I am currently trying to use influxDB to collect tick data. However, these data only have millisecond time stamp. That being said, it is extremely easy to have duplicate data points that gets overwritten. One of solution I am currently use is to manually add counters to time stamp. As you can see, this method has 2 drawbacks:
- it cannot handle data with time stamp like x.999999999 because it shifts time interval and potentially shifts its position (say there’s incoming data at 1+x.000000001).
- if the incoming data contains an older than current time stamp, it is hardly to know how many data in the older interval thus potentially overwritten those data.
Therefore, I am hoping if anyone can shine some lights on this issue? Thank you very much!!