Runbook-Based Troubleshooting of Voice Transient Issue.

Anurag Singh Maurya
40 steps
5 minutes

Netbraintech

**Objective:** Show how NetBrain’s Runbook-based automation accelerates end-to-end troubleshooting by quickly isolating misconfigurations and performance issues across multiple devices—reducing manual effort, improving consistency, and lowering mean time to resolution (MTTR).

***Scenario:*** Recently, users reported intermittent slowness in a critical voice application. When we mapped the application path between the source and destination, we observed that the traffic traversed several routers and switches—any one of which could introduce delay or packet loss. Using NetBrain’s Runbook-based automation, we quickly diagnosed the issue and identified a misconfigured QoS policy on the **US-BOS-R1** router. The voice class was policed at a significantly lower rate than the expected standard, resulting in call degradation during peak periods.\ \ **In this lab we will be:** - **Leveraging a Runbook** to automate manual troubleshooting steps, eliminating repetitive data collection and ensuring a consistent diagnostic workflow. - **Using the Overall Health View automation node** to overlay key operational and capacity metrics directly on the application map, providing an instant snapshot of device health. - **Executing CLI automation** to gather device logs and capture transient or time-sensitive symptoms that may not be visible during live troubleshooting. - **Using a Pre-built Intent** to automatically analyze all devices along the application path, checking whether any device exceeds the configured threshold for packet handling within the voice class-map—an important indicator of potential QoS degradation. - **Validating QoS configurations with the Golden Assessment Library (GAL),** comparing live configurations against golden templates to quickly surface misconfigurations that may impact voice performance. - **Completing the workflow with a Documentation Node** to generate an automated summary of all findings for ticketing, escalation, or post-incident review. **By the end of this lab, you'll see how Runbook-based automation streamlines root cause isolation, reduces manual effort, and ensures consistent QoS across critical voice paths—ultimately improving service reliability and reducing MTTR.**

Navigate to <https://class.netbraintech.com/desktop.html>

Let’s begin by mapping our application path. We’ll start from the source device with IP **10.2.1.100** and trace the connection to the destination device **10.103.1.100**. This represents the actual path taken by the voice application across the data network, traversing multiple intermediate devices, including **US-BOS-R1**. To create an **A-to-B Path** on the map, click the **Path** icon located to the right of the **Search Anything and Create Map...** field. This will open the A and B input fields, allowing us to enter the source and destination addresses and generate the full traffic path.

Click "Path"

Click "Runbook"

Now that we have the application path mapped, we’ll begin troubleshooting the voice issue using Runbook-based automation. This approach mirrors the same experience-driven workflow an engineer would normally follow manually. The idea is simple: **we apply the same CLI-based troubleshooting logic, but NetBrain allows each step to be converted into an automated node.**\ To save time in this lab, **we’ll start with a pre-built Runbook template.**

Click "View All Templates"

Click and select "Voice Application Health & RCA Assessment" and then Click "Add"

The first node added is the Overall Health View. This node provides a quick, visual summary of the operational status of all devices involved in the path. It automatically retrieves key metrics such as CPU and memory utilization, interface status, interface error counts.

We can set a frequency to fetch the data at regular intervals. For now, select **Pull once** from the dropdown, then click **Run**.

This high-level visualization gives us an immediate sense of whether any device along the path is exhibiting abnormal behavior — such as high CPU utilization that might impact packet forwarding, or an interface with a large number of input or output errors. It’s an ideal starting point for a health-based assessment before diving into more specific investigations.

Click on any of the information overlays on the map to drill down into each health check — including CPU, memory, interface errors, device status, and more.

Next, one is a CLI Command Node to execute, depending upon the issue we may use any CLI command can provide us with information from one or multiple devices simultaneously.

In this example we have used show log command on the path devices, as running show log helps us capture recent system messages or other device level events that can provide valuable clues about transient performance issues. For instance, we might find log entries indicating packet drops, interface flaps, or routing issue or even QoS-related policy drops..

Click **Run**, and then select the device whose logs you want to review.

Now, to take the analysis a step further, we’ll use a **Network Intent Node**. This allows us to move beyond basic device health checks and perform a deeper, more targeted analysis based on specific operational logic. In this case, we want to validate **QoS behavior for the voice class-map** across all devices in the application path. Voice traffic is highly sensitive to delay, jitter, and packet loss—so even a small number of *exceeded* or *dropped* packets within the voice class can directly impact call quality. That’s why drilling into QoS counters is critical when troubleshooting voice degradation. Using the Network Intent Node, we apply a predefined logic that inspects each device for: - The current QoS policy applied on the interface - The configured policing and shaping values for the **voice class** - Any **exceeded**, **dropped**, or **rate-limited** packets for that class - Variations in configuration that may not match expected standards This automated intent quickly highlights whether packet drops or misconfigurations in the voice class-map are present anywhere along the path. Instead of manually checking device-by-device, the intent evaluates the entire path in seconds and isolates the device exhibiting abnormal QoS behavior.

Let’s adjust the Network Intent so it performs operational checks across all devices on the path to verify whether any Class-Voice traffic has exceeded packet thresholds. \ \ Click "Use Current Intent as Template"

Click "Run"

Click "Show Result on Map"

We can now see that one of the mapped devices has recorded exceeded packets in the Voice class.

Let’s explore additional options to verify whether packet drops in the QoS class-map are continuing to increase. In this scenario, we can rerun the check periodically and review the updated results through the automation dashboard, as shown in the next steps.

Here, we can see that the automation check can either be executed once or scheduled to run at a defined interval based on our monitoring needs.

We have kept the default check in place at Run once and clicked on Build Dashboard.

Click "Today"

Enter Dashboard and Chart name and Click "OK".

Click "Open Dashboard"

We’ve now built a dashboard that aggregates alerts and success results from the Network Intent, enabling rapid validation of Voice class-map packet drops across the entire path.

To deepen our analysis and validate whether configuration issues may be contributing to the problem, we’ll now move to the next Runbook step: the **Automation Insight Node**. From here, we’ll leverage the **Golden Assessment Library (GAL)**. The Golden Assessment Library is a comprehensive repository of automated checks that compare live device configurations against a defined golden baseline — the organization’s standard for configuration accuracy and performance. GAL includes both system-provided assessments and custom checks built from an organization’s own operational expertise. This ensures validations are not generic, but closely aligned with real-world network behaviors and internal best practices.

In this example, we’ll use the **QoS Policy Validation** assessment to confirm that voice traffic policies are correctly applied across all devices on the mapped path.\ Simply click the search field and filter out irrelevant checks to locate the QoS-related assessment quickly.\ \ In the Search field Type "voice and press [[Enter]]"\ \ Once executed, the automation systematically checks each device’s QoS configuration, analyzing class definitions, policing rates, and priority levels to confirm alignment with the golden standard. Click this checkbox to select the automation.

Click "Run"

Click "Show Result on Map"

Immediately, the automation check flags a deviation on the US-BOS-R1 router, where the QoS class for voice traffic is configured with a policing rate of 40500, which is far below the expected value defined in the golden template.

Compared to the Golden Config, the voice class on **US-BOS-R1** is clearly misconfigured, showing an unusually low policing rate of **40,500**, effectively throttling voice traffic and causing the performance degradation observed by users. With this insight, we’ve quickly identified the root cause without manually logging into multiple devices or parsing through large configuration files.

Now that we have our Runbook replication complete, it's time to export the results to documentation using the **Documentation** node.\ \ The Document Node allows us to automatically generate structured documentation of the troubleshooting process — including the steps taken, the results of each diagnostic node, and the identified root cause. With documentation, NetBrain can intelligently summarize these findings, create readable narratives, and even format the information into a ticket-ready report. For instance, it can automatically include the network path diagram, health status summaries, CLI output highlights, and GAL assessment results — all presented in a concise, professional format. This eliminates the need for manual note-taking or after-action reporting, ensuring that every investigation is fully documented and auditable.

Click "Create"

System appropriately prompts you to save your work before continuing with the document creation—ensuring accuracy and preserving the current state of your analysis.

Your document is now ready for download. If you’d like to refine, edit, or summarize the content, you can continue working with it using the AI Document feature.

Click "AI Document"

Type the necessary instruction in the given field and Click "Create Document".

### **Conclusion** This lab showed how NetBrain’s Runbook automation streamlines troubleshooting by replacing manual, device-by-device checks with fast, consistent, and scalable analysis. Within minutes, we validated device health, detected exceeded voice-class packets, verified QoS configurations against golden standards, and generated clear documentation. NetBrain turns traditional CLI-driven troubleshooting into an automated workflow reducing MTTR, improving accuracy, and strengthening overall network reliability.