Skip to main content
Version: 22.10

Installing Centreon Helios

This documentation concerns only MSP customers.

Overview​

Centreon Helios is a light piece of software that monitors its host machine and the services that run on it.

Helios can be used to monitor servers that operate an On-Premise Centreon service (Central, Remote Server, Poller, Map, etc.).

The data is sent to the Centreon SaaS platform. No personal data is collected.

Although the following procedure and the Helios configuration files in general allow for some customization, we strongly advise you to leave the filenames etc. as shown here.

Requirements​

  • In order for the metrics to reach the Centreon SaaS platform (where the monitoring of the monitoring is done), Helios must be able to access our public endpoint at the following URL:

    https://api.a.prod.mycentreon.com/v1/observability (port 443)

    You can test whether your machine can access our endpoint using the following command:

    curl -v https://api.a.prod.mycentreon.com/v1/observability

    You can also go through a proxy using the following command:

    curl -v https://api.a.prod.mycentreon.com/v1/observability \
    --proxy [protocol://]host[:port] --proxy-insecure

    Example:

    curl -v https://api.a.prod.mycentreon.com/v1/observability \
    --proxy http://proxy.local.net:3128 --proxy-insecure

    The following message will be returned in case of success:

    "Missing Authentication Token"

    If you receive a different answer or no answer, your machine cannot reach our endpoint, likely due to your network rules (firewall, proxy, etc.).

    If a proxy access is configured on the host machine, you need to copy the address and port of the proxy to Helios’s configuration file (see section Network).

  • If a host machine does not have direct access to the outside, two options that complement each other are provided: proxy configuration and gateway configuration.

  • The RPMs are available in the Centreon official repositories for the currently supported versions. The official Centreon repository must be installed:

dnf install -y dnf-plugins-core
dnf config-manager --add-repo https://packages.centreon.com/rpm-standard/22.10/el8/centreon-22.10.repo
  • You must be in possession of your unique token that allows you to send data to our platform. This token is provided to you by our Support team.

Installing Helios​

All Centreon components you wish to monitor (Central, Poller, Remote Server, Database, etc.) must each have Helios installed on their host machine.

On a Centreon Central Server​

  1. Install Helios:

    yum install centreon-helios
    yum install centreon-agent

  2. If this is the first time you are installing Helios on the server, generate the yaml configuration file with the following Shell command:

    You need to carry out this step only if Helios has not been previously configured, otherwise you will overwrite your previous configuration.

    /usr/sbin/centreon-agent config \
    --token [your-token] \
    --type central \
    --output /etc/centreon-agent/centreon-agent.yml

    Example:

    /usr/sbin/centreon-agent config \
    --token aaaa-aaaa-aaaa-aaaa \
    --type central \
    --output /etc/centreon-agent/centreon-agent.yml

    Some settings have default values. Edit the file /etc/centreon-agent/centreon-agent.yml and check the following values:

    • centreonengine_stats_file : are the name of the file and its path correct (i.e. have you customized them on your platform)?

    • centreonbroker_stats_files : are the name of the file and its path correct (i.e. have you customized them on your platform)?

    • centreonweb : are the database settings ok? This is the correct format:

      collect:
        centreonweb:
          config_dsn: [user]:[password]@tcp([dbhost])/[centreondbname]
          storage_dsn: [user]:[password]@tcp([dbhost])/[centreon_storagedbname]

      Example:

      collect:
        centreonweb:
          config_dsn: admin:UzG2b5wcMf8EqM2b@tcp(172.28.2.60)/centreon
          storage_dsn: admin:UzG2b5wcMf8EqM2b@tcp(172.28.2.60)/centreon_storage

      This example is correct only if the database is on the same machine as the central server. If you have a deported database, see Remote database.

      The Topology function uses the centreon-agent.yml file to gather the information it needs: this is hard-coded. If you change the name of this YAML file, the function will fail.

  3. Add an environment tag:

    Open the /etc/centreon-agent/centreon-agent.yml file generated at installation (cf. --output option configured earlier) and add the following instructions under the collect section.

    collect:
      tags:
        environment: [staging|preproduction|production|your-custom-value]

    Example:

    collect:
      tags:
        environment: production

    If you have multiple environments of the same kind, you can suffix your type of environment (for instance: "production_client1").

  4. Enable the centreon-agent service:

    systemctl enable centreon-agent.service

  5. Start the centreon-agent service:

    systemctl start centreon-agent.service

  6. Enable the topology scheduling: edit the cron file /etc/cron.d/centreon-helios and uncomment the following line (i.e. delete the # character):

    0 0 * * * centreon /usr/sbin/centreon-helios.phar

    If you already have a previous version of Helios installed, your file may contain a different line to uncomment, in which case you need to replace said line with the one provided above.

    The Topology function uses the centreon-agent.yml file to correctly gather needed pieces of information: this is hard coded. If you change the name of this YAML file, the function will fail.

  7. You can now configure Helios (gateway, proxy etc.) and then test your overall configuration.

On other host machines (Remote Server, Poller, MAP, etc.)​

  1. Install Helios:

    yum install centreon-agent

  2. If this is the first time you are installing Helios on the machine, configure the centreon-agent.yml file:

    You need to carry out this step only if Helios has not been previously configured, otherwise you will overwrite your previous configuration.

    /usr/sbin/centreon-agent config \
    --token [your-token] \
    --type [system|remote|poller|map] \
    --output /etc/centreon-agent/centreon-agent.yml

    Example:

    /usr/sbin/centreon-agent config \
    --token aaaa-aaaa-aaaa-aaaa \
    --type poller \
    --output /etc/centreon-agent/centreon-agent.yml

  3. Add an environment tag:

    Open the centreon-agent.yml file generated at installation (cf. --output option configured earlier) and add the following instructions under the collect section.

    collect:
      tags:
        environment: [staging|preproduction|production|your-custom-value]

    Example:

    collect:
      tags:
        environment: production

    If you have multiple environments of the same kind, you can suffix your type of environment, for instance: "production_client1".

  4. Enable the centreon-agent service:

    systemctl enable centreon-agent.service

  5. Start the centreon-agent service:

    systemctl start centreon-agent.service

  6. You can now configure Helios (Gateway, proxy etc.) and then test your overall configuration.

Configuring Helios​

Network​

If Helios does not have direct access to the outside, two options allow you to circumvent this: access through an HTTP proxy and/or access through the Gateway mode. For Gateway mode, a Helios that needs access (called “Gateway Client”) can get through another Helios instance (called “Gateway Server”) that does have access to the outside.

Example

Your infrastructure is protected within a closed system and you have a proxy Server to manage all outgoing traffic. The Helios installed on the machine hosting the Centreon Central Server is the only one you want to grant access to the outside. In this case, you could configure your network as such:

  • Configure the proxy option on the Central Helios to grant it access to the outside

  • Configure this same Helios as a Gateway Server

  • Configure all other Helios instances (installed next to Pollers, Remote Servers, MAP, etc.) as Gateway Clients

Proxy Configuration​

If you have a proxy access configured on the host machine, copy the proxy settings to the /etc/centreon-agent/centreon-agent.yml file under output:

output:
  token: [your-token]
  proxy_url: [your-proxy-address]:[your-desired-port]
  proxy_ssl_insecure: [true|false]

Example:

output:
  token: aaaa-aaaa-aaaa-aaaa
  proxy_url: http//proxy.local.net:3128
  proxy_ssl_insecure: false

You then need to restart Helios:

systemctl restart centreon-agent.service

Gateway Configuration​

  • Gateway Server: copy the following code to the /etc/centreon-agent/centreon-agent.yml file of the Helios that will act as a Gateway server. To strengthen the security of communications between the gateway client and the gateway server, you can define an authentication token (auth-token), i.e. the character string you want (this is not the same token as the one you used to configure the centreon-agent.yml file).

    gateway:
      enable: true
      listen_port: [listening-port]
      auth_token: [your-gateway-token]

    Example:

    gateway:
      enable: true
      listen_port: 54321
      auth_token: azerty1234

    You then need to restart Helios:

    systemctl restart centreon-agent.service

  • Gateway Client

    In a Gateway configuration, the Gateway Client delegates the configuration of its main token to the Gateway Server (since only the latter communicates with our platform). As a consequence, the token line needs to be commented with the yaml comment operator “#”. If you have defined an authentication token (auth_token) on the gateway server, you need to add it to the configuration of the gateway client too.

    output:
    #token: [your-token]
      gateway:
        url: http://[gateway-server-ip-address]:[listening-port]
        auth_token: [your-gateway-token]

    Example:

    output:
    #token: aaaa-aaaa-aaaa-aaaa
      gateway:
        url: http://172.28.6.145:54321
        auth_token: azerty1234

    You then need to restart Helios:

    systemctl restart centreon-agent.service

Enabling the Collection of Centreon Logs​

Starting from version 2 and up of Helios, logs generated by the monitored Centreon component can be collected.

To define which logs must be collected, you need to create yml configuration files in the following folder: /etc/centreon-agent/conf.d. To collect a specific log, the configuration file must contain the following arguments: path, type and pattern of the target log file. Example:

- path: /var/log/centreon-gorgone/gorgoned.log
  pattern: "%{CENTREONGORGONE}"
  type: file

You can have several configuration files - each file is parsed and its target log files are added to the collection.

Using the Templates​

To make log collection configuration easier, pre-configured templates are provided. Each template covers a specific scope depending on the target Centreon component, its version, etc.

Templates are located in the following folder:

/usr/share/centreon-agent/examples

Based on your monitored Centreon component you can simply copy/paste the corresponding template to your /etc/centreon-agent/conf.d folder.

Finalize Templates Configuration​

For a Centreon Poller, log files are prefixed with the Poller’s name so you need to adapt the Poller template: open the Poller template and replace all POLLERNAME placeholders within the “path” section with the actual Poller’s name.

The provided templates will work out of the box with a standard Centreon installation. In case of doubt, you can locate the actual targeted log file and compare its path to the one written in your “path” section of the template.

In case of errors, you will find detailed explanations of what happened within Helios's own logs in /var/log/centreon-agent/centreon-agent.log.

Start Logs Collection​

Once your log collection is properly configured, you need to restart Helios with the following command:

systemctl restart centreon-agent.service

Tags​

Helios can contextualize data collection with your own custom tags to define the perimeter in which it is in action. This is used later on to aggregate the monitoring data around your tags and create dashboards or reports in relevant contexts.

We strongly advise the first tag you define to be “environment” in order for us to be able to establish a common baseline between all users.

Tags can be configured in the YAML /etc/centreon-agent/centreon-agent.yml file generated at installation. Tags are case-sensitive (production and Production are seen as two different tags).

collect:
  tags:
    environment: [staging|preproduction|production|your-custom-value]
    [tag2]: [your-custom-value2]    
    [tag3]: [your-custom-value3]

Example:

collect:
  tags:
    environment: production
    City: Paris

You then need to restart Helios:

systemctl restart centreon-agent.service

Remote Database​

If the Centreon component monitored by Helios is configured with a specific or remote database, you can configure Helios to access the database in the YAML /etc/centreon-agent/centreon-agent.yml file generated at installation.

collect:
    centreonweb:
      config_dsn: [user]:[password]@tcp([dbhost])/[centreondbname]
      storage_dsn: [user]:[password]@tcp([dbhost])/[centreon_storagedbname]

Example:

collect:
    centreonweb:
      config_dsn: admin:UzG2b5wcMf8EqM2b@tcp(172.28.2.60)/centreon
      storage_dsn: admin:UzG2b5wcMf8EqM2b@tcp(172.28.2.60)/centreon_storage

You then need to restart Helios:

systemctl restart centreon-agent.service

Logs Rotation​

Helios logs all activity (nominal as well as erroneous) in the /var/log/centreon-agent/centreon-agent.log file.

A default /etc/logrotate.d/centreon-agent file has been created at installation and configured as follows:

/var/log/centreon-agent/centreon-agent.log {
  daily
  copytruncate
  rotate 7
  compress
}

You can leave it as such or further adjust the log rotation policy to best fit your needs using the parameters of logrotate.

Use the following command to apply the changes immediately:

logrotate /etc/logrotate.d/centreon-agent

Testing Helios​

Testing the centreon-agent service​

At this stage, the centreon-agent service should be running and set to launch at system start. The following command checks that the service has been correctly configured:

systemctl status centreon-agent

If all went well, the command will return results similar to the following example:

systemctl status centreon-agent
● centreon-agent.service - Helios collect metrics and send them to Centreon SaaS Platform
   Loaded: loaded (/etc/systemd/system/centreon-agent.service; enabled; vendor preset: disabled)
   Active: active (running) since ven. 2019-11-08 14:52:26 CET; 5 days ago
 Main PID: 22331 (centreon-agent)
   CGroup: /system.slice/centreon-agent.service
           └─22331 /usr/sbin/centreon-agent run

Testing data collection​

Once installation and configuration are done, the following command can be used to force a collection and return a full sample of collected data:

centreon-agent sample

The output should look like this:

1624977737000000// centreonengine_uptime_seconds{_cmaas=cmco,hostname=val-central.centreon.io,os=linux,osfamily=rhel} 693583
1624977737000000// centreonengine_command_buffers_used{_cmaas=cmco,hostname=val-central.centreon.io,os=linux,osfamily=rhel} 0
1624977737000000// centreonengine_command_buffers_high{_cmaas=cmco,hostname=val-central.centreon.io,os=linux,osfamily=rhel} 0
1624977737000000// centreonengine_command_buffers_total{_cmaas=cmco,hostname=val-central.centreon.io,os=linux,osfamily=rhel} 4096
1624977737000000// centreonengine_external_command_1m{_cmaas=cmco,hostname=val-central.centreon.io,os=linux,osfamily=rhel} 0
1624977737000000// centreonengine_general_external_command_5m{_cmaas=cmco,hostname=val-central.centreon.io,os=linux,osfamily=rhel} 0

In case of errors while testing the collection, the logs in the /var/log/centreon-agent/centreon-agent.log file can give you further information for troubleshooting purposes.

Testing that you can access the Centreon SaaS platform​

Once installation and configuration are done, the following command can be used to test the connection between Helios and the Centreon SaaS platform:

centreon-agent ping --config [path to your centreon-agent.yml file]

Helios will then return one of the following:

  • Unable to reach the Centreon Cloud Platform, check your network configuration

  • Centreon Cloud Platform reached successfully but your token is not recognized

  • Centreon Cloud Platform reached successfully and authentication was successful: Helios is properly connected to our platform.

Help​

If you want to know more about usr/sbin/centreon-agent, enter:

centreon-agent --help

Updating Helios​

To update Helios, enter:

yum clean all --enablerepo=*
yum update centreon-agent