It's no secret that cloud usage continues to rise, with 84% of enterprises utilizing a multi-cloud strategy. But with this rise also comes a greater level of complexity in managing hybrid multi-cloud environments, causing headaches for firms. Most enterprises are not able to hire enough network experts, yet they are being forced to run complex networks to connect to their many cloud services. Thus, enterprises are leaning heavily on their data center and network providers to help in removing complexity in their cloud and network architectures, ultimately allowing these enterprises to concentrate their precious resources on their primary business goals.
Many of the traditional conversations about enterprise IT architecture are focused on the server hardware with key considerations like storage, memory, and I/O processing. That being said, the cloud market has matured so much that any of these resources can now be distributed on demand. Furthermore, the costs associated with most of these resources continue to decline as they services become ubiquitous. For example, the average cost of an entry-level public cloud instance saw a 66% decline in the two years leading up to 2016.
While that's great news and it means that more organizations will have the resources to access and use cloud services, the evolution of the cloud itself means that the challenges associated with using it have also evolved. For example, many might not think that performance can become an issue because businesses can just scale to meet the demands of their workloads at cheaper costs. Yet, low performance remains a central concern-among 49 percent of organizations are considering moving analytics to the cloud.
So, What Happened to Cloud Performance?
In today's cloud world, scaling server resources like storage and processing power is easier than it has ever been. But scaling cloud resources isn't the only factor that contributes to cloud performance. Other elements like security or network architecture aren't scalable without reconsidering cloud architecture from the perspective of a comprehensive multi- and hybrid-cloud supply chain strategy. This cloud supply chain strategy must account for numerous vendors, governance of different types of cloud services from these vendors, and performance not only within the cloud environments themselves, but also in the networks that deliver all of these services. To achieve the highest cloud performance, it's critical to implement a customized, comprehensive cloud supply chain strategy since there is no one-size-fits-all solution. Creating a conceptual design based on your business goals, IT environments, and budget can help with achieving the highest performance per dollar spent. Through this process, you'll probably discover that not all your workloads belong in the public cloud. That's why a hybrid cloud strategy can enable you to find the perfect balance between on-prem and public cloud that best suits your business needs.
But first, it’s critical to understand the state of cloud architecture today.
The State of Modern Cloud Architecture: It’s a Multi-Cloud World
For most organizations, it's no longer a question of whether to adopt the cloud, but how and how many cloud services should be leveraged. As noted in RightScale's 2019 State of the Cloud Report, users leverage 4.9 different cloud services on average.
An analogy I like to use is that hybrid and multi-cloud solutions are sort of like trying to feed a family in today's fast-paced world. Decades ago, it was feasible for most families to cook their own dinners every night. However, as societal demands and expectations on our careers, activities, and social lives have changed, many people and families now rely on food delivery services to prep their meals for them, allowing more time to be spent on work, school, hanging out with friends, activities, etc.
This is very similar to how businesses are deploying their hybrid cloud architectures, with enterprises finding it much easier to offload some of their workloads to third parties while keeping some in-house. This hybrid approach allows enterprises to find the right balance between “preparing their own meals” and having 3rd party cloud providers “cook” for them.
In order to fully optimize your cloud architecture, it’s essential to account for this complexity and create a strategy that both streamlines your deployment and user experience while still allowing for a certain degree of freedom for business units to consume the services that best meet their needs.
For example, let's consider the three traditional cloud delivery models and how businesses leverage them today. From there, we can explore how they contribute to complexity in a dynamic multi-cloud environment.
While the primary definition of IaaS cloud architecture is focused on provisioning hardware resources that would typically be found in a data center (storage, processing power, etc), IaaS vendors have also started to provide some complimentary services, such as backup and recovery.
PaaS cloud services remain a popular option for application development in the cloud; by leveraging cloud infrastructure and limited software environment and resources such as databases, businesses gain the freedom to build and test new applications without investing heavily in on-premise resources that may go underutilized once development ends.
SaaS cloud services host full-feature applications within the cloud. This eliminates both the upfront hardware investment and opens up different pricing models based on need (e.g. per-user pricing). This flexibility has made SaaS a popular resource for business units to leverage accessible and flexible cloud services.
Investment across all of these models continues to grow, as Gartner's cloud forecast from April 2018 shows. Another key fact to note about this growing cloud adoption is that the expectations and specific services within each cloud model are also expanding. According to Gartner Research Director Sid Nag, for example, SaaS customers are demanding increasingly purpose-built services, and, within the PaaS market, database-as-a-service offerings are growing alongside environments optimized for application development. Furthermore, as we've noted in the past, business cloud adoption can create challenges in terms of managing many different services, especially without a cloud strategy that accounts for the multi-cloud business world.
Ensuring the reliability and high performance of all cloud services is key, but it isn’t the only thing to solve for when considering cloud strategy. User experience should also play a role in strategic decisions; investment in cloud services will only pay off if they are used optimally and there is buy-in among the key stakeholders using those services. Otherwise, businesses will not get the return on their cloud services investment and waste the precious resources they have.
When all of the above factors come together, it's clear that organizations are increasingly turning to the cloud for purpose-built solutions that meet very specific needs. While this can lead to dramatic benefits, it can also lead to considerable waste if these resources are not governed properly.
Moving Forward: Optimizing Your Cloud Strategy
While RightScale's research shows that organizations are already leveraging multiple clouds, this doesn't necessarily mean organizations can effectively manage all of them. Network connectivity, central management, and seamless resource provisioning are often what separates mature, successful cloud organizations from those that struggle.
As Gartner has noted, cloud adoption remains a key challenge even after a decade of explosive growth across most cloud services markets. An effective multi-cloud strategy starts with the basic question of what business outcome needs to be achieved. From there, a cloud strategy can be formed that unifies technology decisions to those outcomes.
There are several key considerations to make as you move forward with determining your appropriate cloud architecture, mix of cloud services, and how you’ll optimize your overall cloud strategy.
1. Understanding Multi-Cloud and Hybrid Cloud
First, it's important to understand the nuances of the different ways organizations approach adopting cloud services. Hybrid cloud refers to leveraging a mixture of public and private cloud infrastructure and services. This one-two punch mixture of architectures ensures that businesses can always service their steady state IT demands using private resources while using on-demand public resources to services any spikes or test environments. Additionally, hybrid clouds can ensure private data never touches the public sphere while enabling the scalability of public cloud for resources with less stringent privacy requirements if there are any regulatory concerns.
Multi-cloud architecture can encompass any number and variety of cloud services, including public and private clouds as well as IaaS, PaaS and SaaS services. A multi-cloud strategy enables a firm to take a best-of-breed approach when it comes to cloud services instead of being stuck to the services and development cycles of a single cloud provider.
2. Centralized Cloud Management and Cloud Provisioning Flexibility
One of the key challenges of having so many different cloud services within a business is often the lack of a centralized policy in how these services are managed. With SaaS, for example, it's possible for individual departments to manage their own cloud applications, leading to IT having little visibility into where data is flowing, how users are leveraging each service, and the performance of the various cloud resources across the businesses.
Organizations adopting a multi-cloud strategy must therefore determine their expectations regarding interoperability and cloud network connectivity between these services. Some less centralized multi-cloud elements can be beneficial when individual departments have specific needs that can be solved easily via the cloud, where an organization-wide strategy would be too cumbersome or too costly to implement. However, the governance, mix of services, and types of deployment models utilized should be re-evaluated on a regular basis as needs evolve to maximize cloud performance.
3. Balancing Governance, Security and User experience
Governance and security can sometimes be at odds with user experience, particularly when demands on privacy and data security are especially stringent. This can be mitigated to some degree by implementing streamlined processes for adding users to different cloud services (e.g. through single sign-on platforms or by setting up credentials prior to new employee starts).
However, the architecture behind how different cloud services are connected also plays a vital role, especially in use cases where milliseconds can make all the difference in information exchange. For example, cross-connection services are designed to improve data and performance governance as well as the speed of communicating between multiple cloud services.
Putting It All Together: Optimizing Your Hybrid Multi-Cloud Architecture
The bottom line: A hybrid multi-cloud strategy enables a firm to take a best-of-breed approach when it comes to public and private cloud services and is the only effective way to truly maximize modern cloud performance.
However, cloud optimization is no longer just about scaling more compute, storage, and other server resources. The entire cloud architecture, including the networks that deliver these services to end users and connect different cloud providers to each other must be considered. This will only become more important as data volumes grow due to emerging technologies like artificial intelligence, virtual reality, and internet of things.
Want to learn more about optimizing cloud performance? Read our three-step cloud performance guide, which details how we achieved a 55% increase in web application rendering.