faas-request-pattern

Framework to generate request pattern, based on the faas-topologies repository generated topology datasets.

data

We use the generated topology datasets from https://github.com/edgerun/faas-topologies and the public available taxi trip data from https://chriswhong.com/open-data/foil_nyc_taxi/

• place the .csv topology data file into the data/topologies folder
• place the .csv trip data file into the data/trips folder

prepare dataset

To prepare the trip dataset, execute following command and specify the required parameter. python prepare_trip_dataset.py --path data/trips/example.csv --topology data/topologies/example.csv --name example_output

It creates a new csv file and names it like the given --name parameter. The --trip data got filtered by the maximum boundaries of the given --topology file.

generate request pattern

To generate the pattern use python generate_request_pattern.py --name example --trips data/trips/example.csv --topology data/topologies/example.csv --start "2013-10-01 18:00:00" --end "2013-10-01 19:00:00" where the start and end parameters define the time range in which the pattern will be created. The ouput files are saved in a seperate folder in output/${name}/1_pickups.csv where the number in front of the filename indicates the cloudlet number in the selected topology. The file contains the time deltas of each request in the defined time range. [1.5, 1.5, 1.5, 1.5] means e.g. that 4 requests were sent to the specific cloudlet within an interval of 1.5 seconds.

trip input:

• medallion: string
• pickup_datetime: string
• passenger_count: float
• pickup_longitude: float
• pickup_latitude: float

topology input:

• cell: string
• lon: float
• lat: float
• cloudlet: string