There is no difference in setup or services available when the two requests go through. Node Red is running happily all the time. Have you checked the serial monitor output from you core during the request? I suspect you are receiving the same time out during your transactions also.
What has gone wrong is:
- the failure of the first attempt
- the second attempt (and all others) time out.
I’ve written a .net client that works fine (not timing out) and used the Chrome extension (also doesn’t timeout).
The Spark Core HTTPlibrary is seemingly to blame for the failures I just don’t know JS well enough to fix it yet. Node Red is working fine as previously stated.