[Home]
[People]
[Projects]
[Publications]
[Software]
[Sponsors]

Performance Analysis of Computer Systems Lab

Computer Science Department, Stony Brook University

Location: Room 336, New CS Building. Lab PI: Anshul Gandhi  (anshul (at) cs.stonybrook.edu).

Code/Data/Scripts/Software/Tools

• Code for our online configuration tuning of applications work, OPPerTune.
  
  
• Code and traces for our ML power modeling evaluation work.
  
  
• Code and multi-bottleneck traces for our multiple bottleneck detection work, GAMMA.
  
  
• Code for our discrete-event simulator for evaluating selective pushdown for SparkNDP.
  
  
• Code for our microservices bottleneck detection work, B-MEG.
  
  
• Code for our microservices configuration tuning framework.
  
  
• Code for our Lyapunov function based Markov chain truncation method.
  
  
• Code for our serverless scheduler + resource manager system, named ENSURE.
  
  
• Serverless workload suite for evaluating resource management in serverless environments.
  
  
• Tensorflow-based implementation for our custom operation placement and scheduling of DAGs, Pesto.
  
  
• Code for our Model Driven Rehoming System (MERIT) for optimally rehoming VNF chains.
  
  
• A modified Memcached version with support for user specified auto-move algorithms
  
  
• Code for ElMem - a modified Memcached 1.4.31 version that supports autoscaling.
  
  
• Code for MLscale - an application agnostic autoscaler using black-box machine learning.
  
  
• Load generation tool for a Memcache tier backed with a database tier
  
  
• Code for Memcache command to provide additional item stats
  
  
• Simulation framework for performance analysis of a Memcache cluster
  
  
• Scripts for logging android resource usage and estimating smartphone power consumption
  
  
• FCFS client-server load generation tool for benchmarking
  
  
• Hadoop 2.x installation and cluster management scripts