Most of today’s businesses are in the process of modernizing their IT infrastructure and incorporating the hybrid cloud into their processes to enable increased flexibility, scalability and cost savings. At the same time, there is a need for increased availability even though the infrastructure continues to become more complex.
Consumers expect web services to always be available and in response, many organizations are striving toward 24.7.365 availability. However, true continuous availability is extremely difficult to achieve and it can also be very expensive. A more practical goal for most organizations is 99.99% availability. However, to achieve 99.99 percent availability you can still only have a little more than 52 minutes of annual downtime which includes both scheduled and unscheduled downtime. To deliver maximum availability for your hybrid cloud applications you need to incorporate availability concepts into both the application’s design and its on-going data protection plans.
Designing for Hybrid Cloud Availability
Hybrid cloud availability can be difficult because most businesses are running a mixture of many different technologies that all need to be available in order to support their business-critical applications. Availability in the hybrid cloud needs to begin at the design phase of your application infrastructure.
Modern design needs to account for infrastructure diversity, unpredictable demands, and the ability to recover from failure. To help accommodate for infrastructure diversity, it helps to design for a complete separation of compute and data persistence. This can facilitate deployment and scaling flexibility.
Next, if the demand for services can be unpredictable it’s important to provide enough resources to handle peak workload demands. To help reduce costs many companies plan to use the private portion of their hybrid cloud to run the bulk of their workloads and then use the public cloud portion to handle sudden and unanticipated increases in demand. Designing for failure can also help increase application availability. Although you never want to experience application failures, in the real world components do fail and services can become unavailable due to a multitude of causes.
Today’s stateless microservice based application designs are intended to increase resiliency by componentizing applications and allowing the separate components to be automatically restarted in the event of a failure. In addition, to improve the availability of hybrid cloud applications planning to host your applications across multiple cloud availability zones can reduce to possibility of being impacted by a regional outage.
Planning for Hybrid Data Protection
Data protection in the hybrid cloud is more complex than on-premise data protection. The hybrid cloud uses different storage mechanisms and many businesses make use of multiple cloud providers. There are a few important points to consider when you’re planning for hybrid cloud data protection.
- Use modern data protection solutions — Many traditional data protection solutions are not cloud-aware and not able to provide complete protection for modern hybrid applications. Today’s heterogeneous systems require flexible and diverse data protection tools to meet your uptime requirements. Your data protection solutions need to provide local and cloud visibility as well as centralized management capabilities.
- Agility and flexibility — While the hybrid cloud can definitely enhance an organization’s flexibility and scalability, hybrid cloud architectures can also add more complexity and more moving parts. IT environments are increasingly mixed — comprised of physical servers as well as VMs running across multiple hypervisors and multiple clouds. Your data protection solution needs to be able to accommodate and protect the wide range of today’s infrastructure components.
- Orchestration and automation — Production hybrid cloud environments are becoming increasingly more complex. Leveraging orchestration and automation as part of your data protection plan can make the difference between critical application availability and extended downtime. Orchestration and automation can enable you to provide fast recovery times in a complex IT environment. Policies can define priorities and relationships between different components and can enable the execution of predefined disaster recovery scripts.