Back to top

Groups with abnormal latency events

Last updated June 26th, 2024

This insight alerts you when device groups in your fleet report events with abnormally high latency times compared to their historically normal latency trends, and lets you know if those groups’ outlier latency event counts have increased or decreased since the previous week, so that you can take actions to mitigate or rectify potential issues on your network.

Prerequisites

This insight has several dependencies on the Network latency response times insight. Specifically, you must set up one or more latency test servers, and specify a latency test frequency before data will appear in this insight.

Additionally, this feature requires at least 4 weeks of latency test data in order to accurately report abnormal events. Please ensure that you’ve set up your latency tests ahead of time, prior to using this insight.

Main tile view

The main tile view for this insight lists the top 5 groups with the highest abnormal latency event counts in the fleet, the number of abnormal latency events per group, and the change rate indicating whether there is a higher or lower percentage of events compared to the previous week.

A negative change rate (for example, -27.4%) means there are fewer abnormal events than the previous week, whereas a positive change rate (for example, +31.5%) indicates that there are more abnormal events.

Click the Expanded view button (>) to view abnormal latency event counts for all device groups, or click a group name on the tile to see the Drill-down view data for that particular group.

Expanded view

Groups abnormal latency expanded view

On this page, you’ll see a list of every group that reported abnormal latency events. Use the Date selector to view data from Today, Yesterday, or the Last 7 days. Click the Search field to filter results for a specific group.

In the expanded view table with the following information:

  • DEVICE GROUP — Name of each group in the fleet with abnormal latency events. Click the group name to go to a drill-down view of the group’s abnormal latency event data.

  • ABNORMAL LATENCY EVENTS — The total count of all abnormal latency events over the reporting period,

  • CHANGE RATE — The ratio calculated between the abnormal latency event count during the current reporting period, and the count during the previous reporting period.

    • If current reporting period is Today, then the previous reporting period is the yesterday.

    • If current reporting period Yesterday, then the previous reporting period is the day before yesterday.

    • If current reporting period is Last 7 days, then the previous reporting period is the 7 days before this period (8 to 14 days ago).

  • IMPACTED DEVICES — Total number of devices in the group that experience abnormal latency event times.

To see the Drill-down view for a specific group, click the group’s name. You can also enter the Drill-down view for the top five groups by clicking any of their names on the main tile.

Drill-down view

Groups abnormal latency expanded view

On this page, you’ll see information related to each abnormal latency event for a specific group.

Use the Date selector to view data from Today, Yesterday, or the Last 7 days. Use the IP/DNS server filter if you only want to view details related to a specific server. By default, all servers are shown on the chart and in the table.

The chart displays the total number of abnormal events per day, over the selected reporting period. Hover over each segment to view the total event count for each day.

In the table below, you’ll see a breakdown of each abnormal latency event. To download a CSV version of the table, click DOWNLOAD AS CSV.

The table contains the following information:

  • IP/DNS SERVER — The server that reported abnormal latency times.

  • TEST TYPE — Specifies the latency test type on the server, either PING, DNS, or TCP.

  • ABNORMAL LATENCY — The time in milliseconds it took to perform the latency test.

  • % OF DEVIATION — The % that the abnormal latency time deviated from the median latency time.

  • EVENT TIME — The time at which the abnormal latency event was captured by Knox Asset Intelligence, reported in UTC.

Is this page helpful?