Predict and Optimize Cluster Performance

Optimize and plan your big data cluster performance and network behavior with Intel® CoFluent™ technology for big data.

Intel® CoFluent™ technology for big data is a planning and optimization solution that predicts cluster performance and network behavior for big data challenges. Intel® CoFluent™ technology for big data helps address common big data cluster design challenges: predicting system scalability, sizing the system, determining maximum hardware utilization, minimizing costs, and predicting system performance. This technology allows you to optimize and plan according to your business needs, and provides a solution to help you minimize IT spending on big data clusters.

Intel® CoFluent™ technology for big data

Use Cases

Learn how to meet real-world challenges using Intel® CoFluent™ technology for big data.

Video Streaming

Streaming chart

Video Streaming Using HDFS*

Challenge: Create a simulation model for a video streaming system that optimizes the usability and stability of a network during deployment. This includes estimating the network, disk, and node requirements to support 1,000 concurrent users, and identifying the average throughput rate.

Solution: Based on a baseline system simulation model, use an Intel® CoFluent™ technology for big data simulation model to determine the node count requirements; identify possible network bottleneck issues and possible solutions; and determine optimal system disk options.

Video Analytics

Hbase* chart

Video Analytics Using HBase*

Challenge: Determine the optimal hardware and software configurations for typical camera-based use scenarios.

Solution: Use an Intel® CoFluent™ technology for big data simulation model to estimate the number of nodes required to support 1,000 cameras based upon insert intensity, query intensity, and a balanced scenario of intensities. The simulation model also estimates the ratio of HD cameras to server with Intel® Xeon® processors required for the video analytics system.

TeraSort

Chart reduce simulations results

TeraSort Using MapReduce*

Challenge: Estimate the configuration requirements to complete a sort for a 60 GB file within 200 seconds.

Solution: Use Intel® CoFluent™ technology for big data to build a simulation model that can determine the lowest cost configuration with a minimum number of hardware/software components to meet the 200-second execution time.

Online Banking

Online banking results

Online Banking Using Impala*

Challenge: Optimize offline banking, including storage, compressing, partitioning, and caching processes.

Solution: Use Intel® CoFluent™ technology for big data simulation model of reporting and deep analytic query processes. This includes making use of latency, throughput, hardware utilization, software status, job- and application-specific data, and custom metrics.

Related Content

Customer Success Stories

White Papers