Troubleshooting and the Future of Networking

Computer networking can be an incredibly complicated business, and sometimes things just don't work properly. Functionalities are known as error detection and error recovery. Error detection is the ability of a program to determine that something went wrong. Error recovery is the ability of a protocol or program to attempt to fix it.

Cyclical redundancy checks make sure the correct data is received by the receiving end; else, the data is discarded. Even with all of these safety features, errors still often occur. You can use tools from Windows, Linux, and macOS. 

The most common issue is the inability to establish a connection to something. Being able to diagnose connectivity issues is an important part of network troubleshooting. You can use troubleshooting tools to resolve these issues.

When a network error occurs, the device that detects it needs to communicate to the source of the traffic (router can't reach the destination, TTL expired, etc.). ICMP (Internet Control Message Protocol) is used to communicate these issues. 

There are cloud clients and cloud storage and cloud servers too and the cloud is the future, and IT support specialists need to keep up with the latest innovations in tech.

The cloud is just a concept. Cloud computing is a technological approach where resources are provisioned in a shareable way so that lots of users get what they need when they need it. It leans heavily on the idea that companies provide services using these resources. Cloud computing technologies rely on hardware virtualization, where physical and logical machines are abstracted away from each other. With virtualization, a single physical machine (host) can run many virtual instances (guests). 

Hardware virtualization platforms employ a hypervisor, a piece of software that runs/manages virtual machines while providing a platform that's indistinguishable from the actual hardware. A single physical computer can host many virtual instances. The cloud takes this concept one step further. A huge cluster of machines got a system that allows you to share resources among all of these resources.

Let's say we need an email server and expect it needs 8GB of RAM. We also need a name server that barely needs any resources, but you cannot run it on the physical machine on the email server (different OSs). Finally, you have a financial database with 32GB of RAM. Then, there's a second server with the same specifications as a backup, now you have to purchase 80GB of RAM but most of the month you're using much, much less.

Let's imagine a huge collection of interconnected servers that can access virtual servers. Under this level, the company can charge you to use virtual servers, costing much less than physical servers. 

The cloud computing company also has dozens of services. You can just use their solution with a load balancer and moves machines automatically when hardware doesn't work. You just need to click a few buttons on a web browser. A public cloud is a large cluster of machines run by another company. A private cloud is used by a single large corporation and physically hosts its own premises. 

Another term, a hybrid cloud, is a term used to describe situations where companies might run things on a private cloud while trusting less sensitive servers on a public cloud. Cloud computing is a new model in computing where large clusters of machines let us use the total resources available in a better way. The cloud lets you leverage lots of existing services. 

The IANA is out of IP addresses. A 32-bit number was initially chosen for IPv4 because no one expected the exploding popularity. By the mid-1990s, it was more obvious that we were going to run out of addresses in some way. 

IPv5 was an experimental protocol that introduced the concept of connections. Even though IPv5 is a relic of history, the consensus was to not reuse the IPv5 name. The biggest difference between IPv4 and IPv6 is the number of bits in an address. IPv6 addresses are 128 bits in size. 2^128 is a 39 digit number, and undecillion, which equals around the total number of atoms that make up earth, 340,282,366,920,938,463,463,374,607,431,768,211,456. This is like if we can get every number on earth its own IP address. IPv6 addresses are just 128-bit binary numbers. IPv6 is 8 bits of 16 bits each, or 4 hexadecimal numbers:

there's a notation method for IPv6. 2000:0db8:0000:0000:0000:ff00:0012:345

Anything with 2000:0db8 has been reserved for documentation. There are 2 rules when it comes to shortening an IPv6 address.

1. You can remove any leading zeros from a group.

2. Any number of consecutive groups composed of just zeros can be replaced by 2 colons. 

2001:db8::ff00:12:3456.

With IPv6, the loopback address is 0000:0000:0000:0001or ::1. 

The IPv6 address space has several other reserved address spaces. Anything for FF00:: is used for multicast, addressing groups of hosts at once. FE80:: is used for link-local unicast, allowing for local network segment communication, and is configured upon a host's MAC address. 

The host's MAC address is run by an algorithm to turn it from a 48-bit number to a 64-bit one. It's then inserted into the address's host ID. The IPv6 is so huge there's never any need to split up into address classes, having a very simple line. The First 64 bits are the network id, and the second 64 bits are host id, 9 quintillion hosts.

IPv6 subnetting uses CIDR notation used to define a subnet mask against the network id portion of an IPv6 address. 

An IT support specialist loves networks that perform well. One of the best improvements is made to an IPv6 header. The first field is the version field, define what header is in use. The next field is a traffic class field, an 8-bit field that defines the address in the IP datagram. The next field is Flow Label, used in conjunction with the traffic class field. Then, there's the payload length field, then the next header field. IPv6 addresses are 4x as long as IPv4, with more 1's and 0's taking longer to transmit across the link. The IPv6 header was built to be as short as possible, abstracting optional fields. The next header field defines the header after the current one. Each optional header contains a next header field, allowing for a chain of headers to be formed. Then, there's an 8-bit hop limit field (TTL field) then 128 bits for the source and destination field. Then a data payload would follow afterward. 

It's just not possible for the internet to switch to IPv6, with too much coordination. IPv6 and IPv4 need to coexist at the same time. This would allow organizations to make the transitions when they can. IPv6 specifications set aside a number of addresses correlated to an IPv4 address. any IPv6 with 80 0's with 16 1's in IPv4 space. The remaining 32 bits are just the same 32 bits of the IPv4 address it's meant to represent. More important is for IPv6 to travel over IPv4 networks. It's easier to move to IPv6 than the networks. It'll need a way to travel through the old IPv4 remnants. IPv6 tunnels consist of servers on either end of a connection, take incoming traffic, and encapsulate it within traditional datagrams, where received by another IPv6 tunnel server, passing the IPv6 traffic further along with the network. 

The concept of IPv6 tunnel broker emerged, companies that provide IPv6 endpoints for you. There are a lot of competing protocols, and it's not clear who the winner will be. It doesn't really matter which tunneling technology will dominate, the future of networking is the adoption of IPv6, and one day, we won't need any tunnels at all. 

That's it! here, we learned about the network layer, the transport, application, physical layer, networking services, connecting to the internet, and the future of networking. Networking is involved with every little bit of data across the internet.