+1 Good idea CalvinKirs <[email protected]> 于2022年12月13日周二 14:16写道：
> Hi guys, > > > In our Seatunnel WEB planning, we need to access the capability of a data > source that users can use to work with seatunnel. > > The data source specifically manages the metadata information of various > basic link parameters in a unified manner, and has the ability to persist > and retrieve the corresponding information. It also has connectivity > testing and basic metadata capabilities, such as the internal structure of > the data source. > Users can also use the API or query the data structure of the > corresponding data source on the web page, which minimizes the cost of user > configuration errors and improves the ease of use of the product. > > > Let's take jdbc connector as an example, if the user uses mysql, then in > our previous use, each task needs to configure its own mysql account > password, url and other parameters. > And if we use the data source, that is, we can use the data source to > manage this metadata information, the task only needs to be associated with > the corresponding data source can be, and then need to be individually > configured only the specific parameters of their own jobs, we can also > automatically get the table schema on the page, the user selects the field > needed only to synchronize, without manual configuration. > > > Of course, this is a general demand, AFAIK, Apache DolphinScheduler also > has corresponding demand, so I request to create a new repository, we > provide the corresponding SDK, users can use our SDK > So I request to create a new repository, we provide the corresponding SDK, > users can use our SDK, and then complete their own requirements (this > process includes some permission checks, and some product business > requirements). > > > What do you think? > Best wishes！ > Calvin Kirs > >
