@Mohammed El Moumni Currently there is no way to delete a counter, only reset. There was no delete option because if you deleted the counter, it would be recreated immediately if the processor was still running. The use case of a deleting a processor and wanting to remove the counter associated with it was not considered. Moving forward, the use case will be a clear/delete icon. If the counter is deleted and the processor still exists and is running, the counter will be recreated immediately. If the processor has been removed then the counter will be deleted. ... View more

@Simon Jespersen You need to clear state, this can be done by doing the following: Right click on the processor, and click "View state" Then a new window will pop-up, Then click "Clear state" and then your ListSFTP processor will start fresh. ... View more

@Jatin Kheradiya The NiFi node running on your local system should be configured to use the public IP address of your local system, instead of localhost, so the NiFi running on the EC2 instance is able to resolve the address of you local system. Using localhost is an unresolvable host name to the EC2 NiFi instance, so it cannot find the NiFi node to join the cluster. ... View more

@Simon Jespersen You can use the rest-api to issue a command to stop the processor. But, if you already stopped it via the UI, then issuing a command using the rest-api probably wouldn't make any difference. It sounds like the processor is allowing the last thread to complete before releasing the thread. If it does not complete is a reasonable amount of time, roughly a few minutes, you will have to restart NiFi to force it to release the thread. Try reloading the page and see if that makes a difference. Has this processor not stopped previously when making changes to your dataflow? ... View more

@Bala Vignesh N V There aren't any official documents that cover installing NiFi in Windows. But there are a few blogs that cover the process. Here is link to one of them: Install NiFi on Windows 7 Then, once NiFi is installed, you can come here to find several posts about getting data into HDFS. ... View more

@Adda Fuentes When I have seen this error, it was caused by a corrupt flow file repository file, but that was caused by filling the partition to 100%. Or, the zookeeper servers running embedded in NiFi and the systems were unable to respond within the timeout period. As a test/workaround, try dropping the node out of the cluster and see if it can run standalone. If that does not work, stop NiFi, move the current data in the flow file, content and provenance repositories and restart NiFi, it then should join back into the cluster. ... View more