App runner deployment results in failed service that is inaccessible with timeout

The previous ~6 months, this has not happened, but now 2-3 times in the last two days, we see the following events:

- Successful build of project
- Successful health check
- Dead service

Dead service eventually shows an error message which I will look for, but something like upstream service inaccessible. For reference this is a snippet of the terraform:
`code_configuration {
code_configuration_values {
build_command = "PYTHONUSERBASE=/app pip3 install --user -r requirements.txt && pip3 install pipenv"
port = "8080"
runtime = "PYTHON_311"
start_command = "/app/bin/gunicorn --chdir /app --bind 0.0.0.0:8080 -k uvicorn.workers.UvicornWorker main:app"
runtime_environment_variables = merge(var.backend_env["${var.stage}"],{STAGE = "${var.stage}"})
}
configuration_source = "API"
}
`

I have been pausing and resuming the service to fix the issue. On one occasion, I had to do this twice to get the service running.