What is Big Data ? How Facebook,Google,Instagram Manages Big Data?

What are the problems Faced by this Companies While Dealing with Big Data

Big data is a term that describes the large volume of data — both structured and unstructured — that inundates a business on a day-to-day basis. But it’s not the amount of data that’s important. It’s what organizations do with the data that matters. Big data can be analyzed for insights that lead to better decisions and strategic business moves.
The term “big data” refers to data that is so large, fast or complex that it’s difficult or impossible to process using traditional methods. The act of accessing and storing large amounts of information for analytics has been around a long time. But the concept of big data gained momentum in the early 2000s when industry analyst Doug Laney articulated the now-mainstream definition of big data as the three V’s:

Volume: Organizations collect data from a variety of sources, including business transactions, smart (IoT) devices, industrial equipment, videos, social media and more. In the past, storing it would have been a problem — but cheaper storage on platforms like data lakes and Hadoop have eased the burden.

Velocity: With the growth in the Internet of Things, data streams in to businesses at an unprecedented speed and must be handled in a timely manner. RFID tags, sensors and smart meters are driving the need to deal with these torrents of data in near-real time.

Variety: Data comes in all types of formats — from structured, numeric data in traditional databases to unstructured text documents, emails, videos, audios, stock ticker data and financial transactions.

Star Performers Working Behind Facebook’s Big Data
There is a combined workforce of people and technology constantly working behind the successful implementation of this platform. Though the platform is continuously being enriched, below are the prime technological aspects:

“Facebook runs the world’s largest Hadoop cluster” says Jay Parikh, Vice President Infrastructure Engineering, Facebook.

Basically, Facebook runs the biggest Hadoop cluster that goes beyond 4,000 machines and storing more than hundreds of millions of gigabytes. This extensive cluster provides some key abilities to developers:

The developers can freely write map-reduce programs in any language.
SQL has been integrated to process extensive data sets, as most of the data in Hadoop’s file system are in table format. Hence, it becomes easily accessible to the developers with small subsets of SQL.
Hadoop provides a common infrastructure for Facebook with efficiency and reliability. Beginning with searching, log processing, recommendation system, and data warehousing, to video and image analysis, Hadoop is empowering this social networking platform in each and every way possible. Facebook developed its first user-facing application, Facebook Messenger, based on Hadoop database, i.e., Apache HBase, which has a layered architecture that supports plethora of messages in a single day.

With a huge amount of unstructured data coming across each day, Facebook slowly realized that it needs a platform to speed up the entire analysis part. That’s when it developed Scuba, which could help the Hadoop developers dive into the massive data sets and carry on ad-hoc analyses in real-time.

Facebook was not initially prepared to run across multiple data centers and a single break-down could cause the entire platform to crash. Scuba, another Big data platform, allows the developers to store bulk in-memory data, which speeds up the informational analysis. It implements small software agents that collect the data from multiple data centers and compresses it into the log data format. Now this compressed log data gets compressed by Scuba into the memory systems which are instantly accessible.

According to Jay Parikh, “Scuba gives us this very dynamic view into how our infrastructure is doing — how our servers are doing, how our network is doing, how the different software systems are interacting.”

“The amount of data to be stored, the rate of growth of the data, and the requirement to serve it within strict SLAs made it very apparent that a new storage solution was absolutely essential.”
- Avinash Lakshman, Search Team, Facebook
The traditional data storage started lagging behind when Facebook’s search team discovered an Inbox Search problem. The developers were facing issues in storing the reverse indices of messages sent and received by the users. The challenge was to develop a new storage solution that could solve the Inbox Search Problem and similar problems in the future. That is when Prashant Malik and Avinash Lakshman started developing Cassandra.

The objective was to develop a distributed storage system dedicated to managing a large amount of structured data across multiple commodity servers without failing once.

After Yahoo implemented Hadoop for its search engine, Facebook thought about empowering the data scientists so that they could store a larger amount of data in the Oracle data warehouse. Hence, Hive came into existence. This tool improved the query capability of Hadoop by using a subset of SQL and soon gained popularity in the unstructured world. Today almost thousands of jobs are run using this system to process a range of applications quickly.

Hadoop wasn’t designed to run across multiple facilities. Typically, because it requires such heavy communication between servers, clusters are limited to a single data center.

Initially when Facebook implemented Hadoop, it was not designed to run across multiple data centers. And that’s when the requirement to develop Prism was felt by the team of Facebook. Prism is a platform which brings out many namespaces instead of the single one governed by the Hadoop. This in turn helps to develop many logical clusters.

The ‘Scary’ Seven: big data challenges and ways to solve them
Challenge #1: Insufficient understanding and acceptance of big data
Oftentimes, companies fail to know even the basics: what big data actually is, what its benefits are, what infrastructure is needed, etc. Without a clear understanding, a big data adoption project risks to be doomed to failure. Companies may waste lots of time and resources on things they don’t even know how to use.

And if employees don’t understand big data’s value and/or don’t want to change the existing processes for the sake of its adoption, they can resist it and impede the company’s progress.


Big data, being a huge change for a company, should be accepted by top management first and then down the ladder. To ensure big data understanding and acceptance at all levels, IT departments need to organize numerous trainings and workshops.

To see to big data acceptance even more, the implementation and use of the new big data solution need to be monitored and controlled. However, top management should not overdo with control because it may have an adverse effect.

Challenge #2: Confusing variety of big data technologies
Variety of big data technologies

It can be easy to get lost in the variety of big data technologies now available on the market. Do you need Spark or would the speeds of Hadoop MapReduce be enough? Is it better to store data in Cassandra or HBase? Finding the answers can be tricky. And it’s even easier to choose poorly, if you are exploring the ocean of technological opportunities without a clear view of what you need.


If you are new to the world of big data, trying to seek professional help would be the right way to go. You could hire an expert or turn to a vendor for big data consulting. In both cases, with joint efforts, you’ll be able to work out a strategy and, based on that, choose the needed technology stack.

Challenge #3: Paying loads of money
Big data on-premises vs. in-cloud costs

Big data adoption projects entail lots of expenses. If you opt for an on-premises solution, you’ll have to mind the costs of new hardware, new hires (administrators and developers), electricity and so on. Plus: although the needed frameworks are open-source, you’ll still need to pay for the development, setup, configuration and maintenance of new software.

If you decide on a cloud-based big data solution, you’ll still need to hire staff (as above) and pay for cloud services, big data solution development as well as setup and maintenance of needed frameworks.

Moreover, in both cases, you’ll need to allow for future expansions to avoid big data growth getting out of hand and costing you a fortune.


The particular salvation of your company’s wallet will depend on your company’s specific technological needs and business goals. For instance, companies who want flexibility benefit from cloud. While companies with extremely harsh security requirements go on-premises.

There are also hybrid solutions when parts of data are stored and processed in cloud and parts — on-premises, which can also be cost-effective. And resorting to data lakes or algorithm optimizations (if done properly) can also save money:

Data lakes can provide cheap storage opportunities for the data you don’t need to analyze at the moment.
Optimized algorithms, in their turn, can reduce computing power consumption by 5 to 100 times. Or even more.
All in all, the key to solving this challenge is properly analyzing your needs and choosing a corresponding course of action.

Challenge #4: Complexity of managing data quality
Data from diverse sources

Sooner or later, you’ll run into the problem of data integration, since the data you need to analyze comes from diverse sources in a variety of different formats. For instance, ecommerce companies need to analyze data from website logs, call-centers, competitors’ website ‘scans’ and social media. Data formats will obviously differ, and matching them can be problematic. For example, your solution has to know that skis named SALOMON QST 92 17/18, Salomon QST 92 2017–18 and Salomon QST 92 Skis 2018 are the same thing, while companies ScienceSoft and Sciencesoft are not.

Unreliable data

Nobody is hiding the fact that big data isn’t 100% accurate. And all in all, it’s not that critical. But it doesn’t mean that you shouldn’t at all control how reliable your data is. Not only can it contain wrong information, but also duplicate itself, as well as contain contradictions. And it’s unlikely that data of extremely inferior quality can bring any useful insights or shiny opportunities to your precision-demanding business tasks.


Big data quality

There is a whole bunch of techniques dedicated to cleansing data. But first things first. Your big data needs to have a proper model. Only after creating that, you can go ahead and do other things, like:

Compare data to the single point of truth (for instance, compare variants of addresses to their spellings in the postal system database).
Match records and merge them, if they relate to the same entity.
But mind that big data is never 100% accurate. You have to know it and deal with it, which is something this article on big data quality can help you with.

Challenge #5: Dangerous big data security holes
Security challenges of big data are quite a vast issue that deserves a whole other article dedicated to the topic. But let’s look at the problem on a larger scale.

Quite often, big data adoption projects put security off till later stages. And, frankly speaking, this is not too much of a smart move. Big data technologies do evolve, but their security features are still neglected, since it’s hoped that security will be granted on the application level. And what do we get? Both times (with technology advancement and project implementation) big data security just gets cast aside.


The precaution against your possible big data security challenges is putting security first. It is particularly important at the stage of designing your solution’s architecture. Because if you don’t get along with big data security from the very start, it’ll bite you when you least expect it.

Challenge #6: Tricky process of converting big data into valuable insights
Valuable insights with big data

Here’s an example: your super-cool big data analytics looks at what item pairs people buy (say, a needle and thread) solely based on your historical data about customer behavior. Meanwhile, on Instagram, a certain soccer player posts his new look, and the two characteristic things he’s wearing are white Nike sneakers and a beige cap. He looks good in them, and people who see that want to look this way too. Thus, they rush to buy a similar pair of sneakers and a similar cap. But in your store, you have only the sneakers. As a result, you lose revenue and maybe some loyal customers.


The reason that you failed to have the needed items in stock is that your big data tool doesn’t analyze data from social networks or competitor’s web stores. While your rival’s big data among other things does note trends in social media in near-real time. And their shop has both items and even offers a 15% discount if you buy both.

The idea here is that you need to create a proper system of factors and data sources, whose analysis will bring the needed insights, and ensure that nothing falls out of scope. Such a system should often include external sources, even if it may be difficult to obtain and analyze external data.

Challenge #7: Troubles of upscaling
The most typical feature of big data is its dramatic ability to grow. And one of the most serious challenges of big data is associated exactly with this.

Your solution’s design may be thought through and adjusted to upscaling with no extra efforts. But the real problem isn’t the actual process of introducing new processing and storing capacities. It lies in the complexity of scaling up so, that your system’s performance doesn’t decline and you stay within budget.


The first and foremost precaution for challenges like this is a decent architecture of your big data solution. As long as your big data solution can boast such a thing, less problems are likely to occur later. Another highly important thing to do is designing your big data algorithms while keeping future upscaling in mind.

But besides that, you also need to plan for your system’s maintenance and support so that any changes related to data growth are properly attended to. And on top of that, holding systematic performance audits can help identify weak spots and timely address them.