Java Modelling Tools
JMT is a performance modelling suite for capacity planning and system performance prediction. The suite is based on queueing network models. JMT features, among others, an advanced discrete-event simulator for queueing models and Petri nets (jSIM) and mean value analysis algorithms (jMVA).
- Download: http://jmt.sf.net (GPL 2 license)
- Reference: JMT: performance engineering tools for system modeling, ACM SIGMETRICS Performance Evaluation Review, Mar 2009.
LINE is a MATLAB toolbox for solving extended and layered queueing network models. LINE can numerically compute percentiles of response times for service-level agreement (SLA) assessment, describe uncertainty about an operational environment using random environments, and solve queueing networks through model-to-model transformations to native algorithms (CTMC, Fluid ODEs, matrix-analytic methods, …) and external solvers (e.g., JMT, LQNS).
- Download: http://line-solver.sf.net (BSD-3 license)
- Reference: Line: Evaluating Software Applications in Unreliable Environments, IEEE Trans. Reliability, 2017.
KPC-Toolbox is a MATLAB toolbox for fitting empirical datasets into a class of stochastic processes known as Markovian Arrival Processes (MAPs). MAPs can fit non-exponential and long-range dependent time series data. Given a time series, KPC-Toolbox automatically returns a MAP that fits statistical descriptors such as moments, correlations, and bi-correlations.
- Download: https://github.com/kpctoolboxteam/kpc-toolbox (BSD-3 license)
- Reference: KPC-Toolbox: Best Recipes for Automatic Trace Fitting Using Markovian Arrival Processes, Elsevier Performance Evaluation, 2010.
DICE is an open source framework for quality-aware DevOps for Big Data applications. The framework has been created by the DICE Consortium under my coordination. With DICE, software vendors and developers can efficiently prototype new data-intensive applications at low cost, quickly creating business cases and proofs-of-concept for Big data technologies (e.g., Hadoop/MapReduce, Storm, Spark, Cassandra). Check out the DICE YouTube playlist for demos.