What is Distributed Computing: Examples, Pros, and Cons Explained

Distributed computing is playing a major role these days, especially for businesses that depend on superior computing power to serve their customers. It is also increasingly being used for machine learning and artificial intelligence applications. While this model of computing can be more complex than traditional computing, it offers many benefits as we’ll discover shortly.

This concept is shaping many industries, powering valuable services with robust computing power that yield many advantages that would never have been realized. 

In the next sections, we define distributed computing, demonstrate how it works, the benefits and drawbacks.

What is distributed computing?

Simply put, distributed computing is the connection of several computer servers to form a unified network that shares computing resources. 

While each individual computer is autonomous i.e. physically separated from the rest of the computers in the network, they work together in a synchronized system where the task is divided up. It's important to note here that what is shared is computing resources but not physical resources such as memory, since the computers are physically separate.

The participating computers will exchange data through an intranet or internet via internet protocols like UDP and TCP/IP. To the eyes of the common user, distributed computing appears as one functional unit, thanks to the principle of transparency. So if you are searching for a product in an ecommerce store, you view the experience as a single smooth process and not the distributed architecture that powers this experience. In short, distributed computing allows for dividing up a task and spreading the work among multiple machines. The ultimate goal is to improve task execution.

How distributed computing works?

In distributed computing, a group of computers work together on a shared task. This can be done in a number of ways, but the most common method is to divide the task into smaller pieces, which are then assigned to different computers. 

There are several steps involved in distributed computing. First, the task is divided into smaller sub-tasks. Each sub-task is then assigned to a different computer in the network. The computers then work on their assigned tasks independently. Once they have completed their tasks, they send the results back to the central server. The server then combines all of the results to produce the final answer.

Distributed computing is often used for large-scale tasks but it is also increasingly being used for more everyday tasks such as search engines and social networking sites.

Examples of distributed computing in action

Here are some examples of where distributed computing is used. As you will discover, this concept is so common that we are all beneficiaries in one way or the other. Many of us encounter it without realizing it.

Social media

You're scrolling through your Facebook feed when a message pops up letting you know that your friend has just liked a post that you're also interested in. How does that work? Behind the scenes, Facebook is using distributed computing to quickly show you the post. This means that they've divided up the task of monitoring posts into small pieces and assigned them to different computers all over the world. When one computer finds something of interest, it sends a message to Facebook headquarters. There, they compile all of the messages from different computers to show you the post on your newsfeed.

Online banking

Online banking is a convenient way to manage your finances without having to visit a physical bank branch. But have you ever wondered how your device such as a computer or mobile phone is able to communicate with the bank's computers? The answer lies in distributed computing. In the case of online banking, your device is able to send and receive data from the bank's computers using this type of network. This allows you to check your account balance, transfer funds, and pay bills all from the comfort of your own home. So next time you log into your online banking account, remember that you're taking advantage of one of the many uses of distributed computing.

Price comparison

In online shopping, the distributed computing process is used to bring together the prices of products from different stores. It compares the prices of identical or similar products in order to find the best deal for the customer. This system can also be used to find discounts and coupons for customers. The distributed computing process speeds up this comparison by using multiple computers rather than just one computer. This allows customers to quickly compare prices and find deals.

Ride sharing

Companies like Uber and Lyft actually use a distributed system to match drivers with passengers. This system relies on thousands of individual computers (or nodes) to process requests. When you request a ride, your computer sends a signal to the nearest node, which then forwards it to the next closest node until it reaches the driver. This system is incredibly efficient and allows ride sharing companies to offer their services in hundreds of cities around the world.

Streaming services

Streaming services like Netflix, Amazon Prime Video, HBO GO,  and Hulu use distributed computing to power their video-streaming platforms. In fact you might not know it but distributed computing is used extensively by streaming services to provide a scalable and reliable experience. By using multiple computers around the world, a streaming service can provide a much larger degree of scalability and reliability than would be possible with a single computer. For example, when Netflix first launched its streaming service, it used a distributed system to handle the enormous amount of data that was being streamed. This allowed them to provide a smooth and uninterrupted experience for their users. Today, Netflix continues to use distributed systems to power its massive worldwide operations.

Search engines

Think about the last time you searched for something online on a search engine. When you typed in a query and hit «enter,» your request was sent to a massive network of computers around the globe. These computers—known as servers—processed your request and sent back the results. Distributed computing is used heavily by search engines in order to process the large volume of data that they need to index and provide search results. By distributing the search tasks across multiple machines, search engines are able to quickly index and serve results for their users.

File sharing

When you download a file from the internet, your computer is doing the work of downloading the file. But, at the same time, you're also sharing that file with other people who are downloading it as well. So, in a way, you're all distributing the workload.

Email

When you send an email, your computer is sending it to the server. But, at the same time, your computer is also receiving emails from other people. So again, you're distributing the workload.

Why distributed computing?

So why is distributed computing so important? These are the key reasons:

1. Distributed computing can make better use of resources. By using multiple computers, tasks can be completed faster, which makes better use of available resources. This can be especially important when dealing with large data sets.
2. It can provide a higher degree of fault tolerance. If one computer in a distributed system fails, the others can continue working, which reduces the risk of downtime.
3. It can improve performance. By spreading the load across multiple computers, individual machines are not overloaded and performance can be improved.
4. It can make it easier to scale up. When more computing power is needed, it is often easier to add additional computers to a distributed system than to try to upgrade a single machine.
5. It can allow for geographically dispersed teams to work together. Distributed systems make it possible for people in different locations to share data and work on tasks together.

Distributed vs. Parallel Computing

The main difference between distributed computing and parallel computing is that distributed computing uses a network of computers to divide the workload, while parallel computing uses multiple processors to complete tasks simultaneously.

To put it more succinctly, distributed computing involves breaking up a large task into smaller parts that can be processed by different computers.  Parallel computing, on the other hand, involves running multiple tasks or processes simultaneously. In distributed computing, each system has its own memory. In parallel computing the systems share memory. 

As for which type to choose depends on what you want to achieve.  Distributed computing is perfect for organizations that need to speed up the execution of large tasks. Parallel computing is great for organizations that need to run a large number of calculations. Parallel computing is usually faster than distributed computing, but it requires more powerful hardware and special software.

Distributed vs Cloud Computing

There is often confusion between distributed computing andcloud computing. However, there are some important distinctions between the two. Distributed computing involves breaking a problem down into smaller pieces that can be solved simultaneously by different computers. Cloud computing, on the other hand, is a type of online service where users can access remote servers to store or run applications. Examples of cloud computing offerings include pay as you use products such as SaaS. 

Look at it this way: distributed computing is a way of performing the same task using multiple computers, while cloud computing is a way of delivering services to the end user. One is about performing tasks, the other is about storage space and delivery of services plus data. In other words cloud computing can rely on distributed computing to fasten the delivery of cloud services whereas distributed computing can rely on cloud computing for storage functions . 

Distributed Computing Environment

The term distributed computing environment refers to the system in which the resources (CPU, memory, and disk space) are physically dispersed among the nodes of a network. This contrasts with a centralized computing environment, in which all of the resources are located at a single computer.

DCEs can provide a higher level of security than traditional systems, as each computer in the network is isolated from the others. As a result, DCEs are increasingly being used in a variety of settings, from small businesses to large enterprises.

Advantages and benefits of Distributed Computing

Here are 7 key advantages and benefits of distributed computing;

1. Cost-effective

There are many reasons why distributed computing is a cost-effective solution. For starters, it allows businesses to use the resources they already have, rather than investing in new infrastructure. Additionally, it can help reduce energy consumption and server load, making it more environmentally friendly.

2. Increased storage

Distributed computing allows for increased storage. This is because the data is not all stored in one central location, but is instead spread out among different computers.

This means that if one computer fails, the data is still accessible from the other computers. It also means that if you need more storage space, you can simply add more computers to the network.

3. Enhanced security

When data is distributed across multiple machines, it becomes much harder for hackers to gain access and steal information. This is because the data is not centralized in any one place, making it more difficult to hack into.

Additionally, by using multiple machines, you can create a more diverse and secure network. If one machine is compromised, the others will still be safe. This ensures that your data is always protected.

4. Improved performance

When tasks are distributed across multiple machines, the overall execution time is reduced. This is because each machine can work on a portion of the task at a time, and when the tasks are combined, the overall execution time is reduced.

This improved performance can be especially beneficial in situations where large amounts of data need to be processed. For example, if you're trying to analyze a large set of data, distributing the workload across multiple machines can result in a significant reduction in processing time.

5. Increased reliability

One of the biggest advantages of distributed computing is its increased reliability. By using multiple computers to complete a task, the chances of an error occurring are quite minimal. This is because if one computer fails, the task can be completed by the remaining computers.

6. Greater flexibility and scalability

With a distributed system, you can easily add or remove nodes (computers) from the network, making it easy to adapt to changing needs. You can also scale the system up or down as needed, either temporarily or permanently, ensuring that you always have the resources you need. This also makes it possible to handle larger workloads and accommodate more users without any slowdown or interruption.

This is unlike a centralized system where all of the data and processing power is located in one place, making it cumbersome to scale. 

7. Low latency

Latency is defined as the duration it takes for a packet of data to travel from one point to another. When it comes to distributed computing, low latency is a key benefit as this system can move large amounts of data within a short timeframe. The faster data can be worked on and sent back out, the faster the entire system will run.

Thanks to advances in technology, most distributed systems now have a latency of less than 100 milliseconds. This ensures that your applications will run smoothly and without any glitches. In fact, according to Stanford University, low latency is one of the key goals of distributed computing. 

Drawbacks of Distributed Computing

While distributed computing delivers many benefits as we have seen, it also has its fair share of drawbacks and disadvantages. Here is a look at some of the drawbacks of distributed computing;

1. Network transfer bottlenecks

When data is distributed across multiple nodes, it needs to be transferred between them. This can create a bottleneck if the network connection between the nodes is slow or congested.

For example, consider a company that uses a distributed computing system to process sales transactions. If the network connection between the nodes is slow, it can delay or even prevent transactions from being processed.

2. System complexity

Distributed computing can lead to increased system complexity. With multiple systems working together, it can be difficult to track and manage all of the moving parts. This can lead to system failures and data inconsistencies.

3. Standardization issues

Because there are so many different types of distributed systems, it can be difficult to find one that fits your specific needs. Additionally, each system has its own set of commands and procedures, which can make it difficult to learn and use.

4. Maintenance costs

Distributed computing can be expensive to maintain. Since the systems are spread out across multiple locations, it can be more difficult to troubleshoot and fix problems when they arise. In addition, the increased number of components also means that there is a greater potential for hardware and software failures. As a result, companies often need to allocate more resources to maintaining their distributed computing systems.

5. Potential for component failure due to human error

There's always the potential for human error when dealing with multiple components and systems. This occurs when someone who is not familiar with the system makes a mistake that leads to component failure. For example, if a programmer accidentally deletes a critical file, it could render the entire system inoperable. Another example might be if an administrator incorrectly configured the network settings, preventing users from being able to access the system. While human error can lead to component failure, there are steps that can be taken to minimize the risk. For example, developers can create comprehensive documentation and training materials to help reduce the likelihood of users making mistakes. Additionally, you can design systems with built-in checks and balances to help prevent accidental deletions or other types of mistakes.

6. Bandwidth limitations

In order for all nodes in the system to communicate with each other, they need to share a common bandwidth limit. When this limit is reached, the system becomes bottlenecked and performance will start to degrade.

This is often a problem for large-scale distributed systems, as the amount of data that needs to be transferred can quickly exceed the available bandwidth. As a result, careful planning and design is needed to ensure that all nodes in the system can still communicate effectively with each other.

7. Deploy challenges

In order for a distributed system to work effectively, all of the nodes (computers) in the network have to be configured in the same way and be able to communicate with each other. This can be a challenge for organizations with more complex IT infrastructure but inadequate skill sets among their IT staff.

Final Thoughts

As you can see the benefits of distributed computing are vast, and they continue to grow as systems call for greater computing power especially with the growth of AI. The technical side of distributed computing may not be visible to the ordinary user, but it is a powerful tool that is helping businesses to complete tasks faster and provide stellar experience for their customers. When used correctly, it can be a valuable part of your company’s technology stack.

Of course as we have seen, there are some drawbacks to this type of system that you will always have to consider in order to get it right, especially at implementation and maintenance.