OS_PROJECT_NAME : name of the Openstack project to connect to.OS_PASSWORD : password corresponding to the user entered.OS_USERNAME : user with access rights to the project.OS_AUTH_URL : authentication URL address (Keystone).In this file, the following information is provided, as an example : The variables are filled in as follows: OS_AUTH_URL= OS_PASSWORD=password In the file containing the environment variables, the necessary information is provided to export it to connect to the Openstack infrastructure. Possible environment variables to use when launching the container (here, the variables are contained in an external file).The name you want to give to the container.The name of the image to be used for the container, specifying the version.Image: rakeshpatnaik/prometheus-openstack-exporter:v0.2 To deploy the att-comdev Openstack-Exporter, a compound file can be used to quickly deploy the Export service by applying a possible configuration. Despite this, it is the solution developed by att-comdev that is used in our case, because this solution is more stable and easy to deploy thanks to its Docker image. The second is developed by Canonical Ltd, and is somewhat unstable, as it is not finished, but it is Canonical’s solution that offers the widest range of recovered metrics. There are two different versions of this export: the first, developed by att-comdev, offers a containerized version. To retrieve the metrics of an Openstack infrastructure, it is necessary to use an exporter specific to this task: the service called prometheus-openstack-exporter. Schematic of retrieving Openstack metrics. Schematic of retrieving metrics from the host machine. Once this information has been retrieved, we would like to be able to display it in graphical form using two separate Grafana dashboards.

With regard to the information to be retrieved and displayed, we want to retrieve information about the host machine on which Openstack runs (RAM, CPU, network,…) on the one hand, and on the other hand, we want to retrieve information about the main Openstack components (Nova, Neutron,…). For this purpose, the choice is made to use Prometheus and Grafana, two services already tested for the supervision of an NGINX service. The objective is to set up a supervision service to monitor an Openstack infrastructure (RAM, CPU,…) by providing a graphic display of the various information. Below, I propose you the approach followed.ġ) Dashboard for the metrics of the host machine. To do this, I started from the research carried out upstream, to refocus everything around Openstack. After a period of research and familiarization with the tools (see previous article), my main objective was to evaluate the Prometheus and Grafana tools for the supervision of an Openstack infrastructure.

Vous chercher la version française ? Cliquez ici.ĭuring a work placement as part of my second year of BTS SIO, I had the opportunity to discover new tools for supervision and metrology.

Node exporter prometheus docker compose