Module 9 Flashcards
Learning Unit 5 (62 cards)
WAN Characteristics
• Connects LANs over large geographical areas • High data throughput, long-distance support • Uses routers, modems, and specialized transmission media • Often leased from telcos (NSPs): AT&T, Verizon, Spectrum, Comcast • Corporations pay based on bandwidth used/reserved • May connect directly to IX (Internet Exchange) to reduce costs • Uses OSI Layers 1 & 2 differently than LANs • Same protocols as LANs for Layer 3 and above • WAN site = individual location • WAN link = connection between WAN sites o Point-to-point: 1-to-1 connection o Multipoint: 1-to-many connection
Differences Between LAN, CAN, MAN, WAN
LAN (Local Area Network) • Connects devices within small geographic area • Owned/operated by a single organization • Ethernet-based • Uses switches and routers CAN (Campus Area Network) • Multiple LANs in a campus/school area • Typically owned by a single organization • Confined to single or nearby properties MAN (Metropolitan Area Network) • Collection of LANs across a city/province • Network media is leased, often across public property • May serve one customer or general public • Examples: o Police station networks o Hospital to regional medical centers o Home office to branches WAN (Wide Area Network) • Large geographic scope (e.g., global) • Leased from ISPs or NSPs • Mix of tech and media types • Example uses: o Bank connecting offices and global partners o Sales reps uploading to HQ o Manufacturer working with remote contractors o Global e-commerce
Entry Point Equipment & Terms
• Modem: Modulates/demodulates analog ↔ digital • Router: DTE (Data Terminal Equipment), customer-owned • Modem: DCE (Data Circuit-Terminating Equipment), ISP-owned • DTE = customer LAN-side device • DCE = ISP WAN-side device • Sometimes DTE & DCE combined in one unit
CPE & Demarcation
CPE (Customer Premises Equipment) • All equipment on customer site • Owned by customer or ISP • Examples: Router, modem, CSU/DSU, line driver • ISP services only their own devices Demarc (Demarcation Point) • Division between ISP responsibility and customer responsibility • Equipment beyond demarc = customer-managed • Equipment before demarc = ISP-managed
Devices at/near Demarcation Point
NIU / NID (Network Interface Unit/Device) • Connects ISP’s local loop to customer’s network • Smartjack / INID = intelligent NIU o Diagnostic features (e.g., loopback testing) o Sends signal back to ISP CO (Central Office) Line Driver • Repeater to boost signals over long distances • Copper or fiber versions • Can be located on either side of demarc • May be customer- or ISP-owned CSU/DSU (Channel Service Unit / Data Service Unit) • Endpoint for dedicated WAN connection • Converts digital signals for transmission • Usually located between demarc and router • Can be owned by either ISP or customer
Routers – Core Concepts
• Connects 2+ networks, routes packets between them • Makes forwarding decisions using Layer 3 (and often Layer 4) data • Components: o Processor o OS o Memory o Network ports/interfaces o Console interface • All routers can: o Connect dissimilar networks (e.g., LAN ↔ WAN) o Interpret L3/L4 headers o Calculate & select best path (most efficient) o Reroute if path fails
Routers – Optional Functions
• Filter broadcasts to reduce congestion • Basic firewalling (block certain traffic) • Support local & remote connections • Redundancy (e.g., power supplies, NICs) • Monitor/report traffic statistics • Diagnose issues and trigger alarms
Router Types (by Network Location)
• Core/Interior Router o Operates within a single AS (Autonomous System) o Communicates only inside the same organization/network o Trusted, internal domain • Edge/Border Router o Connects an AS to an external/untrusted network o E.g., ISP connection • Exterior Router o Outside any specific AS o Often on Internet backbone o May also refer to other organizations’ edge routers
Router Installation
• Small/Home Networks o Simple setup: plug modem into WAN port, devices into LAN port o Config via web utility • Enterprise Routers o Multiprotocol & high-powered o Complex configuration o Requires networking expertise • L3/L4 switches can also route (similar functionality)
Routing Table – Overview
• Database for host location & best path info • Router = member of all connected networks • Each interface → unique network • Table = list of route entries: o Destination network (IP + netmask) o Gateway (next hop IP) o Interface (router port to use) o Metrics (route preference ranking)
Routing Table – Decision Process
- Packet arrives at router 2. Router checks destination IP 3. For each table row: o Compute IP range from IP + netmask o If destination IP fits range: Read gateway IP (next hop) Read output interface 4. If multiple matching routes → use lowest metric (better route) 5. If no match → look for 0.0.0.0 entry (default route) o Gateway in default = Gateway of last resort 6. If no default route → drop the packet
Routing Table Example: LAN A → LAN D
• Workstation sends job to printer (different LAN) • Steps: 1. IP forwards to default gateway (Router A) 2. Router A searches routing table → finds 2 possible routes Chooses one with lower metric Identifies next hop = Router C Identifies interface to use Decreases packet TTL 3. Forwards to Router C 4. Router C decreases TTL again Searches routing table Sees destination is on its LAN Forwards to Switch D 5. Switch D: Looks up printer MAC via ARP (if needed) Sends to printer
Routing Path Types
• Static Routes o Manually configured by network admin o Fixed path between networks (e.g., small biz to ISP) o No adaptation to: congestion, failures, relocation o Requires manual updates • Dynamic Routes o Router calculates best path automatically o Routing table updates in real-time o Can reroute during congestion/failure o Auto-updates when routers are added
Route Utility & Routing Table Access
• View Routing Table (OS-specific commands): o Linux/UNIX: route o Windows: route print o Cisco IOS: show ip route (Privileged EXEC Mode) • Routing Tables: o Workstations: Few entries (default gateway, loopback) o ISP Backbone Routers: Hundreds of thousands of entries
Routing Metrics (Best Path Criteria)
• Hop Count – Number of network segments crossed • Bandwidth / Throughput – Theoretical vs actual • Delay (Latency) – Time delay along a path • Load – Traffic/processing burden on a router • MTU (Max Transmission Unit) – Largest IP packet size allowed without fragmentation (excludes frame) • Routing Cost – Admin-assigned; lower = more desirable • Reliability – Based on historical path performance • Topology – Network layout/design
Routing Protocol Evaluation Factors
• AD (Administrative Distance) o Default number per protocol o Lower AD = higher priority o Can be manually changed by admins • Convergence Time o Time to identify a new best path after change/outage o Faster convergence = more desirable • Overhead o Burden on network (processing + data exchanged) o Higher overhead = more resources used
Routing Protocols Overview
• Routing Protocol ≠ Routable Protocol (e.g., IP) • Used by routers to exchange network status data • Typically operate at OSI Layers 3, 4, or 7 • Help populate and update routing tables
Common Routing Protocols Summary
Protocol Type Algorithm RIP IGP Distance-vector RIPv2 IGP Distance-vector OSPF IGP Link-state IS-IS IGP Link-state EIGRP IGP Advanced distance-vector BGP EGP Advanced distance-vector / Path vector
IGPs vs EGPs
• IGPs (Interior Gateway Protocols): o Used within autonomous systems (AS) o Used by core and edge routers o Types: Distance-vector, Link-state, Hybrid • EGPs (Exterior Gateway Protocols): o Used between autonomous systems o Used by edge and exterior routers o Only protocol used: BGP
Distance-Vector Protocols (e.g. RIP, RIPv2)
• Determine best path by number of hops (sometimes latency) • Routing by rumor: routers trust neighbors’ info • Periodically exchange entire routing tables, even without changes • Slow convergence, higher risk of persistent errors • Examples: RIP, RIPv2
Link-State Protocols (e.g. OSPF, IS-IS)
• Determine best path using state of links, not just hops • Routers map the entire network themselves • Share info only when changes occur • Faster convergence, more scalable • Require more CPU and memory • Examples: OSPF, IS-IS
Hybrid Protocols (e.g. EIGRP)
• Combine features of distance-vector and link-state • Example: EIGRP o Shares updates only on change o Fast convergence o Lower CPU/memory requirements than OSPF o Proprietary to Cisco (partially open since 2013) o Supports multiple protocols o Preferred on Cisco-only LANs
RIP, RIPv2, RIPng
• Type: Distance-vector IGP • Hop limit: 15 (if exceeded, destination unreachable) • Simple, quick to configure • Broadcasts table every 30 sec • Limited metrics: only hop count considered • Slow convergence • Not suited for large networks • RIPv1 (1988): original, less secure • RIPv2 (1994/1998): less broadcast traffic, more secure • RIPng (1997): adds IPv6 support • All versions considered outdated
OSPF (Open Shortest Path First)
• Type: Link-state IGP • Works on core or edge routers • No hop limit • Uses complex algorithms for best path • Recalculates paths when links fail • Fast convergence, low bandwidth overhead • Prevents routing loops • Supported on multi-vendor routers • Common in mixed environments