A Unified Cloud Metering FrameworkTechnology #inv225052015
Questions about this technology? Ask a Technology Manager
- Image Gallery
- Karim Sobh, PhD. Candidate
- Amr El-Kadi, Professor of Computer Science
- Managed By
- Ahmed Ellaithy Director, Technology Transfer +20226153112
- Ingy Darwish Licensing Officer +20226153130
- Patent Protection
- Provisional Patent Application Filed
- US Patent Pending 15/088,476
A Framework for metering cloud resources that is based on an Extensible Object Oriented Cloud Metering Markup Language (CMML), Multi-Tier Scalable Architecture, and network protocols specifications. The framework is based mainly on data modeling and distributed processing of metering indicators through the following metering data processing propagation stages:1- Collection Engines: Operate on the cloud resources to collect metering indicators from their raw sources and convert them to Autonomous Cloud Metering Objects (CMOs) modeled in CMML.2- Correlation Engines: Receive autonomous CMOs from collection engines and correlate related CMOs with respect to data relations and time.3- Back-end Engines: Long living back-end metering engines that process correlation CMOs, such as Storage, Billing, SLA, etc.
Cloud computing utilizes the integration of different computing technologies to achieve a utility computing model. Computing resources are consolidated and shared among different applications transparently. Cloud environments are like a market place, and an accurate cloud metering framework is needed for fair charge back, and accurate responsive SLA policies.
CMMLThe adoption of Autonomous Cloud Metering Objects (CMOs) is the main design decision that unlocked a lot of features that we targeted through our proposed framework. The main idea is how to couple metering data with their operations in an object oriented model that provides high levels of mobility. An extension to the object oriented model adopted by CMML is introduced through the concept of receptors which encapsulates correlation information, and based on which dynamic multi-perspective correlation takes place. The autonomous CMOs can propagate throughout the metering architecture, and through the encapsulated operations corresponding architectural tiers can perform a subset of the overall metering task.CMOs are CMML objects that represent resources, either primitive or composite. The adopted Object Oriented model is extended such that CMOs encapsulate corresponding resource operations and their target architectural tiers of execution through predefined markup tags. Through CMML class definitions primitive resources, such as CPU and RAM, can be presented and consolidated into wrapper CMML classes to represent more composite cloud resources such as virtual machines or even web farm environments.
Distributed Proc File SystemThe Distributed Proc Filesystem is a unified pseudo filesystem that can transparently consolidate the distributed storage of metering indicators in a virtually unified location. The underlying low overhead network protocol is based on a packet reorder flow control mechanism with group packet acknowledgments to efficiently utilize the underlying network medium as well as the network processing resources. The Linux Netfilter Hooks were used to build the Distributed Proc Filesystem kernel modules in a way the cuts of the network stack into half, which contributes to the low overhead performance of the transport layer.
- Extensible Representation: Ease of interpretation and shareability between federated clouds.
- Autonomous Metering Data: Coupling metering data with their corresponding operations.
- Correlation Capabilities: Correlation of metering data extracted from different architectural layers.
- Programmability: Flexibility of defining metering constructs through writing code.
- Standard Metering Transport: Transporting metering data over simple standard mechanisms and APIs; within the framework and from/to the outside world.
- Elastic Multi-Tier Architecture: Can scale with the metering needs.
- Metering Services Redundancy: For fault tolerance and recovery.
- Low Probe Effect: Exhibits low overhead as a result of probe insertions and other metering framework operations.
- Online Metering: Fast and responsive metering data processing through distributed system techniques.
- Ease of Integration: Ease of integration with different cloud environments irrespective of their type, topology, underlying technologies, and service nature.