hi, song wang
What does flink's preprocessing mean? Does it mean the data of HLL generated by
flink?
I understand that flink does not need to care about the HLL type. The HLL data
is generated by HLL_HASH(column) when it is imported. So you only need to
import the existing data in flink into doris through streamload, and then set
the corresponding hll field in the header.
for example:
curl --location-trusted -u root: -H "label:label_test_hll_load" \
-H "column_separator:," \
-H "columns:dt,id,name,province,os, pv=hll_hash(id)" -T test_hll.csv
http://fe_IP:8030/api/demo/test_hll/_stream_load
> 2022年7月11日 下午12:43，song wang <[email protected]> 写道：
>
> Thank you for your reply, The introduction in the link is very simple,
> I didn't find the pre-processing related to Flink, Could you explain it in
> more detail?
