问题
I have created a partitioned collection on a long field (playerId) and also added a hash index on that field (DataType.Number). When I insert records most of the time it works, but sometimes it gives me a PartitionKey extracted from document doesn't match the one specified in the header.
After I tested this in the Azure Data Explorer I found out there's a rounding problem with long numbers. If I insert 183548146777950021 through Data Explorer it will save it, but then return that same record to me as 183548146777950000. Is this a known issue?
I'm using the latest 1.23.2 of the .NET client, in Direct/TCP mode.
回答1:
If I insert 183548146777950021 through Data Explorer it will save it, but then return that same record to me as 183548146777950000. Is this a known issue?
As far as I know, Azure DocumentDB uses IEEE754 standard for numbers, which could cause the truncation or loss of precision for large integers or higher precision decimal numbers. If possible, you could try to modify and store the playerId field as string "183548146777950021".
And you could refer to this similar issue: Azure DocumentDB decimal truncation.
来源：https://stackoverflow.com/questions/43764755/partitionkey-extracted-from-document-doesnt-match-the-one-specified-in-the-head