Flurry CSV下载（API）事件计数不匹配

Question

我正在使用API​​测试我的应用程序事件的CSV下载。我注意到CSV在同一时间段内针对不同的呼叫具有不同的事件计数。所有数据（每次下载）对于我的应用和请求的时间段都是正确的。有谁知道他们是否对数据进行采样以创建要下载的文件？

编辑为包括样本调用，提取代码和相同时间段内两次调用的结果。

通话

str_init = '20191101'
str_end = '20191102'

# Call data extraction for Flurry from IOS app
get_csv_from_flurry(str_init, str_end, 'IOS')

提取代码

from datetime import datetime
from dateutil import parser
import requests
import json
import time

from functions.ribon_path import ribon_root_path_join
from functions.ribon_s3_integration import ribon_upload_to_s3

"""
Make CSV extraction from flurry based on initial date (yyyy-mm-dd), end date (yyyy-mm-dd) and platform
Save Uncompressed CSV locally for processing
Save compressed file (parquet) to S3 for backup
"""
def get_csv_from_flurry(str_ini, str_end, str_platform):

    # Convert time period to datetime format
    dt_ini = parser.parse(str_ini)
    dt_end = parser.parse(str_end)

    def unix_time_millis(dt):
        # Convert date periods to unix milisecon epoch
        epoch = datetime.utcfromtimestamp(0)
        return (dt - epoch).total_seconds() * 1000.0

    epoch_ini = unix_time_millis(dt_ini)
    epoch_end = unix_time_millis(dt_end)

    #print(epoch_ini)
    #print(epoch_fim)

    if str_platform == 'IOS' :
        Flurry_apiKey = 'XXX' 
    else :
        Flurry_apiKey = 'YYY'

    # Build the parameters of the post request to the flurry API
    url = 'https://rawdata.flurry.com/pulse/v1/rawData'
    payload = {"data": {
                "type":"rawData",
                "attributes":{
                       "startTime": epoch_ini,
                       "endTime": epoch_end,
                       "outputFormat": "CSV",
                       "apiKey": Flurry_apiKey
                       }
                            }
               }
    headers = {"accept": "application/vnd.api+json",
               "authorization": "Bearer ZZZ",
               "cache-control": "no-cache",
               "content-type": "application/vnd.api+json"
               }
    #print(payload)
    # Make the request
    print('Make Request to Flurry')
    r = requests.post(url, data=json.dumps(payload), headers=headers)
    #print(r.content)

    # Test the return, get the status, download url and request id
    test = r.json()
    #print(teste['data']['attributes']['s3URI'])
    #print(teste['data']['id'])
    r_s3URI = test['data']['attributes']['s3URI']
    r_id = test['data']['id']

    # Check if the download link is ready
    url = 'https://rawdata.flurry.com/pulse/v1/rawData/' + r_id + '?fields[rawData]=requestStatus,s3URI'
    #print(url)
    payload = {}
    headers = {"accept": "application/vnd.api+json",
               "authorization": "Bearer ZZZ",
               "cache-control": "no-cache",
               "content-type": "application/vnd.api+json"
               }
    print('Request OK')

    # Check each minute if the download link is ready
    print('Start Pooling to Check if the File is Ready for Download')
    while r_s3URI == None:
        time.sleep(60) 
        # Make the request
        r = requests.get(url, data=json.dumps(payload), headers=headers)
        print(r.content)
        test = r.json()
        #print(test['data']['attributes']['s3URI'])
        r_s3URI = test['data']['attributes']['s3URI']

    # When the download is ready, get the file and save 
    # Set local folder to save file
    flurry_filename = str_ini + '_' + str_end + '_' + str_platform + '.csv.gz'
    flurry_path_gz = ribon_root_path_join('data', 'Flurry_Download', flurry_filename)

    # Download the file
    print('Start Flurry Download')
    myfile = requests.get(r_s3URI)
    open(flurry_path_gz, 'wb').write(myfile.content)

On the link there is an image with the 2 files I got, they are not the same size and don't have the same number of records

Answer 1

在Flurry支持的帮助下，我发现了差异。对于15天以上的API下载，API调用每次都会给出相同的数字。最多15天的API调用通常会获得不同的结果（具有更多记录的更新调用）。通话时间越长，差异就越小，因此我同意可以将其归因于迟到事件的支持。Flurry不在线，并且通过在移动设备上排队并将数据转储到服务器来工作。