Identify and troubleshoot a node that has a problem

Before you begin:

By default, devices monitored by NPM are polled for data every nine minutes. It might take some time before all the nodes you added have data you can review.

Step 1: Determine there is a problem

The easiest way to identify a problem is to have an alert notify you. Some alerts are enabled by default. You can enable additional alerts described later in this guide.

The Node down alert is enabled by default. Therefore, if a node goes down (that is, it does not respond to a ping), you will see it immediately in the Active Alerts resource on the Home page.

Down nodes appear in resources as red (down) or yellow (warning).

If you have configured your alerts to send email, you will get an email when a node goes down.

If you do not see any alerts, click My Dashboards > Network > Network Top 10.

The resources on this page help identify nodes that respond to a ping but have other health problems.

Step 2: Get more details about the node

When you find a node with a problem, click the node name in any resource to open the Node Details page.

If a node is down (red), this means it does not respond to a ping. To resolve an issue of this severity:

  1. Check the power. Is it plugged in?
  2. Check the LAN link light. Is it connected to the network?
  3. Log in to the device and begin troubleshooting it.

    If a node responds to a ping but shows signs of health or performance issues, use the information on the Node Details page to help troubleshoot.

    • Check the Response Time, Packet Loss, CPU load, and Memory Utilization. Usually, those statistics are the first indicators of a problem. In our example, the CPU load on this node is high.

    • Use the Network Latency & Packet Loss, as well as the Min/Max/Average Response Time resources to see if this is a momentary problem or a continuing issue.

      Min/Max/Average Response Time & Packet Loss

    • Depending on what type of node you are monitoring, you may see additional resources specific to that type of device. For example:

      Hardware health: Reports on physical elements of the hardware for Cisco, Dell, F5, HP, and Juniper.

      Routing table information: For routers and switches, multiple resources show a variety of route-related information. Look under the Network subview for these resources.

      Routing Neighbors

      Routing Table

      Default Route Changes

Step 3: Get more details about the alert

When a problem causes an alert to be issued, that alert appears on the Node Details page in the Alerts for this Node resource. Click the alert name to go to the Alert Details page. Use the resources on this page to investigate the cause of the alert.

Alert Status Overview: Tells you when the alert happened, its importance, and whether or not it was acknowledged.
History: If the same alert is triggered repeatedly, there may be a systemic problem. For example, if a device frequently goes up and down, it may be a sign of a flapping route.
Other Objects: Sometimes the same alerts occur on multiple nodes because of a single trigger. For example, if an edge device is having problems, any devices that are dependent on the edge device might also report problems.
Acknowledge: Acknowledging an alert indicates that you are aware of the issue and the problem is being investigated.
Alert Notes: Each person troubleshooting an issue can enter notes about their activities and any discoveries. The Acknowledge and Notes features are helpful when multiple people are troubleshooting a problem.