Cloud or Data Center: Workload Placement Decision Checklist

“The location of service deployment doesn’t matter to us. We’re only interested in what’s most practical and efficient in terms of cost, security, and so on.” This stance, echoed by a Senior VP of Infrastructure at last week’s CloudBolt customer event, resonates with a significant number of IT leaders. For a variety of organizations, the deployment location of a new business service, either on-premise or in the cloud, is inconsequential.

However, in reality, the majority of developers, commercial off-the-shelf software vendors (COTS), and third-party service providers tend to lean towards the public cloud when given a choice. Often, the simplest answer is also the most fitting – the cloud simply offers a superior experience! It provides a user-friendly, feature-packed platform with a myriad of service options. Nonetheless, it’s not always the ideal choice for every business, and it can bring about certain disadvantages.

To prevent the public cloud from gaining an unfair advantage, tech leaders need to set limitations on its usage. One straightforward solution is creating a checklist or questionnaire to help evaluate if the public cloud is indeed the best choice. A well-constructed pre-deployment questionnaire proves instrumental in comprehending the necessity and advantages of utilizing the public cloud.

Here are some essential questions we suggest including in your assessment to confirm whether your workload requires the benefits of the public cloud, versus using your in-house data center:

1. Business Goals

• What are the main business objectives driving this project or service? How will the business quantify ROI?
• How is your current infrastructure impeding these business objectives?

Field Notes: Each proposed service should flaunt a rock-solid ‘return on investment’ model and crystal-clear criteria for business value. While these often connect to revenue, don’t shy away from linking them to customer or employee experience metrics like NPS or CSAT.

2. Performance Requirements

• What does the proposed architecture look like? Are there components that need technologies not available on-premise? If so, is it possible to refactor the application for on-premise operation?
• What are the performance requirements (latency, throughput, etc.)? Can the public cloud offerings exclusively meet or exceed these?
• What performance testing has taken place to confirm the proposed requirements?

Field Notes: Beware, a COTS vendor or service provider might peddle specifications loftier than needed or introduce shiny new technology as performance insurance. Uncertain? Roll up your sleeves and use your own testing methods to confirm the performance claims.

3. Cost Considerations

• Have you carried out a detailed cost analysis comparing maintaining your existing data center infrastructure to transitioning to the public cloud?
• Are there financial considerations or constraints to take into account, such as licensing or others?

Field Notes: It’s alarming how many organizations gloss over the total cost of ownership. Remember, the decision-making process must account for all relevant factors: resource & hosting, maintenance & operations, licensing, commitments, and more.

4. Workload Characteristics

• Are there possibilities to host parts of the workload on-premise, like storage or compute?
• How predictable is your workload? Could a pay-as-you-go model be more cost-effective than maintaining capacity during peak loads?
• What’s the nature of the applications/data being hosted? Are they stateless, stateful, or a combination?
• Do you foresee a need to quickly scale your resources up, down, or out?
• How rapidly must you be able to respond to changes in demand? Why?

Field Notes: Keep in mind, cloud bursting stands ready for workloads expecting predictable capacity surges. If your workload consistently experiences seasonal capacity spikes, hosting on-premise could still be the go-to choice most of the time.

5. Data Security and Compliance

• What are your data security needs? Can the public cloud offerings meet these with the current configuration?
• Are there regulatory or compliance requirements (such as GDPR, HIPAA, PCI DSS) influencing your decision?

Field Notes: Sure, on-premise data centers often get the limelight for easier customization of security measures and compliance protocols. However, as industry standards and regulatory obligations morph, many security teams find public cloud adaptability refreshing.

6. Disaster Recovery and Business Continuity

• What’s your current disaster recovery and business continuity plan?
• Would a public cloud solution enhance your resilience and recovery time objectives (RTO) and recovery point objectives (RPO)?

Field Notes: Considering the public cloud? Know that options to power down unused resources can trim down disaster recovery infrastructure costs. But beware of the catch – slower failover and potential business continuity disruption. If your business demands a constantly operational DR infrastructure, the public cloud can turn pricey, and a fixed-cost service might be the smarter move.

7. Integration and Interoperability

• How will your existing systems merge with a public cloud solution?
• What’s your strategy for cloud data management, including data integration, data migration, and interoperability?

Field Notes: Embarking on a multi-cloud, hybrid-cloud journey can feel like navigating a labyrinth. With different cloud providers using unique APIs, management tools, security models, service offerings, pricing structures, and so forth, complexity is inevitable. Keep your compass handy! Many companies should employ enterprise architect leaders to be the savvy navigators capable of ensuring that all these puzzle pieces not only fit together but also align seamlessly with the overall strategic direction.

8. Staff Skillset

• Does your team possess the necessary skills to manage and optimize the architectural components of the deployment?
• If not, what’s your strategy for training your team or recruiting new members with the required skills?

Field Notes: Silos between development and operation teams often lead to the use of new services or SKUs where internal capabilities are missing. Stay sharp – maintain an updated “fly / no-fly list” of approved cloud offerings. Pair this with a robust proof-of-concept process to pragmatically test, learn, and operate new technologies.

9. Exit Strategy

• How long does the workload and its components need to stay on the public cloud?
• If you decide to transition the workload to a different platform or bring it back to your in-house data center in the future, what steps would be necessary?

Field Notes: We might dub this list a pre-deployment framework, but don’t box it in. Use it as a routine spot check. Regular audits of both on-premise and cloud solutions remain critical to keeping your service deployment strategy on its toes.

Choosing between on-premise and public cloud deployment is not a one-size-fits-all decision. It requires a thoughtful analysis of various aspects, including business goals, performance requirements, cost considerations, workload characteristics, security and compliance, disaster recovery, integration, staff skills, and exit strategies. Ultimately, the decision should align with your organizational objectives and optimize resource usage, security, and performance. So, don’t be daunted by the intricacies of the questionnaire – remember, armed with even a simple checklist can go a long way – something is always better than nothing!

Ready to learn more? See how you can get the best out of a hybrid cloud environment.