Timers: update-90s, invalid-270s, holddown-270+10s, Flush-630s, Admn Dist:100
IGRP routes : Interior: routes between subnets, System: routes within AS, Exterior: routes out side AS (1-65655).

Enable IGRP: Enable IGRP with AS no. (config t, # router igrp 10, # network 172.16.0.0, # wr mem )
Monitoring IGRP: R# sh ip route igrp : sh like RIP+ delay, bw, reliabilty, load, R# sh ip route igrp-shows IGRP routes,

R # sh ip protocol-sh all timers, broad freq, etc, R# sh ip int-shows int config, status & global parameters,

R# un / debug ip igrp events / transactions : sh IGRP traffic live, config t # no route igrp-stops IGRP.

OSPF: Link state proto, used within AS, a grp of routers that share a routing proto. Each router retains own dbase.
Huge metric limit 65533, assign cost to a link, load balancing among equal cost links, update-only for a change, pkts carry info for multi routers (tell a friend), can communicate out side AS, supports VLSM, TOS (types of svce routing)

EIGRP: Cisco prop, Balanced Hybrid Proto (best path by dist vect, update by link state), fast conv, support VLSM, partial updates, multi proto support (IP/IPX/Apple talk), update-only for change only for neighbors.

Exterior Routing proto : used outside an AS. More complex, stores more info about more routers.

EGP (Exterior g/way proto): dist vector proto, Dynamic, doesn’t use metrics like interior routing proto, can’t detect or correct routing loops, communicates thro?a core backbone routing net that connects AS together, establishes a set of neighbors to share routing info, send polls to neighbors to check ‘alive? updates other routers within an As.

Message types: neighbor acquisition message to check ‘alive? not, neighbor reachability message to check down/not, Poll messages to collect info about remote hosts?network, Error message to identify various errors.

BGP (Border g/way proto): An inter AS routing proto used on internet, used between & within AS, Determines loops, two BGP neighbors must be on same physical net to communicate, doesn’t use a core, all areas can comm w/o going thro?a core backbone, incremental updates for optimal routes only, metric with degree of preference.

Message Types: Open-conn established, Notification: error det, Keepalive: keeps the hold timer alive,

Novell IPX proto stack : ( SPX Sequence pkt exch, IP Internetwork pkt exch)

7/ 6/ 5-SAP, NCP; 4-SPX; 3-IPX, RIP, NLSP; 2+1-Thro Open Data Link ODL Interface Novell supports all media types.

IPX: connless proto, uses sockets, controls assignment of address ( software addressing), governs pkt delivery, makes routing decisions based on info from RIP or NLSP proto, SPX: adds conn oriented comms to the connless IPX. Works by creating virtual ckts or connections between machines with each conn having a conn ID included in the IPX header.

Novell RIP: dist vect routing proto, uses ticks (1/18 sec) & hop count (max 15) as metrics, split horizon, Aging timer,

RIP Broadcast @ 60 sec, SAP: svce advt proto, servers use it to advt svcs @ 60 sec, nodes use it to locate svcs,

Netware server: SAP(svcs table)& RIP (routing table) info xge between servers by broadcast @ 60 sec,

NLSP: Netware Link svce proto, advanced link state routing proto, might replace RIP & SAP. NCP: Netware core proto, provides file access, printing, synch, & security. GNS Request : Loc server, Cisco router + Rem server (book)

IPX Addressing: use 80 bits (10 bytes), first 4 B-Net add; 6 B-Node add & is assigned autoly, mostly MAC add (3B vendor code +3B sl no- so no need ARP). Since hw MAC add is included as part of software add IPX , no need of ARP eqvlnt.

上一页  [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] 下一页