Discrepancies in Binance Historic data

API Futures API
1

There appear to be significant discrepancies in historic data for Binance 15m BTCUSDT Kline futures.
Data downloaded from data.binance.vision/?prefix=data/futures/um/monthly/klies/BTCUSDT/15m

  1. March 2 2021 (2021-03-02 01:00:00.000Z) the number of trades and volume differ greatly from the prior kline (ending at 00:45:59.999Z). Then the next 3 klines all have no price changes and zero volume (the final ignored field is not shown here).
    00:45 kline 1614645900000 49681.98 49782.22 49310 49401.01 4229.44 1614646799999 209252269.9 39210 1920.517 95032171.34
    01:00 kline 1614646800000 49401 49413.38 49361 49361.01 63.932 1614647699999 3157740.221 457 28.1 1388152.24
    01:15 kline 1614647700000 49361.01 49361.01 49361.01 49361.01 0 1614648599999 0 0 0 0
    (01:30 and 01:45 are identical to 01:15 kline other than the time stamps)

  2. for 1 January 2021, data from the historic file does not fully match either the kline from
    fapi/v1/klines OR fapi/v2/continuousKlines (PERPETUAL contract type):
    File (organized to correspond to API return):

	[
		1609459200000,
		28948.19,
		29045.93,
		28706.00,
		28786.75,
		3181.960,
		1609460099999,
		91834596.81370,
		24166,
		1277.389,
		36899008.19096
		0
		]

fapi/v1/klines:

    [
        1609459200000,
        "28948.19",
        "29045.93",
        "28706.00",
        "28786.75",
        "42236.270",
        1609460099999,
        "1220393434.49536",
        1895,
        "19324.997",
        "558571428.81492",
        "0"
    ]

continuous klines:

    [
        1609459200000,
        "28907.76",
        "29011.21",
        "28697.39",
        "28750.21",
        "42236.270",
        1609460099999,
        "1220393434.49536",
        1895,
        "19324.997",
        "558571428.81492",
        "0"
    ]

Prices from historic file match Prices from fapi/v1/klines but Volume does NOT.

Volume from fapi/v1/klines and fapi/v1/continuousKlines matches but Prices do NOT.
These glaring data discrepancies at the least do not bolster confidence in Binance data integrity, and without data integrity everything else is also suspect.
I do note that there appears to be consistency with current klines (fapi/v1/klines and fapi/v1/continuous klines, retrieved today December 25).
Questions:

  1. Will these issues (e.g., the zero values in March data) be fixed?
  2. Is the URL I used for historic data (manual downloading zip files by month) correct?
  3. What is the difference between Futures klines (fapi/v1/klines) and fapi/v1/continuousKlines with a contract type of PERPETUAL?
    On the face of it, it seems to me they are/should be the same.
2

Please raise an issue here.