We work with a trigger mechanism using wait/notify to ensure that a sequence will be maintained. Basically a new flowfile will only be worked upon if a corresponding special triggerflowfile has been received. As soon as the flowfile has been 'finished' the flow itself creates the trigger (kind of like a receipt) for the next expected data waiting in line. This works in our use case as we have available a sequence or generation identifier as metadata on the incoming files without any gaps.
-----Ursprüngliche Nachricht----- Von: Van Autreve Dries <[email protected]> Gesendet: Donnerstag, 1. April 2021 11:44 An: [email protected] Betreff: Re: Strict order of flow files in a cluster Hello Harald Thanks for the swift reply. If I read about this feature, I see in the user guide: "This means that in a 5-node cluster, for example, there may be up to 5 incoming FlowFiles being processed simultaneously." In the tooltip: "Only a single FlowFile is to be allowed to enter the Process Group at a time on each node in the cluster." So it's not accros the cluster. This is also the behaviour I notice after a quick test. On 01/04/2021, 10:24, "Dobbernack, Harald (Key-Work)" <[email protected]> wrote: I suppose the feature 'Ability to specify group level flow file concurrency - for instance run a single flow file end to end before running another for traditional job handling' available from Version 1.12 upward should be helpful here (have not tried myself yet) Harald Dobbernack Key-Work Consulting GmbH | Kriegsstr. 100 | 76133 | Karlsruhe | Germany | www.key-work.de<https://www.key-work.de> | Datenschutz<https://www.key-work.de/de/footer/datenschutz.html> Fon: +49-721-78203-264 | E-Mail: [email protected] Key-Work Consulting GmbH, Karlsruhe, HRB 108695, HRG Mannheim Geschäftsführer: Andreas Stappert, Tobin Wotring
