Method and apparatus for providing batch-oriented execution environments as services
Publication Date: 2010-Sep-09
The IP.com Prior Art Database
This article describes how batch-oriented workloads can be executed and provided as-a-Service based on a dynamic cloud infrastructure.
Method and apparatus for providing batch -oriented execution environments as services
1 Technical Background
1.1 Cloud Computing and IT Services
The present article relates to the field of Cloud Computing, which is a style of computing where massively scalable IT-enabled capabilities are provided "as a service" to external customers using Internet technologies, where the consumers of these (Cloud) services need only care about what the service does for them, not how it is implemented.
As states above, a Cloud service (or more generically - an IT service) represents any type of IT capability, which is typically made up by a well-defined aggregation of
1..n resources like servers, storage, network, operating systems, SW licenses, etc. Examples for IT Services are clusters of servers like a WebSphere Application Server Cluster, an SAP System or also more simply resource compositions like a single virtual machine, a data volume or an end-to-end network connection. Such different types of Cloud Services are referred to as service types. Typically, there is one so-called "Service template" capturing all service type specific management knowledge in a formalized fashion.
The special characteristic of an IT Service is that it is treated as a single unit of management where all complexities associated with managing such an aggregation of resources are shielded away from the administrator.
In a Cloud Computing environment, IT Services are managed from a well-defined management infrastructure being responsible for instantiating, managing and terminating these IT services. This management infrastructure has knowledge all IT services existing within its domain, which resources belong to them, who owns this IT Service, etc. In the following, this management infrastructure will be referred to as the Service Lifecycle Center (SLM).
The SLM is generic in a sense that it doesn't contain any 'hardcoded' knowledge about any service type. All service type specific management knowledge is plugged into the SLM as service templates, i.e. whenever a new type of service should be supported by the SLM, the respective service template must first be registered to the SLM. Once this is done, the SLM becomes enabled for creating new instances of the respective service type, managing those instances and terminating them (all based on the management knowledge coming with the service template).
2 Current State
In many IT environments it is a very typical task to schedule certain "pieces of work" (in the following referred to as a "job") on a specific group of servers, so that this job gets executed in a timely fashion. This type of workload is referred to as "Batch"-Workload. In contrast to classical Online Transactional Processing (OLTP) type of workload (which gets executed immediately), batch workload normally includes the execution of a larger piece of work which does not have to be completed immediately, but by a well-defined point-in-time....