Custom Workload Data¶
By default, Green-DCC includes workload traces from Alibaba and Google data centers. These traces are used to simulate the tasks that the DC needs to process, providing a realistic and dynamic workload for benchmarking purposes.
Data Source¶
The default workload traces are extracted from:
Expected File Format¶
Workload trace files should be in .csv
format, with two columns: a timestamp or index (must be unnamed), and the corresponding DC Utilization (cpu_load
). The CPU load must be expressed as a fraction of the DC utilization (between 0 and 1). The workload file must contain one year of data with an hourly periodicity (365*24=8760 rows).
Example Workload Trace File¶
,cpu_load
1,0.380
2,0.434
3,0.402
4,0.485
...
Integration Steps¶
Place the new workload trace file in the
data/Workload
folderUpdate the
workload_file
entry inDEFAULT_CONFIG
dictionary inhierarchical_env.py
with the path to the new workload trace file