The Programmable Benchmarker and Graphing Tool
The problem:
Identifying bottlenecks and problems within a system is essential when developing larger systems. Being
able to quickly produce simple graphs helps see these bottlenecks more clearly as well as possibly see degenerative issues within a system.
progbg is a simple framework that exposes a planning API to allow you to plan execution of
your benchmarks through the registering of workloads and backends. Data produced by these classes are parsed to produce data that can be used by graphs and figures.
progbg utilizes matplotlib uses for its graphing library and tries to reduce the amount of work
needed to create custom graphs.
More information and docs can be found here
progbg can be installed through pip
$ pip install progbgTake a look at a working example plan. Try running it with the command
progbg plan.py -p 8080 and go to localhost:8080 to view the example graphs!
A step by step guide through an example can be found here.
Graphs are saved by default in the graphs directory of the current working directory.
By passing the argument -p PORT to progbe, this will expose figures and graphs to be viewed by your browser
at http://localhost:PORT
Kenneth R Hancock krhancoc <https://github.com/krhancoc>