSUs DSUs with dial backup capability the HP routers switched backup function provides more flexibility in selecting the type of switched circuit Figure 5 Backup and Primary Circuits in a Circuit Group a b d c e CIRCUIT SWITCHING NETWORK a Circuit Group DSU CSU Terminal Adapter Primary Circuit Leased Line to Remote DSU CSU To Switched Netwo...

Page 207: ...ecommended for use with all of the routable protocols Since a backup circuit is normally inactive a switched dial up circuit is the logical choice to implement as a backup circuit since it is the most cost effective alternative When a switch from primary to backup circuits must be performed quickly ISDN and Switched 56 are preferred to analog modems ISDN and Switched 56 connections can typically b...

Page 208: ...y sites are networked together using switched services Figure 4 above shows a network utilizing pool circuits for this purpose The second situation is the use of pool circuits to provide a backup path for failures of leased lines or packet switching networks Figures 7 and 8 illustrate these cases a Circuit Group Terminal Adapter Terminal Adapter Termnial Adapter To Switched Network To Switched Net...

Page 209: ...Figure 8 Packet Switching Network with Circuit Switching Backup b c d e a2 a1 Circuit Switching Network b a c d Packet Switching Network Circuit Switching Network Application Notes Case Studies 3 Improving Network Availability Circuit Group Considerations 3 15 ...

Page 210: ... to effectively eliminate routing protocol control traffic on switched circuits This ensures that switched circuits are opened only for the transfer of user data Switched Circuit Types The routers Configuration Editor allows the definition of two types of switched circuits Manual adapter V 25 bis adapter These two circuit types can provide access to virtually any circuit switching facility includi...

Page 211: ... shows the four possible combinations Of the four combinations only the two shaded combinations are recommended Thus one side opens the circuit when data is available and the other side always waits to receive data Table 3 Parameter Combinations for Initiating Manual Adapter Connections Side A Router Side B Router Circuit is enabled Circuit is enabled Circuit is enabled Data is available Data is a...

Page 212: ...calls and router B always receives calls In other words applications on systems attached to router A always initiate the connections to systems attached to router B This is frequently the situation for client systems at a sales support location connecting to servers at a central or regional office for database access credit verification electronic mail etc When circuit switching network connection...

Page 213: ...n this example for the first three minute period and Connect Inactivity Time could be used for the incremental one minute periods These parameters when set to the correct billing periods cause the router to monitor data transmission and reception They serve two important functions First the user will incur the minimum connect time charge for data transmissions The router will automatically disconn...

Page 214: ...al capabilities provided include Bilateral communication initiation and establishment Each side of the switched connection can dial the other party and establish communication V 25 bis supports connecting when data is available or on incoming calls rather than when data is available as is the case using manual adapters Enhanced security A router may be set up to accept calls only from a predefined...

Page 215: ...r tone is detected to disable further V 25 bis dialog The call is then placed On the receiving side the adapter that previously raised CTS sends the router an INC to indicate that an incoming call has arrived The originating phone number and the first part of the data packet is forwarded to the router this assumes that ANI automatic number identification is available Thus the ring indicator contro...

Page 216: ...all Circuit is enabled Data is available or on incoming call Data is available or on incoming call The valid combinations of connection initiation parameters for circuit group members and circuit group backup members using V 25 bis adapters are shown in the shaded fields of table 4 Pool circuit group members can only be configured to connect when data is available or on incoming call As in the cas...

Page 217: ...Network Services The switched circuits at the remote sites are configured as primary circuits The switched circuits at the central site are pool circuits Note that there are more remote circuits than there are central site circuits The balance between remote site dial access circuits and central site access circuits is determined by the frequency of access and by users tolerance for waiting for an...

Page 218: ...ffic that will hold switched circuits open permanently N o t e Pool circuits may be used only with IP protocol NxN Backup NxN backup refers to a configuration in which every leased line circuit has its own dedicated backup circuit for example in a central site Dedicated backup circuits are provided using the backup circuit type Either manual adapters or V 25 bis adapters can be used to provide the...

Page 219: ...igured as backup members of a circuit group In this example two switched circuits at the central site are used to back up four leased line circuits 1 for 2 backup Ra1 and Ra2 are attached to the central site LAN and connected to routers at four remote sites The backup circuits become active only in the event of a primary link failure Additionally the backup circuits are only used to connect from a...

Page 220: ...be configured to map IP addresses to phone numbers Switched circuits must be configured using static routes Each leased line is favored over its switched counterpart by ensuring that leased line circuit costs are set to result in the lowest cost path All switched circuits are members of the same IP network or subnet work Each leased line is a separate network or subnetwork 1xN Backup Meshed Networ...

Page 221: ...ntral site from the perspective of the dial backup scheme Ra and Rc have static routes to each of the other Ethernets Rd is the next hop for all of the static routes to the remote Ethernet networks The point to point circuit between Rb and Rd has been installed to provide a backup path in the envent of a failure in Rd s packet switching access circuit A new feature in the next HP router software r...

Page 222: ...networks The inverse multiplexer parameters that can be controlled by HP routers include Per channel bandwidth 56 Kbit s 64 Kbit s etc Minimum channels to aggregate the minimum number of ISDN channels to open when data is available Maximum channels to aggregate the maximum number of channels to open regardless of the amount of data to transmit Figure 15 Inverse Multiplexers Used for Bandwidth on D...

Page 223: ... be able to detect failures and establish alternate routes quickly The routing protocols as well as the Spanning Tree Protocol meet the fast switching criteria Second the network architecture on which the application runs must have recovery mechanisms usually in the transport layer to tolerate brief communication failures Applications running at the time of the failure may or may not recover trans...

Page 224: ...Improving Network Availability Application Recovery 3 30 ...

Page 225: ... been a small single school district It was established in 1876 and maintained a stable population of between 100 and 200 students for most of its first 100 years of existence Around 1988 significant residential development occurred within the district and the student population began to increase by 400 to 600 per year The current enrollment is almost 2 300 students With this rapid increase in stu...

Page 226: ... network must support office automation applications such as electronic mail word processing graphics and electronic spread sheets as well as host and server based business and student administration applications The wide area network also must handle the additional load of instructional traffic such as electronic mail and Internet access when those services are brought online Network Topology The...

Page 227: ...Cost Installation Cost 56 Kbit s point to point 4 locations 324 30 3 720 00 ISDN 4 locations 110 50 849 75 The traditional point to point extended LAN topology that was considered is shown in figure 1 Figure 1 Point to Point Network Design Option R R R R R Hub Hub Hub Hub Bridge Application Notes Case Studies 3 ISDN Wide Area Network Design Dry Creek Joint Elem School District Network Topology 3 3...

Page 228: ...of a particular local exchange carrier or between carriers In many cases a digital link may be available between two end points but the call may start out as ISDN with out of band signaling be converted to Switched 56 with in band signaling and be converted back to ISDN In this example there would be a loss of through put because of the Switched 56 in band signaling Signaling System 7 is a standar...

Page 229: ...o usage charges ISDN meets all of the district s design criteria The network as implemented is shown in figure 2 Figure 2 ISDN Network Design Option R R R R R Hub Hub Hub Hub Pacific Bell Supplied WAN Link Bridge Application Notes Case Studies 3 ISDN Wide Area Network Design Dry Creek Joint Elem School District ISDN 3 35 ...

Page 230: ...nt than do bridges With HP router operating system release 5 74 and later ISDN can be configured for manual or automatic operation The automatic operation provides for very sophisticated call control using the CCITT V 25 bis call connect protocol The Dry Creek implementation uses point to point connections that are semi permanently established requiring only simple manual operation to be configure...

Page 231: ...acing 2 Connect Retry Count 3 Connect Wait Time sec 15 Delay After Connect Failure min 5 District Office Terminal Adapter TA AT T 7500B Data Module operating in synchronous DCE mode acting as the originator Configuration Set Autodial YES Set busy out OFF Data Rest OFF Set DTR FOLLOW Set Duplex FULL Set mode CS Set speed 64000 The ISDN telephone number for the remote site is configured in the termi...

Page 232: ...en Circuit is enabled Connect retry count 3 Connect wait time sec 0 1 Delay after connect failure min 5 1 It was discovered during testing that this parameter should be set to zero to ensure that the answer end point is ready to reconnect immediately in the event of a power loss or component failure Remote Office Terminal Adapter TA AT T 7500B Data Module operating in synchronous DCE mode acting a...

Page 233: ...apters that provide inverse multiplexing are becoming available The inverse multiplexing feature will allow both B channels to be combined for 128 Kbit s of bandwidth This provides a growth path that will be more than adequate for Dry Creek s needs for at least three years Application Notes Case Studies 3 ISDN Wide Area Network Design Dry Creek Joint Elem School District Performance 3 39 ...

Page 234: ...ISDN Wide Area Network Design Dry Creek Joint Elem School District Performance 3 40 ...

Page 235: ... increasing bandwidth security and fault tolerance These are examples of some of the network requirements that are met by FDDI At 100 Mbit s FDDI allows high speed interconnection of all the LANs in an organization s network FDDI easily overcomes the performance limitations of 10 Mbit s Ethernet and 16 Mbit s token ring networks In addition FDDI provides the security of fiber optic cabling as well...

Page 236: ... standard dual attached ring Network access is negotiated between stations at initialization and at every time a new node is added to the network Figure 1 FDDI Ring The network backbone is constructed of stations interconnected by two counter rotating rings These rings are two pairs of fiber optic cable to which each device is attached Cable lengths between stations can be anywhere from 2 kilomete...

Page 237: ...equivalent to the physical layer of the OSI model see figure 2 The PMD and SMF PMD correspond to the lower portion of the physical layer The PMD defines the media requirements for multimode fiber such as fiber optic cable connectors and driver receiver operation for FDDI stations Cables connectors and receivers are discussed below in the Cabling section The SMF PMD defines similar requirements for...

Page 238: ...the ring and also to the extended LAN FDDI also has a process that defines protocols for managing the PMD and SMF PMD PHY and MAC called Station Management SMT SMT defines facilities for connection management station configuration error recovery and the encoding of SMT frames MAC and SMT frames carry data and control information for the operation and management of the FDDI network The MAC PHY and ...

Page 239: ...s DAS Single Attachment Stations SAS and concentrators See figure 3 Dual Attachment Stations are devices that connect to both the primary ring and the secondary ring simultaneously Single Attachment Stations are devices that connect to the ring through the primary ring only and are usually attached to the ring through concentrators Dual Attachment Stations have two ports one labeled port A the oth...

Page 240: ... link state characters upon completion of the tests Stations initiate CMTs between each upstream and downstream neighbor Upon successful completion of the initialization phase the ring begins the claim process to generate a token The claim process begins with all stations negotiating a target token rotational timer known as TTRT Each station makes a bid for the timer and the one with the lowest va...

Page 241: ...the three techniques that FDDI uses in order to ensure fault tolerance If any station on the dual ring fails or if the cable is damaged the dual ring automatically wraps into a single ring as shown in figure 4 Figure 4 Ring Wrap Multiple failures within a single dual ring network present other challenges The likelihood of multiple failures greatly increases as the number of nodes within an FDDI ne...

Page 242: ...ommended when designing a multi station FDDI backbone In a fault condition a DAS will cause the ring to wrap over the secondary ring isolating the bad section of the ring If this happens to more than any single section of the FDDI ring the network is sectioned into two independent rings Stations that have an optical bypass device are physically switched out when a fault condition occurs They do no...

Page 243: ...que used for redundancy in concentrator environments where continuous uptime is crucial Dual homing allows a DAS to use one link as a backup link for redundancy purposes Thus one of the two attachments is active at any given time Figure 6 shows a dual homed device attached to concentrators Figure 6 Dual Homed Station A station can be dual homed to the same concentrator or to two different concentr...

Page 244: ...ical signals into light and transmit the light into the fiber optic cable The HP routers do not support single mode fiber They do support multimode fiber only at 1300 nanoseconds Attenuation and Optical Budget Signal attenuation through optical fiber is important with FDDI It describes the amount of energy optical power that is lost as the light signal travels from the transmitter through the cabl...

Page 245: ...alculating the dB loss of a particular link it is important to compare the resulting power with receiver sensitivity If the receiver is not sensitive enough poor link quality will result Optical Transmitters and Receivers Transmitters convert data from electrical signals to light The receiver converts the light signals back to electrical signals Receivers contain photo detectors that convert incom...

Page 246: ...ator HP s FDDI link interfaces only support multimode 50 micron or 62 5 micron graded index fiber optic cable FDDI Routing Since routers are media independent they encapsulate packets to conform to the media type in use RFC 1188 specifies 802 2 LLC with SNAP subnet work access protocol when routing TCP IP traffic over FDDI media RFC 1188 conforming routers provide IP routing between source and des...

Page 247: ...EEE 802 3 frame into a new message type in this case an FDDI specific packet for travel across the FDDI ring At the destination the message is removed from the FDDI packet and transmitted in its original form Figure 7 Encapsulation Bridging It is important to note that encapsulation is not the preferred method of transporting data across an FDDI network In addition no standard method of encapsulat...

Page 248: ...at IEEE 802 3 802 2 LLC and 802 3 LLC SNAP remain unchanged when trans ferred across an intervening LAN Table 1 summarizes translation rules Table 2 summarizes protocol specific translation Table 1 Translation Rules for Encapsulation Source Network Encapsulation Translation Rules Destination Network Encapsulation Ethernet RFC 1042 SNAP OUI 00 00 00 Ethernet 802 3 LLC SNAP FDDI LLC SNAP 802 3 LLC S...

Page 249: ...n 5 addition of the FDDI MAC level trailer Standalone FDDI Rings Network designers can take advantage of FDDI s high speed capabilities without an FDDI backbone in a network topology In the absence of a ring a network designer can use the HP 27290A Router BR as a standalone FDDI network interconnecting a maximum of two SAS devices or a single DAS LAN 1 FDDI Ring LAN 2 LAN 1 Frame Translated Frame ...

Page 250: ...r BR as shown in figure 9 two benefits would be The traffic between segments would decrease draftically effectively increasing the bandwidth of the Ethernet internetwork Each server would have 100 Mbit s bandwidth with which to serve all incoming segments In general a device can be connected to the HP Router BR either as a DAS creating a dual attached ring or as an SAS creating a single attached r...

Page 251: ...oftware versions 5 70 and later that allows synchronous traffic HDLC SDLC LAPB to be consolidated with LAN to LAN traffic on HP router based internetworks This application note describes how sync pass through works and examines several sync pass through applications Figure 1 shows an extended LAN internetwork where sync pass through is used to connect synchronous interactive and printing terminals...

Page 252: ...us Traffic Conveyed Through a High Speed Extended LAN Network IBM PC SITE A SITE B IBM 3745 IBM 3090 SITE C IBM 3174 PC Router Router Router HP 3000 Using Synchronous Pass Through to Consolidate Synchronous Traffic 3 58 ...

Page 253: ...addresses are assigned to sync pass through ports by the user In figure 1 station addresses are assigned to each of the synchro nous circuits marked with a circled number When configuring the site C sync pass through interface for example the station address assigned to the local sync pass through circuit marked is entered as the local LAN address The station address assigned to the sync pass thro...

Page 254: ...raffic Using Sync Pass Through HOST A HOST B SITE B SITE A X 25 PACKET SWITCH X 25 PACKET SWITCH ROUTER ROUTER ROUTER SITE C PAD TERM 1 TERM 2 Using Synchronous Pass Through to Consolidate Synchronous Traffic How It Works 3 60 ...

Page 255: ...s Through Encapsulation The router constructs LAN packets by encapsulating the synchronous frames received on a sync pass through port as shown in figure 3 The local and remote station addresses assigned by the user are used as source and destination addresses A special type field recognized only by HP and Wellfleet routers is added to mark the frame as a sync pass through packet The original CRC ...

Page 256: ...ting protocol is removed a new CRC is computed and the synchronous frame is transmitted to the destina tion synchronous device The frame transmitted to the destination device is identical to the frame received by the first router SYNCHRONOUS FRAME SYNCHRONOUS PASS THROUGH PACKET Using Synchronous Pass Through to Consolidate Synchronous Traffic Sync Pass Through Encapsulation 3 62 ...

Page 257: ... G address The I G bit is the first address bit transmitted although it is not the most significant bit as shown in figure 4 The address selected for sync pass through must be an individual address I G bit 0 The next bit U L bit specifies whether the address is globally administered U for universal or locally L administered This is the handy mechanism referred to above Computer manufacturers alway...

Page 258: ...1600 bytes Thus the maximum allowable synchronous frame is 1600 bytes minus the encapsulating protocol size The size of the encapsulating protocol varies by link type but never exceeds 22 bytes Generally a frame size greater than the maximum is not a problem since most synchronous protocols do not use frame sizes greater than 1024 bytes Additionally the synchronous traffic must use the CRC CCITT f...

Page 259: ... by Pin outs for some null modem WAN cables are shown in figures 6 and 7 for the HP routers Figure 6 is for cables that can be attached directly to the router s 62 pin connector and have either a V 35 or RS 232 V 24 V 28 interface Making these cables requires a DB 62 connector on the router end Figure 7 is for a V 35 cable that can be attached to the V 35 WAN cable that can be ordered with the HP ...

Page 260: ...Figure 6 Pin Outs For V 35 and RS 232 V 24 V 28 Null Modem Cables for 62 pin Direct Connection Using Synchronous Pass Through to Consolidate Synchronous Traffic Physical Connections 3 66 ...

Page 261: ...ugh Extended LAN traffic is charac teristically bursty Bursts of LAN to LAN traffic can cause variable response time delays on interactive terminals According to Gurugé variable response time delays are one of the most annoying problems for interactive terminal users To compensate for this delay he recommends increasing backbone capacity by up to four times the sync pass through line rate Thus up ...

Page 262: ...ic can help reduce recurring monthly communication costs and costs associated with network support Synchro nous frames are transformed into LAN packets and bridged from source to destination Synchronous devices must employ the HDLC data link protocol or a derivative to transfer data using sync pass through Using Synchronous Pass Through to Consolidate Synchronous Traffic Conclusion 3 68 ...

Page 263: ...ances to neighbor adjacent routers When a router receives its neighbor s routing update it compares the update with its own routing table and changes its routing table if necessary Distance vector protocols are susceptible to two main problems First they can form routing loops and second they can be slow to converge Conver gence is the time required for the routing tables in all of the connected r...

Page 264: ...at refers to a collection of routers that all use the same IGP interior gateway protocol IGP is also an IP term It refers to a routing protocol run within an AS IGPs are a subset of routing protocols They are referred to as IGPs to distinguish them from EGPs exterior gateway protocols another type of routing protocol EGPs are used to route data between ASs The Exterior Gateway Protocol also known ...

Page 265: ...r The topological database is discussed in more detail below It also helps to minimize routing control traffic An AS may consist of one or more areas Small networks may consist of a single area large networks may contain many areas Each area comprises a group of networks or subnets Each area in the AS is attached to the OSPF Backbone An OSPF AS is logically a star areas extend in a ra dial fashion...

Page 266: ...in office 1 of region 1 of area 1 receives the router ID 1 1 1 1 In any event it is desirable to select identifiers that have some significance Area Sizing A frequently asked question is How large should an OSPF area be Or What is the optimal number of routers in an OSPF area An area could contain as few as one router or as many as hundreds of routers Carving a network up into smaller sized areas ...

Page 267: ...ed into area A Virtual Links In figure 1 area C is not directly attached to the Back bone Instead it is attached to area B through a backbone virtual link Virtual links allow areas to be configured where it would otherwise be inconvenient to do so due to the distance or cost to attach to the Back bone Metrics The OSPF routing algorithm calculates the shortest path to each destination network A pat...

Page 268: ...120 19 2 Kbit s 150 9 6 Kbit s 200 Figure 2 Multi Area OSPF Internet Area 1 0 0 0 Area 2 0 0 0 NET 2 Router ID 1 1 1 1 NET 8 Router ID 1 2 1 1 Area 0 0 0 0 NET 1 1 544 Mbit s NET 12 NET 13 NET 7 NET 6 NET 5 1 544 Mbit s 1 544 Mbit s NET 11 NET 3 56 Kbit s NET 4 56 Kbit s NET 9 56 Kbit s NET 10 1 544 Mbit s Router ID 2 2 1 1 Router ID 2 0 0 1 Router ID 1 0 0 1 Router ID 2 1 1 1 Routing with OSPF Ro...

Page 269: ...routers in its area Flooding is a way to send a message that will be relayed by all routers receiving the message Received LSAs are used to build and maintain a topological database from which each router builds its routing table There are several types of LSAs Consider the network in figure 2 Each router in area 1 0 0 0 sends a router links LSA to all other routers in area 1 0 0 0 at 30 minute in...

Page 270: ...build and maintain their topological databases Topological Database Routers in each area have identical routing databases This topological database describes which routers are connected to which networks Attributes of the networks and routers such as subnet masks metrics etc are also maintained in the database Each router constructs its routing table from the database The routing table contains th...

Page 271: ...out the internet must be identical on all subnets The information about individual networks in cluded in a RIP update includes the network or subnet number and metric hop count RIP updates do not include the subnet mask associated with a network This often results in the over allocation of IP address space especially on synchronous point to point links NET 1 Metric 40 NET 8 Metric 130 NET 9 Metric...

Page 272: ...iable Length Subnet Masks The mechanics of conserving IP address space using variable length subnet masks are straightforward Again consider the internet in figure 4 With OSPF the subnet mask can be specified per network or subnetwork The subnet masks for the LAN subnets can be specified as before 255 255 255 0 This allows 254 addressable nodes per subnet The point to point links however require o...

Page 273: ...14 The following method of selecting subnet numbers and point to point link addresses illustrates one method of allocating addresses that seemingly maintains the address structure used in allocating LAN subnets The first point to point subnet defined in figure 4 was subnet 170 200 5 x the second was 170 200 6 x and so on To maintain consistency with that addressing scheme the point to point subnet...

Page 274: ...cast and those that are all zeros any subnet Variable length subnet masks introduce extra complexity to this issue since there may now be several different subnet definitions in a single network For example consider the internet in figure 5 IP subnets in the range 170 200 0 4 through 170 200 0 248 appear to be valid IP subnets when used with the WAN subnet mask 255 255 255 252 However these addres...

Page 275: ...e network by a second network In figure 6 the class B network 128 1 x x is partitioned by network 192 1 1 x Figure 6 A Partitioned Network To understand why network partitioning might be useful refer again to figure 6 Suppose network 128 1 x x is a large network with very few remaining subnets Subnets of the class C network 192 1 1 x can be used for synchronous links so that none of the remaining ...

Page 276: ...tween areas Thus the topology of an area is hidden from routers outside the area Type of Service Routing Type of service routing is not yet available on HP routers Currently HP routers implement TOS 0 routing only The following discussion is intended only to explain the type of service routing concept Type of service routing is based on the type of service defined in the Internet Protocol Specific...

Page 277: ...the TOS requested in the header of a datagram Thus a separate set of routes can be calculated for each IP type of service Conceptually this will allow networks where file transfers using FTP can be routed over high throughput routes such as satellite circuits and time sensitive data using Telnet can be routed over low delay terrestrial circuits see figure 7 High Throughput Shortest Path FTP Low De...

Page 278: ...nk metrics to ensure that multiple paths are used to forward data when that is desired Conclusion OSPF provides support for large networks using hierarchical network organization improved metrics and a link state protocol Additional features that help conserve IP address space make addressing more flexible and improve network security make OSPF attractive for use on many small and medium sized int...

Page 279: ...rame Relay network is shown in figure 1 Frame Relay networks are composed of fast switching nodes connected by high speed digital trunks User devices routers in this case are connected to the switching nodes through digital circuits such as T1s The Frame Relay standards define the interface between the user device and the Frame Relay network How data is routed within the network is determined by t...

Page 280: ...d the packets must retransmit the packets that were discarded which typically occurs at the transport layer in a system s networking software Frame Relay allows for the transmission of large LAN sized frames 1600 bytes at minimum This means that network software on both hosts and routers does not have to segment and reassemble packets Frames may be up to 4500 bytes on some Frame Relay networks Fra...

Page 281: ...ique throughout the network the network would be limited to 1024 addresses too few addresses for large public networks However DLCIs are not globally unique instead they are locally significant which means they are unique only on the boundary Frame Relay switch Thus there are duplicate or reused DLCIs in a Frame Relay network Also routers on either end of a PVC typically have different DLCIs to re...

Page 282: ...pted and implemented worldwide by the early 1980s X 25 allowed a user to communicate with any other X 25 user worldwide simply by knowing the other party s X 121 address X 121 is a global data communication addressing standard In addition X 25 pricing was usage based X 25 was ideal for companies with many locations and with low to moderate communication requirements X 25 could be implemented quick...

Page 283: ... T1 617 ISDN Signaling Specification for Frame Relay Bearer Service and T1 618 ISDN Core Aspects of Frame Protocol for use with Frame Relay Bearer Service In 1992 CCITT released Recom mendation Q 922 ISDN Data Link Layer Specification for Frame Mode Bearer Services and Draft Recommendation Q 933 Digital Subscriber Signaling System No 1 DSS 1 Signaling Specification for Frame Mode Basic Call Contro...

Page 284: ...t 1 Flag 7E hex Octet 2 Q 922 Address Octet 3 Q 922 Address Octet 4 Control UI 03 hex Octet 5 Optional Pads 00 hex Octet 6 NLPID Octet 7 Octet N 3 Data Octet N 2 Frame Check Sequence Octet N 1 Frame Check Sequence Octet N Flag 7E hex A standardized procedure for transmitting multiprotocol data over Frame Relay was defined in Internet RFC 1294 Multiprotocol Interconnect over Frame Relay The shaded ...

Page 285: ...Connection Identifier The six bit upper DLCI and four bit lower DLCI are concatenated to form a ten bit DLCI This provides for up to 1024 DLCIs per Frame Relay switching node again consider ing only the two octet address format Control Field and Command Response Bit Frame Relay frames are sent with the control field octet 4 in table 1 set to 03 hex This encoding indicates that the frame is a UI un...

Page 286: ...e B may discard the frame from R5 if the discard eligibility DE bit in the frame is set The discard eligibility bit would typically have been set by node A in frames that exceeded the CIR of the PVC from R5 to R6 HP routers never set the DE bit in frames transmitted to the Frame Relay network To notify R5 that a congestion condition exists on the route to R6 node B sets the backward explicit conge...

Page 287: ...e affected PVCs Optional Pads Pad characters may optionally be included in the Frame Relay header Pads allow the frame to be extended to an even or odd number of bytes or words in length to suit the needs of the frame s transmitter NLPID The Network Layer Protocol Identifier NLPID identifies the type of network on which the packet was generated for example TCP IP Novell IPX AppleTalk etc NLPID val...

Page 288: ...rame Relay frame contains Table 4 shows the encoding of a Novell IPX frame The encoding of the three octet SNAP header is 00 00 00 hex This indicates that an Ethernet type field follows 8137 hex is the Ethernet type field for Novell IPX The Novell header and data follow in the remainder of the Frame Relay frame Table 4 Routing Novell IPX NLPID Continuation 00 hex SNAP Header Indicating EtherType F...

Page 289: ...dging 00 hex IEEE 802 1 Organization Code 80 hex C2 hex xx hex Protocol ID PID xx hex Complete Bridged Frame The PID Protocol ID field identifies the media type and whether or not the LAN FCS is appended to the bridged frame The PID values are shown in table 6 Table 6 PID Values for Bridged Frames With Preserved FCS Without Preserved FCS Medium 00 01 00 07 802 3 Ethernet 00 02 00 08 802 4 Token Bu...

Page 290: ...e mechanisms for providing a user device with network configuration and status information Both ANSI T1 617 Annex D and CCITT Q 933 Annex A have the same capability for providing configura tion and status information for PVCs as the LMI has These three standards are collectively referred to here as LMI unless otherwise indicated It should be noted that HP routers do not yet support CCITT Q 933 Ann...

Page 291: ... not The default polling interval is ten seconds A number of keep alive polls are sent in between full status polls Link Integrity Verification Keep Alive Both the router and the network verify the integrity of the link with every Status Enquiry Status exchange Send and receive sequence numbers are included in each LMI message exchanged with the network The network and attached rout ers verify tha...

Page 292: ...ation returned by the network indicates which PVCs are available for use The full status reply lists each of the DLCIs available and provides two bits of status information per DLCI The first status bit is called the New bit the second status bit is called the Active bit Table 7 shows how these bits are used Frame Relay Network Poll Interval Interval Between Full Polls R Keep Alive Poll Full Statu...

Page 293: ...n empty list of DLCIs no DLCIs and accompanying status information This is because none of the other routers are yet active Figure 5 PVCs Connecting Routers Now let s suppose that R2 powers on and its Frame Relay interface is initialized and enabled Like R1 the first message it sends to the network is a full status poll As before the network responds with a full status reply This time however the ...

Page 294: ...o all members of the multicast group This is a useful feature for the transmission of routing updates and other types of broadcast messages The FRF s LMI reserved four DLCIs for this use The multicast feature is not available in ANSI T1 617 Annex D Without multicast support by contrast whenever a broadcast message such as a routing update must be sent it is transmitted on each DLCI In figure 2 for...

Page 295: ...outer s Configuration Editor Protocol Analysis Testing The HP PT502 and HP PT302 are high performance wide area network protocol testers which provide powerful monitoring analysis emulation and conformance testing capabilities for a variety of protocols including Frame Relay at speeds up to 2 Mbit s The PT502 is a multiport protocol tester designed for lab use and high level network maintenance Th...

Page 296: ...s information can be used to help eliminate design or implementation flaws before the development of the equipment is completed thus reducing the time to market as well as costs The product numbers for the PT502 and PT302 protocol testers and associated Frame Relay software packages are PT502 HP E3910B PT302 HP E3939B Frame Relay Analysis software package HP E3946A Frame Relay Emulation software p...

Page 297: ...I ANSI T1 617 Annex D Null Management Interface PVCs must be statically configured LMI Switch and ANSI T1 617 Annex D Switch These interfaces are included for test purposes CCITT Rec 933 Annex A not yet supported Multicast Support Frame Relay Forum LMI Multicast Extension Flow Control LMI XON XOFF DE Bit Setting The DE bit is never set Response to FECN BECN Received FECNs and BECNs are counted No ...

Page 298: ...6 Interval Between Full Polls 7 Monitored Events 8 Error Threshold Circuit Table 1 per physical circuit 1 DLCI 1 first table entry DLCI actual DLCI used DLCI creation time Circuit state Time since last circuit state change Frames received Frames sent Octets received Octets bytes sent BECNs received FECNs received 2 DLCI n next table entry etc Error Table 1 per physical circuit 1 Error data string ...

Page 299: ...aging and data services to customers In February of 1991 Fleet Call received authorization from the FCC to build Digital Mobile networks This will allow Fleet Call to begin the process of building a single nationwide all digital network using compatible technologies and providing compatible services across all geographic areas The move to digital technology will provide enhanced service and superi...

Page 300: ...gation Users will access the host applications from PC and Macintosh workstations distributed across the network The work stations will be equipped with network software and hardware to allow Telnet access to the host systems over TCP IP for terminal emulation Host printing will also be spooled over the network to TCP IP equipped HP LaserJet printers It is Fleet Call s goal to provide a completely...

Page 301: ...allows connection of the local PC network to the backbone through wide area links Routers were chosen rather than bridges because routing technology provides a greater level of network traffic filtering The Fleet Call network will grow to incorporate many hundreds of nodes and optimizing the bandwidth utilization of the wide area links is essential to preserve high performance and minimize cost TA...

Page 302: ...he cost of the wide area links can be optimized The tail site links will fall principally within a LATA Local Access and Transport Area the service area boundary allowing a local operating company such as Pacific Bell to provide the link service The link costs are usually based on distance between connection points and the short intra LATA links from the backbone to the tail sites will be much mor...

Page 303: ...hieve any kind of link redundancy A sample mesh network is shown in figure 2 Figure 2 Point to Point Meshed Network Circuit costs are based on distance and Fleet Call s wide geographic dispersion would mean sizable monthly circuit costs to effectively run a meshed network between the backbone sites In addition Fleet Call would be solely responsible for the troubleshooting and problem isolation in ...

Page 304: ... single point of entry to the frame relay network and all network routing happens transparently This is depicted in figure 3 Figure 3 Frame Relay Network Frame relay is also very suitable for bursty data networks where irregular traffic patterns cause variable bandwidth demands A frame relay backbone network would provide Fleet Call with built in redundancy and high perform ance Compared to a mesh...

Page 305: ...d with 10Base T networks running Novell NetWare to provide local file and disk sharing Except for electronic mail transfer the NetWare traffic is localized within each office preserving the bandwidth of the WAN for the TCP IP host directed traffic Fleet Call will use the HP OpenView Network Management system It allows Fleet Call to monitor and manage all the network devices that have IP ad dresses...

Page 306: ... planning a frame relay network make sure the network equipment has been completely certified by the circuit provider This will aid in troubleshooting and support 2 When purchasing newly released equipment plan the rollout of the equipment after suitable testing has been performed 3 Never underestimate the time and effort required to bring up a new WAN link 4 In remote site installations perform s...

Page 307: ...1 6 2 110 node address 2 110 2 113 2 114 routing service 1 3 1 6 1 8 1 19 2 107 2 108 2 110 2 114 topology example 2 107 2 109 zone 2 111 AppleTalk Address Resolution Protocol 1 8 2 113 AppleTalk Echo Protocol 2 113 2 114 application filtering for IP 2 81 architecture 1 21 1 34 area number 2 117 2 118 areas DECnet 1 7 2 116 2 117 2 121 2 123 OSPF 1 6 1 32 2 66 ARP Address Resolution Protocol 1 6 1...

Page 308: ...6 1 28 2 85 2 86 See also topology utilities 1 17 1 19 Zone Name list 2 111 2 112 Configuration Editor 1 18 2 112 2 119 2 123 console 1 9 1 15 1 20 2 85 2 93 2 124 See also RS 232 console port controller chip 1 22 1 23 1 25 1 27 1 28 convergence 2 66 coprocessor See processor chip cost 1 32 1 33 2 65 2 67 2 69 2 71 2 74 2 102 2 119 2 122 count to infinity 2 66 CPU See processor chip CSMA CD 2 107 ...

Page 309: ...tocol 1 6 1 8 1 32 2 65 2 67 2 74 2 94 electronic mail 2 56 2 96 encapsulation for adjacent host routes 2 76 2 78 encapsulation for Novell 2 99 engine 1 24 1 26 1 28 EPROM 1 22 1 28 Error Protocol 1 7 Ethernet 1 3 1 7 1 10 1 23 1 24 1 27 2 3 2 77 2 93 2 107 2 108 EtherTalk 2 107 event log 1 16 1 17 2 93 2 124 export route filter 1 32 2 79 2 80 2 94 Exterior Gateway Protocol See EGP F factory defau...

Page 310: ...HEX 2 47 Internet 1 5 2 58 2 59 2 67 2 93 2 94 Internet Activities Board 2 94 Internet Control Message Protocol See ICMP Internet Datagram Protocol See IDP Internet Packet Exchange See IPX Internet Protocol See IP internetwork 1 3 1 5 2 56 2 58 2 63 2 93 2 96 2 98 2 101 2 104 2 105 2 107 2 110 2 111 2 114 2 116 2 117 2 119 2 121 Internetwork Packet Exchange See IPX intra area routing 2 121 2 123 i...

Page 311: ...mory 1 22 1 23 1 25 1 28 memory bus 1 25 Message Handling Service MHS for NetWare 2 96 MHS Message Handling Service NetWare 2 96 MIB 1 17 1 20 2 85 2 124 MIC media interface connector for FDDI rings 1 11 MIC medium interface connector for token rings 1 10 modem 1 11 1 12 1 15 1 17 modules for 650 1 24 1 25 MOP Maintenance Operations Protocol 2 115 motherboard 1 21 multicast address 1 6 multimode f...

Page 312: ...90 2 94 2 114 planning 2 117 2 119 point to point WAN link 1 11 1 13 1 19 2 72 2 123 poison reverse in RIP 2 66 port controller 1 22 1 23 1 25 1 27 1 28 filter 2 81 2 94 POTS See PSTN power supply 1 9 1 21 1 24 PPP Point to Point Protocol 1 13 preference 1 32 2 74 2 75 print server 2 96 2 102 2 116 printed circuit assembly See PCA printer queue 2 96 2 116 prioritization 1 30 prioritization of WAN ...

Page 313: ... 94 868 2 94 877 1 8 1 12 2 94 Rgetat command 2 114 Rgetis command 2 103 ring FDDI See FDDI ring token See token ring RIP Listen 2 70 2 72 RIP Supply 2 70 2 72 RIP Routing Information Protocol 1 5 1 7 1 8 1 31 1 33 2 65 2 67 2 69 2 70 2 74 2 79 2 80 2 94 2 101 2 102 RISC 1 25 1 27 1 28 ROM read only memory 1 22 1 27 routable protocols and suites See routing services route filter 2 79 2 80 Route Le...

Page 314: ...umber IPX 2 96 2 97 2 102 2 103 software architecture 1 16 1 29 1 30 1 32 1 33 distribution 1 16 See also operating system or configuration SONIC 1 23 1 27 source routing 1 5 1 29 2 93 2 115 source routing bridge 1 4 1 5 2 3 2 93 spanning tree 1 3 1 4 1 29 1 31 2 3 specifications 2 94 2 124 split horizon in RIP 2 66 SPX Sequenced Packet Exchange 1 6 2 95 SRAM static random access memory 1 22 1 27 ...

Page 315: ...ating system 1 16 1 17 1 22 1 26 1 28 upgrade memory 1 26 1 28 User Datagram Protocol See UDP V V 24 V 28 See RS 232 V 25 bis adapters 1 11 1 13 V 35 WAN interface 1 11 1 12 V 36 See RS 422 449 VAX 2 115 Videotex 2 116 virtual circuit 1 12 virtual IP host on non IP networks 2 90 2 93 virtual terminal DECnet 2 116 VMS 2 116 voltage auto adjusting 1 9 1 21 1 24 VT100 1 17 W WAN cables 1 11 1 12 1 14...

Page 316: ......

Reviews:

No comments

Related manuals for 600 Series

Kenmore 385.12116690

Brand: Sears Pages: 93

Brand: Singer Pages: 34

Brand: Singer Pages: 12

Brand: Ricoh Pages: 109

Brand: Baby Lock Pages: 2

Brand: baalbaki Pages: 8

Brand: Singer Pages: 8

HAKOMATIC B 90 CL

Brand: HAKO Pages: 42

1888 Foot-Bar Set of attachments

Brand: Singer Pages: 20

Brand: Singer Pages: 24

SWEEPER VAC V-036

Brand: Vax Pages: 12

AMS-210E Series

Brand: JUKI Pages: 243

Brand: DS Produkte Pages: 68

GBC Ricoh StreamPunch Ultra

Brand: ACCO Brands Pages: 11

Brand: 1byone Pages: 13

Brand: Tunstall Pages: 20

Brand: ZOJE Pages: 27

Brand: ZOJE Pages: 86

Brands by name

0 1 2 3 4 5 6 7 8 9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Popular brands

Load more brands