Big Data

Big Data are becoming a major driver of all businesses. Why? Simple. All successful businesses are online. Not so simple and somewhat technically put, big data are large sets of information that have been produced for analytical purposes in order to find trends and patterns or associations in user behavior. In other words, big data are how companies turn existing customers into brand-loyalists who want to be there at every second and who want to stick with them all the way. Companies also take new users and give them exactly what they need and more. The thing is, is that companies cannot let this ever growing amount of data go to waste anymore. They need to do something with it, otherwise they are losing out on potential which leads to losing out on money, then their jobs, and all of a sudden the doors of the company are closed.

New inductees to the world of Big Data might be overwhelmed and understandably so. The word “big” is not used lightly when companies like Boeing bring in hundreds of terabytes of data during one flight of a Boeing 737. Knowing that now try to guess how many terabytes Facebook brings in per day.

With that being said, we all know that big data are here to stay. Big Data has a bright future. Why? Because big data are how companies grow and big data are why new companies are starting. The information they are collecting is growing exponentially and the best way to keep up with it is to grow with it. To be ahead of the curve of technology/big data.

With such large amounts of data and with different users and different creators, there are different types of databases. The main databases are relational, operational, distributed, and database warehouses. Each have their own functions. There are different databases with different functions because different companies have different uses for each.

MongoDB

MongoDB, formerly known as 10gen, is the operational database that is going to be discussed. This is not a promotional piece of content though. It is a piece that covers when and why MongoDB should not be used.

MongoDB is a great open-source database used primarily for operational purposes. Take a website for example. A user will go to a website and click on a category page, say the electronics category, and the next page will bring you back information on everything included in the electronics category. Then clicking on a product page on the category page will bring you information on that product.

This is the best use case for MongoDB. A user makes a request for a specific set of data and it will supply you with that data. Ask for data, get that data. Simple as that.

It sounds like an ecommerce website or other similar websites could be great clients of MongoDB. There are other instances where different companies could be great MongoDB users and use cases. There are large companies out there and there are some great use cases that the company shares. Some of those companies are Disney, Intuit, and MTV Network. Based on what I have just stated, do you think that MongoDB is good enough for them? Is it scalable? Final question, is it your use case?

Using MongoDB

If you are trying to create something with more intricacies and a bright future with a plan to scale, this will not be the case. Use the often cited Sarah Mei example of the Diaspora social network for instance. They thought it was going to be enough. They thought it would be enough because it was shiny, new, and supposedly easy to set up. All of those reasons are true but that does not necessarily make it the right choice. And what does easy to set up even mean?

One of the big reasons their company says it is easy setup is because you are able to get started without setting up the schema. This removes time for planning and implementation. That is great on the front end but can be deceiving. Later, as everything starts to scale, you want that schema set up.

The reason companies use it? A simple reason, companies want shiny and new. Technology is an industry based off of shiny and new and we all get it! It is almost as if you can almost define technology to the person sitting next to you as it is shiny and new and it could possibly benefit your life in some way.

Don’t Use MongoDB

Just because technology is shiny and new doesn’t make it the best product to work with. SQL databases that use similar technology, such as MySQL, have been around since the 1970s. This is mature technology that has been through a lot, has grown with the times, and has always found a reason to stick around. In order for something to stick around in the years of fast growing technology, it must mean it does something incredible, or in the very least, it completes tasks the right way on a consistent basis. If it is doing something right, the old idiom, if it ain’t broke, don’t fix it, can be used.

It is so hard to disagree with that statement. On the other hand, MongoDB is new and sure has bugs. If they haven’t been found, they will be in the same amount of time as other databases. There are bugs in MongoDB that have been found, fixed, and never heard of again. Then there are some that haven’t been found. Some of those might not be experienced by all users, many users, or even half of their users, but that does not mean that they are not there. This is known because it is a newer technology that is not strongly tested.

Due to the industry and the needs of companies, databases are put through rigorous tests. If MongoDB does not have a high level of testing and they still are not found, that shows that they are not performing strong quality assurance tests. Trusting a company with your data that is not even up to the task of meeting the rest of the industry on quality assurance testing is incredibly difficult.

Let’s move onto another point why MongoDB isn’t really what should be used. MongoDB is a database type that is functional when it comes to the day-to-day work, as expressed with an ecommerce website as an example. The thing is, is that it doesn’t let you break down the data into smaller, easier to digest and presentable sets. Users want to slice and dice their data. They want to see their data. They want to query everything. They want to see the trends, the behavior of whoever is using that data. When working with big data, it is inevitable to require analysis and querying. If you cannot see a trend, what is the point? How would it be beneficial for a company if analyzing the data can’t be completed?

The user wants their data quickly. If that cannot happen in the short amount of time that is expected, the user will become upset. With MongoDB, this will be a problem. The performance speed is already slow and will only get longer as more users start to query at the same time. When this begins to happen, their minds change. They get data somewhere else, the user has moved on, wrote a negative review, and became a loyal customer somewhere else.

It also does not let you scale vertically, only horizontally, and what horizontally means is that you have one master and as many slaves/replicas as you want. Every database that can be added in MongoDB is readable only. Only some people want that and we know that the trend of users do not want readable only. They want to have their own experience. They want to have a readable and writeable experience.

Again, if that is all that is needed, fine. Do it. But, with the future that technology is headed in, that is not what you need.

Your Goal

To put it simply, there are very few instances that selecting MongoDB over anything else is a good choice. Just a couple of key points that anyone can understand is that its data selection took twice as long as another database and it consumed more disk space. These are both big marks in the cons section when trying to find the best database.

That is not the situation here. Sometimes you want the mature technology to remain in control. The technology that is tried and true. It might cost more but in the end, being able to scale to and beyond what was originally imagined is the ultimate goal.