# ALOJA: A Framework for Benchmarking and Predictive Analytics in Big Data Deployments

**Authors:** Josep Ll. Berral, Nicolas Poggi, David Carrera, Aaron Call, Rob Reinauer, Daron Green  
**arXiv:** [1511.02037](https://arxiv.org/abs/1511.02037)  
**Date:** 2015-11-06  
**Summary page:** https://www.summarizepaper.com/en/paper/1511.02037v1/

## Key points

- The ALOJA project is a collaboration between BSC and Microsoft
- It has developed analytics tools to interpret Big Data benchmark performance data and tuning
- The project focuses on Hadoop, which presents a complex run-time environment where costs and performance depend on numerous configuration choices
- ALOJA has created an open vendor-neutral repository featuring over 40,000 Hadoop job executions and their performance details
- The repository is accompanied by a test-bed and tools to deploy and evaluate the cost-effectiveness of different hardware configurations, parameters, and cloud services
- The predictive analytics extension ALOJA-ML provides an automated system that models environments from observed executions allowing knowledge discovery
- The resulting models can forecast execution behaviors such as predicting execution times for new configurations or hardware choices enabling model-based anomaly detection or efficient benchmark guidance by prioritizing executions.
- This work was partially funded by European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement No 639595) - HiEST Project.

## Summary

The ALOJA project is a collaboration between BSC and Microsoft that has developed analytics tools to interpret Big Data benchmark performance data and tuning. It focuses on Hadoop, which presents a complex run-time environment where costs and performance depend on numerous configuration choices. To address this challenge, the ALOJA project has created an open vendor-neutral repository featuring over 40,000 Hadoop job executions and their performance details. The repository is accompanied by a test-bed and tools to deploy and evaluate the cost-effectiveness of different hardware configurations, parameters, and cloud services. To further improve the study of large resource-constrained search spaces, the predictive analytics extension ALOJA-ML provides an automated system that models environments from observed executions allowing knowledge discovery. The resulting models can forecast execution behaviors such as predicting execution times for new configurations or hardware choices enabling model-based anomaly detection or efficient benchmark guidance by prioritizing executions. The community can benefit from ALOJA datasets and framework to improve the design and deployment of Big Data applications. This work was partially funded by European Research Council (ERC) under the European Union's Horizon 2020 research and innovation program (grant agreement No 639595) - HiEST Project.