Today, I was developing a dashboard about DEX activities on NEAR in 2024. when I came across a peculiar case. That peculiar case is a chart that you see on the right. As you can see, on April 3, 2024, the dollar volume of trading on the NEAR network shows a very strange increase.

After sorting the transactions by dollar volume on April 3, I realized that the price of the BLACKDRAGON token might not have been recorded correctly. The table on the right shows the price of the BLACKDRAGON token on April 3. It seems that the calculated dollar price in the table ear.defi.ez_dex_swaps was obtained using the table crosschain.price.ez_prices_hourly. At 18:00, 19:00, 21:00, and 22:00, the prices were not recorded correctly.

The image on the right shows the price of BLACKDRAGON on April 3, 2024. As you can see, the price peaked around 14:00, which is still about 60 million times smaller than the price of BLACKDRAGON recorded at 18:00 on the flipside tables.

Upon further observation, I realized that this bug also exists for the PEM token, which you can see in the chart. It also seems that the price of the NEKO token has a similar bug.

My observations indicate that this issue also exists on other dates such as April 4, 2024, and May 3, 2024. On April 4, there is at least one transaction where the dollar amount of the PARAS token was not recorded correctly(tx_hash : 3FSo4FC6vbiJBk6NUbrUeHEU2fnegwLX4fV8sDfoqU ).

Price records for BLACKDRAGON and PEM tokens on April 3, 2024, exhibit notable inconsistencies. Specifically, a dramatic variation in BLACKDRAGON's price at 14:00 was observed, being approximately 60 million times less than its price at 18:00. These anomalies are not isolated events, as further investigation revealed similar patterns across other dates and tokens. This recurring issue highlights a critical concern regarding data reliability on the NEAR network, potentially impacting trading decisions and market analysis.