The reason this isn't working is because AbstractDatabaseFetchProcessor is in nifi-standard-nar which is not on the classpath of your custom NAR at runtime. You could add a NAR dependency in your NAR pom on nifi-standard-nar (you already have the dependency but its currently marked provided): <dependency> <groupId>org.apache.nifi</groupId> <artifactId>nifi-standard-processors</artifactId> <version>1.2.0</version> <type>nar</type> </dependency> However, the better approach here would be to refactor things such that AbstractDatabaseFetchProcessor lived inside of some utility JAR that could be re-used by standard NAR and your NAR. There could be a nifi-db-utils module here: https://github.com/apache/nifi/tree/master/nifi-nar-bundles/nifi-extension-utils That would be a cleaner approach and follow the pattern used for other abstract processors. ... View more

You have to implement a custom reporting task that pages through the events and does something with them... A ReportingTask has access to a ReportingContext which has access to EventAccess which has a method for paging through the flow change events. Here is an example: https://github.com/bbende/incubator-atlas/blob/NIFI/addons/nifi-bridge/nifi-atlas-bundle/nifi-atlas-reporting-task/src/main/java/org/apache/nifi/reporting/atlas/AtlasReportingTask.java#L117-L139 ... View more

When a processor is started, it shows green which means it is scheduled to run accordingly to the scheduling strategy, and when its stopped it shows red which means it is not scheduled to run. So even if a processor is on a CRON schedule for once a day, it will be green all the time because its still scheduled, it will only be red if you specifically say to stop the processor. When the processing is triggered to run on a CRON schedule, it doesn't run for a certain amount of time, it runs once (one call to onTrigger of the processor), so it depends what the call to onTrigger does... GetHDFS has a Batch Size property which specifies how many files to pull in one execution, so you would need your batch size to be greater than however many files are going to be in the directory so that it can grab them all in one execution. Alternatively, ListHDFS should list all files newer since lasting, so you could use ListHDFS + FetchHDFS. ... View more