this is not a memory leak, this is the correct behavior of NPrinting 17.
I try to explain how it works:
Multiple postgresql connections:
When NPrinting 17 Server starts, it opens up a pool of 20 or more connections to the underneath database (postgresql) in order to have always a quick connection available (those 20 connection are no resources-consuming until the server uses them).
Multiple QV.exe instances:
When you run a preview or a task within the designer or the web console, the NPrinting 17 Server calculates the number of objects (charts, images, levels) needed in order to produce the report. When it's done, it sends the order to the NPrinting 17 Engine to open-up as many QV.exe instances ( less or equal the number of CPU cores of the engine machine ) as needed in order to perform as many parallel jobs as possible.
"Zombie" QV.exe instances:
This is a known issue. With the next SR1 of NPrinting 17 (May) we'll optimize this process and you should no longer see "zombie" QV.exe instances.
I hope that i gave you all the needed info. But if you need more details don't hesitate to ask.
Another thing is that NPrinting is not recommended to be installed on the same server as QlikView. Help document says that it will consume as much resources as possible, so it should be different box!
Qlik NPrinting Engine and Server, and QlikView Server should not be installed on the same machine. There are no known incompatibilities between these two products, but in a production environment, Qlik NPrinting requires a separate machine (virtual or physical) from QlikView Server. The reason for this is that Qlik NPrinting Engine and Server will consume all the available resources of the machine they are installed on. In a production environment this may lead to QlikView Server crashing because of lack of resources.
regarding zombie QV.exe - as long as you use QVP connection they will not harm, even if they stay open! also as Gianluca Perin says there are updates coming - so stay tuned!
When applicable please mark the appropriate replies as CORRECT. This will help community members and Qlik Employees know which discussions have already been addressed and have a possible known solution. Please mark threads as HELPFUL if the provided solution is helpful to the problem, but does not necessarily solve the indicated problem. You can mark multiple threads as HELPFUL if you feel additional info is useful to others.