jBEAM Cluster: enhanced capability for Big Test Data analytics

Multiple instances of jBEAM located in a jBEAM Cluster simultaneously analyze different sets of measurement data within one data lake. The partial results are then aggregated to generate one final result.

How jBEAM Cluster works:

  • An analysis generated in one jBEAM undergoes parallel processing in a jBEAM Cluster that contains multiple instances of jBEAM.
  • A jBEAM cluster comprises multiple instances of jBEAM (nodes) that run on one or more computers.
  • Each cluster node processes one file at a time, returns the result to the aggregator, and notifies the cluster manager of its current state.
  • This method corresponds to a MapReduce with files. There is no need to split the files.
  • Once all files are processed, the aggregator combines all the partial results and returns them via the network to the user (client), as one final result.
  • Both conventional file systems (such as Windows, NAS and Linux, etc.) and distributed file systems can be used.

No need to transfer vast amounts of data

Calculations are performed where the data is located

Can be integrated into the MaDaM measurement data management system

A jBEAM Cluster uses multiple instances (nodes) of jBEAM to further enhance the processing speed for huge amounts of test data.


To efficiently process ever-increasing amounts of measurement data, it is important to run the calculation at the actual location of the data. This reduces the effort of transferring vast amounts of data.

We offer a complete solution for processing large amounts of data. jBEAM is our full-state application for parallel processing and analysis of Big Test Data. Combine jBEAM with MaDaM – our measurement data management system – to handle your measurement data with maximum efficiency.

jBEAM Cluster for Big Test Data analytics extends jBEAM's processing capabilities – the key to truly global analyses.

Multithreading of multiple calculation components

Analyses of suitable calculation components are parallelized in jBEAM to take full advantage of the possibilities offered by modern multicore CPUs.

An analysis chain consists of multiple calculation instances and a number of graphic components. Each component can perform its tasks in parallel, provided that the type of operation allows this. For example, a histogram calculation can be split into independent tasks, but this is not possible for an FFT.

Efficiently connect and process multiple data lakes

Global companies with multiple research centers often operate several data storage centers, with a local measurement data management system in place for each one. Engineers normally use their 'local' MaDaM for most of their analytical tasks. If a more global analysis is required, one local MaDaM communicates with all the others (via MaDaM Connect) to distribute the search request as well as the definition of the analysis. Each MaDaM handles the request independently, returning the results to the requesting MaDaM where all results are aggregated for the final report. Within each data lake (managed by MaDaM), analyses are processed close to the data and then aggregated to generate a global result. The analyses in each MaDaM are supported by their local jBEAM cluster.

Max 25 MB