Glossary

average core utilization

• the average utilization of all cores on a platform while executing a given workload

bandwidth

• rate of data transfer across a network link, measured in bytes per second

cluster compute service

• a compute service that manages multiple machines

compute node

• the worker nodes in a cluster compute service that perform work as requested by the frontend node

compute service

• a software service that manages compute resources and handles requests for executing computational tasks

cyber infrastructure

• a set of hardware resources (hardware infrastructure) managed by software services (software infrastructures) that make it possible to execute application workloads, such as workflows

data locality

• the concept of moving computation closer to where data resides as this can be more efficient than sending large amounts of data to where a compute resource is available

directed acyclic graph (DAG)

• a directed graph containing no cycles, often used to represent a workflow

floating point operation (flop)

• a standard unit of “amount of work to do” used for scientific applications (as many of such applications perform mostly floating point arithmetic); computational hardware resources then have speed measured in flop/sec, i.e., how much “work” they can perform per time unit

frontend node

• the entrypoint to a cluster compute service that handles job requests from a WMS and dispatches those requests to compute nodes in the cluster

horizontal scaling

• augmenting a compute service by upgrading it with more machines

join (Workflow structure)

• a workflow structure where multiple tasks “join” into a single task

latency

• the time it takes for a single packet to travel from one end of a network path to the other end

makespan

• the overall execution time of an application execution (i.e., the duration of the time internal between the beginning of the execution and the completion of the last task).

parallelism

• the ability for multi-core architectures to execute multiple instructions in parallel

persistent storage

• a storage device, such as a hard drive, that retains data even after its power source has been shut off

staging (a file)

• the process of storing certain application files (i.e., initial input files) on some persistent storage prior to the application execution

storage service

• a software service that manages persistent storage and handles requests for creating, deleting, reading, and writing files

turnaround-time

• a tasks end time minus the time it had been submitted to be run

utilization

• the usage of a core while executing a given workload

vertical scaling

• augmenting a compute service by upgrading it with a faster processor or one with more cores

workflow

• an application comprised of individual computational tasks that must execute in some particular sequence to produce a final desired output

workflow management system (WMS)

• a software tool that automatically orchestrates the execution of a workflow on a cyber infrastructure