Move over Hybrid, the Multi-Cloud version is going to be the standard very soon for enterprises. With workloads and complexity in the cloud architecture increasing at a regular pace, the CIOs are riddled with the issue of how to maximize their ROI on cloud and ensure seamless functioning of the business processes. This is where multi-cloud architecture makes sense. In an interview with DynamicCIO, Jeff Harris – VP, Product Portfolio Marketing, Keysight Technologies explains the various aspects of adopting multi-cloud architecture and how CIO’s can optimize the new cloud architecture.
What is the state of multi-cloud adoption among enterprises today?
Worldwide, analysts expect multi-cloud adoption to soon be the norm for enterprise IT environments. In the Asia-Pacific region, analyst firm IDC is predicting that more than 70% of enterprises will have a multi-cloud strategy by the end of 2018(1). Multi-cloud is typically pursued to avoid vendor lock-in and maintain a good negotiating position if cloud providers raise their fees too much. If recent reports on accelerating fee structures at major providers are true, we may see more of this in 2018.
Why is this model of cloud computing growing over hybrid cloud architecture?
Hybrid cloud architecture is often used to extend a private cloud with greater scalability to increase flexibility and responsiveness. This is often used by organizations with fluctuating or seasonal demand that want to avoid the expense of over-investing in infrastructure that is only needed some of the time. This architecture is often used when the organization is hesitant or not ready to move the entire application or service to the public cloud. Hybrid cloud lets them blend the control they get with private cloud with the efficiency and cost-saving of public cloud.
Multi-cloud solutions are shaping up to be the next step in building truly dynamic infrastructures as they help enterprises take advantage of the operational benefits of public cloud, while also reducing the risk of rising costs. Running workloads dynamically across multiple cloud providers, taking advantage of different service level agreements and pricing structures, gives enterprises the ability to ensure workloads are truly optimized.
What is the major difference between hybrid and multi-cloud?
In multi-cloud, you combine and use cloud services from different providers, often to meet precise workload needs, but you do not necessarily connect or orchestrate between them. Hybrid cloud is a mix of on-premises private cloud and third-party public cloud with orchestration between the two. Theoretically, a hybrid cloud could be implemented using more than one public cloud provider and would therefore become a “multi-cloud hybrid architecture.”
Are the organizations really clear about the difference between the two architectures?
Organizations may not agree on the definitions of these labels, but that is not surprising because standardization of nomenclature always lags implementation of a new IT approach. Someone solves a problem in a new way and others follow suit. Eventually, certain terms come to mean certain things. Organizations are just evolving organically to keep their customers happy and their employees productive. They use these strategies, not because of what they are named, but because they solve a particular problem, whether it is vendor lock-in, or burst protection, or performance management.
What kind of challenges is an enterprise expected to face as they adopt cloud architecture?
With IT services transforming into a mix of self-hosted applications and cloud-based services on multiple cloud platforms, providing end-to-end visibility extending across the hybrid environment is becoming a key challenge. The lack of granular access to cloud traffic, regardless of the cloud service provider, creates blind spots in the network which could compromise security or application performance, resulting in data breaches or a degraded customer experience.
When managing a multi-cloud architecture what are the key factors the CIO should keep in mind to ensure seamless business process?
Multi-cloud has a certain level of management complexity. While a multi-cloud strategy can address the problem of escalating public cloud expenses, it comes with its own set of issues. Although multi-cloud is gaining popularity, the practice is still new enough that there are not many solutions commercially available to help determine when to shift a workload to maximize performance and minimize costs or how to ensure safe and cost-efficient transfer of workloads from one provider to another. This either means hiring consultants or assigning senior IT staff to deal with these issues. The CIO would be well advised to consider these additional “costs” when deciding to use a multi-cloud strategy.
The second thing to look at is maintaining flexibility. Whatever the organization uses to manage the shifting of workloads between providers, management should keep in mind that commercial solutions will eventually be introduced. Any solutions they develop should remain lightweight and be easy to discard and replace once a better solution comes along. Lock-in costs are not only a problem with cloud providers, but also a problem with legacy solutions that are too difficult or costly to replace.
How can CIOs ensure a secure architecture in the face of attacks?
Opting for multi-cloud means expansion of the attack surface. Complex multi-cloud environments have many potential vulnerabilities that attackers can exploit. The data breaches we hear about remind us that building an impenetrable environment is not realistic, and we cannot rely solely on cloud providers to handle our defense. Security in the cloud is a shared responsibility between the cloud provider and the user—and enterprises need to take ownership by deploying data access and security solutions, as well as testing those solutions to make sure they are working as expected and provide adequate protection. In some organizations, the pressure to achieve the cost advantages and speed associated with cloud deployments or having a “cloud first” mandate can divert attention away from the basics of security enforcement. For example, data breaches reported in 2017 at Dow Jones and Verizon were attributed to faulty implementation of Amazon Web Service (AWS) security options. Persistent, multi-layered security is, therefore, the best way to defend enterprises against security breaches, data loss, and business disruption.
How to ensure that the multi-cloud framework is secure, and the workloads are managed optimally? Do share few pointers on this
Multi-cloud environments create challenges for security teams who need to manage multiple segments, and no longer have the packet visibility to secure their network. Cloud environments have many potential vulnerabilities that attackers can exploit, and attacks are frequently conducted in multiple stages that may not be caught by intrusion detection systems or next generation firewalls. To stay ahead of would-be attackers, security analysts increasingly use data correlation and multi-factor analysis to find patterns associated with illegitimate activity. These sophisticated solutions require granular, packet data to work effectively. Most organizations have solutions like these deployed on-premises to evaluate packet data captured from physical infrastructure.
Ixia CloudLens provides comprehensive visibility across cloud environments—public, private, hybrid and multi-clouds, and is the first to be delivered as a pure Software-as-a-Service (SaaS) solution. CloudLens was designed from the ground up to leverage the benefits of the cloud – elastic scale, flexibility, and agility, while providing security, analytics, and forensics tools the data they need, whether the tool is located on-premises or in a public cloud.
How should one chart out the policies for this architecture? Do share few pointers/examples for this
As organizations adopt multi and hybrid cloud environments, complexity will be the biggest challenge and should be avoided, minimized, and managed as much as possible.
Avoid unnecessary duplication. For example, implementing a security monitoring solution for cloud-based traffic will add complexity if it differs from the solution you use in your data center. Choosing to use the same monitoring solution lets you avoid the complexity of mastering a new solution and retraining staff.
Minimize required tasks. Even in a complex environment, automation can minimize the effort involved in certain tasks. For instance, you can deploy a solution to access data inside your clouds that automatically embeds a container-based sensor inside every cloud instance that is created. With this design, the complexity of accessing data as clouds scale is minimized, as are the errors associated with manual configuration.
Manage from a single pane-of-glass. Complexity is unavoidable, but you can proactively seek more efficient ways of managing hybrid and multi-cloud environments. Accessing and filtering data is done using different technologies in private and public cloud environments. However, you can work with a vendor that provides the same user interface for both environments to simplify management tasks.
(1) Reported by DataCenter News on April 12, 2018,