Deepak Jeswani

Adaptive Monitoring: Application of Probing to Adapt Passive Monitoring

Availability of good quality monitoring data is a vital need for management of today’s data centers. However, effective use of monitoring tools demands an understanding of the monitoring requirements that system administrators most often lack. Instead of a well-defined process of defining a monitoring strategy, system administrators adopt a manual and intuition-based approach. In this paper, we propose to replace the ad-hoc, manual, intuition-based approach with a more systematic, automated, and analytics-based approach for system monitoring. We propose an adaptive monitoring framework where end-to-end probing-based solutions are used to adapt the at-a-point monitoring tools. We present a systematic framework to use probes for adjusting monitoring levels. We present algorithms to select and analyze probes and to dynamically adapt the monitoring policies based on probe analysis. We demonstrate the effectiveness of the proposed solution using real-world examples as well as simulations.

Minimizing Latency in Serving Requests through Differential Template Caching in a Cloud

In Software-as-a-Service (SaaS) cloud delivery model, a hosting center deploys a Virtual Machine (VM) image template on a server on demand. Image templates are usually maintained in a central repository. With geographically dispersed hosting centers, time to transfer a large, often GigaByte sized, template file from the repository faces high latency due to low Internet bandwidth. An architecture that maintains a template cache, collocated with the hosting centers, can reduce request service latency. Since templates are large in size, caching complete templates is prohibitive in terms of storage space. In order to optimize cache space requirement, as well as, to reduce transfers from the repository, we propose a differential template caching technique, called DiffCache. A difference file or a patch between two templates, that have common components, is small in size. DiffCache computes an optimal selection of templates and patches based on the frequency of requests for specific templates. A template missing in the cache can be generated if any cached template can be patched with a cached patch file, thereby saving the transfer time from the repository at the cost of relatively small patching time. We show that patch based caching coupled with intelligent population of the cache can lead to a 90% improvement in service request latency when compared with caching only template files.

A Practical Approach for Target Tracking in Sparsely Deployed Binary Sensor Network

In this paper we propose a novel real time approach to localize and track a target using sparsely (non-overlapping) deployed binary sensors. Each of these sensors reports 1-bit information regarding a targets presence or absence within its sensing range. Our technique estimates the distance covered by a target on the basis of the time it spends in the sensing region. The trajectory of the target is approximated by a piece-wise linear path, where each piece is an estimated tangent segment to a pair of circular sensing range of two adjacent sensors that sense the target. The formulation of the problem as estimates of tangent segments allows us to convert the original problem into a quadratic programming problem. We compare the results of the proposed method with that of the other existing methods and demonstrate that even with a sparse deployment, our approach tracks the target with competitive accuracy.

Columbus: Configuration Discovery for Clouds

Low-cost, accurate and scalable software configuration discovery is the key to simplifying many cloud management tasks. However, the lack of standardization across software configuration techniques has prevented the development of a fully automated and application independent configuration discovery solution. In this work, we present Columbus, an application-agnostic system to automatically discover environmental configuration parameters or Points of Variability (PoV) in clustered applications with high accuracy. Columbus uses the insight that even though configuration mechanisms and files vary across different software, the PoVs are encoded using a few common patterns. It uses a novel rule framework to annotate file content with PoVs and a Bayesian network to estimate confidence for annotated PoVs. Our experiments confirm that Columbus can accurately discover configuration for a diverse set of enterprise and cloud applications. It has subsequently been integrated in three real-world systems that analyze this information for discovery of distributed application dependencies, enterprise IT migration and virtual application configuration.