[jira] [Resolved] (NIFI-10022) InvokeHTTP. Retry Relationships

["David Handermann (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/NIFI-10022?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

David Handermann resolved NIFI-10022.
-------------------------------------
    Resolution: Information Provided

The framework retry relationships configuration will work based on the configuration, but after eventual failure, FlowFiles will be routed to the appropriate relationship. In this case, an HTTP 404 would be routed to the No Retry relationship after exceeded the configured number of retry attempts. That may not be the most intuitive approach for every processor, such as InvokeHTTP where there are multiple types of failures, but that is the intended behavior.

> InvokeHTTP. Retry Relationships
> -------------------------------
>
>                 Key: NIFI-10022
>                 URL: https://issues.apache.org/jira/browse/NIFI-10022
>             Project: Apache NiFi
>          Issue Type: Bug
>          Components: Core Framework
>            Reporter: Zaytsev Sergey
>            Priority: Major
>         Attachments: Test_case_InvokeHTTP.xml
>
>
> Hi,
> I observe incorrect operation of repeated requests in the InvokeHTTP processor.
> Send a request to the address, ex. GET www.google.com/1. 
> Get response 404.
> I go to repeat on the No Retry relationships.
> Setting retry:
> "Number Of Retry Attempts": 5
> "Retry Back Off Policy": Penalize
> "Retry Maximum Back Off Period": 10 sec
> If "Number Of Retry Attempts" is exceeded, attribute retryCount.\{processor_id} is reset.
> Requests will be repeated until a 2xx response is received or a timeout expires.
> I assume that when the connection limit is exceeded, the executable file of the stream should be sent along the Failure relationships.
> NiFI 1.16.1



--
This message was sent by Atlassian Jira
(v8.20.10#820010)