Network Connectivity & Service Monitoring: Tools, Protocols & Troubleshooting

```figure
{"type":"tree","root":{"v":"Network Connectivity & Service Monitoring","children":[{"v":"OSI Model","children":[{"v":"Layer 1: Physical (Cables, Hubs, Signals)"},{"v":"Layer 2: Data Link (Switches, MAC, Frames)"},{"v":"Layer 3: Network (Routers, IP, Packets)"},{"v":"Layer 4: Transport (TCP/UDP, Ports)"},{"v":"Layers 5-7: Higher Layers (Apps, Sessions, Presentation)"}]},{"v":"Connectivity Tools","children":[{"v":"ping (ICMP)","caption":"Checks reachability"},{"v":"traceroute (Path Analysis)","caption":"Maps network hops"},{"v":"netstat (Ports)","caption":"Shows active connections"},{"v":"nslookup (DNS)","caption":"Resolves domain names"}]},{"v":"Service Monitoring","children":[{"v":"SNMP (Traps/Queries)","caption":"Device alerts"},{"v":"Syslog (Logs)","caption":"Centralized logging"},{"v":"Nagios/Zabbix (Dashboards)","caption":"Visual monitoring"}]},{"v":"Real-World Use Cases","children":[{"v":"eSewa: TLS + ICMP for payment routing"},{"v":"Ncell: SIP + SNMP for call monitoring"},{"v":"Daraz: HTTP + TCP for order tracking"}]}]},"caption":"Hierarchical breakdown of network connectivity components and tools"}

Core Concepts: How Networks Connect and Monitor Themselves

1. The OSI Model: Layers That Build Connectivity

Networks are organized into 7 layers (Open Systems Interconnection model). Each layer handles a specific task, and problems often appear at a single layer. Below is a layered breakdown with real-world examples:

Why it matters:

• Troubleshooting: If a website loads slowly, check Layer 7 (Application) for server issues or Layer 4 (Transport) for packet loss.
• Security: Layer 6 (Presentation) handles encryption (e.g., HTTPS in eSewa).
• Performance: Layer 4 (TCP/UDP) determines speed vs. reliability.

2. Connectivity Verification: Tools to Check If Networks Work

Before monitoring services, you must verify connectivity. Use these command-line tools (Linux/Windows):

Worked Example: Diagnosing a Slow eSewa Payment

1. Step 1: Run ping eSewa.com → High latency? Problem at Layer 3 (Network) or Layer 4 (Transport).
2. Step 2: Run traceroute eSewa.com → If packets drop at 10.0.0.5, the ISP’s router is faulty.
3. Step 3: Check netstat -ano on the server → See if port 443 (HTTPS) is listening.

3. Service Monitoring: Keeping Networks Healthy

Once connectivity is confirmed, monitor services to ensure they run smoothly. Key methods:

A. SNMP (Simple Network Management Protocol)

• How it works:
  • Manager (e.g., Zabbix) queries devices (e.g., routers, switches) for stats.
  • Devices send traps (alerts) when issues occur (e.g., high CPU).
• Example:
  • Ncell monitors 3G towers using SNMP to detect signal drops.
  • Command: snmpwalk -v 2c -c public router-ip system

sequenceDiagram
    participant Manager as SNMP Manager (Zabbix)
    participant Device as Router/Switch
    Manager->>Device: SNMP GET-REQUEST (CPU Usage?)
    Device-->>Manager: SNMP RESPONSE (85% CPU)
    Device->>Manager: SNMP TRAP (CPU > 90%)
    Manager->>Admin: Alert! High CPU on Router-1

B. Syslog (System Logging)

• How it works:
  • Devices (servers, routers) send log messages to a central server.
  • Example: A Daraz server logs ERROR: Database timeout at 14:30.
• Command:

  tail -f /var/log/syslog | grep "error"
  

C. Nagios/Zabbix (Monitoring Dashboards)

• Features:
  • Graphs for bandwidth, CPU, memory.
  • Alerts via email/SMS (e.g., "NTC’s router down!").
• Example:
  • NTC uses Zabbix to monitor fiber optic links in Kathmandu.

In the Real World

1. eSewa’s Payment Routing

   • Idea Used: Layer 4 (TCP) + ICMP (ping)
   • How? When you pay via eSewa:
     • Your phone checks connectivity with ping eSewa.com (ICMP).
     • The payment uses TCP (port 443) for secure data transfer.
     • If ping fails, eSewa shows "Network Error."

2. Ncell’s Call Monitoring

   • Idea Used: SIP (Session Initiation Protocol) + SNMP
   • How? When you call a friend:
     • Ncell’s SIP server (Layer 5) sets up the call.
     • SNMP monitors tower CPU to prevent drops.
     • If a tower’s CPU hits 95%, SNMP sends a trap to technicians.

3. Daraz’s Order Tracking

   • Idea Used: HTTP (Layer 7) + TCP (Layer 4)
   • How? When you track an order:
     • Daraz’s server sends HTTP requests (Layer 7) to your phone.
     • TCP ensures the order status arrives intact.
     • If ping daraz.com fails, you see "Service Unavailable."

Common Network Issues and Fixes

Worked Example: Fixing Kathmandu Traffic (Network Analogy)

• Problem: Heavy traffic on Thapathali-Mahaboudha route (like a network bottleneck).
• Solution:
  1. Monitor: Use traceroute to see where packets delay (like traffic cameras).
  2. Fix: Add a new route (Layer 3) or upgrade bandwidth (Layer 1).

Exam Tip: How to Score Full Marks

1. For "Why is service monitoring important?" (5 marks):

   • Structure: Use the 5W1H formula:
     • What? (Definition of monitoring)
     • Why? (Prevent downtime, detect attacks)
     • How? (SNMP, Syslog, Nagios)
     • Where? (Data centers, ISPs like NTC)
     • When? (24/7, real-time)
     • Who? (Network admins, IT teams)
   • Example Answer:

     "Service monitoring ensures networks like Ncell’s 4G towers operate without failures. Tools like SNMP detect hardware issues (e.g., overheating routers), while Syslog logs security breaches. Without monitoring, outages (like during Dashain) would go undetected, costing millions. Real-time alerts (e.g., 'CPU > 90%') allow admins to act fast."

2. For "Basic commands to check connectivity" (5+5 marks):

   • Must include:
     • Linux: ping, traceroute, netstat, nslookup, ifconfig
     • Windows: ping, tracert, netstat, ipconfig
   • Bonus: Add a real-world scenario (e.g., "To check if eSewa’s server is up, I’d run ping eSewa.com and then traceroute eSewa.com to find delays.").

3. For "Explain the OSI model" (5 marks):

   • Draw the 7 layers (use the Mermaid diagram above).
   • Link to real examples:
     • Layer 1: Fiber cables in NTC’s backbone.
     • Layer 3: Routers directing traffic to Daraz.
     • Layer 7: WhatsApp messages.

4. Avoid common mistakes:

   • ❌ Saying "OSI has 5 layers" (it’s 7).
   • ❌ Confusing traceroute (Linux) with tracert (Windows).
   • ❌ Forgetting real-world ties (exams love eSewa/Ncell examples).

Final Checklist Before Exam: ✅ Can you draw the OSI model and label 3 layers with examples? ✅ Do you know 5 tools to check connectivity and 2 for monitoring? ✅ Can you explain SNMP with a sequence diagram? ✅ Can you relate ping/traceroute to eSewa/Ncell/Daraz?