Context aware dynamic composition of migration plans to cloud

摘要

Context aware dynamic composition of migration plans may be provided. A request for application or image migration may be received. Target machines and associated configuration may be identified. Resources and a schedule may be allocated. An appropriate tooling for each migration action may be selected. An artificial intelligence aspect of the migration planning process may continuously replan migration based on monitored changes in the context of source or target environment.

主权项

1. A computer readable storage medium storing a program of instructions executable by a machine to perform a method of providing context aware dynamic composition of migration plans, the method comprising:
receiving a request to migrate at least one of application and computer system image from a source environment; identifying a target environment and associated configuration to define target definition; defining one or more contextual actions comprising a sequence of migration steps; allocating resources associated with the source environment and a schedule for triggering the one or more contextual actions; selecting automatically one or more migration techniques to use to implement the one or more contextual actions based on current context of the source environment and the target environment; monitoring an execution of migration performed according to the allocating and the selecting; responsive to detecting one or more exceptions in the execution of migration, regenerating a migration plan, the one or more exceptions comprising a component defined in the target definition missing from the target environment, and the regenerating the migration plan to include a contextual action comprising installing the missing component in the target environment; and automatically generating a migration template comprising learned patterns associated with a particular context learned based on semi-supervised machine learning observing historical data and user input, wherein the template captures one or more steps that apply in a given context, which can be reused in future; wherein the semi-supervised machine learning involves interactively querying a user to obtain outputs at new data points.