Cisco Certified Internetwork Expert

Internetworking Fundamentals

Cisco’s Fundamentals Online: Click here

Cisco’s Internetwork Design Guide: Click here

OSI

TCP/IP Addressing: IP basics Documentation by Cisco

Private address blocks

Ports

Ports by number

Routing and Routed Protocol Basics: Routing basics by Cisco

• Routed Protocols are IP, IPX and AppleTalk (AT).

• Routing protocols are OSPF, RIP, RIP II, EIGRP, NLSP, RTMP, etc.

• These are Internal Routing protocols where BGP would be external.

• Distance vector protocols are RIP, RIP II, IGRP, and IPX RIP.

• Link State Protocols are OSPF, NLSP, etc

• EIGRP is considered a Hybrid Routing protocol

• Important routing Protocols based on TCPIP, NOVELL or AppleTalk:
  

  1. IP- RIP

  2. IP – OSPF

  3. IP – EIGRP

  4. IPX – IPX RIP

  5. IPX – NLSP

  6. IPX – EIGRP

  7. AT – RTMP

  8. AT – EIGRP
     

• Notice EIGRP (the Hybrid) works with them all

• Scalability restraints
  

  1. IP – 500 Workstations

  2. IPX – 300 Workstation

  3. AT – 200 Workstations

  4. NetBios – 200 Workstations

  5. Mixed – 200 Workstations
     

• MTU – You should be careful and avoid changing the size of the Maximum Translation Unit, but if you do, you can improve network performance by manipulating that size to be the largest possible.

• Ipv4 Header size: 20 Bytes W/Options

• Ipv6 Header size: 40 Bytes Fixed

• Standard administrative distances for IP routes (the lower the number the more preferred the route will be)

• Update Timers for distance vector protocols
  

  1. IP RIP 30 seconds

  2. IP IGRP 90 seconds

  3. AppleTalk RTMP 10 seconds

  4. IPX RIP 60 seconds

  5. IPX SAP 60 seconds (SAP is not a routing protocol)
     

• Other Baseline and network health facts
  

  1. Ethernet segments should not exceed 40% Network Utilization

  2. Token Ring should not exceed 70% Network Utilization

  3. WAN Links should not exceed 70% Network Utilization

  4. Response time should be less than 100ms

  5. Broadcasts/Multicasts should be no more than 20% of all network traffic

  6. On Ethernet there should be no more than 1 CRC error per 1million bytes of data

  7. Cisco Router CPU Utilization should not exceed 75%

Routing (Dynamic – Static)

• Static Routing – manually assigned by the Admin user entering the routes

• Dynamic Routing – generated/determined by a Routing Protocol

Dynamic

• With Dynamic Routing, routers pass information between each other so that routing tables are regularly maintained.

• The routers then determine the correct paths packets should take to reach their destinations.

• Information is passed only between routers.

• A routing domain is called an Autonomous System, as it is a portion of the Internetwork under common admin authority.

• Consists of routers that share information over the same protocol. Can be split into routing areas.

Distance Vector

• Used in smaller networks that are have fewer than 100 routers.

• Easy to configure and use.

• As routers increase in number, you need to consider CPU utilization, convergence time, and bandwidth utilization.

• Convergence is due to routing updates at set intervals; e.g., 90 seconds.

• When a router recognizes a change it updates the routing table and sends the whole table to all of its neighbors.

• Routing loops or counting to infinity occurs because of the delay in sending updates. This problem can be fixed with:
  

  1. Split Horizon - Info cannot be sent back on the interface it was received from

  2. Route Poisoning - When the network goes down, route gets set to 16 or unreachable until it’s back up

  3. Hold–Downs - Prevents routes from changing too rapidly in order to determine if a link has really failed, or is back up

Link-state

• Maintains Topology Database.

• Routers have formal neighbor relationship.

• Exchanges LSA (Link State Advertisement) or hello packets with directly connected interfaces.

• These are exchanged at short intervals (typically 10 sec).

• Only new info is exchanged.

• Scales well: only downside is that link–state protocols are more complex.

Appletalk: Cisco’s Documentation on AT

• RTMP – Routing Table Maintenance Protocol – a Distance vector protocol that has a default update timer of 10 seconds. Updates only when changes occur.

• AURP – AT Update-based Routing Protocol that allows the creation of a tunnel to interconnect two AT based networks through TCPIP to form an AT WAN. AURP uses UDP. It does not send periodic updates through the link.

• With AT, your client devices will automatically select a network address and will then broadcast a probe to ensure that it is unique.

Also, with AT – your workstations have the ability to remember the router address that sent the last RTMP packet

EIGRP: Cisco’s Documentation on EIGRP

• Enhanced Interior Gateway Routing Protocol

• A hybrid Routing protocol

• Proprietary to Cisco

• Uses metrics: BAN-DEL-REL-LO-MAX (How I remember it)
  

  1. BANDWIDTH

  2. DELAY

  3. RELIABILITY

  4. LOAD

  5. MTU (Maximum Transmission Unit) Notes on MTU

Note: Documentation for IGRP

WAN and LAN protocols

Ethernet: Cisco documentation on Ethernet, Fast Ethernet and Gigabit.

You must know this-

Make sure you visit the hyperlinks of each one and do extra research.

ATM: Cisco’s Documentation on ATM

• Asynchronous Transfer Mode

• FIXED length, 53 byte cells (48 payload +5 header).

• ATM Layers are at layer 2 while the Adaptation Layer is Layer 3.

• AAL’s have the responsibility of isolating the upper layer protocols from the ATM process details.

• Frame types (or better to be called Adaptation layers)
  

  1. AAL1 – will take a continues bit stream and will place it within the ATM cells, between source and destination – ATM will Require timing synchronization.

  2. AAL3/4 – supports connectionless and connection-oriented data. Adds a header and a trailer whereas AAL5 does not. (Messages can be interleaved)

  3. AAL5 – Also supports connectionless and connection-oriented data but also supports LANE. (LAN Emulation) AAL5 can also be referred to as SEAL (Simple and Efficient adaptation layer)
     

• A knowledge of ATM and its framing should be reviewed.

PPP: Cisco’s Documentation on PPP

• Point to point protocol.

• Major benefit is the use of more protocols then just TCPIP.

• There are other protocols used
  

  1. LCP – Will establish, then configure (and test) the connection

  2. NCP – (A family of NCP’s) will establish and configure the upper layer protocols.
     

• PPP uses HDLC as a basis to have encapsulation of Datagrams over Point-to-point links.

• HCLC –High level data link control.

• Make sure you are comfortable with the whole PPP process.

X.25: Cisco’s Documentation on X.25

• A WAN protocol that operates at Layers 1-3 of the OSI

• Offers Error checking but becomes slower due to that extra checking

• X.25 Protocol suite
  

  1. Layer 3 – PLP (Packet Layer Protocol) will manage the packet exchange between the DTE devices across a virtual link

  2. Layer 2 – LAPB (Link Access Procedure B) Data Link Layer Protocol that will deal with the packet framing between the DTE and DCE devices. Operates at Layer 2. Integrated into X25. Router can be DTE or DCE. HDLC confined to ABM transmission. Job is to make sure that frames are error free. There are three different frame types: Information frames - flow control, error detection; S Frames - requesting and suspending communications; and U Frames - link setup, disconnecting, error reporting. Has high overhead, but good error checking

  3. Layer 1 – (X.21bis) X-21bis will handle the Layer 1 aspects of activation / deactivation at speeds of up to 19.2 Kbps.
     

• X.25 uses other physical layer serial interfaces: EIA-TIA 232, EIA-TIA 449, EIA-530 and G.703.

SDLC (Synchronous Data Link Control) Cisco’s Documentation on SDLC

• Main SNA link layer protocol. PTP, half, full duplex. Two node types: Primary Stations control other stations, setup and manage links. Secondary stations can only transmit to the primary and only after permission.

HDLC (High-Level Data Link Control) Cisco’s Documentation on HDLC

• Link layer protocol for Serial links. Cisco Default. Supports the following modes: Normal Response Mode – as per Secondary under SDLC; Asynchronous Response Mode allows secondary to communicate without permission; Asynchronous Balanced mode combines the two stations. Has lower overhead than LAPB but less error checking.

ISDN Cisco’s Documentation on ISDN

Cisco’s Documentation on Designing ISDN networks: Click here

Supports data, text, voice, music. BRI 2 B and 1 D Channel. PRI 23B + 1D or in Europe 30 B + 1D.

ISDN Terminals – TE1 – Terminals that understand ISDN Standards; TE2 precedes ISDN standard have to use a terminal adapter. ISDN has four reference points to define logical interfaces R = TE2 to TA, S = Terminal and NT2; T= NT1 to NT2; U = NT1 and line termination equipment

ISDN Protocols

• E = Existing network

• Q = Switching and signaling

• I = Concepts, terminology and service

Frame Relay Cisco’s Documentation on Frame Relay

• Establishes a Frame relay Virtual Circuit that is a connection between two DTE devices.

• Two circuit types: Permanent (PVC) and switched (SVC) identified by DLCI.

• Multi–protocol support: IP, DECnet, Appletalk, IPX, XNS, ISO.

• More efficient and faster than X.25 because of less error checking.

• Default encapsulation on CISCO is CISCO or can be IETF. Use IETF if connecting to non–Cisco devices with frame relay.

• DLCI – Data Link Connection Identifier – IP addresses need to be mapped to DLCI's to communicate over a virtual circuit. Can be done dynamically with IARP or manually though the map command.

• LMI - Local Management Interface – gives DLCI global rather than local significance. Makes entire frame relay network appear as typical LAN. Manages status-providing info on keep–alives, multicasting, addressing and status of virtual circuit. With version 11.2 of IOS, auto–sensed.

• Three LMI types Cisco (default) ANSI and q933a.

• Subinterfaces allow you to route IP on one virtual circuit and IPX on the other.

• Some routers have limits - 2500 can handle max of 255. Two types of sub interface Point–to–point and multipoint.

Frame Relay Congestion Control

• DE – Discard Eligibility used to identify traffic importance

• FECN (Forward Explicit Congestion Notification) – To tell others the path is congested

• BECN (Backward Explicit Congestion Notification) – Goes back to sending router to tell it to slow down

• CIR (Committed Information Rate) – Minimum bandwidth guaranteed. Choose realistic level; can choose zero if retransmission is acceptable. Can be by BC or committed burst size that allows customers to exceed CIR for limited time
  
  

IBM networks – Source Route Bridging

For in depth explanations on Cisco’s site click here: SRB | SNA | IBM

• Bridging basics: Creates a single data–link, flat network
  

  1. Transparent Bridging – Connects two or more Ethernet segments. Learns MAC address of all devices and then starts filtering.

  2. Integrated Routing/Bridging – Allows you to route and bridge the same protocol by using a virtual bridge–group interface

  3. Source–Route Bridging – Knows the entire route to destination before it sends data. Not designed for large networks.

  4. Source–Route Transparent Bridging – Use this when you have to go across bridging domains. Affects spanning tree, as packets cannot cross over domains and therefore you cannot have multiple paths between these domains.
     

• Source Route Bridge (also known as –SRB)

• Route descriptors – They are bridge/ ring #’s fields in a RIF. They are used to set the path that a frame should take on a SRB network.

• SRB frames contain a RIF that has routing descriptors to the destination

• SRB networks use DLSw+ to establish a TCP link so that it can reduce NetBios / broadcast traffic queries occurring between peers.

• In SRB networking, the SOURCE determines the route to arrive at the destination node BEFORE sending information frames to it.

• In SRB networking, the source node will acquire the routes to destinations using something called EXLORER FRAMES.

• SRB can use STP, but does not rely on it because it is usually loop free. It would use STP when sending explorer frames to reduce the traffic on the line during its route discovery process.

• SRB is locked down to a hop count of 7 bridges.

Cisco’s Documentation on Designing SRB networks: Click here

Cisco Documentation on Designing DLSw+ network: Click here

Router Basics

Router Switching

• Process Switching
  

  1. Packet gets copied to process buffer, address is retrieved and the packet is encapsulated and forwarded on the appropriate outbound interface. Cache is updated and subsequent packets to the same address are handled on cached info. Most processor intensive.
     

• Silicon Switching
  

  1. Only 7000 Series + SSP6. The SSP is a dedicated switch processor that takes over from the router processor. Fast solution.
     

• Optimum Switching
  

  1. Faster than both Fast and Netflow Switching. Replaces fast switching on high-end routers.
     

• Fast Switching
  

  1. Is used when no entries exist in more efficient caches; on by default in low-end routers; sometimes necessary to disable due to memory limits or to aid troubleshooting.
     

• Autonomous Switching
  

  1. Compares packets against autonomous switching cache. When a packet arrives the interface checks the switching cache closest to it. Only found on 7000 and AGS+ series routers.
     

• Distributed Switching
  

  1. Happens on VIP (Versatile Interface Processor), very efficient. Gets more efficient as more VIP cards added. No need to use router processor.
     

• Netflow Switching
  

  1. Admin tool increases overhead; gathers stat data, port, protocol, and user info that can be sent to a management station.

Commands

• Show interface - Will show router Layer 2 errors (CRC, collisions, etc.)

• Ping | traceroute - Enables you to ping and perform Route Tracing functions

• Show access-lists - Will display your access lists and you can also specify by number

• Debug - Shows real time - should be used with caution - very CPU intensive

• Show {protocol} route - Will show routing table

• Show processes - Will show CPU usage and CPU time

• Show buffers - Will show usage and misses

Route Summarization

• Contiguous networks are grouped together and advertised as a single entity called a supernet.

• Move network prefix to the left (i.e., borrow bits from network portion of address) to describe a single route to contiguous block of IP addresses (Classless Inter–Domain Routing or CIDR).

• This can only be done using contiguous IP addresses.

Note: Classful routing uses Class A, B, C addresses.

• IOS Software: Familiarize yourself with IOS feature sets: Click here

• Offers a rich set of features
  

  1. Access lists – to filter / security

  2. Proxy services – see below

  3. Encryption - never use if CPU is at 65% utilization

  4. Compression – to compress / WAN-serial link optimization

  5. Queuing – FIFO, Weighted fair, Priority, custom
     

• Proxy services
  

  1. IPX GNS Request

  2. IPX Watchdog Spoofing

  3. Proxy ARP

  4. IP Helper

Design Fundamentals

Hierarchical Topologies (Cisco defines a three–layered approach)

Cisco's Small/medium Sized Business Solution Framework

Broken into three Categories

Analyzing Customer Requirements falls into two areas:

Network Management: (FCAPS)

Network Management straight from Cisco: Click here

Constraints to Design

Twelve steps to document the customers existing network

Fields for documenting the customer’s existing applications: (Step 1)

Fields for documenting the customers existing Protocols: (Step 2)

Three Part Firewall

Note: A PIX Firewall is Cisco’s Firewall based product, but remember that the IOS has a Firewall / NAT based feature set.

Design Document Components

Responding to an RFP (Request for Proposal)

Network Management

Cisco documentation on SNMP and RMON

More Products

For a detailed list of most of Cisco’s Network Management tools: Click here

Routers and Switches

• Make yourself familiar with the Product lines: Click here

• Click here for a Switches index: click here

• Core High-end routers: Click here

• Routers: Click here | Switches: Click here

Test your design: (Determining the Appropriate Testing Plan)

• Pilot – Very small implementation, used to get your point across

• Prototype – A larger scale and costly test of your design.

Notes for study and test

• Make sure you have studied thoroughly; you will be asked in depth questions from every corner of your study guides.

• Make sure you have enough practice with case studies. This is not a test for you to just memorize all these facts. You need to know how to implement them. Therefore, practice the case studies. Click here

• Make sure you are comfortable with the Cisco Product line. You will be expected (like any good designer) too offer your advice on what products to implement for the best price.

• Use the Cisco Site – you can look at most of your information right from the Documentation provided to you. Click here