Orchestration: Importance and Use Cases
XaaS Environment, Composite Service
A service in the service catalog may be composed of multiple functions and requires cloud orchestration across multiple service providers for delivery/fulfillment. We see these examples in business all the time. Travel service providers, for example, offer multiple services such as airline, hotel and car rental bookings. In IT operations, especially in XaaS models like PaaS and SaaS, provisioning requires cloud orchestration across multiple elements.
In a typical scenario, an enterprise would procure for its employees e-mail as a service (say Microsoft), personal storage service as a service (say Google Drive or Drop box), and provide a virtual desktop, while retaining domain authentication and directory services in-house. HR becomes a service integrator for “New Hire” services. The implementation requires sequential execution of individual service providers – domain account creation, triggering a new email account, which is then used to provide the new hire with a virtual desktop and access to Google Drive. Manual service execution could also be provided if a physical laptop needs to be delivered and added to the domain.
A prominent example of orchestration is Service Integration and Management (SIAM) and research analysts like Gartner have repeated emphasized that the multivendor model for IT services is complex and requires orchestration between vendors who are fiercely competitive. This orchestration requires the enterprise to establish a multi-sourcing delivery model that provides visibility and transparency into multiple vendor systems and processes. A SIAM service provider orchestrates the operational activities across multiple service providers. (Method of activity does not matter—activities can be manual, automated or both). A typical example would be critical incident management processing, where critical incident manager is an orchestrator and directs different service provide to act together to restore the services.
Choreography: Importance and Use Cases
Landscape monitoring consists of systems, networks, composite applications, containers running microservices etc., have significant potential in using choreography-driven models. Today monitoring systems are controlled centrally via a management server which becomes the bottleneck in a dynamic microservices environment. Having monitoring itself running as microservices which choreograph themselves with other monitoring services across the landscape and communicate using RESTful APIs could make the overall delivery of the service more agile. When we combine this with an Orchestrated Service Layer we get the benefit of Scalability and Control achieved
Implementing microservices fundamentally means that at the business technical service layer we bring in quite a significant layer of autonomy to allow services to choreograph themselves based on the service boundaries coded into the system and the top level composite service catalog is the only orchestrated entity which then let the respective microservices operate and manage their I/O in a self-managed way. When we start to deploy microservices within containers as fully enclosed runtimes then the multiplier effect of choreography becomes exponential.