~ Only the Hardware addresses change when packets go through routers.
~ Half duplex Ethernet - One station can only send or receive at any time.
~ Ethernet Frame - 64bytes Min 1518bytes Maximum.
~ ISL frames are 1522bytes long, this can be mistaken for Giants and lost. Have to use ISL NIC cards. On router interface use 'encapsulation isl 2' to use ISL frames on VLAN 2.
~ FX and SX are fibre media, 100VG-AnyLAN is twisted pair copper media.
~ Spanning Tree is IEEE 802.1d - created by DEC (Digital Equipment Corp).
~ BPDUs are Multicast frames, sent every 2 seconds. Blocked ports still receive BDPUs.
~ Forward delay - Time taken from listening to learning (approx 50 seconds)
~ Default IEEE bridge priority 32,768, used to select root bridge. If these are identical then switch with lowest MAC address is used.
~ ISDN Protocols - E = Telephone network standards, I = Concepts, Terminology, Q = Switching, Signalling methods.
~ ISDN Reference Points - R = non-ISDN device and TA, S/T = references point between NT1 and NT2, U = NT1 and ISDN network (US only)
~ TE1 = Device compatible with ISDN, TE2 = Device NOT compatible with ISDN, TA = Converts non ISDN signals to ISDN signals, NT1 = Converts 4 wires into 2 wire local loop, NT2 = Providers equipment (Switch, PBX)
~ BRI - 2 * B-channel 64kbps, 1 * D-channel 16kbps (D-channel - LAPD)
~ PRI (Europe, Aus) - 30 * B-channel 64kbps, 1 * D-channel 64kbps (20.48Mbps)
~ PRI (EUS, Japan) - 23 * B-channel 64kbps, 1 * D-channel 64kbps (1.544Mbps)
~ ISDN supports IP, IPX, Appletalk...
~ ISDN can use PPP, HDLC, LAPD, each B-channel needs a SPID
~ Use static routes for ISDN otherwise it will keep link open.
~ MAC address 48 bits (12 Hex), IPX address 80 bits.
Netware 3.11 (1983-) - ethernet_802.3/novell-ether (Cisco default on~ Ethernet networks), Netware 3.12 or later (1985-) - Ethernet_802.2/sap - includes LLC, Ethernet_II - arpa, Ethernet_SNAP - snap, Netware 4.11 - use sap, Netware 5 uses IP
~ Novell RIP - Metrics = ticks and hops (15 max), 60 sec updates (tick = 55ms / 1/18 sec)
~ Novell 4.11 > uses NLSP (Netware Link Service Protocol) Link State Routing
~ SAP - Updates 60 Secs - 4 = Netware file server, 7 = Print server, 24 = Remote bridge server
~ Ping Responses - U = unreachable, C = congestion, I = user interrupt,? = unknown packet type, & = lifetime exceeded
~ Trace Responses - N = Network unreachable, !H = Not forwarded due to ACL restriction, P = Protocol unreachable, U = Port could not be reached
~ Ethernet 5-4-3 rule = Between 2 nodes there can only be max 5 segments, 4 repeaters and only 3 segments must have users.
~ 80/20 rule - 80% of traffic should be local 20% across backbone
~ Class 1 repeater (translational) - delay 140 secs, number you can use 1
~ Class 2 repeater (transparent) - delay 92 secs, number you can use 2
~ CSMA/CD - Used on half duplex devices
~ Auto-negotiate on FastEthernet checks link speed and duplex of line.
~ Protocol field in IP header - TCP = 6, UDP = 17, ICMP = 1, IGRP = 9
~ Ports - 20 FTP data, 21 FTP program, 23 - telnet, 25 - SMTP, 69 - TFTP, 53 - DNS, 80 - HTTP
~ Loopback address - 127.0.0.1
~ ACL - Standard ACL as close to destination as possible, Extended ACL as close to source as possible
~ IP = 1-99, Ex IP = 100-199, AppleTalk = 600-699, IPX = 800-899, Ex IPX = 900-999, IPX SAP = 1000-1099
~ Remember that there is an explicit ACL of 'deny all' if no statements match.
~ Multiprotocol routing supports more than one routing protocol, allows a router to deliver packets from several routed protocols.
~ Core Layer - High speed switching - free from filtering or anything which will slow packets etc.
Distribution Layer - Packet manipulation, address area segregation,~ broadcast domains, VLANs, security (ROUTERS), WAN access, queuing, firewalls, multicast domains, ACLs
~ Access Layer - End users, ACL/filters, remote access, shared bandwidth (SWITCHES), segmentation, DDR
~ HSSI - 52Mbps max
~ ATM cell size - 53bytes
~ Cisco LMI - DLCI - 16-1007, ANSI LMI - DLCI 16-992 (DLCI = 10bits)
~ LMI is a special DLCI = 1023
~ LMI Multicasting reserved for 1019-1022
~ LMI extensions - Virtual circuit status, multicasting, global addressing, simple flow control
~ LMI types Cisco (default), ansi, q933a. From IOS 11.2 LMI is auto-sensed
~ Class A - 1-126
~ Class B - 128.1-191.255
~ Class C 192.0.1-
~ Class D - (1110 highest order bits) - remaining bits for multicasting
~ Class E - (1111 highest order bits) - Reserved for future use
~ RIP 1 (Classful), single subnet, periodic updates of full routing table, max hop count 15
~ RIP 2 (Classless addressing), triggered updates, full routing table updates
~ Directed Broadcast - All host bits set to 1 received by all hosts on local broadcast domain.
~ Local Broadcast (255.255.255.255) - All bits set to 1 received by all hosts on local and remote broadcast domains.
~ Synchronous serial links default to HDLC on Cisco routers
~ VIP cards - type slot/port adapter/interface (e.g e/1/0/2) (remember first interface is 0 not 1)
~ IGRP Metrics - Delay, Bandwidth
~ Default route - ip route 0.0.0.0 0.0.0.0 172.16.20.1 - need to use 'ip classless' (Classless is enabled by default on IOS 12.x) (Only work on stub networks)
~ ip default-network 172.16.1.0
~ CDP timer default 90 secs, holdtime 240 secs
~ Trunked links - FastEthernet or GigabitEthernet only
~ Frame Tagging - ISL = Adds it's own FCS, Cisco propriety (default), IEEE 802.1q
~ LAN Emulation (LANE) - Used for multiple VLANS over ATM
~ 802.10 - FDDI Frame Tagging
~ Hosts can only communicate between VLANS using Layer 3 devices
~ VTP Modes - Server (Default for Catalyst switches) Need at least one server in a VTP domain. All changes are advertised. Client - Sends and receives updates. To make a switch a server make it a client first then promote it once it's VTP database has received the latest revision. Transparent - Does not participate in VTP domain, but forwards VTP ads through trunked links. They keep their own database.
~ VTP adverts sent every 5 mins or when a change occurs, changes only kept by other switches if higher rev no than their current version.
~ VTP pruning - If a switch does not have any ports configured for VLAN 5 then it won't receive updates for it. Disabled by default. Enabled across entire domain if configured. VLAN 1 is not pruning eligible.
~ Config Reg - 00 Rom Monitor, 01 Boot Image from ROM, 02-F NVRAM, Bit 6 set to 1 to ignore NVRAM. Register is 16 Bits.
~ 1900 Switch Config - enable password level 1 - usermode, level 15 - enable password.
~ 1900 switch can have up to 64 VLANS. You cannot telnet from a switch but you can telnet into it.
Administrative Distances
Routing Protocol
Administrative Distance
Connected Intf 0
OSPF 110
Static route 1
RIP 120
EIGRP 90
UNKNOWN 255
IGRP 100
~ RIP - Updates 30 secs, Max Hops 15, Invalid 90 secs, Flush 240 secs, metrics hops, load balance 6 equal cost links
~ IGRP - Updates 90 secs, max hops 255 (default 100), invalid 3x90 secs, holddown 3x90+10 secs, flush 7x90 secs, metrics bandwidth, delay, load balance upto 6 unequal cost links.
~ When routers are converging no data is sent.
~ Frame Relay - 64 kbps - 1.544 Mbps, non-broadcast multi-access encapsulation (NBMA), dynamic bandwidth allocation, congestion control. Can use PVC and SVCs, PVC more common. Virtual circuit established before data sent. Encapsulation Cisco (Default), IETF (use when connecting non-cisco routers). Static routes are more stable than IARP.
~ Routers are DTE devices by default, DCE interfaces need a clock rate.
~ Telneting uses layers 1-4 so a good test of functionality. If you type a command the router doesn't know or type and IP address it will try to resolve the name and telnet.
~ Bandwidth command sets cost for serial links. This is only used by routing protocols so they can 'cost' paths. Default = 1.544kbps (T1) Command is in Kbps.
~ Clock rate command is in bps.
~ HDLC - Connection-orientated, operates at the datalink layer, small overhead, no way of distinguishing network protocols. Every vendors implementation is different, NO authentication, CISCO Default over serial lines.
~ LAPB - Connection-orientated, datalink layer protocol, HUGE overhead, uses windowing, used instead of HDLC for error prone links.
~ PPP - industry standard, used when connection between different vendors devices. NCP to identify network protocol, authentication, compatible with async + sync links, operates at physical + datalink layers only. PAP - insecure authentication, CHAP auth provides initial + periodic auth. PPP compression uses stacker and predictor methods. Error detection - PPP uses quality and magic number methods. Multilink - IOS 11.1 only, spreads the load over 2 parallel circuits (bundle).
~ Ethernet 0 is up, line protocol is down - keepalive or framing issue, check keepalives on both sides should match, check clocking on DCE, check encapsulation on both ends.
~ Ethernet 0 is down, protocol is down, - carrier detect is not present, other end maybe administratively shutdown or interface or cable problem.
~ Ethernet 0 is administratively shutdown - the 'no shutdown' command has not been issued on the interface.
~ Show interface serial 0 - shows bandwidth, MTU, keepalives.
~ MTU default = 1500bytes.
~ Bandwidth default = 1.544Kbps (T1)
~ Keepalives default = 10 seconds.
~ Use a cross over cable to connect devices of the same type (e.g router Ethernet intf to router Ethernet intf)
~Cross over cables swap pins 1 and 3 RD, and pins 2 and 6 TX
~ STP - 10-100Mbps - 100metres
~ ScTP - 10-100Mbps - 100metres
~ UTP - 10-100Mbps - 100metres
~ Coax - Coaxial - 500metres
~ Fiber - Single Mode upto 3000metres
~ Fiber - Multimode upto 2000metres
~ Connectionless protocols rely on application layer protocols for error handling and delivery.
~ EIGRP holds separate routing tables for IP,IPX,Appletalk, but only uses one protocol to distribute the updates.
~ CDP uses SNAP (Subnetwork Access Protocol) to enable neighbouring devices to exchange data.
~ IPX NLSP - link-state routing protocol intended to replace IPX RIP and SAP
~ NCP - Netware Core Protocol - Provides clients with access to server resources
~ IPX SAP - Sent every 60 seconds - includes all known services.
~ sap is Cisco default for Token Ring networks, SNAP is default for FDDI networks
~ VTP allows VLANs to be trunked over Ethernet, ATM, LANE or FDDI
~ Gigabit Ethernet using Multimode Fibre can run up to 260m
~ 100BaseFX up to 400m
~ VLAN Management Policy Server - Must be configured with all hosts' MAC addresses for dynamic allocation.
~ Standard ping - 5*100 byte ICMP echos, time out 2 seconds
~ DHCP uses UDP packets
~ Passive interface command stops interface sending routing updates, but still receives them.
~ 2 ways to configure VLAN membership, statically or dynamically through VLAN Management Policy Server.
~ ISL and Trunk protocol used to configure trunking on a switch.
~ Pre 10.3 IOS commands Config Net - copy config from tftp to DRAM Config Mem - copy NVRAM to DRAM
~ IP routing table [administrative distance/composite metric]
~ IPX routing table [ticks/hops]
