How do you approach designing a scalable and resilient technology architecture? Are there any guiding principles you follow?

With foundation support from underlying storage we can develop patterns which enables HA, Site redundancy and disaster recovery for applications meeting RTO, RPO requirements.

1. The first step would be to ensure your organization has a defined Enterprise Storage and Data Protection Strategy which supports scalable and resilient technology architecture for applications and its data.

2. Design applications with each app tier highly available

3. Incorporate site redundant features using GSLBs as required based on app RTO/RPO requirements.

Our strategy is to manage a set of proven patterns, then re-use those patterns as new solutions are needed.  Selecting the right pattern based on the requirements and roadmap becomes the critical decision point.

All starts with the functional and non-functional requirements, in particular the latter category is fundamental to guarantee the levels of services as per expectations, instead the former make sure that the right Architecture is built or better the one that does fulfil the business needs/requirements.

When it comes to scalability, non-functional requirements should expressly call it out when it comes to the different level of volumes that the infrastructure and platform implementing the Architecture should undergo during its operational lifecycle. Clearly, there are multiple ways to achieve it by design, like the principles of elastic scalability would require the Architecture to provision an orchestrator which is driven by active metadata and works on specific scale up/down rules according to specific expectations in terms of thresholds. 

When it comes to resiliency, again non-functional requirements should help with estimating the level of reliability and availability expected, as well as the disruption in case of downtimes. A resilient Architecture is one that is able to undergo failure without visible impacts on the service levels, therefore redundancy and ability to re-route traffic is fundamental to guarantee failure masking from an end user perspective. 

Worth mentioning that is of a vital importance to define specific indicators tracking the key metrics to measure scalability and resiliency to make sure that the Architecture has a companion with guaranteed SLI and SLO. 

Create a roadmap. Start with the requirements and build a plan based on business objectives — why do you need a resilient technology architecture if there's no requirement for it? Then you create strategies to achieve those goals. You might have multiple initiatives that come out of those strategies and the initiatives may be to find resilient technology architectures. In other cases, you might not need it, so it's a matter of understanding what is fit for purpose and how it aligns to the business objective. I can build the most redundant, resilient environment you want, but if there’s no business requirement for that, you're just spending money needlessly.