Physical Flow of data

When the data arrives at its intended destination it will travel up the 7 layers to pop out onto the
Users application as a for example an email or webpage.

When the user on the left hand machine press the send button on their email program the email is sent down the stack starting at layer 7 and through the layers. There is no way of avoiding passing through a layer. As the data moves through each layer the functions of the layers will determine exactly what happens to the data.

Logical flow of data
The data will move physically down through the layers until the data is placed onto the physical wire as 1’s and 0’s, even though the data moves down the stack for example i.e. layer 4 passes the data to the next layer down to layer 3 the actual communications is logical from peer layer to peer layer.

In the diagram above the arrows indicate the logical communication between layers. Layer 3 for instance is only interested in its corresponding peer layer’s information, and layer 4 is only interested in its corresponding peer layers information.

Encapsulation of data
User data moves down the stack from 7 to 1 on the left hand stack, from Layer 4 down to layer 2 the data is packaged or encapsulated in the lower layer data.

1. Layer 4 is encapsulated into Layer 3.
2. Layer 3 data is encapsulated into layer 2
3. Layer 2 data is transmitted onto the wire
4. Data arrives at the destination
5. Receiving layer 1 pulls the 1’s and 0’s off of the wire to reconstruct the data and passes it up to layer 2
6. Layer 2 removes it’s peer layer data and passes the remaining data to layer 3
7. Layer 3 removes it’s peer layer data and passes the remaining data to layer 4

The data enters the stack from the top and makes its way from layer 7 down to layer 1. From Layers 7 to 5 the data is encoded and or encrypted, from layer 4 down to layer 2 new fields are added to the data. These fields are for addressing, sequencing, length, type of data, error correction, error detection, flow control to name a few (We will cover all of these terms).

Once the data now wrapped up nicely inside all the other information a bit like when you send a parcel in the post you would wrap the parcel and apply a to and from label and maybe a sticker which says “fragile” or “This way up”

De-Encapsulation
On the receiving end, the data is pulled off of the wire and reassembled, the parcel of data is then passed up the stack. Each layer will read and remove and discard its corresponding peer layer data.

When layer 2 receives the data parcel from layer 1, it will check the information contained inside the layer 2 header for example at this layer it will check layer 2 hardware addresses, if it is happy with the info it will remove the layer 2 header and tail then pass the rest up to the next layer above.

Each layer will do the same until the data is passed up to layer 7 then presented to the user as an email or webpage etc.

Naming the data parcels
The data parcel otherwise referred to as the PDU (Protocol Data Unit) which is produced at each layer has a specific name and it is important to learn the names so you can become a more efficient fault finding network work engineer.

1. The PDU at layer 4 is referred to as a “Segment”
2. The PDU at layer 3 is referred to as a “Packet”
3. The PDU at layer 2 is referred to as a “Frame”
4. The PDU at layer 1 is referred to as a “Bits”

How Layers are used

In a layered system, a layer is considered to be a service provider to the layer above it. The upper layer is considered to be a service user by the lower layer. The service user avails itself of the services of the layer below by sending a transaction to the provider. This transaction informs the provider as to the nature of the service that is to be provided.

A layer cannot be by passed, even if the end user does not require the services of that layer, the user must still “pass though” the layer on the way to the next adjacent layer, this pass though will only invoke a small set of code and therefore translates to some processing overhead.

The OSI Model refers to the layers with the terms N, N+ and N-. The layer with is the focus of attention is designated layer N. The layer above is N+1 and the layer below is N-1.

If the Network layer is the focus of attention then the Network layer is N, the Transport layer is designated N+1, and the data link layer is designated N-1.

Each layer contains entities that exchange data and provides functions (horizontal communications)

With peer entities at other computers.

For example Layer N in machine A communicates logically with Layer N in machine B, layers N+ communicate logically with one another likewise with layers N-.

At each layer at a transmitting station adds header information to data. The headers are used to establish peer-to-peer sessions across nodes (horizontal communications). At the receiving side, the layer entities use the headers to implement actions created by the peer entity at the transmitting side.

The figure below shows how a machine (a) sends data to another machine (b). Data is passed from the upper layers or the user application to layer N+1. Layer N+1 adds a header to the data (labelled N+1 in the figure). Layer N+1 also performs actions based on the information in the transaction that accompanied the data from the upper layer.

Layer N+1 passes the data unit and its header to the layer N. This layer performs some actions based on the information in the transaction, and adds its header N to the incoming traffic. This traffic is passed across the communications line (or through a network) to the receiving machine B.

At ‘B’, the process is reversed. The headers that were created at machine ‘A’ are used by machine ‘B’ to determine what actions are to be taken at the peer layer. As the data and the existing headers are sent up the layers, the respective layer removes its header, performs the defined actions, and passes the traffic on up to the next layer.

The user application at machine ‘B’ is presented only with user data, which was created by the sending user application (machine A). These user applications are unaware of the many operations in each layer that were invoked to support the end-user data transfer.

The headers created and used at the peer layers are not to be altered by any non-peer layer. As a general rule, the headers from one layer are treated as “Transparent data” by any other layer.

There are exceptions to the rules, as examples, data may be altered by a non-peer layer for the purposes of compression, encryption or other forms of syntax alteration. This type of alteration is allowed if the data are restored to the original syntax when presented to the receiving peer layer.

An exception to the exception, the presentation layer may alter the syntax of the data permanently, because the receiving application layer has requested the data in a different syntax (ASCII instead of a bit string)

NETWORK MODELS AND PROTOCOLS

This next chapter will describe two very important concepts in networking known as protocol stacks. There are two very important protocol stacks which as a network engineer you must know the OSI model and the TCP/IP model. Knowledge of both these models will help you understand the role of various devices that go to make up a network also help you design, configure and fault find a network more efficiently.

The OSI model
One of the very first concepts I learnt when I entered this networking arena back in the 90’s was that of the “Protocol Stack”, I still recall how I was a little confused by this term, what was this protocol stack?, where does it live, can I see it?, can I hold it?.

What I would like to do over the course of this section is to explain the purpose of the “Protocol Stack” and in particular the OSI protocol stack long with the IP/TCP protocol stack.

Note: It is a good idea to note terms down. I.T is awash with weird and wonderful TLA (Three letter acronyms) even if right now the terms do not mean anything, just note them down.

Here we go, so what is the OSI model, well to put it in its simplest terms, the OSI model is a conceptual model which determines what should happen to your data when it moves from your “Screen” to the “Wire”. The emphasis is in the “what should happen” since the OSI model is a conceptual model or rather a reference model and not an actual working model

Another way to explain it is like this, imagine you wanted to build a car, how would you do it?, you would learn all the necessary skills required to build the car, from welding the body panels, to stitching the fabric of the seat, building the engine and wiring the electrics, this is going to take a little while as you have to learn all the skills needed, so instead of doing all the work yourself you decide to hire seven (7) people to build the car for you, but rather than teach each person all the skills needed to build a complete car you we teach each person an individual skill, so one person does one job only.

Your car is a huge success so you decide you are going to give up the day job and build cars for a living, for this you need a bigger space like a factory, as you have no experience in setting up a car factory you bring in a “Car factory consultant”, who reports that a car factory will need the following:
1. Goods in
2. Storage and inventory
3. Conveyor system to transport the cars down the production line
4. Robots to do the heavy lifting
5. Paint shop
6. Interior station
7. Goods out

Note I have never built a car let alone built a car factory so these are just guesses

The consultant has not and will not specify where to purchase any of these items but will lay out a design how the factory ought to be laid out, the specifics are up to you, where you get the robots is your decision, the paint shop is your decision, all you have from the consultant is the basic frame work or a reference of what should happen in the car factory and what at which stages certain actions and process ought to occur.

The OSI model is the same as the car factory production line. Your car factory you will have a production line, the basic car frame will be placed at the top of the of the production line and will move down the line, as it moves down the line each of your 7 faithful workers will each perform their individual tasks.

1. Body shop to press the panels
2. Panels are welded together to form the chassis
3. Doors, hood and boot lid are fitted
4. Paint shop
5. Assembly were the electrics and mechanicals are fitted
6. Assembly were the interior is fitted
7. Driven out of the factory into the showroom

Why have a production line? Simple, since we are able to change how any of the individual processes is performed without affecting the process before or the process to follow, therefore if you decided the paint shop process is to be changed what will not affect step 3 or step 5 provided whatever happens are step 3 presents the car to step 4 in a manner that step 4 can perform it’s task. Whatever happens at step 4 is of no concern to step 5 provided step 5 is presented a car with which is recognises and can work with.

Now going back to the OSI model, just like our production line, the OSI model is a layer model which will take our data say for instance an email and move the “1’s and 0’s” that make up our email down through 7 layers of its production line, but unlike our production line the OSI layered model starts at layer 7, layer 7 is the top layer and the lowest layer is referred to as layer 1.

As our email goes down through the layers each layer will perform certain actions on the email, from encoding using ASCII to encryption, placing addresses, sequence numbers, error detection and so on and so forth. When the e-mail now wrapped up in layers and layers of communications information to deliver the email reaches layer 1 it is then transmitted onto the wire.

Having a layered approach like our car production line makes it much easier to fault find as we can narrow problems down to specific layers, it also means that any layer can be changed without affecting the next layer above or the next layer below.

The diagram above shows the 7 layers of the OSI model and their respective names

CABLING AND TOPOLOGIES

Topology: Physical Bus
A bus network is a single length of cable to which work Stations, printers, servers, routers etc. connect.

There have been two major physical bus technologies:
• 10Base5
• 10Base2
• 10Base5

This technology capable of 10Mbps and allowed a cable run of 500m with the maximum number of nodes connected to the wire limited to 100 devices at a distance of 2.5meters apart. 2.5 Meters is due to the wavelength of the signal used to transmit the data.

At each end of the cable run there was a terminator (Resistor of 50ohms), if this terminator was removed the network would fail as the AC signal rather than being dissipated would be reflected back down the wire, always a barrel of laughs that one.

Stations would connect to the network by having to pierce the actual cable. This type of connection is referred to as a vampire tap,

10Base2
This technology also referred to as “Thinnet” or “cheapnet”, is capable of 10Mbps and allowed a cable run of 185m with the maximum number of nodes connected to the wire limited to 30 devices per segment, with no more than 5 cable segments separated by 4 repeaters with no more than three of the 5 cable segments being mixing segments.

These mixing segments can either be 10Base2 or 10Base5, a mixing segment is one with connected stations, only a mixing segment can be populated with stations i.e. P.C’s etc.

Two of the segments can be Link segments either 10Base5 or 10Base2, these link segments connect the repeaters together

The minimum distance for cables between workstations must be at least a half-meter. Drop cables should not be used to connect a BNC connector to the network interface card (NIC) because this will cause signalling problems unless the NIC is terminated. The entire cabling scheme, including all five segments, can’t be longer than 925 meters.

The terminator (Resistor of 50ohms) is built into the connector of the last station on the bus network, if this terminator was removed the network would fail as the AC signal rather than being dissipated would be reflected back down the wire and the reflected signal is impossible to distinguish from a collision.

Terminators had a metallic chain used for grounding the network. Only one end of the network needed to be grounded, if both ends were grounded it would result in a “grounding loop” which results in network outages.

The Disadvantages of a physical Bus network are:
1. If there is a break anywhere along the cable run the entire network would stop functioning
2. One large broadcast and collision domain, the stations connecting to the wire the slower the network becomes
3. 10Base5 stations need to connect to the network by piercing the cable
4. Lots of collisions on the cable
5. No security
6. The number of work stations and cable length is limited

CCNA TRAINING: THE NETWORK INTERFACE CARD AND BIA ADDRESS

There are a few ways to connect to a network for this moment we will concentrate on a wired only connection.

To connect to a network it is common to use a cable attached to a port on your P.C called an Ethernet port using an RJ45 connector.

Every interface has been assigned a unique address called a “Burned in Address” which identifies the interface uniquely on the network.

Every Frame which your device sends will include the unique BIA (Burned in Address) as a source address.

The BIA

The BIA is a 48bit value written in a format called “Hexidecimal” or just “Hex”

The BIA is made of two parts, the OUI (Organisational Unique Identifier) and the Vender Assigned Part. The 48bit BIA is split down the middle 24bits make up the OUI and 24bits define the Vendor assigned part.

The OUI identifies the manufacturer of the interface and is assigned to the maker by global standards body called the IEEE (Institute of Electrical and Electronics Engineers), more on this later.

To discover the MAC address of your own NIC (Network Interface Card) simply type in the following command into the command prompt of your machine:

Step 1: Go to start, type “cmd” in the box indicated by the red arrow, then hit the return key. A new screen will appear

Step 2: Type into the command prompt the command “ipconfig /all” following by the return key. A series of messages will appear

The machine used in this example has a unique MAC of “68-F7-28-33-2D-67” MAC address is shown is unique to this particular machine, it is expected that no other Ethernet interface in the entire world has the same address as this particular interface.

HEXIDECIMAL
All MAC address are written in HEX.
1. Each character in the MAC address is a HEX character.
2. Each HEX character is composed of 4 binary bits
3. These four bits are sometimes referred to a “Nibble”
4. Each Binary character in the 4 bit nibble has a value. The value of the binary bit is dependent on its position in the nibble.

5. By changing the binary from off “0” to on “1” will cause the corresponding value to take effect
Example One:

The binary bits with the values of 2 and 1 have been switched “on”, whilst the binary bits with the values of 8 and 4 have remained in the “off” position. The value of the four bits is 3

Example 2:

The binary bits with the values of 8 and 1 have been switched “on”, whilst the binary bits with the values of 4 and 2 have remained in the “off” position. The value of the four bits is 9

Example 3:

The binary bits with the values of 8 and 2 have been switched “on”, whilst the binary bits with the values of 4 and 1 have remained in the “off” position. The value of the four bits is 10

In HEX any value from 10 to 15 are not represented as the numerical value but are substituted with a letter, the table below show the correlation between the binary, value and letters

Anatomy of a MAC address
Every MAC is made up of 48 bits. Each MAC address has two main parts:
1. The OUI = 24 bits
2. The Vendor assigned part = 24 bits

Say for example you wanted to get into the NIC market one of the very first things you would have to do is get a unique OUI (Organisational Unique Identifier), these are acquired for the IEEE. The IEEE is in charge of ensuring that each NIC card maker has a unique OUI.

The OUI is always the first 24 bits of the MAC address and the Vendor assigned part is always the second 24 bits. Recall that each character is HEX, and each HEX character is composed of 4 binary bits.

It is the responsibility of the NIC manufacturer to ensure that they do not create NIC will duplicate MAC addresses and it is the responsibility of the IEEE to ensure that they do not assign the same OUI to multiple vendors

How Binary Works

Imagine this, you and your friend live across the road from each other, at night you wish to communicate using flash lights, you need to have an agreed code of light flashes.

So you agree on the following code, to represent a single binary 1 you will switch the light on for one second and to represent a binary zero you will keep the light off for one second.

Of course this is going to quickly go to the dogs but you understand the premise that we have agreed on a line code that we both understand, now it’s simply a matter of keeping count of the on’s and offs to interpret the code as one and zeros to reconstruct the message.

When one P.C wishes to communicate with another P.C is will tap out on the wire using electrical pulses a series of binary bits, sequences of these binary bits represent letters, numbers and other characters, the most commonly used character coding is known as ASCII. Most modern character-encoding schemes are based on ASCII (American Standard Code for Information Interchange). Codes represent text in computers, communications equipment, and other devices that use text. Most modern character-encoding schemes are based on ASCII, though they support many additional characters.

The letter A is = 1000001

The letter F is = 100 0110

The letter W is = 101 0111

The % is = 010 0101

Can you decode the following message?

NOTE: To convert drop the first binary character using the ASCII encoding

01101001 00100000 01100001 01101101 00100000 01100001 00100000 01100010 01101001 01101110 01100001 01110010 01111001 00100000 01100111 01100101 01100101 01101011

When a P.C communicates with another P.C on the network it will create a parcel of data called a frame, this parcel of data is a up to 1500 bytes in size, a byte Is another way to state 8 bits or an octet, so a byte is an octet and octet is 8 bits. Therefore 1500 byte is 1500 x 8 = 12000 bits per frame of data.

Each frame your P.C creates has the following fields.

1. Destination Address
2. Source Address
3. Length
4. Data
5. Frame Check sequence

Data Frame Example

The frame is pretty much like any parcel which you would send in the post, it has an address where we would like the parcel to be sent to, the address of the interface which transmitted the frame since data communications is generally a bi-directional operation.

The frame also has dimension just like a parcel in the way of length and type of data it is carrying

HOW DEVICES COMMUNICATE

When I first started in networking the one thing that I could not get my head around was how does one machine communicate with another?, I understood that P.C’s used binary to communicate but what I could not visualise was how a parcel of data was sent from one interface to another in my mind I used to picture a precession of packets on the wire in line one after the other like in the diagram below.

Of course I soon realised that this was not possible, we are talking about a single pair of wires to transmit and receive. The wire can only be in a single state at any one time, like a light bulb can be either off or on, but not both on and off together. On = 1, Off = 0.

It then dawned on me that just like Morse code the wire is carrying a single signal at any one time. Therefore the wire is either on or off, the transmitting interface just like a Morse code operator is “tapping” out the signal one bit at a time, albeit very quickly to achieve the high rates of data throughput we have come to expect.

The diagram above shows how data is sent over the wire as single on pulse or off electrical pulses each one lasting 1 second.

Don’t worry too much about all some of the terms like “Frame” for now, simply make a note of them, writing things down helps recollection later on.

The biggest draw back of using hubs is that they are not “intelligent” devices. When we speak of intelligence in a piece of networking gear we are referring to the ability to examine the parcel it receives  and determine how to best deliver this parcel of data.

A hub will flood all data received out of all ports. Devices receiving data which they know is not for them will dispose of the data.

Going back to our initial question, “What is a Network?” we answered it by saying that a network is a collection of devices that move parcels of information from A to B.

When you are driving you know a few things, you know where you started and where you are going. To get to your destination you will follow road signs which point towards your destination, without these signs you would not be able to find your way.

These road signs where placed there by someone at some point, every so often someone will come along and clean the signs, fix them up and make sure they are still pointing the direction they need to be pointing.

Your network devices operate in the same way a cross road, initially the cross roads have no signs so no one arriving at the cross roads would find any useful information until the person from the “authorities” comes along and puts up some signs for travellers to use.

So in the networking world what are these signs? These signs are entries in tables held by your networking devise full of addresses, remember an intelligent networking device has this list address that it has either been told about by you the administrator or has learnt dynamically from another networking device. The networking devices like switches and routers use this information to forward your parces of data correctly out of the right interfaces towards their intended destinations. By the way your P.C, your laptop, your tablet computer all have addresses with which they use to identify themselves to receive traffic and send traffic.

Instructor Led Cisco CCNA & CCNP Classroom and self paced study courses.

Intro to Networking

What is a network?

How would you answer this? What would you say if you were asked to explain “what is a network” in your job interview?

To put it simply a network is a collection of devices that move parcels of information from A to B.

Do not think of a network as any more complex than this statement and you will find that networking is a complete doddle, I promise.

Of course there will be those who will think that I am trying to down play the sometimes complex nature of networks, sure, yes networks can be complex but only in the sense that there may be many components involved, but If you take each individual component and examine it as a single standalone function or item then the whole will make more sense and will appear far less complex. Than before. It’s a case of study each tree to understand the forest.

Hubs and the physical layer, the very basics toe in the water time

In the beginning the way in which computers were connected to one another was very primitive compared indeed to today’s methods.

Image a long electrical wire like a clothes line with shorter lengths of wires hanging down, at the end of each of these hanging wires you would find that there was a computer attached. That’s it, that’s the way the earliest networks were put together, they were very simple and very crude, but since that was the only method to connect computers together that was fine.

At some point someone looked at this “clothes line” network and decided it was too messy having this long length of cable snaking around the building so they decided to get a cardboard box and stuff the cable into the box with holes in the box to run the cables down to the computers, so now instead of a network which resembles a centipede we now have a network which resembles a box with lots of wires coming out of it, like in the picture below.

This type of network is referred to as a “Star Topology”

If you were to remove the box you would see only the wires, it would resemble the picture below

So in effect the “clothes line” network and the “box” network are one and the same, the major difference is that the “box” is more than a box it has the ability to amplify any data it receives and repeat the data out to all connected computers, this box has a name, and it is called a hub.

Hubs are very simple devices, they were used to connect multiple machines in an office network. One machine has something to say it will parcel it up into a little package called a “Frame” places it onto the wire, and once on the wire all connected “live” computers will receive a copy of the “Frame” even if they are not the intended recipient

When I started Commsupport in the famous garden shed at the end of my garden on that cold January in 2007 many people used to call and ask how I could offer courses at such low prices when every other training company was charging thousands for the same course, well since that fateful January I have proved that a great course can be delivered despite the low cost much to the distress of my competitors.

With almost 4000 students having attended our courses, we have proved that I and my gang of elves at Commsupport have built a company which is a very hard act to beat let alone follow.
Some would say why didn’t I just tow the line and offer the same courses at the same prices and make a killing. The truth is I could have easily done that, but I have always been the sort of person that loves a good scrap and that’s exactly how we like to view ourselves, we’re the underdogs, friday night bar room brawlers, above all we want to be different from the legions of sharks out there. It is this view that has been the driving force behind the idea that we have to be the first to innovate new ideas and concepts, it really annoys the market but benefits our students.

The first, last and every moment in between my mind is occupied by the single thought of how to improve the experience that my students have with us. Ultimately it is you the student which puts the bread on our table and roof over our heads, this is a sentiment held by all of us at Commsupport, it’s only logical you get the best course possible, to date I think we have done damn good job.

1. BREAK/FIX tm We’re the first school in the country to make it part of every CCNA course to run continual break/fix sessions in the classroom after the labs (The 1st and the best), before Commsupport no other company did this, it has proved so popular that students come to us for that alone.

2. FREE TRAINING DAYS: We’re the first school in the country to offer free one day training sessions to help people to decide if networking was for them before parting with any money.

3. FULLY KITTED OUT LABS: We’re the first school in the country to offer 6 pieces of lab equipment per student on the CCNA courses.

4. LIVE ON-LINE INSTRUCTOR LED CLASSES We’re the first school in Europe to offer line on-line Webinars so that our students can keep as current as possible
5. CUSTOM MADE VIDEOS We’re the FIRST school in Europe to have custom made all of our own complete set of CCNA videos, we don’t cheat you by giving you “Free Videos” out the back of a cheap CCNA manual

6. FREE WORK EXPERIENCE IN OUR CLASSROOMS: We were the first school in the country to offer work experience in the way of the LABRAT so that anyone could gain experience in our classrooms as assistants.

7. UNLIMITED FREE RETRAINING: We were the first to offer complete free re-training with no time limit, no having to take exam first conditions, free re-training means free retraining without the strings attached. we do ask that you attend the whole of the original course

8. FULLY KITTED OUT LABS IN THE PROFESSIONAL COURSES: We were the first school in the country to offer a complete lab per student on the CCNP courses.

9. We’re the first school to offer all of the Advanced CCNA courses with one Lab per student, even though the kit cost an arm and a leg.

10. We’re the first school to RECORD all of our classes for the student to take away with them at the end of the course