TCP/IP and OSI Model: Key Networking Concepts
In the TCP/IP model, which layer(s) of the OSI model does the Network Interface layer correspond to?
Physical and MAC layers
Give three examples of link layer protocols.
IEEE 802.3/Ethernet, Frame Relay, ATM, and SONET
In the TCP/IP layer, the network layer is responsible for enabling the routing of data across (logical or physical):
It is across a logical network path as it uses packet format and IP address format.
What is the min/max length of an IP header?
20/65535 octets
Which field in the IP header is responsible for defining the Quality of Service?
It is TOS (Type of Service)
How do you prioritize the forwarding of an IP packet?
The priority is set through the Precedence field of the TOS. The TOS is a part of the DiffServ quality of service model and it is also used by the OSPF protocol.
What does Per-hop-behaviour (PHB) refer to?
It refers to the queuing and the treatment behavior of a router.
How can a router notify its neighbors that it supports congestion?
By setting the ECN field (Explicit Congestion Notification) to 1.
Provide the definition of MTU and state for which purposes it is used.
It defines the maximum size of a frame in the data link layer. If a packet exceeds the tolerated size, it is then fragmented.
In which TCP/IP layer is an IP packet fragmented? Based on which criteria is it reassembled in the correct order?
A packet is fragmented in the network layer. An identifier is added to each fragment.
Is it always possible for a router to fragment a packet (assuming that the packet size is higher than the data link capability of a router)?
The fragmentation rule is not absolute. A router can deny fragmentation by setting the flag Don’t Fragment (DF) to 1 (DF=1).
How does a receiver detect the end of a fragmented flow (a flow is a set of packets from the same source)?
The receiver detects the end of a fragment flow when it encounters a fragment with MF=0.
Is the checksum of an IP packet recalculated at each router?
Yes, because the TTL field is decremented at each hop, and this field is part of the IP packet header.
How does a router or a host determine the network of an IP address?
A router determines the network of a given IPv4 address by using the first octet and the network mask.
What are the two main restrictions that apply for network masks in a classful environment?
An IPv4 host address cannot have all its host bits set to 0 (reserved for routers), and the host bits cannot be set to 1 (reserved for broadcast).
How does a host discover the data-link identifier of another neighbor?
By using the ARP protocol
How is an ARP request sent?
The request is encapsulated into a MAC frame. The source @ of the frame is the MAC @ of the source, and the destination @ is the broadcast address (all-hosts address).
Which Cisco command debugs the activity of the ARP protocol?
debug arp
Which Cisco command deletes all the entries of an ARP table?
clear arp-cache
What is a gratuitous ARP request?
It is an ARP request issued by a host using its own IPv4 address as the target @. It is used to detect address duplication.
For which purposes is HSRP (Hot Standby Router Protocol) used?
It is used for fault tolerance. One router is configured to be active, and another one is a standby. When the active router fails, the standby takes its place.
What is a “Source Quench” message? When is it used?
A destination host may send a source quench message if datagrams arrive too fast to be processed.
How do hosts discover local routers?
By listening to Router Advertisement and Router Selection messages (ICMP messages, type 9 and 10)
By which means does the TCP protocol provide reliability?
TCP uses a system of acknowledgment, checksums, retransmissions, and timers to provide reliability.
How does the TCP protocol regulate the flow of IP packets?
The TCP protocol uses a mechanism called windowing to control the flow between two pairs and avoid congestion.
What are the essential differences between TCP and UDP?
Correct sequencing, reliability, and flow control through the windowing mechanism.
Instead of ARP, Novell NetWare uses a network address that includes a device’s MAC address at the host portion. Why can’t IP do this?
If IP used MAC identifiers as the host part of the IP address, subnetting would not be possible.
Select a subnet mask for 10.0.0.0 so there will be at least 16000 subnets with at least 700 host addresses available on each subnet.
A subnet mask of 255.255.252.0 provides 16,382 subnets of class A, and 1022 host addresses on each subnet.
Select a subnet mask for 172.27.0.0 so there are at least 500 subnets with at least 100 host addresses available on each subnet.
255.255.255.128 provides 510 subnets with 126 host addresses on each subnet.
Use a 28-bit mask to derive the available subnets of 192.168.147.0? Then derive the available host addresses of each subnet.
192.168.147.16, 192.168.147.32, 192.168.147.48, …, 192.168.147.224