Software:Jumbune

Jumbune

Developer(s)	Impetus Technologies
Stable release	1.3.0
Written in	Java
Operating system	Linux
Website	jumbune.com

Jumbune [juhm-b-yoon] is a Big Data APM available in enterprise grade and open-source offerings. Jumbune leverages Machine Learning to achieve performance improvement of Big Data Clusters, and workloads like, Spark, Hive, Tez, MapReduce and others.

Enterprises leverage Jumbune to get multi fold business benefits, including:

• Faster analytics at reduced cost
• Maturing operational efficiency of Data Lakes
• Reducing time to analytics with ML-assisted recommendations
• Target-meeting, aggressive application SLAs
• Empowering teams to operate on clean data

Exclusivity brought by Jumbune:

• Proactive monitoring of BigData Clusters
• ML-based, Cost-Based Optimization for Hive, Tez, Spark, MapReduce and other Big Data workloads
• Hundreds of critical recommendations for Business Applications, Infrastructure and Hadoop services
• Pinpointing inefficiencies in applications and clusters

Jumbune provides complete view of whole BigData cluster and jobs at one go. This framework comprises various modules

1. ML based Cluster Analysis: Improve productivity and efficiency of Infrastructure, Engineering teams by proactively showcasing deep insights and recommending optimizations for individual applications, Execution engines, Cluster, Infrastructure and operating Queues.

• Highlighting features:
  • Uncovers Inefficiency at Cluster, Application and Queue level
  • ML based Recommendations
  • Deep Queue Analysis
  • Optimal Tuning of individual workloads
  • Proactive & Customizable Alerts…and many more

2. Optimize Job: Jumbune recommends most optimal configuration based on application execution logistics, resource consumption and real cluster state. It employs best in class ML algorithms on built heuristics to deduce the most comprehensive set of recommendations, which achieves the fastest execution time of your workloads.

• Highlighting feature:
  • Most Comprehensive set of Tuning Parameters for Spark, Hive, MapReduce, Tez
  • Tuning Parameter based on execution history
  • Query recommendations

3. Manage Data Quality: Jumbune provides most comprehensive set of automated data quality management tools which enables Dataops team to easily profile and identify quality of data. We establish data quality metering to analyze Quality KRAs even over incremental ingestions.

• Highlighting feature:
  • Comprehensive suite of Data Quality functions
  • Automated Data Cleansing
  • Scheduled Data Validation
  • Establish Quality metering over multiple ingestions
  • Data Completeness, Accuracy and Correctness checks

See also

• Big data
• Data Intensive Computing
• Apache Hadoop – Distributive computation framework