To reproduce: 1. start up sds services: `sds start all -f` 1. shut down instances using AWS dashboard without stopping sds 1. start up instances again 1. start up sds services: `sds start all -f` Issues with `supervisord.sock` file.