Tom, You have the Minimum Number of Entries set to 2 and Minimum Group Size set to 1 MB. That means that as soon as you have 2 files queued up that total at least 1 MB, it will create an output FlowFile. That output will contain as much data as it can, based on what is queued up at that moment. So if you pull in 2 files, then a few seconds later pull in 2 more, then a few seconds later 2 more, then 2 more, you're going to end up with 8 output FlowFiles, each containing 2 files.
As I understand it, the idea is to periodically pull in everything in the directory, merge them together, and then move on. Later, you'll do another listing, merge those together, and move on, correct? The problem here is that, as you said, you don't know how many files are in the directory, so MergeContent doesn't know how many files to wait for, before merging. So one possibility would be to just set the Minimum Number of Entries and Minimum Group Size to something much larger. Then set "Max Bin Age" to say 30 seconds or 60 seconds. That way, as soon as MergeContent sees a single file, it will wait 30 seconds or 60 seconds or whatever you have set, and then merge together all of the files that it has queued up. Does that help? Thanks -Mark On May 31, 2019, at 8:27 AM, Tomislav Novosel <to.novo...@gmail.com<mailto:to.novo...@gmail.com>> wrote: I forgot to mention that I put Correlation Attribute Name as attribute name which holds directory name from which that 8 files coming from. And there is not always 8 files, this is just for example. The number of files changes always as the files come in the folders. On Fri, 31 May 2019 at 14:24, Tomislav Novosel <to.novo...@gmail.com<mailto:to.novo...@gmail.com>> wrote: Hi all, I need to create one flowfile from multiple flowfiles which are of files in one directory. So if I have e.g. 8 files in directory, I want to merge them and the output I want is 1 flowfile for further processing ( I want to extract folder name and path of that 8 files). I tried with MergeContent processor and this is my setup. <image.png> But every time i run the flow I get multiple flowfiles in the output queue. Sometimes 2, sometimes 3 or even 4. What am I doing wrong? Or is there some other way to do this in Nifi? Thanks in advance, BR. Tom
