1. Aliyun Services involved
- Tablestore
- MaxCompute
- DataWorks
- QuickBI
2. Prepare data
Create Tablestore instance and generate fake user data. step by step document is here.
3. Create tables (in Dataworks)
first, create a business process in Dataworks; create tables:
- create an external table ots_user_trace_log(use ODPS SQL node, to sync data from table store);
- create table ods_user_trace_log;
- create table dw_user_trace_log;
- create table rpt_user_trace_log;
Add resource jar/file, which will be used to translate ip address to region(province/city).
4. Work Flow Design
In DataStudio, add a virtual node and 3 ODPS SQL node to business process; and link them like: start -> ods_user_trace_log -> dw_user_trace_log -> rpt_user_trace_log;
The logical work flow looks like:
graph TD;
A2[Start] -->|ODPS ods_user_trace_log| T1[fa:fa-table ods_user_trace_log];
T1 --> |ODPS dw_user_trace_log| T2[fa:fa-table dw_user_trace_log];
T2 --> |ODPS rpt_user_trace_log| T3[fa:fa-table rpt_user_trace_log];
save the work process, submit and run.
verify data in a temp query builder.
1
select * from rpt_user_trace_log where dt=20220413;
5. Present in QuickBI
This practice was based on 云原生大数据计算服务 MaxCompute > 产品教程 > 搭建互联网在线运营分析平台
