@hargav  Please create a new community question for your queries around MergeRecord processor.  This is the best way to get attention and best for community to have a separate thread for each specific query.   I am not clear on your use case for using "cron driven" scheduling with the MergeRecord.  This would not be a common thing to do. Best to explain your use case in a new community thread along with sharing your MergeRecord processor configuration.  Feel free to @MattWho in the new community post to notify me. Thanks, Matt ... View more

This is a kind of bug of Apache NiFi. I've been noted that when you create a flow that, for some reason, does not return a proper HTTP response for every HTTP request (this is the case in the flow you created), that specific thread got stuck forever (it doesn't timeout). You can force threads to terminate (right clicking on HandleHttpRequest), but this will lead you to a situation where the port insist in being binded by NiFi (you can confirm this by running "netstat" in a terminal). The only way I am aware that un-stuck the binding ports when this happens, is restarting the whole NiFi process. If this is not possible/desirable, you can also change the port that your processor is supposed to listen to. I successful avoided this from happening assuring there is no path that doesn't leads a HandleHttpRequest to a HandleHttpResponse. If you also use restricted components (e.g ExecuteScript), you should also assure your processor does not raise uncatched exceptions. ... View more