Trends and predictions for cloud systems
2020, and partly also 2021, is a special period due to the pandemic and accelerated virtualization and digitization of many aspects of our lives. The shift to remote learning and work, as well as even greater activity of online business and services, has led to a dramatic increase in resource needs for IT infrastructure and systems.
According to a report prepared by Flexera (Flexera 2021 State of the Cloud Report), ninety percent of companies said their usage of cloud infrastructure and services had been higher than originally planned for 2020. This has been illustrated in the diagram below.
Organizations continue to rapidly increase their cloud spending. At the same time, many companies report that they have difficulty accurately forecasting rapidly rising cloud costs. Survey respondents said their spending on the public cloud was over budget by an average of 24 percent. This means that cloud competencies and the ability to properly forecast and optimize costs are currently one of the greatest challenges for IT departments.
Nowadays companies most often use a hybrid strategy, i.e. they transfer only part of their resources and systems to public clouds, leaving some applications in their own data centers, e. g. due to security policy or the lack of adequate resources for the simultaneous migration of all systems.
AWS, Azure and Google Cloud remain the leaders of public clouds. Among the top three, the largest increase in 2020 was recorded by Google (49% of running systems)
Respondents estimate that their organizations waste about 30 percent of their spend on the cloud. Sixty percent of organizations plan to optimize the current use of the cloud.
Organizations are increasingly using not only cloud infrastructure but also many services offered by public cloud providers. The three most popular of these are
data warehouses, relational databases, and containerization. Increasing interest in the use of containers is aimed at accelerating the implementation of systems and increasing their scalability. You can read about the benefits of containerization in one of our next articles on cloud migrations.
Migration of infrastructure and on-premise systems to the cloud
Due to the presented, existing trends, more and more companies are more boldly deciding to transfer infrastructure and on-premise systems to public clouds.
So it’s time to answer the question: what are the real benefits of migrating infrastructure and systems to the cloud?
Flexibility and cost optimization
Basing the systems on the cloud enables the solution to be scaled depending on the real needs of the application. With proper configuration, it is possible to adapt to the jumps of this demand and not need to use infrastructure that is prepared for the highest system load.
To illustrate this benefit, imagine an online system for selling tickets to concerts of musical bands. When opening an offer for the purchase of extremely popular bands, we can certainly expect a sharp increase in the number of inquiries from users, and thus an increased demand for resources. However, after some time (e.g. when all tickets are sold out), the system load will return to the average value. In the case of own infrastructure, preparing for this most burdensome use case would involve the need to purchase and maintain expensive resources. In the case of the public cloud, we rent the strongest resources only when we really need them.
Tolerance to errors and failures
By handing over the infrastructure management to the cloud provider, we take the responsibility off from ourselves for the ongoing management and monitoring of the equipment. Administrative activities such as data backup are also performed automatically for us. Thus, restoring the system to life after a failure is then less resource and time-intensive. More and more cloud systems are also built on the basis of the fault-tolerant paradigm, which ensures automatic detection of service failures and their automatic recovery (if possible).
Access to the latest technologies and services
Observing the largest cloud service providers, you can see their continuous development both in terms of the quality of virtual resources and cloud services provided, from efficient databases to elements of artificial intelligence such as machine learning.
Regardless of the industry, almost every company has necessarily become a technology organization, heavily dependent on IT. If we’re limited to on-premises servers, we’re likely giving our competitors an edge when it comes to scalability and being resilient to future technology changes.
One of the main reasons organizations decide to migrate to the cloud is to accelerate growth and increase business agility. In the world of your own infrastructure (on-premise), adding new resources can take weeks due to the entire process of purchasing, and then installing and configuring the equipment in your own computing centers. If you add the time needed to install the appropriate software and its configuration, such a solution is rarely optimal for business.
More and more businesses are run globally. This makes it necessary to ensure sufficiently fast reaction of the systems for users from many different parts of the world. Having one or even several computing centers in a limited territory, we are almost never able to provide equally fast access to all users regardless of their geographic location.
The answer to these needs is the public cloud. Public cloud providers have their hubs scattered across almost all continents. For example, Google Cloud provides fast and reliable connections for its users from data centers in 20 regions, spanning over 200 countries.
From the very beginning, when the public cloud offer appeared, the issue of the security of data stored in such clouds raised a lot of controversies. Many people are skeptical as to why their critical business data would leave company-owned equipment. They fear that criminals may hack into their cloud system, thereby stealing and, for example, publishing sensitive data. Similar concerns are related to access to data by foreign governmental and administrative units when data is stored outside one’s own country.
One way to mitigate this problem is to use strong encryption to keep your data safe. Even if the data as such is stolen, a cryptographic key is needed to decrypt it.
According to the Gartner report, the real security problem does not lie only in the cloud service, but in the people and applications interacting with this infrastructure.
Below are some of Gartner’s predictions regarding cloud security:
By 2025, 90% of organizations that do not control public cloud use will be mis-sharing sensitive data.
Cloud-related security policies and policies tend not to keep pace with its use. This leaves most organizations to deal with a large amount of uncontrolled use of the public cloud, creating unnecessary exposure to risk.
Until 2024, most enterprises will still face the problem of adequately measuring cloud security risks.
Questions about the security of public cloud services are legitimate, but overestimating the risks associated with the cloud can potentially result in missed opportunities. While in the past companies have tended to overestimate cloud risk, there has been a recent shift – many organizations now underestimate cloud risk. This could be as harmful as overestimating the risk.
By 2025, 99% of cloud security failures will be customer-driven.
In summary, organizations must enforce cloud policies and create appropriate cloud interaction rules to prevent critical security flaws. It is not the cloud that is the problem itself, but the way it is used.
Summary of part one
Infrastructure and system migration is not a binary process. Organizations can start experimenting with the cloud by moving only parts of sensitive systems and data to it. As the presented reports show, such a hybrid strategy is very popular nowadays. The most common cloud approach is to combine multiple public and private clouds. You also do not have to decide on a single cloud service provider.
For an increasing number of organizations migrating to the cloud is becoming a long-term, continuous evolution because projects, priorities and requirements for resources and services also evolve.
Reports also show that organizations are becoming bolder in putting even sensitive data in the cloud. The two biggest challenges facing organizations are optimizing cloud costs and ensuring high security of data and applications placed in the cloud.