Network Detection Abnormal Data Upload

The Network Detection - Abnormal Data Upload insight displays anomalies detected in external data uploads for IoT clients. For each client, the baseline for historical data uploads is first established over a 30-day period. Based on the client-specific baseline, this feature automatically detects and flags abnormally large data uploads. If no anomaly is detected, this insight is not displayed in the Insights page.

To access the insight, complete the following steps:

1. Access the Insights page. For more information, see Accessing Insights Page.

2. Click the Network Detection - Abnormal Data Upload tile.

   Figure 1: Network Detection - Abnormal Data Upload Insight

   

   The Network Detection - Abnormal Data Upload metrics displays the client information where the anomaly was triggered.

3. Select the client to see more details.

   Figure 2: Behavior Mapping

   

   In the Behavior Mapping chart, you can view the details of the anomaly. Hover over the charts to see the following metrics:

   • Usage—The value of uploaded data.

   • Historical Baseline—Average of historical data values over a 30-day period.

   • Peer Baseline—Displays the average of peer data values over a 30-day period, where peer is defined as same model OS.

   The Data Uploads table lists all the anomalies detected for the client. You can view the details of the anomaly:

   • Source Role—Displays the role assigned to the client.

   • Usage—Displays the value of uploaded data.

   • Historical Baseline—Displays the average of historical data values over a 30-day period.

   • Peer Baseline—Displays the average of peer data values over a 30-day period, where peer is defined as same model OS.

   • Date—Displays the date when the anomaly was detected.