Commit 5eed4743 authored by Waleed Akbar's avatar Waleed Akbar
Browse files

Changes in Analytics. 

- UNSPECIFIED option added in the "AnalyzerOperationMode" enum as a best practice.
- In SparkStreamer, changed the thresholds parameter from optional to compulsory.
parent 990395f4

proto/analytics_frontend.proto

+3 −2
Original line number Diff line number Diff line
@@ -30,8 +30,9 @@ message AnalyzerId { 
}



enum AnalyzerOperationMode {

  ANALYZEROPERATIONMODE_BATCH     = 0;

  ANALYZEROPERATIONMODE_STREAMING = 1;

  ANALYZEROPERATIONMODE_UNSPECIFIED = 0;

  ANALYZEROPERATIONMODE_BATCH       = 1;

  ANALYZEROPERATIONMODE_STREAMING   = 2;

}



message Analyzer {

src/analytics/backend/service/SparkStreaming.py

+1 −2
Original line number Diff line number Diff line
@@ -73,7 +73,7 @@ def ApplyThresholds(aggregated_df, thresholds): 
        )

    return aggregated_df



def SparkStreamer(kpi_list, oper_list, window_size=None, win_slide_duration=None, thresholds=None, time_stamp_col=None):

def SparkStreamer(kpi_list, oper_list, thresholds, window_size=None, win_slide_duration=None, time_stamp_col=None):

    """

    Method to perform Spark operation Kafka stream.

    NOTE: Kafka topic to be processesd should have atleast one row before initiating the spark session.
@@ -86,7 +86,6 @@ def SparkStreamer(kpi_list, oper_list, window_size=None, win_slide_duration=None 
    if window_size        is None: window_size        = "60 seconds"    # default

    if win_slide_duration is None: win_slide_duration = "30 seconds"    # default

    if time_stamp_col     is None: time_stamp_col     = "time_stamp"    # default

    if thresholds         is None: thresholds         = {}              # No threshold will be applied

    

    try:

        # Read data from Kafka