I'm running tap-braintree with target-stitch on PythonAnywhere.com which is a simple Python hosting platform (with a simple bash command scheduler) built on AWS EC2.
Apparently because they operate on a shared environment with a large cluster, normal Linux process management doesn't really work.
Maybe a command line option to disable memory reporting would be helpful if anyone runs into this? It might be too much of a fringe case to worry about - although other than this issue, PythonAnywhere is actually a great environment for running singer because you can schedule bash commands daily or hourly.
I'm running tap-braintree with target-stitch on PythonAnywhere.com which is a simple Python hosting platform (with a simple bash command scheduler) built on AWS EC2.
Apparently because they operate on a shared environment with a large cluster, normal Linux process management doesn't really work.
Maybe a command line option to disable memory reporting would be helpful if anyone runs into this? It might be too much of a fringe case to worry about - although other than this issue, PythonAnywhere is actually a great environment for running singer because you can schedule bash commands daily or hourly.
If anyone else runs into this issue, I created a fork here with the extremely inelegant solution of commenting out the MemoryReporter.start() line. You can install pip install git+https://github.com/chrisgoddard/target-stitch