In BigQuery I'm making a partitioned data table (partitioned by hour) and while data are going into it, it doesn't appear that the results have a _PARTITIONTIME
pseudocolumn; when I do
SELECT
_PARTITIONTIME AS pt,
*
FROM
[my_dataset.my_partitioned_table]
LIMIT
1000
I get all the regular columns of my table, but _PARTITIONTIME
is null for every entry. The data are being sent in from a call to the Go BigQuery API the same way they were when I was sending the data to an unpartitioned table, and they're queried from the BigQuery console. Would the data be more likely to be being inserted wrong or queried wrong?
Currently only DAY is supported as a partitioning type!
See more details in timePartitioning property
And see more about Partitioned Tables in general