Mastering Network Ops: 36 Strategies and Real-World Case Studies

Zhang Yongfu Solution architect at Dahe Cloud Union, veteran network engineer with experience in telecom, finance, government, and transportation projects, focusing on SDN design, network architecture, and automation since 2016.

Network engineers constantly face diverse technical problems and faults. Based on experience, we summarize the “36 Strategies for Network Operations” into three categories:

Knowledge‑driven troubleshooting : Continuously learn, improve technical skills, and extract lessons from each incident to sharpen logical thinking.

Automation and process governance : Shift from manual to automated operations, reduce cost and complexity, and boost efficiency through well‑defined workflows.

Cross‑department collaboration : Coordinate with upstream and downstream teams to resolve issues faster and more effectively.

Case 1 – Developing the “peeling‑the‑onion” skill in fault isolation

Effective troubleshooting requires solid fundamentals and extensive hands‑on experience. The following story illustrates how a lack of basic knowledge can lead to unclear fault‑resolution paths.

Fault scenario recreation

During a night shift, an alarm sounded indicating a problem on a PE router. Initial checks showed BGP session flapping and high CPU usage on one line‑card.

Further inspection revealed the fourth line‑card’s CPU at ~80% and occasional IPC error logs.

Recognizing the ipc_send_rpc_blocked warning, the engineer attempted a line‑card reboot, which did not resolve the issue. Subsequent port‑by‑port shutdown identified the fifth port as the trigger; disabling it stopped BGP flapping and restored CPU levels.

Investigation showed that the port was a VLAN‑based customer gateway forming a Layer‑2 loop, generating massive ARP broadcasts. After the customer corrected the loop and the PE router was re‑connected, all services returned to normal.

Post‑mortem reflections

Insufficient knowledge of legacy equipment necessitates expert assistance.

Fault analysis must combine log review with configuration verification.

When multiple symptoms appear, a holistic view prevents tunnel‑vision on a single cause.

Case 2 – Leveraging automation to boost operational efficiency

Traditional fault‑handling model

Even in an SDN‑centric company, engineers still need to manually log into controllers and switches to locate faults, assess impact, and verify service health, making the workload comparable to traditional network operations.

Building an automated operations platform

Adopting a DevOps mindset, the company formed a cross‑functional team (network engineers, system engineers, developers, and data analysts) to design and deliver an SDN‑based automation platform within two months.

The platform’s alarm module aggregates and filters alerts using big‑data analysis, routing high‑priority incidents to phone calls, medium‑priority to messaging apps, and low‑priority to a searchable log.

After deployment, on‑call staff no longer need constant screen monitoring; they receive concise impact reports on their phones and can coordinate resources efficiently.

Continuous feedback from operators drives iterative development: new requirements are coded, tested, and released, while a knowledge base and self‑healing modules evolve from each case.

To date, the automation system has reduced manual workload by 60%, freeing engineers to focus on innovation.

Full version of the 36 network‑operation strategies is available in the accompanying images.