8 Mins Read  October 14, 2019  Cuelogic Insights

Why is Cloud Waste so Huge?


With 2019 about to end, the Gartner cloud waste forecast for the running year has been about USD 14.1 Billion. Even though the actual number could be far higher due to several reasons, yet this figure is humongous. If you think it has been a one-off event, we would like to tell you that it was USD 12.9 Billion in 2018. 

There is no stopping this phenomenon as well; we expect it to hit a whopping USD 21 Billion in 2021. Due to inefficiency and several limitations, we are wasting around 35% of our cloud spending. 

Cloud computing service providers have also validated the rising scale of cloud waste and their thesis remains that organizations or individuals accumulate more cloud space than they will ever need. Considering how we see cloud as one of the energy-saving resources, the amount that is going down the drain is more concerning. 

So what exactly is Cloud Waste, and why are the numbers sky-rocketing? In this article, we will discuss why cloud waste is so vast and what we can do to cut down the number. Let us delve deeper.

What is Cloud Waste?

Like any regular waste, Cloud Waste takes place when you acquire more cloud than you can utilize.

There are several forms of cloud waste –

  • Many data centers are guilty of letting resources run full-time even though they remain idle for a part of the day. These can be for development, demo, test, or training environments that forget to switch the cloud off after their work is complete.

We cannot blame any single party for this fault. There are several PaaS (Platform as a Service) providers who are also at fault along with the general cloud users. These Paas providers have failed to give the option of switching their services off when not in use. It not only increases the cost but also leads to a massive chunk of cloud spend going down as waste.

  • Development procedures are one of the main reasons behind cloud waste. Developers, while forecasting how much of a cloud instance is required, often choose a bigger-than-required instance. It ensures that there is no-huddle to prevent smooth operation.

We also tend to think that bigger is always better and end up getting more cores than required. It can be associated with uncertainty or inexperience, but the result is the same. More often than not, the databases remain over-provisioned than their requirements, and the extra storage remains unused. Is there anyone to blame for this? Well, none that we can think of!

What is Cloud Spending?

Cloud spending is the cumulative expense of cloud infrastructure by an organization. Most of the cloud services work on a metered basis, and we pay as per usage. Typically most providers charge on an hourly basis to all the clients, and the total amount is known as cloud spending. 

For the year 2019, Gartner has suggested that cloud spending will go up by 3.2 percent from the last year. The figure is a staggering USD 3.8 Trillion as compared to USD 2.7 Trillion in the previous year. IaaS (Infrastructure as a Service) is leading the troops with a growth of 27.6 percent YoY basis. 

We are seeing a new organization or two stepping into this lucrative market every day, which has further hastened the increase. Not to mention the tremendous rise in resource-heavy tasks, cloud spending is going to rise soon.

Who are the ones suffering due to it?

The suffering due to cloud waste is like a chain, and it hurts the consumers most. The implications are multi-dimensional, and there are several ways this phenomenon hurts. The most visible impact is the lowering of returns. Due to entities being unable to utilize the cloud to its brim, they have a lower return on assets.

The lower ROA also comes down to lower Net Income, and this affects the dividends payable to the consumers. We are at the receiving end and receive a smaller dividend due to cloud waste. If you think that the providers are at sea and it doesn’t impact them, it affects their business as well. Azure and AWS (Amazon Web Services) have introduced Auto-Scaling groups and several other options to ensure that both parties benefit due to the reduction in cloud waste generation. 

Public Cloud providers earn the maximum when they have oversubscribing consumers. It means that they benefit from more and more people adopting their existing data centers. As soon as cloud waste creeps in, they are forced to build newer, more expensive products that reduce their profitability. 

Reasons for Increasing Cloud Waste

The benefits of cloud computing are endless, but that doesn’t take away the perils of increasing cloud waste. With more and more organizations preferring cloud-based services over the traditional ones, the future seems worrying. 

Several reasons add up to the enormous cloud waste – 

Lack of General Understanding

Like a kid doesn’t understand the perils of wasting water, we adults don’t know how cloud waste can hurt us. As per Syncsort, there are around 62 percent of people report higher than anticipated costs when it comes to cloud services.

More often than not, the billed amount for clouds is higher than expected. It is because we lack the understanding of how to manage cloud resources. A friend of mine uploads everything regarding his business on the cloud and is unapologetic of that fact. We must know that the cloud is like a reserve; we keep what is necessary and let the rest be away.

Most of us pay for how much we use, and this is another reason why caution must be adopted. People often forget that there are strategies required to be in place about how you manage your resources. You must remember to be adaptive and organized to use the cloud efficiently. 

Just because it is easier to upload doesn’t mean that you will have to upload it! 

Over Provisioned Resources

If you know, there are several instances, and all of them are for specific purposes. Owing to different purposes, they cost differently and come in different sizes. Organizations have tended to hoard unnecessary instances which find no use in the ongoing project. Not only it causes storage issues, but it also increases the cost of the project.

Another example is the hoarding of the cloud itself. As per recent data, we spend around half of the total spend on instances. Now you may think what is wrong with spending so much on instances. I would have agreed to it had it been for the right sizes. As per reports, around 40 percent of these are up to two sizes bigger than required. Why take your helicopter for the shop in the next lane? 

If we convert this into figures, around USD 5.3 Billion are spent every year on oversized instances. Why buy that we don’t need? Instead, let the needy have it and be happy about it. With or without the figure, the amount is staggering and must ring a bell in our head.

Unscheduled VMs

Most of the organizations are guilty of letting their VMs (Virtual Machines) run throughout the clock. We are not complaining when it is necessary to do so. But more often than not, these machines rack up cloud waste by doing nothing.

People are yet to understand cloud computing completely, and the inexperience shows. Cloud is expensive or cheaper depending on regions. Several organizations have clouds all over the world but fail to utilize the cheapest ones in the best possible manner. 

It leads to the creation of cloud waste and latency issues. If you utilize a server in India for any operation in the USA, you are bound to incur a higher cost. If we don’t step up our game today, the prices of these unused units will go up as time passes. It would be great if you have the habit of shutting down those unused VMs after their work is complete. The most surprising part is that most such organizations don’t even care to take any significant step to improve it.

Idle Resources

Idle resources are like those flats which remain unused, but we pay rent regularly. Even though we aren’t using the extra funds, we bear the cost. Cloud is by no means cheap, and the figures will be difficult for you to fathom.

Idle resources are mainly an occurrence in development centers, where testing, staging, and other procedures take place. Occasionally it takes place in production, but chances of it are rare. The cost of idle resources form a significant part of Cloud Waste being generated globally; they stand at USD 8.8 billion every year.

Why buy a parking space for a truck when you have a scooter only?

Unused Capacity

Cloud works differently, and most organizations feel it works exactly like their old data centers. When they shift to the cloud, they tend to pick up the same storage as they had earlier or higher. 

As per a report by Koomey, around 80 percent of in-house data centers occupy more server capacity than required. It not only translates to higher costs for the organization, but the service providers are also suffering. The same report also states that after migrating to the cloud, around 36 percent of companies pay more for clouds than required.

When we have a pay-per-use scenario for the cloud services that we take, why can we not improve our efficiency and utilize it as per our needs? Providers have long been asking companies to handle cloud services better. Providers like AWS and Azure always push updates that improve cloud capacity management. 

Even though the cloud is not a fossil resource, but it has its limitations too. So why not use it efficiently so that others can use it without having to bear for your inefficiencies. It is time that we realize that the cloud should not have the same fate as other non-renewable energy sources.

The problem of too many Snapshots

Snapshots can be created in three ways in the cloud environment – VMware’s way, copy-on-write, and redirect-on-write. As soon as VMware takes a snapshot, there is a new location created for the new writes. To our misery, all reads will scan both the existing location and the original location before it supplies the current block versions.

How these snapshots operate is unknown to most of its users, and they treat it like the regular ones. They will take snapshots but fail to delete them when they are no longer required.

Let us examine the case of an OS upgrade. Users will take a snapshot before upgrading to a newer version. It lets them rollback to the previous one if required. After testing the latest version for days or even months, they forget to delete the old snapshot which keeps on running. It merely creates unwanted waste and leads to more resource utilization.

Misusing On-Demand Services

The 21st Century people have seen the growth of on-demand services across various platforms. With cloud providers providing on-demand services to its clients, a new problem has surfaced. Many organizations find such services lucrative and easy to deploy. So they tend to utilize more of it without accounting for its start and stop expenses.

On-demand services are for emergencies and not for general use. They cost much more than the reserved resources and even spot resources. It is time we pay heed to these costs and stop deploying unwanted resources just for the sake of doing it. If it is necessary, you must do it. Why not use the available resources at your disposal instead? 

Orphaned Resources

Orphaned resources are those resources that are no longer useful. It mostly occurs when you have turned off your computing machine but forgets to switch off the storage. The storage will have no other work to do, and it will keep on piling the cost. It will lead to you ending up paying significantly more than what you should have.

It is not only increasing your costs, but it is also leading to cloud waste. To curb such happenings, we must be careful enough and take extra care to ensure that no additional resource is exhausting in the background.

Inefficient Pricing

In general, you will agree that it is easier for you to choose if there are two options. What if you have thousands of options and you are asked to choose one? Your mind will go on a mental ride, and you will have a hard time getting the right one for yourself.

It is what cloud users suffer mostly. Even though most providers charge on an hourly basis, you will get different prices without any significant change in offerings. There are thousands of such options to choose from. Experienced people may know what they are looking for and will get the most appropriate one.

What about the majority who stand inexperienced and are in there just because their competitors are shifting to cloud? These organizations end up choosing the wrong instance sizes and add-ons that are much more than they need. They also lead to the creation of cloud waste and higher costs. 


Human beings are guilty of driving several resources to near-extinction, primarily due to their carelessness. We are spending, instead of wasting billions every year without even realizing our mistakes. What if we have a better cloud management strategy? What if we make sure to switch off the machines when not in use? 

Having a suitable working strategy will reduce your costs and also lead to a decline in the amount that is going down the drain. The price may not be significant when you see it individually, but the cumulative figures are fear-inducing. We are repeating the same mistake that our forefathers did, taking a resource for granted. It may seem widely available today, but what if once it gains enough traction? We are the ones who are going to suffer and be ready for the day. It is not too far off.

Recommended Content

Go Back to Main Page