Skip to main content

Troubleshooting your pollers

This document provides procedures to troubleshoot issues with the Centreon pollers in the Centreon Cloud environment.

What's inside a poller?​

A poller basically consists of two services:

  • Gorgone (the gorgoned process) is responsible for managing the monitoring configuration. It receives and applies configurations on the poller when a user deploys the configuration for this poller from the Centreon Cloud interface.

  • centengine or centreon-engine/Engine is responsible for scheduling and executing checks on monitored resources and for sending check results to the Centreon Cloud platform.

Both of them are connected to services hosted by Centreon.

Installation issues​

Check the following log file for any errors (located in the folder created by the installation script in the directory where the installation script was executed).

less ./log/register-poller.log

Monitoring changes are not taken into account when I deploy the configuration​

Typically, checks are not executed. Perform the following checks: if a step has worked for you, then go to the next step.

Step 1: Test Gorgone's Connection to the platform​

Check that Gorgone can connect to your Centreon Cloud platform (you may have to install the nc command: on EL, use dnf install nc and on Debian use apt install netcat-traditional).

nc -zv -w 5 gorgone-<organization_code>.euwest1.centreon.cloud 443

A message like this one should be printed:

Ncat: Connected to n.n.n.n:443.

If you get a timeout instead, this means that the network configuration doesn't allow the poller to talk to the platform. Check whether you need a proxy, or whether you need to open some flows or allow some IP ranges.

Step 2: Check that Gorgone is running​

Run the following command:

systemctl status gorgoned

If the status is not active (running), restart Gorgone:

systemctl restart gorgoned

Step 3: Check Gorgone logs​

Gorgone logs are written in the following file: check them for any errors.

/var/log/centreon-gorgone/gorgoned.log

Step 4: Set Gorgone log level to info or debug​

The default log level for Gorgone is error. If you need more details about what Gorgone is doing:

  1. Edit the following file:

    /etc/sysconfig/gorgoned

  2. Change the --severity parameter value to info or debug.

  3. Restart the gorgoned process:

    systemctl restart gorgoned

Step 5: Check Gorgone's connection​

Once you have set the log level to info or debug, this INFO message should appear in Gorgone logs:

INFO - [pullwss] websocket connected

If you get another message, this means that there is a problem with the network configuration or an authentication problem.

Step 6: Check that the deploy command works​

Check the log messages generated by Gorgone while the configuration is being deployed. The log should contain the following lines, which mean that the poller has received the configuration.

2024-09-13 14:28:50 - INFO - [action] Copy processing - Received chunk for '/etc/centreon-engine//'
2024-09-13 14:28:51 - INFO - [action] Copy processing - Copy to '/etc/centreon-engine//' finished successfully
2024-09-13 14:28:51 - INFO - [action] Copy processing - Received chunk for '/etc/centreon-broker/'
2024-09-13 14:28:51 - INFO - [action] Copy processing - Copy to '/etc/centreon-broker/' finished successfully

If all previous steps were OK but the log does not show that the poller has received the configuration, contact our support team.

Step 7: Check that configuration files are correctly updated​

Check the modification dates of configuration files: if you see that the files were not updated when you deployed the configuration, check that the gorgone user has rights on this folder. The centreon-gorgone user must belong to the group that owns the directory and the rights for the group should be rwx.

ls -al /etc/centreon-engine

Step 8: Check that Engine reloads or restarts correctly​

Check Engine's log file according to the method you have selected when deploying the configuration.

  1. Open the log file:
tail -f /var/log/centreon-engine/centengine.log

  1. Deploy the configuration from the interface using either the reload or restart method.

  2. Check the log file for the following messages:

  • If you have selected the reload method, the log file should contain:
[process] [info] [xxx] Starting to reload configuration.
(...)
[process] [info] [xxx] Reload configuration finished.
  • If you have selected the restart method, the log file should contain:
[process] [info] [3142] Centreon Engine yy.mm.p starting ...

If it doesn't, check the logs for errors.

The poller does not send any data to the platform​

Step 1: Test network connectivity​

Check that the poller can send real-time data to the Centreon Cloud platform. Execute the following command (replace <org-code> by your organization code):

nc -zv -w 5 broker-<org-code>.euwest1.centreon.cloud 443

The command should print a message like this one:

Ncat: Connected to n.n.n.n:443.

If you get a different message, check your network configuration. For instance, check that no domain name filtering blocks the broker process.

Step 2: Check that Engine is running​

Run the following command:

systemctl status centengine

If the status is not active (running), restart centengine:

systemctl restart centengine

Step 3: Check whether any retention files exist​

Inspect the folder containing retention files to know if there is currently some retention on the poller.

ls -l /var/lib/centreon-engine

If the previous 2 tests are OK but there is retention on the poller, contact our support team.

Step 4: Enable and review Engine logs​

Modify log levels to get more information about what Engine is doing.

  1. Edit the centengine.cfg file.

    vi /etc/centreon-engine/centengine.cfg

  2. Locate and adjust parameters prefixed with log_level_. The default values are:

    log_level_functions=warning
    log_level_config=info
    log_level_events=info
    log_level_checks=info
    log_level_notifications=info
    log_level_eventbroker=warning
    log_level_external_command=info
    log_level_commands=warning
    log_level_downtimes=info
    log_level_comments=info
    log_level_macros=warning
    log_level_process=info
    log_level_runtime=warning

  3. Restart Centreon Engine.

    systemctl restart centengine

Note: This change will be overwritten next time the configuration is deployed.

Step 5: Enable and review cbmod logs​

  1. Edit the Centreon Broker module configuration (replace <hostname> by the correct value):

    vi /etc/centreon-broker/<hostname>-module.json

  2. Adjust the centreonBroker.log.loggers object to the desired log level.

  3. Restart Centreon Engine.

     systemctl restart centengine

Note: This change will be overwritten next time the configuration is deployed.

Step 6: Check export statistics​

Some statistics about real-time data exported to Centreon Cloud are available in a JSON file. Check this file for any errors.

cat /var/lib/centreon-engine/*-module-stats.json

Note: This file is a UNIX pipe file and is only accessible when the centengine service is running.

If all troubleshooting steps have failed, contact our support team.

Poller log files​

Use to debug what?ProcessFileSymptoms
Installation of the poller<scriptdir>/logs/register-poller.log
Centreon Enginecentengine/var/log/centreon-engine/centengine.log
  • Checks are not performed
  • Objects that have been deleted are still monitored
Gorgonegorgoned/var/log/centreon-gorgone/gorgoned.logActions are not triggered: acknowledgements, downtimes, force checks, autodiscovery