Topics in this article
Cloud infrastructure (IaaS or Infrastructure-as-a-Service) is a highly commoditized space today. All major cloud providers offer Compute, SSD, Elastic-IP, Firewall and Bandwidth. Choosing the right cloud infrastructure for your business often tends to be tricky, and involves multiple areas of assessment. Here are a few important metrics to help you evaluate cloud vendors and their ability to meet your business needs.
Price to performance ratio
This is the most basic of all parameters and is the primary financial metric when selecting cloud vendors. Performance needs and goals (of the organization, departments or specific use cases) need to be clearly understood to ensure that you pick the vendor with the best price to performance ratio for your business. In a multi-cloud scenario, you would choose different cloud vendors for different business needs, depending on the optimal price to performance ratio. This ratio also needs to be tracked regularly to stay above pre-defined performance benchmarks.
There are a number of pricing models available. Amazon, for example, has a very simple model defined by specific rates for different types of uses (e.g., per GB of storage, hours of server usage or GB of bandwidth usage). Tier 1 providers like NTT can offer customized pricing models, or even upfront billing for a committed amount of usage. Understanding which pricing models work best for your business is very important when choosing a cloud vendor, and can potentially help you save (or lose) a lot of money in the long run.
This is a very important factor to ensure that your IT environment in the cloud is able to handle the pressures of a fast changing business. Naturally, greater service maturity would result in greater flexibility and lower risk. Here are some examples of cloud services with high levels of maturity:
- Built in cloud governance and management platform/dashboards
- Support for seamless vertical/horizontal scaling with zero downtime
- Auto-scaling of compute nodes for handling unexpected loads and cloud bursting if needed
- SaaS support (DBaaS, DRaaS, Analytics-as-a-Service, CI/CD-as-a-Service)
- Integrated application/data migration tools/APIs
- Pre-built synchronization/orchestration capabilities
- Secure connection to other/on-premises applications (hybrid cloud support)
- Network service features such as load balancers, direct connect DNS support
- Instance/service definition
- Different cloud vendors define standard instances or standard services in different ways. For example, a vendor may offer a pre-defined instance with 512MB RAM 1 vCPU to 976 GB RAM and 128 vCPUs. They also provide different configurations for storage (object storage, block storage, cold storage, etc.) bandwidth and network services (private or public networks).
Compliance and security needs
Regulatory compliance and data security concerns often govern your selection of cloud vendors. There are a number of key criteria that impact vendor choices, such as:
- Compliance certification, e.g., ISO, PCI, PDPA, etc.
- Ability to provide strong security and access control – encryption, firewalls, 24/7 SOC, etc.
- Strong suite of managed security services
- Strong audit tracking capabilities for all transactions and users
The breadth and depth of partner ecosystems is a key part of long-term success. AWS is a great example for a powerful partner ecosystem, and has thousands of partners and a strong development community that builds tools and accelerators for AWS users to consume. Global IaaS vendors like Netmagic have a strong partner ecosystem that enables companies to access an extensive pool of third-party platforms and tools to build and implement cloud applications.
The in-house DC model thus doesn’t seem like a strategically viable model in the near-term. Building and maintaining their own DC is likely to be very low on the CIOs’ list of priorities, going into the future. CIOs will need to reassess their technology strategy and move away from in-house data center management, if they want to stay competitive in a highly connected and hyperscale future.
In many cases, organizations would end up choosing different IaaS vendors (resulting in a multi-cloud environment) for different needs based on the above parameter, and many other parameters, depending on things like geographical proximity, configurability, OS support, etc. Eventually the overall effectiveness of your cloud strategy will depend on not just the initial vendor decisions, but also on your ability to effectively govern and optimize your resources to stay aligned to business needs over a long period of time.