Currently writing a Windows service in .NET and I'm using named pipes to have other processes communicate with my service. In the more complicated constructors of NamedPipeServerStream, there's a parameter with a descriptive name of maxNumberOfServerInstances. Awesome. But what does that mean?
MSDN's documentation is also helpful at explaining:
The maximum number of server instances that share the same name.
Okay. That still doesn't really tell me what this does for me, or how I utilize it. It would make sense to be if NamedPipeServerStream also accepted some delegate for "run this code when I receive a connection", so then each "Server instance" would then run that code. But that's not the case.
In practice it limits the number of NamedPipeServerStream instances you can create for the same pipe name. This will throw System.IO.IOException: All pipe instances are busy.:
var pipeStreamA = new NamedPipeServerStream("mypipe1", PipeDirection.InOut, 1);
var pipeStreamB = new NamedPipeServerStream("mypipe1", PipeDirection.InOut, 1);
If you change to -1 or 2, then it won't throw an exception. Also, it only respects the value on the first instance you create. If you specify 1 on the first call and 2 on the second, it will still throw an exception.
It's probably a good idea to set this to the maximum number of simultaneous servers that you expect to try to run, because as Hans mentioned, it may be using the number as a hint for the amount of resources to allocate.