Hi, We had another server that was not experiencing this bug, but when we changed to the WriteAheadProvenanceRepository (like on the server where we first noticed this) we did see it. Have there been changes in that implementation? Could this cause what we are seeing?
Thanks! Rotem On 28 Dec 2017 3:35 am, "Koji Kawamura" <[email protected]> wrote: Hi Rotem, When I tested it with NiFi 1.5.0-SNAPSHOT, the REPLAY event has its FlowFile UUID as the parent (original) FlowFile UUID as expected. Type REPLAY FlowFile Uuid 8c61fdd7-c084-4756-946c-f5669dc4442d File Size 4 bytes Component Id 9abc21e3-0160-1000-6d6f-a1c408f75b7a Component Name UpdateAttribute Component Type UpdateAttribute Parent FlowFiles (1) 8c61fdd7-c084-4756-946c-f5669dc4442d Child FlowFiles (1) 0428b60c-44cb-46df-85ab-3787a38693e9 The implementation has not been modified. It's been using parentUUID. https://github.com/apache/nifi/blob/master/nifi-nar- bundles/nifi-framework-bundle/nifi-framework/nifi-framework- core/src/main/java/org/apache/nifi/controller/FlowController.java#L4241 Is this different than what you are seeing? Thanks, Koji On Thu, Dec 28, 2017 at 2:46 AM, Rotem Mayo <[email protected]> wrote: > Hi! > > I have been using nifi for a while, and have recently upgraded from 1.2.0 to > 1.4.0. > > I am having problems with replay events. In version 1.2.0, the UUID of a > replay event is the id of the original flowfile that is being replayed. This > is the behavior that is documented in the developer guid ( > https://nifi.apache.org/docs/nifi-docs/html/developer- guide.html#provenance_events). > > But, in nifi 1.4.0, the replay event's UUID is that of the child flowfile > that is created. This change is not documented in the changes log, and I > could not find any ticket related to the subject. In addition, other events > that create child flowfiles, like clone and fork, still behave according to > the documented way. > > Was this change done on purpose? Will other similar events change in the > future? > > Thanks! > Rotem
