•The primary uses of Prometheus, an open-source Linux server monitoring program, are alert management, event monitoring, and metrics monitoring.
•Because Prometheus has revolutionized system monitoring, the Cloud Native Computing Foundation (CNCF) has named it as one of its top projects.
•Prometheus supports dashboarding and graphing using a variety of modalities.
The initial part of the Prometheus architecture is the Prometheus server. The central component of the Prometheus architecture, which is further split into storage, Prom, HTTP server, and so on, is the Prometheus server. Data is scraped from the target nodes in the Prometheus server and then saved in the database.
•The Prometheus server includes local on-disk storage .
•Numerous interfaces in Prometheus enable integration with external storage systems.
•The services that need to be scraped are found with the use of service discovery.
•using the service discovery we can check what targets are activley being scraped aswell as monitoring all entitys
•We can extract metrics and scrape metric from the target after the services have been determined and the targets are prepared.
•Node exporters allow us to export the endpoint's data.
•Prometheus keeps the metrics or other data in a local storage once it has been fetched.
Another crucial element is the user interface, which creates a link between the user and the system. Grafana, a visualization tool, and Prometheus work together to create excellent unique dashboards. Grafana dashboards show data using tables, pie charts, line charts, and data graphs of network load, CPU and RAM consumption, among other things, along with the ability to display alerts. Grafana uses prom ql in order to extract data from Prometheus
•Grafana, which is an open-source, free visualization tool that is typically utilized in conjunction with Prometheus to track metrics.
•Grafana offers a range of dashboards, diagrams, charts, and notifications for the specific data source.
•One of Grafana's key advantages is that we may share and store the dashboard with every member of the group.
An agent called the Node Exporter collects system metrics and makes them available in a manner that Prometheus can read. One project that is managed by the Prometheus project is the Node Exporter. You can skip this step if system stats are not needed. We are using node exporter to scrape metrics from the different nodes and then relay them to Prometheus. The metrics scraped from port 9100 are for the head-node and the ports 9101,9102,9103 and 9104 are for the sub-nodes
In this section I configured the system user, created directories where the configuration files will be stored and set the ownership of the Prometheus directory
sudo useradd --no-create-home --shell /bin/false Prometheus -- Create a system user for Prometheus
sudo mkdir /etc/prometheus
sudo mkdir /var/lib/Prometheus – stores the libraries and configuration files of the directories
sudo chown prometheus:prometheus /var/lib/prometheus -- sets prometheus as the owner of the directory
In this section I downloaded the latest version of prometheus for our operating system from the prometheus home page under downloads
cd /tmp/--enters tmp
wget. https://github.com/prometheus/prometheus/releases/download/v2.46.0/prometheus-2.46.0.linux-amd64.tar.gz –uses wget command to download the prometheus setup tar -xvf prometheus-2.46.0.linux-amd64.tar.gz – extracts the file form the setup
cd prometheus-2.46.0. linux-amd64 –- enters the Prometheus download
sudo mv console /etc/prometheus* –- moves the configuration file
sudo mv prometheus.yml /etc/prometheus –- moves the configuration file
sudo chown -R prometheus: prometheus /etc/prometheus –- sets the user of prometheus
In this section I confirmed the Prometheus yml file was present and modified as needed. i set the evaluation time and scrape intrval time to 15 seconds then
I added all data points that I wanted to scrap in the scrap configs section with the main prometheus job in the scrape config section. An alert manager can also be implemented if necessary to alert the user of certain readings that exceeds recommeneded values
Figure .1 shows the prometheus yml file
In this section I created a service file that sets up its startup and states its service type and users. I then reloaded the system to confirm the setup. The code used will be included in the appendix.
Figure .2 system.MD file for prometheuese
./prometheus –config.file=prometheus.yml (or) ./prometheus – launches the Prometheus server
As the default configuration for the prometheus the local host on port 9090 we can verify that its up by accessing http://server-ip-address:9090.on our web browser. When the prometheus interface was displayed I could see if was configure correctly
Grafana was then installed by following a similar process which was adding the Grafana repository to the system and then installing it and updating the packages by using the get-update and apt-get commands. Followed by enabling the server using the enable command. The password for Grafana was then changed to using the sudo grafana-cli admin reset-admin-password command
I added the prometheus connection my adding the URL of prometheus local host 9090 as a new data source and then saving the configuration. From here I created a new dashboard using the data source based of a template on the grafana download page
The first graph displays the CPU core usage which tells us the percent of processing power that is being used as well as the memory usage which show the amount of ram being utilized
Figure .2 cpu and memory graph readings
The second graph used displayed disk space which determines the total data the hard disks can store. It also displays the load carried by the board
Figure .3 load and the disk usage readings
The third graph shows the disc utilization per device which is the percentage of your hard disk your computer is currently using to run programs and carry out tasks. It also displays the rate the disc can read or write to a device
Figure .4 disk usage per device and sta/stw readings
The fourth and final graph displays context switches which are the process of switching the CPU from one process, task or thread to another. It also shows the traffic over the network
Figure .5 context switch and network readings