Network Routing and Protocols: A Comprehensive Guide
Network Routing and Protocols
Bridge Potency and Packet Switching
1. Two networks are connected by two bridges, A and B. If bridge A is more potent and packet switching is limited to 1000 bits long, and bridge B is less potent but the exchange of packets is 1500 bits, which bridge works more efficiently?
Multihoming
2. Can you explain the reasons why alternative gateways (multihomed) are necessary?
A: Fault Tolerance. A multihomed site is immune to faulty traffic or access from any of the ISPs that serve it. This means that if a flaw is detected that affects an ISP, traffic should be able to be attended to through another ISP that is not affected by the problem.
Load Balancing. Multihoming allows for the distribution of incoming and outgoing traffic between different ISPs to which it is connected, in a way that tends to maximize the utilization of resources.
Traffic Engineering. The site may decide, to some extent, what type and volume of traffic is sent to each ISP based on aspects such as service level agreements, the cost of sending certain types of traffic to each ISP, or any other policy that has been defined.
Independence from ISPs. Normally, a site does not rely on ISPs to enjoy the benefits of multihoming. That’s why multihoming solutions usually take into account that no special cooperation from ISPs is required to implement them. This implies that each site can recruit different ISPs in an independent and self-deploying multihoming setup.
IP Datagram Size
3. In the IP protocol, what is the maximum size of a datagram? 65536
ARP Protocol
5. How does the ARP protocol work?
The ARP protocol (Address Resolution Protocol) is responsible for converting IP addresses into physical network addresses. The operation of the ARP protocol is quite simple. When a machine wants to send a message to another machine that is connected through an Ethernet network, there is a problem: the IP address of the machine in question is different from the physical address of the same. The machine that wants to send the message only knows the IP address of the destination, so it will have to find a way to translate the IP address to the physical address. This is done with the ARP protocol.
IPv4 Datagram Size
9. What is the size of an IP datagram in IPv4? 576 bytes
IPv6 Fragmentation
10. Is there fragmentation in IPv6?
IPv6 allows routers to fragment packets. The issuer is always informed when ICMP fragmentation will be needed. Thus, the sender can lower its packet size for this connection, and fragmentation is no longer necessary.
Frame Bursting and Carrier Extension
14. What is Frame Bursting?
A station can transmit a burst of frames by cable without having to relinquish control.
What is Carrier Extension?
The minimum CSMA/CD carrier and time slot have been extended from 64 to 512 bytes.
Gateway
41. What is a gateway?
Gateways are responsible for linking inter-domain routing with intra-domain routes. They translate internal and external routes.
Backbone Router
45. What is a Backbone router?
Backbone routers belong to area 0 and are responsible for the spread of networks between different areas.
Routing Protocol Used by Backbone Routers
46. What routing protocol does a backbone router use?
OSPF
Distance Vector vs. Link State Routing Protocols
47. Write the differences in routing protocols: distance vector and link state.
Distance Vector
- Given the network topology from the perspective of the neighbor.
- Add vector of distances from router to router.
- Frequent periodic updates, slow convergence.
- Pass copies of the routing table to neighboring routers.
Link State
- Get a common view of the entire network topology.
- Calculates the shortest route to other routers.
- Event-triggered updates, fast convergence.
- Spend updates link state routing to other routers.
Routing Algorithm Metrics
48. What metrics are used in routing algorithms? Give four examples.
- Delay: Amount of time required to move a package through the inter-network.
- Bandwidth: Traffic capacity available from a link.
- Freight: Refers to the degree of occupation of a resource on the network (router).
- MTU (Maximum Transfer Unit): Maximum size of a packet that can flow through the network.
- Communication Cost: These are the operating costs of the links.
- Error Rate:
- Hop Count:
Spanning-Tree Routing Algorithm
49. Describe the routing algorithm spanning-tree.
The idea is to immerse branches of the graph to transform it into a tree of shortest paths. It is used in static routing, does not consume bandwidth, and is suitable for very reliable networks.
Phases of the Distance Vector Routing Protocol
50. Describe the phases of the routing protocol distance vector.
- Each node has a different identifier.
- Each node knows the cost to reach its neighbors. At home, the vector of distances (VDD) contains only one entry for itself with distance 0.
- Each node transmits its VDD toward its neighbors (periodically or when there are changes).
- Each node stores the latest VDD received from each neighbor.
- Each node recalculates its own VDD depending on what its neighbors report. This is done when: a VDD is received from a neighbor other than the one that has been stored, and when a link drops.
Convergence of Dynamic Routing Protocols
52. Explain the convergence of routing protocols dynamic.
Condition of an international network in which all routers have an accurate picture of the topology and/or resources available, and all routes are updated on the network.
Autonomous System
53. What is an autonomous system?
A set of routers with a common administration running the same routing protocol.
IGP Protocols
54. What protocols are classified as IGP?
Distance vector, link state, OSPF.
IPv4 Datagram Looping
55. In IPv4, how is it possible that a datagram remains in an infinite loop?
It uses the TTL for that.
IPv6 Addressing Rules
56. What are the addressing rules in IPv6?
x:x:x:x:x:x:x:x (x = 2 bytes in hexadecimal), addresses of 128 bits.
IPNAT
58. What is IPNAT and what problems does it resolve?
IPNAT is a mechanism used by IP to exchange packets between two networks that are assigned mutually incompatible addresses. It converts the addresses used in the package volume in real time. When the IP addresses used on the private network are public IP addresses in use on another network, the router has a translation table where it specifies the replacement of these with a single public IP address. This avoids address conflicts between the different networks.
IPNAT Address Conversion
59. How do you perform the conversion of addresses in IPNAT?
These address translations are stored in a table to remember what address and port corresponds to each client device and thus, know where they should return the packets. If a packet tries to enter the internal network and does not exist in the translation table, then it is discarded. Because of this behavior, it can be defined in the table that a specific address and port can access a particular device, such as a web server, which is called reverse NAT or DNAT (Destination NAT).
Networks Without Routers
64. Is it possible to implement different networks without using routers?
With VLANs.
ACLs in Cisco IOS
65. What are ACLs and what types are defined in the Cisco IOS of a router?
An ACL is a list of statements that defines the type of packets that a router allows or rejects. There are two types: standard and extended.
Standard vs. Extended ACLs
66. How are standard ACLs different from extended ACLs?
Extended lists are numbered 100/109 and extend the capabilities of a normal ACL. They filter by fate and protocols used.
Scalability of Static Routing
67. Why do we say that static routing is not scalable?
It does not react to changes in load and network topology.
Routing Protocol
What is a routing protocol?
A routing protocol initializes and updates (given) the routing table in routers (network layer) and switches.
- Centralized: It’s simple but not scalable.
- Distributed: Requires collaboration between routers, which makes it more complex and dependent on state transitions.
Static Routing
Static routing is not sensitive to changes in load and topology. It uses one way for all packages.
Dynamic Routing
Dynamic routing reacts to changes in topology and load. Packets can travel different paths to reach the same destination.
Distance Vector
- Each node has a different identifier.
- Each node knows the cost to reach its neighbors. At home, the vector of distances (VDD) contains only an entry for itself with distance 0.
- Each node transmits its VDD toward its neighbors (periodically or when there are changes).
- Each node stores the latest VDD received from each neighbor.
- Each node recalculates its own VDD depending on what its neighbors report. This is done when: you receive a VDD from a neighbor other than the one it has, and when a link drops.
EGP Features
- It is responsible for collecting and distributing information between class networks.
- They participate in both intra-domain and inter-domain routing.
- Translates internal and external routes.
- Example of protocol: Border Gateway Protocol (BGP4).
IGP Functions
- Calculates routes to all destinations within an autonomous system.
- Distributes routes to external destinations.
IGP vs. EGP
- IGP: Used internally within an autonomous system.
- EGP: Used between autonomous systems. Exchange of information available.
Advantages of Distance Vector
- Only required for maintaining a routing table.
- Simple iterative procedure to construct and update the routing table.
- Based on knowledge of the cost of link nodes vecinos.
- Finds minimum cost routes.
- Distributed algorithm for computing the shortest path (Bellman-Ford).
- Reacts to topology changes and cargo.
- The nodes send their neighbors (current) best distance to reach another node in the network.
- Each node receives distance vectors to its neighbors.
- Updates its best route to each destination.
Basic Link State Algorithm
- Each node builds a package called Link State Packet (LSP) that contains a list of its neighbors and the cost of reaching them.
- The LSP from each node is distributed through a diffusion mechanism to the remaining nodes in the network.
- Each node receives the LSP of the remaining nodes and builds a global map of the network from them.
- On the global map of the network, better routes are calculated by Dijkstra.
Benefits
- Fast convergence.
- Easy growth.
- Easier problem detection and correction.
Disadvantages
: Resource consumption high: CPU to run Dijkstra and process LSPs memory to store LSPs ( each node must know latopología complete network) Complex (broadcast algorithm of LSPs too hard)