Yeah to double down on Bryan's comments I know this is frequently done. For instance we often would pull data from a website that we'd use for lookups/enrichments in the flow. We'd use GetHTTP or related processors to grab the contents of the given URL constantly and honor things like etags/etc... If anything new it would pull it and we'd send results to PutFile. If necessary you can decompress/etc.. This is a great way to go because even as you add nodes it will automatically happen.
Thanks On Fri, Sep 20, 2019 at 3:40 PM Bryan Bende <bbe...@gmail.com> wrote: > Currently there isn't really a standard approach for this, but you > might be able to make part of the flow get the file from somewhere. > For example, if you had an SFTP server or HDFS, you could put it there > and have processors like ListSFTP -> FetchSFTP -> PutFile. > > On Fri, Sep 20, 2019 at 3:20 PM Christopher J. Amatulli > <camatu...@technicallycreative.com> wrote: > > > > Looking for ideas on how to put a file on every Nifi node in a cluster. > > > > > > > > Use case: We have a file (approximately 25MB) which needs to be written > for reference to all nodes in the Nifi cluster. Since Nifi inherently > mirrors the flow.xml.gz, does a method exist to add “other” files to that > replication process? > > > > > > > > >
