Graph-based network monitoring or another approach to repetitive low level network troubleshooting possible?