Support

Support for Nexus Telemetry Fleet during the beta is by email to [email protected], one problem per message. The subject names the fleet server and the collector, the software at the site that reports a terminal. See Add a collector for the parts of the product.

Contents of a support message

  • The fleet server’s address, as shown in the browser’s address bar.
  • The collector’s name as the dashboard shows it, or its collector id.
  • When the problem started, in UTC or local time, and anything that changed at the site at that time.
  • The terminal’s state on the dashboard, the cause shown under the state, and any alert.
  • The support bundle, when someone can sign in to the collector’s host.

Warning: never send an enrolment token, a private key or a certificate file. The support bundle leaves all three out.

Collector status

The collector’s own status report says whether the collector is enrolled, which sources it reads, and whether it is connected to the fleet server.

  1. Open a terminal window on the collector’s host.
  2. Run the command below and paste the whole of its output into the message.
bash
sudo nf-collector status

The command exits 0 when the collector is enrolled, 1 when it is not, and 2 when the report could not be produced. On a collector installed from a deb or rpm package, it requires sudo to read the collector’s files, and stops with a permission error and exit code 2 without it.

See Command line for the report’s sections.

Support bundle

The support bundle is one text file with the collector’s configuration, its status and its recent log, with secrets left out.

  1. Run the command below on the collector’s host.
  2. Attach the file it names to the message.
bash
sudo nf-collector diagnose

The command lists what it includes and what it leaves out before it writes the file. If it warns that no log files were found, the collector’s log is in a directory other than the default, set by NF_LOG_DIR in the collector’s environment. Run the command again with the same NF_LOG_DIR set.

See Command line for the bundle’s sections.

Bundle from the setup page

On macOS, Windows and Docker the same bundle can be downloaded from the collector’s setup page. It includes less of the log.

  1. On the collector’s host, open http://127.0.0.1:41700 in a browser.
  2. Press Download support bundle and attach the file.

See Setup page.

Server health

For a problem with the fleet server itself, quote its health page. Only a user in the platform organisation, the one that manages the fleet server, can open it.

  1. Open Server health from the icon rail.
  2. Quote the status, healthy or degraded, the reason shown with it, and any check reading Attention or Problem.

See Server health.

External probe

The fleet server’s health address answers without a sign-in.

bash
curl -s https://<your domain>/healthz

It answers 200 while the fleet server is healthy and 503 while it is degraded. The response contains no organisation data and no collector identifiers, so it can be pasted into a message.

Server logs

The fleet server’s log records what it did and any error, one line per event. For a problem on the fleet server, send the log lines from around the time of the problem.

bash
sudo journalctl -u nf-server --since "1 hour ago"

Under Docker the same lines come from docker compose logs nf-server. Send the minutes either side of the problem rather than the whole day, and say which minutes were sent.

See Log events for the event names.

← Troubleshooting Known limitations →