You turn on your computer, and connect to the WiFi. You open your browser and remember that you have to pay your credit card. You search for the bank’s name and click the first link you see, because the first one is always the correct one. You submit the credit card information and then get an email confirming your payment. But, how are you sure you submitted your information in the bank’s page? Or, how are you sure no one gained access to it?
When you turned on your computer, connected to the WiFi, opened the browser and started searching, your computer started sending little pieces of information in what are called packets. This packets contain the message or information you want to send, but also who sent it and where it is going. We identify the computers communicating using primarily IP Addresses, which are numbers that specify your internet address. However, how this computers communicate, validate each other, and understand the messages varies greatly by the network protocol you are using.
But, why do we as Data Scientists care about how computers communicate? Over the last years there has been a surge in the amount and complexity of data we analyze. This has caused that many projects need to be developed in what is called a ‘cloud’, which is really someone’s else computers. Depending on the project you are working you might need to know which network protocol to use and how to use it to pull data from the cloud or from a database. In this blog post I’m going to mention briefly a couple of protocols and some of their usages. I divided the protocols by whether they are encrypted or not. (We’ll go back to this shortly)
File Transfer Protocol, or FTP, as it’s commonly known, is a client-server protocol, meaning that there is a client (your computer) and there is a server (someone’s else computer) communicating. These protocols tend to communicate through a specific port. Imagine this as if you are calling a company, but already know the extension of the person you want to communicate with. When the client initiates a connection, the default in FTP is to use port 21. This protocol is fairly common in web development.
Some known GUI Clients for FTP are:
Hypertext Transfer Protocol, or HTTP, is also a client-server protocol. HTTP works with a series of methods that define how the communication between the client and the server is going to work. These are:
One way in which organizations share data is through Application Programming Interfaces. These API’s as they are commonly known, are commonly written to communicate through HTTP. If you need to evaluate HTTP resources you can do so through the following methods:
httr package.requests package.The two methods I have mentioned are non-encrypted methods. But what does encryption means and why is it useful? According to the Oxford Dictionaries it is defined as
“The process of converting information or data into a code, especially to prevent unauthorized access.”
This ‘process of converting information or data into a code’ is done by encryption algorithms. Encryption algorithms become necessary when dealing with sensitive or confidential information as it protects it from unauthorized access. They do this in multiple ways and levels, here I discuss TLS and SSH, which are two of the most common encrypted protocols.
Transport Layer Security is an encrypted client-server protocol based on the sharing of some documents called certificates. These are documents validated by an independent authority which validates that the website is owned by who it says it’s owned. Once the client tries to connect to the server, the server sends part of its certificate to the client. This is known as the public key. This public key and a private key which is stored on the server form a cryptographic key which encrypts all the communication between the server and the client. This is now the standard for all web pages in the internet. TLS connections are represented by HTTPS, standing for HTTP Secure and are done by default through port 403.
Secure Shell is also an encrypted client-server protocol based on the sharing of documents called SSH keys. Similar to in Transport Layer Security, these keys are used to form the cryptographic key that masks the communications. However, SSH is mostly used to gain remote access to other computers. It communicates through port 22 and is mostly used from the terminal. If you need to gain access to a server, it will probably be using SSH.
In general, we can see the application of many of these protocols in our career as data scientists. In this blog post I covered only a small amount of the depth of computer networks, if you are interested, I encourage to look for more resources or write to me.