Formalizing the Cloud through Enterprise Topology Graphs

Abstract

Enterprises often have no integrated and comprehensive view of their enterprise topology describing their entire IT infrastructure, software, on-premise and offpremise services, processes, and their interrelations. Especially due to acquisitions, mergers, reorganizations, and outsourcing there is no clear 'big picture' of the enterprise topology. Through this lack, management of applications becomes harder and duplication of components and information systems increases. Furthermore, the lack of insight makes changes in the enterprise topology like consolidation, migration, or outsourcing more complex and error prone which leads to high operational cost. In this paper we propose Enterprise Topology Graphs (ETG) as formal model to describe an enterprise topology. Based on established graph theory ETG bring formalization and provability to the cloud. They enable the application of proven graph algorithms to solve enterprise topology research problems in general and cloud research problems in particular. For example, we present a search algorithm which locates segments in large and possibly distributed enterprise topologies using structural queries. To illustrate the power of the ETG approach we show how it can be applied for IT consolidation to reduce operational costs, increase flexibility by simplifying changes in the enterprise topology, and improve the environmental impact of the enterprise IT.

DevOpSlang – Bridging the Gap between Development and Operations

“…On the IaaS level approaches such as Amazon CloudFormation 18 or OpenStack Heat 19 are used to orchestrate infrastructure resources (VMs, storage, network, etc.). Moreover, middleware and application components can be stacked and orchestrated using application topologies based on Ubuntu Juju 20 , Amazon OpsWorks [16], Blueprints [15], or enterprise topology graphs [3]. The Topology and Orchestration Specification for Cloud Applications (TOSCA) [12] is an emerging standard to define portable application topologies.…”

Standards-Based DevOps Automation and Integration Using TOSCA

“…Related work [23], [29] presents approaches to deploy and manage Cloud applications that are modeled based on TOSCA and our transformation framework. There are several alternatives to TOSCA such as CloudML [30], Blueprints [31], and enterprise topology graphs [32]. Moreover, pattern-based deployment approaches [33] utilize a proprietary meta model based on Chef to transform patterns to deployable artifacts.…”

Automated Discovery and Maintenance of Enterprise Topology Graphs

“…An Enterprise Topology Graph (ETG) [1] is a graph capturing a fine-grained technical snapshot of the complete enterprise IT, i. e., including all processes, services, software, infrastructure, and Cloud offerings as nodes, as well as their relations and dependencies as edges. Figure 1 shows the conceptual model of ETGs on the left.…”

“…In practice, dependencies between components are still uncovered by "pulling the plug" of one server and analyze the impact. To tackle this, the concept of Enterprise Topology Graphs (ETG) as a technically fine-grained, formal, graphbased model to depict snapshots of the complete enterprise IT have been introduced [1]. An ETG includes all kinds of components, from business processes and services to their implementation and infrastructure, connected by typed edges representing the relations between the components.…”

“…An ETG includes all kinds of components, from business processes and services to their implementation and infrastructure, connected by typed edges representing the relations between the components. ETGs have already been applied to adapt [2], analyze [3], manage [4], and optimize enterprise IT, for example, by improving the ecological sustainability of business processes [5] and consolidating duplicate components [1]. Our current research uses ETGs to migrate existing enterprise applications to the Cloud [2], [6].…”