Latest Releases

mdtmGUI

A web-based tool developed to monitor and manage MDTM-enabled DTNs. However, it can also be deployed to monitor and manage any networked computer systems.

Live demo

Version 1.0.1

Release Date: Oct, 2016

OSes supported

  • Linux kernel 3.12 or higher

Support Features

  • Online and real-time monitoring of data transfer status and progress
  • Online and real-time monitoring of DTN system status and configurations
  • Online and real-time monitoring of MDTM-based data tranfer tool status and configurations

Docker container release

Source code release

MDTM Middleware

A user-space resource sceduler that harnesses multicore parallelism to scale data movement toolkits at multicore systems.

Version 1.0.5

Release Date: January 6, 2020

OSes supported

  • Linux kernel 3.12 or higher

Support Features

  • NUMA system topology profiling
  • Online system status monitoring
    • cpu load per core
    • memory load latency per NUMA node
      • This feature allows applications to use system memory intellignently to avoid memory hotspots.
  • NUMA topology-based resource scheduling
  • mdtmFTP generate three types of threads: Network I/O threads, Disk I/O threads, and Management threads. The MDTM middleware will schedule Disk and Network I/O threads to cores close to I/O devices (I/O locality).

  • Supporting core affinity on network and disk I/Os
  • System zoning

    The MDTM middleware will partition system cores into two zones -- a MDTM-zone and a Non-MDTM-zone. A data transfer application will run in the MDTM-zone while other applications will be confined within the Non-MDTM-zone. This strategy will reduce and minimize other applications' interference to the data transfer applicaiton, thus resulting in optimium data transfer performance.

Download links

mdtmFTP Software Package

A high-performance data transfer tool that build upon the MDTM middleware.

Version 1.1.1

Release Date: January 6, 2020

OSes supported

  • Linux kernel 3.12 or higher

Support Features

  • Adotp a pipelined I/O design

    Data transfer tasks are carried out in a pipelined manner across multiple cores. Dedicated I/O trheads are spawned to performed network and disk I/O operations in parallel, with each thread pinned to a particular core.

  • Implement a large virtual file mechanism to efficiently handle lots-of-small-files (LOSF) situations.

    Different from tar-based solutions, the virtual file is logically, instead of physically, created. Data is transferred on a per-dataset basis, instead of on a per-file basis.

  • Use MDTM middleware servcies to fully utilize the underlying multicore system

Singularity container release

Docker container release

Source code release

  • Last modified
  • 01/16/2020