Skip to main content

Troubleshooting Node Monitoring and Alerts

E2E Networks pre-configures monitoring for default node images. Monitoring data should appear in the Node Details page within 5–10 minutes of launch. Alerts are created and managed from the node Alerts tab or the global alert flow. This guide covers what to do when graphs are missing, alert notifications are not sent, or the Zabbix agent — which collects the data — is not running.

Common Issues​

IssueWhat to check
Memory or storage chart is missing on the details pageConfirm the node is Running, then select Test Monitoring Service. If the test fails, follow the troubleshooting link in the agent warning.
Test Monitoring Service is disabledThe node is Powered off. Start the node and retry.
Create Alert is disabledThe node is not Running, or the node is a Disaster Recovery target. Start the node, or manage alerts on the source node for DR targets.
Volume Read Ops or Volume Write Ops is not shownThese triggers require a specific volume. From the global Manage Alerts flow they are hidden when multiple nodes or All nodes are selected.
Attached volume monitoring shows "not enabled"For Linux nodes, run the portal-provided command inside the node and wait 10 to 15 minutes. Windows nodes are handled automatically.
Webcheck create is rejectedA Webcheck with the same URL and required text already exists for the host, or the host has reached its Webcheck limit.
Alert exists but notifications are not receivedCheck the linked alert group members and confirm the alert is not paused. For notification channel setup, see Node Monitoring.
Agent warning shown after a recent changeWait a few minutes after node creation, reinstall, or restart. For custom images or modified firewalls, see Node Monitoring.

Details​

Memory or Storage Chart is Missing​

Charts depend on the Zabbix agent inside the node reporting back to the E2E monitoring server. If a chart is missing on the Node Details page:

  1. Confirm the node is Running. Stopped nodes do not report.
  2. Click Test Monitoring Service on the node details page. The test exercises the round-trip between the portal and the agent.
  3. If the test fails, follow the troubleshooting link inside the agent warning — it points to the right next step based on which check failed.

Test Monitoring Service is Disabled​

The Test Monitoring Service action is only available on Running nodes. If it is greyed out, the node is Powered off. Start the node and retry.

Create Alert is Disabled​

Two cases hide the Create Alert button:

  • The node is not Running. Start the node and retry.
  • The node is a Disaster Recovery target. Alerts are managed on the source node, not on the DR target. Open the source node's Alerts tab to manage triggers.

Volume Read Ops or Volume Write Ops Trigger is Missing​

Volume Read Ops and Volume Write Ops are per-volume triggers — they need a specific volume to attach to. From the global Manage Alerts flow, when you select multiple nodes or All nodes, the portal hides these per-volume triggers because there is no single volume to bind to. To create a Volume Ops alert:

  • Open the node's Alerts tab directly, or
  • Select a single node in the global flow so the available volumes can be enumerated.

Attached Volume Monitoring Shows "Not Enabled"​

When a Block Storage volume is attached to a node, the portal needs to enable monitoring on the new volume:

  • On Linux nodes, the portal will display a one-line command to run inside the node. Run that command, then wait 10 to 15 minutes for the first data points to appear.
  • On Windows nodes, this is handled automatically — no command needs to be run.

Webcheck Create is Rejected​

The two reasons a Webcheck create can be rejected:

  1. A Webcheck with the same URL and required text already exists for the host. Webchecks are de-duplicated per host.
  2. The host has reached its Webcheck limit of 5 URLs. Delete an unused Webcheck and retry.

Alert Exists but Notifications are Not Received​

If an alert is firing in the portal but emails are not arriving:

  • Check the linked alert group members. An alert is linked to a User Group; if the group is empty or has the wrong addresses, no one receives mail.
  • Confirm the alert is not paused. Paused alerts continue to be visible but do not send notifications. Resume the alert from the Actions column.

For setting up notification channels, see Node Monitoring.

Agent Warning Shown After a Recent Change​

The Zabbix agent takes a few minutes to start reporting after the node has been created, reinstalled, or restarted. Wait a few minutes before assuming the agent is broken.

If the warning persists — typically because you imported a custom image, modified the OS firewall, or removed the Zabbix agent — re-install or reconfigure the agent. See the section below.

Zabbix Agent Troubleshooting​

The Zabbix agent runs inside the node and reports metrics to the E2E monitoring service. Default E2E images ship with it pre-installed. If you imported a custom image, removed the agent, or changed the node firewall, follow the steps below.

Step 1: Confirm the Agent is Running (Linux)​

service zabbix-agent status

If the agent is not running, start it:

service zabbix-agent start

If the package is not present at all, install the agent package that matches the operating system version, then configure it for E2E monitoring.

Zabbix Installation on Linux​

Prefer the Zabbix package for the node's operating system from the official Zabbix downloads page, or use the repository already approved for your image. After installation, enable the service:

systemctl enable zabbix-agent

Start or restart it:

systemctl restart zabbix-agent

Verify status:

systemctl status zabbix-agent

On older images without systemd, use:

service zabbix-agent status

Step 2: Connect the Agent to the E2E Monitoring Server​

After installing, point the agent at the E2E monitoring server by editing /etc/zabbix/zabbix_agentd.conf. The E2E monitoring server IP used by the current portal checks is 172.16.103.23:

sudo sed -i 's/^Server=.*/Server=172.16.103.23/' /etc/zabbix/zabbix_agentd.conf
sudo sed -i 's/^ServerActive=.*/ServerActive=172.16.103.23/' /etc/zabbix/zabbix_agentd.conf
sudo sed -i "s/^Hostname=.*/Hostname=$(hostname)/" /etc/zabbix/zabbix_agentd.conf
sudo sed -i 's/^# Timeout=3/Timeout=3/' /etc/zabbix/zabbix_agentd.conf

Restart the agent so the new config takes effect:

systemctl restart zabbix-agent

(On systems without systemd, use service zabbix-agent restart.)

Note

Port 10050 must be open on the node for the source IP 172.16.103.23 (the E2E Cloud One Monitoring server). Zabbix uses port 10050 to pull metrics from the agent.

Zabbix Agent on Windows​

On Windows nodes the Zabbix agent runs as a Windows service. To check it:

  1. Right-click Start and select Run.

  2. Type services.msc and click OK.

  3. Search for Zabbix agent and confirm it is Running.

  4. If the service is not running, click Start.

If the Zabbix agent service is not present at all, install the Windows agent from the Zabbix downloads page. Use the agent version and architecture that match the current E2E monitoring setup and your Windows image.

Also confirm that port 10050 is open in Windows Firewall. By default it is allowed on all E2E nodes — look for an existing rule named "Zabbix". If the rule is missing, follow the Windows firewall guide to allow it.

If the issue is not resolved by the steps above, contact cloud-platform@e2enetworks.com.

Last updated on May 15, 2026.