Error in Parallel Computing Toolbox on Matlab Runtime

I am trying to use the features of the Parallel Computing Toolbox inside a ctf package compiled with Production Server Compiler. The package then is called by a C++ code on a linux server.

The minimal example which fails is:

function parallel()
pool = parpool(5);
delete(pool);
end

The matlab API is intialized with these lines in C++:

std::vector<std::u16string> opts = {u"-startmsg", u"-nodisplay", u"-completemsg"};
auto application = matlab::cpplib::initMATLABApplication(matlab::cpplib::MATLABApplicationMode::OUT_OF_PROCESS, opts);

The script runs without a problem on my local Matlab. But when I run it in on the Matlab Runtime the Error is:

Starting parallel pool (parpool) using the 'local' profile ...

Error using parpool (line 113)
Parallel pool failed to start with the following error.

Error in parallel (line 2)

Caused by:
    Error using parallel.internal.pool.InteractiveClient>iThrowWithCause (line 678)
    Failed to start pool.
        Error using parallel.Job/submit (line 351)
        AppComponentUUID should not be empty

Parallel pool failed to start with the following error.

Can somebody tell me what is going wrong?

Matlab version: R2019a

Matlab Runtime version: R2019a

Answer 1

[Jasper Lammering]

The support answered me and told me that the problem was the MATLABApplicationMode I started the runtime with.

The solution is to start the runtime IN_PROCESS like this:

auto application = matlab::cpplib::initMATLABApplication(matlab::cpplib::MATLABApplicationMode::IN_PROCESS, opts);

The Parallel Computing Toolbox is currently not supported in OUT_OF_PROCESS mode as the support said.