Who is responsible for cloud data resiliency and availability?

Organizations encounter a bewildering assortment of cloud storage platforms that vary not only in terms of underlying hardware and software but also in relation to the management model they employ. The difference between the offerings lies in who is responsible for scaling, resiliency and performance: the provider or the customer.

KEY POINTS

• Cloud providers offer storage and database services for differing use cases. These services vary depending on the underlying technology and who is responsible for architecting and managing scalability and resiliency.
• The choice of database management system (DBMS) is primarily an architectural decision based on the application and data. The choice of management model is mainly an operational decision based on the availability of skills and procedures for effective cost, resiliency and performance management.
• Serverless platforms require the least customer management to meet performance targets but are prone to runaway costs. Traditional instance-based platforms are more likely to be backwards compatible and predictable in costs, but they may not be easily scalable in capacity and performance terms.
• Service level agreements (SLAs) provide few guarantees regarding the availability or durability of data. They also only offer modest compensation in the case of outages that is unlikely to cover direct costs, let alone business losses. By using a cloud provider for mission-critical applications, organizations need to accept that they ultimately place their trust in a third party.