Hello,
Well it is nothing really special. I’ve been testing it on a jupyter notebook using pymongo Version: 4.0
At first, I let mongodb to handle _id field thinking it could handle duplicate data with the same metadata and timestamp. and I think this is really important for a timeseries database. It should not let insert if the metadata and the timestamp are the same. we shouldnt be dealing with generating _id field for this purpose…
using insertMany, inserted couple of hundred documents. Below is one of the documents;
{
"timestamp": {
"$date": "2021-12-01T18:58:00.000Z"
},
"metadata": {
"chart": "candles",
"interval": "1min",
"market": "USD",
"symbol": "REQ"
},
"open": 0.7027,
"low": 0.7015,
"high": 0.7083,
"close": 0.7074,
"volume": 163137
}
then noticed it does not handle duplicate data so I decided to generate _id field on my application side. _id generated using metadata field and timestamp as below;
{
"timestamp": {
"$date": "2021-12-01T18:58:00.000Z"
},
"metadata": {
"chart": "candles",
"interval": "1min",
"market": "USD",
"symbol": "REQ"
},
"open": 0.7027,
"_id": "REQ-USD-1min-candles-1638385080",
"low": 0.7015,
"high": 0.7083,
"close": 0.7074,
"volume": 163137
}
but still, duplicates were allowed… here is a screenshot
