Skip to main content

Workpackages overview

List of work packages

Deliverables:

WP1: Project management (UC3M)

WP2: Ad hoc storage systems (JGU)

  • 2.1: Definition of requirements for ad-hoc storage systems (JGU)
  • 2.2: Design of the ad hoc storage system (JGU)
  • 2.3: Short-lived ad hoc storage system (BSC)
  • 2.4: Resiliece for ad hoc storage system (BSC)

WP3: Malleability management (TUDA)

  • 3.1: Malleability requirements definition (UC3M)
  • 3.2: Base mechanisms for malleability (UC3M)
  • 3.3: Scheduling algorithms and policies (TUDA)
  • 3.4: Malleable storage systems available (UC3M)
  • 3.5: Slurm malleability plugin available (JGU)

WP4: I/O scheduler (BSC)

  • 4.1: I/O Scheduler requirements and API (BSC)
  • 4.2: Software to support I/O scheduling policies (BSC)
  • 4.3: Final report on the I/O scheduler (BSC)

WP5: Sensing and profiling (DDN)

WP6: Intelligent controller (INRIA)

WP7: Application co-design (FZJ)

WP8: Dissemination and exploitation (PARATOOLS SAS)

Description of work

WP1: Project management (UC3M)

WP2: Ad hoc storage systems (JGU)

  • 2.1: Data layer semantics (JGU) BSC, DDN, CINI
  • 2.2: Short-lived ad-hoc storage systems (BSC) UC3M, JGU, DDN, CINI
  • 2.3: Resilience mechanisms for ad-hoc storage systems (BSC) JGU, UC3M, DDN

WP3: Malleability management (TUDA)

  • 3.1: Base mechanisms for malleability (UC3M) BSC, CINI, JGU, TUDA
  • 3.2: Scheduling algorithms and policies (TUDA) JGU, UC3M, INRIA
  • 3.3: Malleable storage systems (UC3M) UC3M, JGU, BSC
  • 3.4: Integration of malleability features into the job scheduler (JGU) PSNC, TUDA, UC3M

WP4: I/O scheduler (BSC)

  • 4.1: Definition of the I/O requirements, QoS metrics and data movement API (JGU) UC3M, BSC, TUDA, DDN, CINI, PSNC
  • 4.2: I/O scheduling policies and algorithms (INRIA) UC3M, JGU, BSC, TUDA, DDN
  • 4.3: In-situ/in-transit data transformations using low-power processors (MPG) JGU, BSC, CINI, PSNC

WP5: Sensing and profiling (DDN)

  • 5.1: Definition of the requirements and design of monitoring and profiling tools (UC3M) JGU, BSC, TUDA, DDN, PARA, CINI, PSNC
  • 5.2: Scalable monitoring tool for exascale systems (DDN) UC3M, PARA, PSNC, JGU
  • 5.3: Tools for application I/O profiling and modelling (PARA) UC3M,INRIA,PSNC,TUDA
  • 5.4: Predictive performance (DDN) UC3M, INRIA, TUDA

WP6: Intelligent controller (INRIA)

  • 6.1 Intelligent controller design (UC3M) BSC, JGU, INRIA, TUDA, FZJ, PSNC
  • 6.2 Novel runtime analytics tools to tune I/O system behaviour (CINI) BSC, JGU, UC3M, FZJ, PSNC, INRIA

WP7: Application co-design (FZJ)

WP8: Dissemination and exploitation (PARATOOLS SAS)