I create deployment, when I try predict data come from AI Catalog and inject the prediction to server, I get success status in job history and all data also successfully inserted to my target server.
But when I change the source to my server I get the following error "Could not configure intake adaptor"
how to solve this?, just for information the server is same server with data in AI catalog that successfully predicted.
ya, connection still work because I still can upload some data.
like I mention before that data in AI catalog use same connection and same table.
I test the job definition with data in AI catalog first (success) and then I test job definition directly using data from server. It's the same table on my server that was uploaded to the AI catalog and predicted directly.
May be a silly question but in your account data-connections tab does your data connection still work when you select test connection? I'm just thinking something about your servers configuration may have changed after making predictions initially.
Are you also reading data and writing predictions to the same to the same table and database? If not does the table your reading from have a data-connection you can test?