Simple Python Script to export the old data from Application Insight without reaching Application Insight's API limit. Exported data can be used for Deep analytics or for Archival purpose. More Details: https://payodatechnologyinc.medium.com/archiving-azure-application-insights-data-for-deep-analytics-cb2a656c1a21

export-data-ai.py

import requests
from datetime import datetime
from datetime import timedelta
import json
import sys
import argparse
import time
import csv

#Arguments Parser
parser = argparse.ArgumentParser(description='Export the data from Application Insight between from_range and to_range in split_range interval.')
parser.add_argument('app_id',  type=str, help='Application Insight\'s Application ID')
parser.add_argument('api_key',  type=str, help='Application Insight\'s API Key with Read Permission')
parser.add_argument('from_range', type=str, help='From Date Range. Format Eg: 2021-01-01T00:00:00Z')
parser.add_argument('to_range', type=str, help='To Date Range. Format Eg: 2021-01-02T00:00:00Z')
parser.add_argument('--split_range', type=int, default=3600, help='Split Range in seconds.To overcome the Application Insights API limits.Default: 3600 seconds (1 Hour)')
parser.add_argument('--save_prefix', type=str, default="data", help='File Name Prefix. Default: data')
parser.add_argument('--sleep', type=int, default=1, help='Sleep time between each API request. To overcome the Application Insights API limits.Default: 1 second')
parser.add_argument('--toCSV', type=bool, default=False, help='Convert the exported data into CSV format. Note:By deafult API returns in JSON format.')

args = parser.parse_args()

# Application Insight's API Endpoint
base_url='https://api.applicationinsights.io/v1/apps/%s/'%(args.app_id)

# URL Headers for API Request with API Key
headers={'X-Api-Key':args.api_key}

# Application Insight Tables
tables=['availabilityResults','browserTimings','customEvents','customMetrics','dependencies','exceptions','pageViews','performanceCounters','requests','traces']

# From and End date range
from_date= datetime.strptime(args.from_range, '%Y-%m-%dT%H:%M:%SZ')
end_date= datetime.strptime(args.to_range, '%Y-%m-%dT%H:%M:%SZ')

while from_date < end_date:	
	
	# Looping through the 10 tables
	for table in tables:

		temp_date=from_date

		# Acutal Application Query - in our case simply the table name
		query={"query": table}

		# Timespan argument for requesting api
		timespan=temp_date.strftime('%Y-%m-%dT%H:%M:%S.000Z')

		temp_date=temp_date + timedelta(seconds=args.split_range) # Add Split Range 

		timespan=timespan+"/"+temp_date.strftime('%Y-%m-%dT%H:%M:%S.000Z')
		
		print(datetime.now(),table,timespan)

		# Forming Query with timespan as paramter to end point
		query_url='query?timespan=%s'%(timespan)		

		# Sending POST request with Query and headers
		r = requests.post(base_url+query_url,json=query, headers=headers)

		# Check the status of API Request
		if r.status_code is 200:

			# File Name to save the exported data
			if not args.toCSV:
				filename='%s-%s-%s.json'%(args.save_prefix,table,timespan.replace("/","_"))
			else:
				filename='%s-%s-%s.csv'%(args.save_prefix,table,timespan.replace("/","_"))
			
			data=r.json()
			if 'error' in data:
				print(datetime.now(),"Error:",data['error'])
			else:
				with open("%s"%(filename),'w') as file:
					if not args.toCSV:
						json.dump(data,file)
					
					else:	# CSV Conversion
						csv_headers=list()
						#Get Columns
						for column in data['tables'][0]['columns']:
							csv_headers.append(column['name'])

						csvwriter = csv.writer(file, delimiter=',',quoting=csv.QUOTE_MINIMAL)
						csvwriter.writerow(csv_headers)

						for row in data['tables'][0]['rows']:
							csvwriter.writerow(row)						

			time.sleep(args.sleep)

		# If api failed just log to stdout with status code and message
		else:
			print(datetime.now(),"API Failed:",r.status_code,r.text)
			sys.exit(1)

	from_date = from_date + timedelta(seconds=args.split_range)