A multi-cloud storage strategy is only as effective as the cloud platforms themselves and the multi-cloud storage...
features they offer. Any platforms that fail to support your strategy can impact storage operations throughout the entire organization. For this reason, each cloud service must be vetted to ensure it meets your organization's specific storage needs.
When evaluating features of a cloud platform, take into account these seven factors to help determine its effectiveness in supporting a multi-cloud storage strategy.
A cloud platform must integrate seamlessly with the systems you have in place for moving, storing and protecting data. So when looking at various platforms and their features for multi-cloud storage, make sure the cloud provider offers standards-based APIs that expose the platform's services, making it possible to link systems regardless of the technology or location. To this end, you must examine the provider's available APIs to determine their capabilities and compatibility with existing services, applications, tools and other systems.
Some cloud platforms might provide additional services to help with this integration. For example, Google Cloud's Cloud Endpoints service provides a framework and tools for developing, managing and monitoring APIs based on the OpenAPI specification. You can use Cloud Endpoints with services such as AWS Lambda to integrate workloads across environments.
Networking and connectivity
Integrating disparate systems relies on a full-featured network infrastructure that facilitates connectivity among users, applications, services and data stores. The network must be reliable enough to handle fluctuating, and sometimes unpredictable, workloads as well as different types and amounts of data. Each cloud service should support industry-standard mechanisms that facilitate secure network communications, such as VPNs or dedicated connections through internet service providers.
Even with these mechanisms in place, managing connectivity with multiple cloud platforms can still be one of your biggest challenges. To eliminate some of the pain points, consider technologies such as network function virtualization, software-defined networking, software-defined WAN or network as a service. If you go this route, you ensure the technology is compatible with the cloud platforms and can deliver the necessary connectivity.
Even without multiple cloud platforms, IT pros have their hands full managing the various systems that support data movement and storage. Teams often rely on a variety of tools for implementing, maintaining, monitoring and securing resources, but these tools are often inadequate to support a multi-cloud environment. That leaves administrators scrambling for ones that can augment or replace what they're already using.
Unfortunately, there's no easy answer to this situation. Legacy tools weren't designed for multi-cloud environments, and newer tools are still relatively young and often limited to only a few platforms. When planning a multi-cloud strategy, you must determine what it will take to manage the platforms, what you have on hand that can help, what new tools might be available to ease your burden and which platforms these tools will actually support.
In all likelihood, you'll still need to do a fair amount of scripting to make everything work, but you should at least know what's coming. Before choosing a cloud platform, get as clear a picture as possible of what it will take to manage its services in both the short- and long-term.
One of the benefits of the multi-cloud strategy is the ability to move and distribute data and applications across cloud platforms. As needs change, so can workflows and the data that supports them. For each cloud platform, you must assess its ability to operate within a distributed environment, while delivering the performance, flexibility and reliability necessary to support changing requirements.
As part of this process, you should evaluate a platform's features for multi-cloud storage in terms of its ability to work with the DevOps tools you already use or plan to use to develop, deploy and manage your applications and data, including orchestration, provisioning and automation tools. The platform should also support any infrastructure technologies you require, such as containerization, virtualization and serverless computing. In general, you must know what it will take to support data storage and its movement across platforms, as well as the effort needed to integrate the cloud platform into your workflow management.
Security and privacy
Each cloud platform must have mechanisms in place to secure data, authenticate users, control access levels and protect against malicious attacks. You should be able to monitor and log user behavior and data access, and respond to security events as soon as they occur. You should also take into account privacy and compliance requirements, which can get tricky if data will be moving across national and regional boundaries.
Unfortunately, cloud providers take different approaches to protecting their systems, making multi-cloud security complex and difficult. Even so, you must be able to apply the same user, data and application policies to the cloud environments that you apply to your on-premises systems, without incurring undue overhead or affecting the user experience. To this end, when vetting features for multi-cloud storage, carefully assess each provider's capabilities so you know what it will take to protect your data across all environments.
Services and features
Even if you can easily integrate, manage and secure a cloud platform, it must still be able to provide the features necessary to support your multi-cloud storage strategy. One of your first considerations will likely be the type of storage -- object, block or file -- the platform supports. However, you must also take into account other considerations, such as data center locations, vendor history and stability, available support models, service-level agreements, ease of use and self-service capabilities, that can impact data storage and application delivery.
Keep in mind one of the advantages of a multi-cloud strategy is being able to pick-and-choose a cloud platform to best serve a specific requirement. Not all platforms need to satisfy all required services. For example, you might choose one or more platforms with global data centers for your application data, but use different platforms for archiving data and supporting disaster recovery strategies. Some of your decisions might also be based on applicable compliance laws and industry regulations.
A multi-cloud strategy can theoretically save money, but only if you can closely manage service fees. Doing that is complicated not only by the use of multiple platforms, but also by the complex cost structures inherent in many of those platforms. You must take into account the cost of storing and transferring data, as well as factors such as billing increments, access frequencies, currency differences, support requirements, service levels and service bundle discounts.
Even under the best circumstances, arriving at exact costs in advance is difficult. To help with this process, some platforms provide pricing calculators for estimating usage. For example, Microsoft offers the Azure Cost Management service for monitoring, allocating and optimizing cloud computing costs. You might also consider third-party tools or consulting services.
Regardless of the approach, you'll want a full understanding of what it will cost to support your data storage needs. For this, you need cloud providers that provide clear and manageable pricing structures.