Hi devs, Yuan and I would like to start a discussion about FLIP-221[1], which introduces an abstraction of lookup table cache and its standard metrics.
Currently each lookup table source should implement their own cache to store lookup results, and there isn’t a standard of metrics for users and developers to tuning their jobs with lookup joins, which is a quite common use case in Flink table / SQL. Therefore we propose some new APIs including cache, metrics, wrapper classes of TableFunction and new table options. Please take a look at the FLIP page [1] to get more details. Any suggestions and comments would be appreciated! [1] https://cwiki.apache.org/confluence/display/FLINK/FLIP-221+Abstraction+for+lookup+source+cache+and+metric Best regards, Qingsheng
