Resource Page CIS 2153 Syllabus Chapter Lesson Notes: 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11
Chapter 9 Lesson Notes
Chapter 9 Terms Installing, Configuring, Managing, Monitoring, and Troubleshooting IP Routing
I. Overview of Windows 2000 IP Routing
A. What is a Router?
Cisco, Bay Networks, Digital, and Cabletron Systems offer technologies dedicated to routing. Windows 2000 Server offers a routing component within the operating system itself called a software routing service.
1. Characteristics of a Software Routing Service
a. Dependant on the underlying OS
b. Benefits:
i. Tight integration with Windows 2000 features and benefits
ii. Built-in monitor and keyboard with standard Windows 2000 GUI interface for ease of use and reduced learning curve
iii. Flexibility of running other applications and services on the same computer
iv. Potentially cheaper than a separate hardware box
v. Same vendor equipment as other workstations/servers
2. Characteristics of a hardware router include speed and support of different protocols.
- Routed vs. Nonrouted Networks
1. Routed Network
One or more networks joined together by a common protocol and a router
2. Nonrouted Network
One or more networks that can be physically joined together but have no means of communicating with each other over the Network Layer—this may be due to a variety of reasons:
a. No common protocol
b. Protocol in use is not routable
c. No router to forward the packets
d. Router not configured to forward packets
- Routing Fundamentals
1. Static vs. Dynamic Routing
a. Static routes are paths which are manually specified by an administrator and do not change without administrator intervention. The routers have no knowledge of any routes that are not specifically configured.
b. Dynamic routes are paths that are updated automatically by the routing software. A router goes through a discovery process to learn the routes and destinations connected to it.
2. The Default Gateway
Gateways are actually routers. A TCP/IP computer sends nonlocal traffic to its default gateway when there is no specific route defined for it.
a. Default gateway issues and troubleshooting tips are discussed on pages 608 and 609.
i. You can only have one current default gateway per machine due to the single-route table in Microsoft’s TCP/IP.
ii. You can have multiple gateways defined, but the other gateways will not be used unless the primary gate is unavailable.
2. Routing Interfaces
a. Entry points into a router in the form of IP addresses, typically a NIC (Network Interface Card)
b. Depends on a logical IP address, not a physical interface, because routing occurs at the Network layer, not the Physical layer
3. Routing Tables
a. Lists are compiled by a router of known routes, or manually by an administrator.
b. When a router receives a packet to forward, it consults its routing table to see which interface should be used to forward the packet.
c. Host routing occurs when a computer forwards a packet to a router instead of sending the packet directly on its own network.
d. Router routing occurs when a router receives a packet that is not destined for another computer, so it must send the packet to either the destination computer (if directly attached to the router) or another router.
e. Perform Exercise 9-1 on page 611: Viewing a Routing Table on a Windows TCP/IP Computer.
f. Single-Route Routers
Software routing components such as those in Windows 2000 use a single routing table.
g. Multiple-Route routers, such as a Cisco router, can be configured to use a different routing table for each interface.
h. Routing Table Structure consists of the following:
i. Destination address and subnet mask
Contains the IP Network Address for a network route or an internetwork address for a host route
ii. Gateway or forwarding address - Identifies the IP Address of the interface
iii. Interface - Network Interface to be used when forwarding packets to the destination address
iv. Metric - Indicates the cost of a particular route
v. Lifetime - Defines how long the route is valid
See Table 9-1 on page 614 for a description of these attributes.
II. Enabling and Configuring Windows 2000 as an IP Software Router
The routing service in Windows 2000 Server must be enabled rather than installed. Perform Exercise 9-2 on page 615: Manually Enabling the Routing Option in RRAS.
A. Do you need to install routing protocols?
1. The RRAS router will deliver packets destined for another network if this router is the only router on your network.
2. No routing protocols are required to route packets from one interface on your router to another interface.
B. Static Routes
Added to Windows 2000 Server through the following:
1. Route Add Command
Used to add static routes to a WIN2000 server acting as a router. The route print command lets you confirm the addition.
2. Netsh Utility
Allows configuration of both local and remote RRAS components
3. RRAS Snap-in
Easiest way to add a static route: The GUI, shown in Figure 9-1 on page 617, makes the process much less error-prone.
C. Configuring the Router for IGMP
IGMP (Internet Group Management Protocol) is an extension to the Internet Protocol. It is used by IP hosts to report their host group memberships to immediately neighboring multicast routers.
1. Unicasting is sending a packet directly to a single TCP/IP client (a one-to-one relationship).
2. Broadcasting is sending multiple packets to all TCP/IP clients.
3. Multicasting is sending a single packet to multiple hosts using a specific Class D IP address (IP addresses in the range of 224.0.0.0 to 239.255.255.255).
4. Multicast Members are those hosts that register themselves with a specific multicast address, specifying that it would like to receive that server’s multicast traffic.
5. Multicasting Across Networks
a. The IGMP routing protocol needs to be installed when a direct connection between an IGMP host and IGMP server is not possible.
b. Enabling IGMP is accomplished by installing the IGMP router and IGMP proxy services as if they were a protocol, then adding two or more interfaces. A typical IGMP configuration is shown in Figure 9-2 on page 622.
D. Configuring the Router for ICMP Router Discovery
1. ICMP (Internet Control Message Protocol) automatically assigns a default gateway to workstations to ensure that nonlocal workstation traffic can be routed.
2. ICMP offers fault-tolerant mechanisms with minimum administrative overhead.
3. Components of ICMP discovery process:
a. Router Advertisements are sent out to advertise a router’s availability.
b. Router Solicitations come from TCP/IP hosts who require a default gateway. They send a Router Solicitation ICMP message to the all-routers multicast address of 224.0.0.2.
c. Perform Exercise 9-3 on page 624: Enabling ICMP Discovery Advertisements.
d. Router Discovery Settings and their explanations are contained in Table 9-2 on page 625.
III. Implementing Routing Protocols
A. Installing Dynamic Routing Protocols
1. There are two dynamic routing protocols you can install with Windows 2000 RRAS:
a. RIP (Routing Information Protocol)
distance vector routing protocol
b. OSPF (Open Shortest Path First)
link state routing protocol
2. The steps to add any protocol with RRAS are the same:
a. Add the protocol.
b. Add the interface(s) you want to use with the protocol.
3. With Windows NT 4.0 there was automatic binding of the protocols to the adapters, which is not the case in Windows 2000.
4. Silent RIP for IP allows your router to automatically construct a routing table with details of other routers, but without advertising its own routing presence on the network. This one-way communication is termed "silent."
5. Perform Exercise 9-4 on page 627: Installing Routing Protocols.
B. Adding Interfaces to the Routing Protocols
1. Do not delete the "internal" routing interface you see displayed in the RRAS snap-in. This represents all Remote Access Services devices, and all RAS clients are part of this interface.
2. Perform Exercise 9-5 on page 628: Adding the Dynamic Routing Protocols to Interfaces.
C. Dynamic Routing Issues
1. Convergence occurs when all the routers on the internetwork have the correct routing information in their routing tables.
2. Convergence time refers to the time it takes all routing tables to be updated.
3. The primary differences between distance vector protocols (RIP) and link state routing protocols (OSPF) are as follows:
a. The routing information exchanged
b. How the information is exchanged
c. How quickly the internetwork can recover from a fault (such as a downed server)
4. Distance Vector Routing (RIP)
a. Advantages:
i.Simple to configure and has low administrative overhead
- Disadvantages:
i. In large networks, produces large and unwieldy routing tables
ii. High bandwidth requirements
iii. Does not scale and a single route has a limitation of 15 routers
iv. High convergence time
5. Link State Routing (OSPF)
a. Advantages
i. Smaller routing tables
ii. Lower bandwidth requirements
iii. Abilty to scale to very large networks
iv. Lower convergence time
b. Disadvantages
i. Complex in design and configuration and therefore susceptible to human error
ii. Potential resource insensitive on large networks
D. Common Problems with Dynamic Routing
1. Rogue Routers
Routers that send conflicting routing information in order to send network traffic to destinations other than their legitimate end point.
2. Routing Loops
If the information in a router is out of date, or invalid, it is possible for a packet to end up at its starting point. This loop is ended or prevented by the TTL (Time To Live).
3. Count to Infinity
In order to prevent hops from router to router infinitely, the maximum hops for RIP is set to 15.
4. Black Holes
Because RIP uses unacknowledged delivery, announcements for downed routers could go unheard. This causes other routers to continue sending packets to routers that are no longer available on a network, thus the Black Hole.
5. Overhead of Large and Complex Routing Tables
When a routing table is full with many routes for a particular end point, the overhead in searching all of the tables for the best route can be very time consuming.
6. High Network Bandwidth and Broadcast Flooding
Distance vector-based routing protocols announce their routes, frequently causing bandwidth on a network to be consumed between routers.
7. Slow Convergence
If a router goes down on the network, it can often take some time for all routers on the network to converge. This slow convergence causes lost packets, and extra network traffic.
E. Options used to reduce convergence time
1. Split Horizon
Forces routers not to send information back to where it was just learned, thus reducing the network overhead.
2. Split Horizon with Poison Reverse
Announces all networks, but all networks already learned are announced with a hop count of 16. This causes them to be ignored since the maximum hop count allowed is 15.
3. Triggered Updates
Allows routers to announce updates immediately rather than waiting for a periodic update.
F. Combining different routing protocols on the same router
1. If a single protocol reports multiple routes to the same network, the one with the lowest metric (shortest path) would be stored in the routing table.
2. The preferred source decides which metric to use—RIP or OSPF—if they both report a metric for the shortest route to the same network.
3. View and set preference levels for route sources on the Preferences Level tab of the General Properties dialog box, shown in Figure 9-3 on page 637.
G. Common routing terminology used with large internetworks
1. AS (Autonomous System)
A group of routers and networks under the same administrative umbrella that uses the same routing protocol
2. IGP (Interior Gateway Protocols)
Routing protocols used within the AS (Autonomous System)
3. EGP (Exterior Gateway Protocols)
A protocol used for communication between different ASs (Autonomous Systems)
IV. RIP for IP
A. RIPv1
Easy to configure, but it has these shortcomings:
1. Uses Broadcast instead of Multicast announcements
2. Subnet mask is not announced with the route
3. There is no protection from rogue routers
B. RIPv2 advantages over RIPv1:
1. Multicast option rather than broadcast announcements
2. Subnet Mask is announced
3. Authentication
4. Supports route tags
C. Windows 2000 supports RIPv2 and has the following abilities:
1. Convergence options for split horizon, poison reverse, and triggered updates
2. Ability to modify the announcement interval
3. Ability to modify the routing table entry timeout
4. Ability to support Silent RIP
5. Peer filtering—the ability to accept or reject RIP announcements from specific routers (by IP address)
6. Route filtering—the ability to accept or reject RIP announcements of specific networks or specific routers
7. RIP neighbors—the ability to send unicast RIP announcements to specific routers that could not normally accept multicast announcements
8. Ability to announce or accept default routes or host routes
9. Perform Exercise 9-6 on page 643: Changing RIP Interface Properties to Support RIPv2 Exclusively with Authentication
V. OSPF
A. Characteristics of OSPF and advantages it has over distance vector routing (RIP):
1. Efficient use of bandwidth, used by directed (multicasts) and acknowledged information only when necessary rather than periodic announcements
2. Only routing changes are exchanged between neighboring routers rather than whole routing tables
3. Smaller and more efficient routing tables
4. Scales well for large networks (accommodates more than 15 routers, and up to 255)
5. Supports authenticated communication between routers as protection against rogue routers
6. OSPF is processor-intensive.
B. OSPF Terminology and Concepts
1. Hop versus routing metric
a. A hop is a distance vector’s implementation of a routing metric; there are no hops with a link state routing protocol such as OSPF.
2. Table 9-3 on page 647 lists equivalent terms in RIP and OSPF.
3. Area
a. An area is the sections of a divided network. Each area has a boundary, which sets the limit on routing announcements. Flooding occurs when announcements are sent. (See the following Beyond the Basics box)
4. Link State Database
a. All routers within a boundary contain the same routing table or Link State Database (LSDB).
5. Default Route
a. Each area can have a default route, which is used when a direct route is not known.
6. Identifying Areas and Routers
a. ID numbers identify areas, and each router is assigned a unique ID number. It is common practice to use the largest or smallest IP address for the Router ID.
7. Backbone
a. The backbone is needed when an OSPF network contains more than one area, and that backbone has the ID of 0.0.0.0.
8. Different types of networks include the following:
a. broadcast
i. A network that can support a hardware broadcast where a single packet sent by the router is received by all routers on the network.
b. point-to-point
i. A network that can be connected by two routers only.
c. non-broadcast multiple access
i. A network that can connect more than two routers but cannot support hardware broadcasts.
9. Different Types of Routers
a. Different types of routers include the following:
i. Areas Border Router (ABR): Has its interfaces in different areas and handles inter-area communication; sends only the summarized routing information instead of individual routes
ii. Internal Router (IR): Sits in its area and handles intra-area routing
iii. AS Border Router (ASBR): Connects different ASs
b. Routers exchange information to form adjacencies with neighbors through the OSPF Hello packet that contains router configuration information.
c. A Designated Router (DR) is elected by OSPF to minimize the amount of routing information exchanged when routers are powered on. This process is similar to Windows NT 4.0’s election process. See Figure 9-4 on page 652.
10. Reducing Routing Information
a. Accomplished with External Route Filters
i. External Route Filters allow you to limit where your AS will accept its external routes.
b. Stub Areas (See page 653and Figure 9-5 on page 654.)
i. To keep the routing table small, an AS can assume that any network point it cannot reach should be sent to address 0.0.0.0. One of the Area Border Routers must be able to accept this request and process it.
C. OSPF Configuration Requirements
1. Table 9-4 on page 656 lists the options for router configuration.
2. Interface configurations require you to verify the OSPF configuration for each of the following interfaces:
a. Areas
b. Priority for DR election
c. Cost based on bandwidth
d. Password for authentication
e. Network type
f. Values for Hello Interval, Dead Interval, and Poll Interval should be set the same as other routers in the area
VI. Demand-Dial Routing
A. What is Demand-Dial Routing?
1. Dial-on demand connections do not have a permanent connection as with interfaces that are always available. Once connected through dial-up, packets are forwarded on a dial-on-demand interface called Demand-Dial Routing.
2. On-Demand Demand-Dial Connections
a. Creates the connection when there is data to send over the interface and terminates it when the transfer of data is complete
3. Persistent Demand-Dial Connections
a. Used when a permanent connection is desired, but the infrastructure will not support it, a dial-up connection is made, but is not terminated when the transmission is complete. Thus it is immediately available for the next connection.
4. Two-way Initiated Demand-Dial Connection
a. Routers on either side of the connection can initiate the connection.
5. One-way Initiated Demand-Dial Connection
a. Only one router can initiate the connection.
6. Connection Authentication
a. Requires that you specify credentials which match a valid account on the answering router.
7. See Table 9-5 on page 661 for demand-dial interfaces, routers, and usernames.
8. Security on Incoming Connections incorporates the following features:
a. Remote access permission
b. Authentication
c. Encryption
d. Callback
e. Caller ID
f. Remote access account lockout
B. Enabling Demand Dial Routing
1. Perform Exercise 9-8 on page 664: Enabling Demand-Dial Routing.
C. Configuring Demand-Dial Interfaces (shown in Figure 9-6 on page 669)
Configurations for demand-dial interfaces include the following::
1. Set Credentials
2. Connect and Disconnect
3. Enable and Disable
4. Unreachable Reason
5. Set IP Demand-Dial filters
6. Dial-Out Hours
7. Demand-Dial Interface Properties
D. Using Remote Access Policies with Demand-Dial Routing
1. Remote access policies control conditions for the account used to authenticate an answering demand-dial interface.
2. Create a remote router accounts group, and then create a policy that allows connection for the Windows-Group condition.
E. Routing over Demand-Dial Interfaces
1. If your demand-dial connection is on-demand, use static routing.
2. Defining Static Routes for Demand-Dial Interfaces (page 673)
a. Auto-static updates
Allows RIP to get a one-time routing table from its neighboring routers and retain that routing table as static and available.
3. If your demand-dial connection is persistent, use dynamic routing with routing protocols.
a. Table 9-6 on page 674 shows some configuration options that should be changed for a dial-up connection.
VII. Troubleshooting IP Routing Problems
A. When is routing troubleshooting required?
1. Be sure that you actually have a routing problem and not a problem due to hardware failure, basic TCP/IP configuration, or name resolution.
2. Common reasons for problems with IP routing:
a. Routing failure: Packets sent from source computer do not arrive at the destination computer.
b. Intermittent routing failure: Packets sent from source computer only sometimes arrive at the destination computer.
c. Routing efficiency problem: Packets sent from source computer often take a long time to arrive at the destination computer.
- Different Troubleshooting Approaches:
1. Random panic mode
2. Methodically checking forward
3. Methodically checking backward
4. Intuition based on knowledge and experience
- TCP/IP Tools and Utilities
1. ARP – a
2. Route Print
a. Ping
i. Used to test connectivity to remote systems, it is a simple tool that is most often used in initial testing of network trouble.
- TraceRT
i. Traces the network route taken by an IP datagram as it travels the network. It gives some detail about each hop the datagram has taken.
- PathPing (new in Windows 2000)
i. Used to test connectivity to remote systems, it is a simple tool that is most often used in initial testing of network trouble. It combines the functionality of Ping and TraceRt.
- Network Monitor or similar network capture utility (page 683)
i. Captures network packets from the network and allows you to analyze them for problems, errors, or patterns.
- Verifying each component
When there is a problem, break the overall routing process into components and check them individually:
1. TCP/IP configuration on hosts
2. Routing tables
3. Router configuration
4. Dynamic routing protocol configuration
a. RIP
b. OSPF
OSPF Terms
Area (Routing Area)
Similar to a site - a group of IP subnets connected by high-speed links. Each routing area is identified by a number, called an area ID, that looks like and IP address, only it has nothing to do with IP addressing - it simply identifies the area.
Backbone Area (Area 0)
The area automatically created when OSPF is installed. This area is the core of OSPF routing. Normally, all other areas are connected to the backbone area. This area’s ID is always 0.0.0.0 (hence the name, Area 0)
Internal Routing
This is routing that occurs within a single routing area.
Internal Router
This is a router that performs internal routing. All of this router’s interfaces are connected to subnets in a single routing area.
Area Border Router
This is a router that, unlike an internal router, has interfaces that are connected to subnets in more than one routing area. Normally, at least one interface of an area border router is connected to the backbone area, but this is not a requirement. Area border routers are used to route packets between routing areas.
Autonomous System
All routing areas under the control of a single organization or company are referred to an autonomous system. In other words - all of a company’s networked routing areas.
Autonomous System Boundary Router
This is a router that connects your autonomous system with either the Internet or another organization’s autonomous system.