SaaS operations overview
Ravello's deep technology is wrapped up and presented to users as an easy to use service. Think of Ravello as an overlay cloud provider that sits on top of Tier-1 leading clouds. Enterprises simply log into Ravello, import their workloads, and deploy them to any supported cloud region.
Ravello does not own and operate its own data centers (a capital extensive endeavor), but rather, operates a cloud service, powered by its unique HVX software infrastructure on top of Tier-1 leading clouds. Ravello leverages Oracle Cloud's, AWS's and Google's best-in-class hardware, scale, economics, and operational practices to provide a robust enterprise grade cloud service to its customers.
Through Ravello, enterprises can deploy their VMware or KVM workloads on any leading cloud region (without requiring migration), in Ravello's global footprint - and benefit from low latency, high throughput access to their application environments.
Uptime and service availability
A big element of uptime is the architecture of the software used to provide the service. As discussed here the Ravello management system is built using industry best practices and includes redundancy and high availability at all layers in the overall stack.
In addition to the architecture, Ravello operations has implemented industry best practices to ensure maximum service uptime. Ravello monitors and manages its entire service offering - system, storage, network, databases and customer usage - 24 hours a day, 7 days a week, 365 days a year. This monitoring is done using internal and external engines from various locations around the world in a distributed and scalable manner. Ravello's redundant software architecture and robust operations have enabled the company to achieve very low unplanned downtime for its service. Over the last one year, Ravello management system downtime has ranged from 0 hours - 0.8 hours per quarter.
Another element of the software architecture is around minimizing planned downtime. Most Ravello software updates and upgrades to the management system do not cause any downtime. In exceptional cases, the management system is unavailable for a maximum of 10 minutes. It is also important to note, that the Ravello service is designed such that even if the management service were to suffer downtime, enterprise workloads that were already deployed will continue to run uninterrupted.
Given Ravello's unique capability of enabling enterprises to deploy their workloads to leading clouds, enterprises can achieve very high availability levels for their application environments. For example, if there is a storm on the US eastern seaboard, enterprises can with a single click deploy that application environment to Google's US central region with zero downtime.
Ravello's development and operations team function as one. The product architecture and the operational practices are built around rolling out new features rapidly while at the same time, maintaining high quality and uptime for the service. Ravello rolls out dozens of software enhancements, updates and bug fixes every quarter. The company has implemented robust release and change management methodologies, procedures and tools to ensure that this level of agility is provided safely to its customers. One of the risk mitigation approaches used, is to gradually roll out new features and key core components to subsets of users.
Since Ravello does not operate its own servers, networking and storage - but rather, leverages Tier-1 leading clouds, theoretically Ravello's capacity is the sum of the capacity of the underlying clouds - basically, a lot. Despite that, the Ravello operations team carefully monitors and manages underlying cloud resources and quotas to ensure that its customers always have enough capacity even for peak workloads.
Transparency is a guiding tenet of Ravello's operations team. Ravello provides a systems health page outlining the status of all service components. Customers and prospective customers can always reach Ravello operations via the support portal. The operations team proactively sends out regular updates to customers when specific actions are required. For example, if Ravello gets notified that instances belonging to certain customers are being retired by Oracle Public Cloud, the Ravello operations team reaches out to the customer and makes sure that there is a plan to transition that workload to a different instance or even a different cloud provider. Storm warnings that have the potential to impact service operations are similarly communicated in advance as are cloud provider rolling upgrades.