Overview
High latency and packet loss are among the most common network incidents in SD-WAN and multi-site environments. This guide walks through a systematic approach to diagnosing these incidents using NetOp.
Step 1: Identify the Incident
- Navigate to Dashboard → Incidents
- Filter by category: Service Performance or Network Utilization
- Look for incidents with main cause: ISP QoE, SLA QoE, Packet Loss, or Latency
- Click → to open the incident detail
Step 2: Review Observations
In the Observations tab:
- Note the start time — does it correlate with a known change or event?
- Check which interface/element is affected (e.g., wan1 on a specific firewall)
- Note severity progression — did it start Medium and escalate to Critical?
Step 3: Check Insights
The Insights tab provides the automated root cause. Common insights for latency/packet loss:
- "ISP Gateway latency exceeded 150ms on wan1 for 20 cumulative minutes"
- "SLA QoE for Zoom was Poor (below 3.1) for 45 minutes"
- "Packet loss on wan1 was above 2% for 30 minutes"
Step 4: Analyze via Statistics
Click the → Statistics shortcut within the incident. NetOp automatically opens the Statistics view with the relevant metrics pre-selected in the correct time range.
Look for:
- Correlation between latency spike and throughput spike (congestion-induced latency)
- Whether packet loss occurred on one WAN interface only (ISP issue) or all interfaces (device issue)
- Whether the issue is periodic (scheduled backup, peak hours) or continuous (ISP fault)
Step 5: Cross-Reference Changes Log
Navigate to Operations → Changes Log and check for configuration changes around the incident start time. Common change-induced causes:
- WAN interface bandwidth limit changed (affects QoE calculation)
- SD-WAN policy modified (traffic rerouted to a worse link)
- Firmware update applied to the firewall
Step 6: Check VPN Graph Topology
If the affected site uses VPN tunnels, open the VPN Graph Topology tab. Look for:
- Red or orange VPN links (degraded or down)
- Affected tunnel's latency and traffic stats in the right pane
Common Root Causes
| Symptom | Likely Cause |
|---|---|
| Latency high, throughput normal | ISP routing issue or distance |
| Latency and throughput both high | ISP link congestion |
| Packet loss only, latency normal | Faulty ISP link or hardware |
| All metrics degrade simultaneously | Firewall CPU/memory saturation |
| Periodic spikes | Backup job, QoS misconfiguration |
Escalation Path
If NetOp Insights point to an ISP issue:
- Run an ISP-provided looking-glass or ping test from the firewall
- Open a ticket with the ISP, referencing the NetOp incident ID and time range
- Use NetOp's Executive Summary Report as supporting evidence
If the issue is device-related:
- Check firewall CPU and memory in Statistics
- Review session count — high session counts can cause latency on lower-end devices
- Consider a firmware update or hardware refresh
Comments
0 comments
Please sign in to leave a comment.