Prometheus is an open resource surveillance, quizing and also notifying device. Initially developed by Soundcloud in 2012, the feature-rich device has actually been embraced by numerous firms to check their IT framework and also make sure all systems are running efficiently. Prometheus enables you to quiz and also draw time-series metrics such as CPU & memory application over HTTP method and also picture them on real-time charts. You can additionally set up Prometheus to press signals in case of a node or solution downtime and also incorporate it with various other third-party surveillance devices such as Grafana for improved information visualization. In this overview, we will certainly take a look at the installment of Prometheus on the CentOS 8/ RHEL 8 system.

Step1) Producing a Prometheus customer and also team

To start, we are mosting likely to develop a system customer for Prometheus. Perform the command listed below to attain this.

[[email protected] |] # useradd -m -s/ bin/false prometheus
[[email protected] |] # id prometheus
uid =1002( prometheus) gid =1002( prometheus) teams =1002( prometheus)
[[email protected] |] #

As you might have kept in mind, the system customer has no login consents as defined in the/ bin/false alternative

Action 2) Producing arrangement directory sites for Prometheus

Once the customer for Prometheus has actually been developed, we are after that mosting likely to develop arrangement directory sites in the/ etc and also/ var directory sites which will certainly save Prometheus arrangement data and also information. So run the commands listed below:

[[email protected] |] # mkdir/ etc/prometheus
[[email protected] |] # mkdir/ var/lib/prometheus

Establish the possession on/ var/lib/prometheus

[[email protected] |] # chown prometheus/ var/lib/prometheus/

Action 3) Downloading and install Prometheus tar data

With the directory sites in position, we can currently download and install the Prometheus. To obtain the most up to date variation, go out to the Download and install web page to get the most up to date variation for your atmosphere. At the time of penciling down this post, the most up to date variation was v 2.140. Conversely, simply run the command listed below

[[email protected] |] # dnf set up wget -y
[[email protected] |] # wget -P/ tmp

Once the download is total, essence the tarball data as revealed

[[email protected] tmp] # tar -zxpvf prometheus-2.140. linux-amd64 tar.gz

This will certainly leave you with a directory site called prometheus-2.140. linux-amd64

Usage tree command to check out directory site framework,


The drawn out directory site has 2 binary data prometheus & promtool which we require to duplicate to the/ usr/local/bin course.

So, browse to the drawn out directory site and also duplicate them making use of the command:

[[email protected] |] # cd/ tmp/prometheus -2.140. linux-amd64
[[email protected] prometheus-2.140. linux-amd64] # cp prometheus / usr/local/bin

Do similarly to the various other binary data

[[email protected] prometheus-2.140. linux-amd64] # cp promtool / usr/local/bin

Action 4) Producing an arrangement apply for Prometheus

To start with the arrangement, develop a data/ etc/prometheus/prometheus. yml and also paste the arrangement in the data

[[email protected] |] # vi/ etc/prometheus/prometheus. yml
# International config
scrape_interval: 15 s # Establish the scrape period to every 15 secs. Default is every 1 min.
evaluation_interval: 15 s # Review policies every 15 secs. The default is every 1 min.
scrape_timeout: 15 s # scrape_timeout is readied to the worldwide default (10 s).
# A scrape arrangement having precisely one endpoint to scuff: # Below it ’ s Prometheus itself.
# The work name is included as a tag ‘work =‘ to any type of timeseries scuffed from this config.
– job_name: ‘ prometheus ’
# metrics_path defaults to ‘/ metrics ’
# system defaults to ‘ http ’.
– targets: [‘localhost:9090’]

This will just check your regional system just (Prometheus Web Server).

Following, readjust the firewall software as complies with to permit outside links to the web server by means of port 9090

[[email protected] |] # firewall-cmd – add-port =9090/ tcp – long-term
[[email protected] |] # firewall-cmd – reload
[[email protected] |] #

Action 5) Producing a Systemd solution apply for Prometheus Web server

For us to take care of Prometheus as a solution making use of systemd, we require to develop a system apply for it. So, develop the data as revealed and also paste the web content,

[[email protected] |] # vi/ etc/systemd/system/ prometheus.service
Summary= Prometheus Time Collection Collection and also Handling Web Server
Needs= network-online. target
After= network-online. target

Customer= prometheus
Team= prometheus
Kind= simple
ExecStart=/ usr/local/bin/ prometheus
– config.file/ etc/prometheus/prometheus. yml
– storage.tsdb.path/ var/lib/prometheus/
– web.console.templates=/ etc/prometheus/consoles
– web.console.libraries=/ etc/prometheus/console _ collections

WantedBy= multi-user. target

For the modifications to work, refill the systemctl,

[[email protected] |] # systemctl daemon-reload

Currently begin and also make it possible for Prometheus to work on boot

[[email protected] |] # systemctl begin prometheus
[[email protected] |] # systemctl make it possible for prometheus

Simply to be particular that Prometheus is running, run the command:

[[email protected] |] # systemctl standing prometheus


From the outcome presented, we can plainly see that Prometheus is running as anticipated without any mistakes. Additionally, you can utilize the netstat energy to examine is the solution is paying attention on port 9090.

[[email protected] |] # netstat -tunlp


That’s excellent! Prometheus is working on port 9090 as anticipated. Currently head to your internet browser as surf your web server’s IP as revealed



Click the ‘Status’ tab and after that click ‘Targets’


Your system will certainly be presented as revealed


Action 6) Install & set up node_exporter

Node merchant is an energy that accumulates and also ships a huge range of Linux system metrics such as CPU, memory use, filesystems and also network data. In this area, we are mosting likely to set up node_exporter on the Prometheus web server and also on a remote CentOS 8 Linux host and also screen system metrics on both hosts.

On the Prometheus node, we are mosting likely to develop a system customer for node_exporter.

[[email protected] |] # useradd -m -s/ bin/false node_exporter

Following, go out to the downloads Web page of Prometheus and also download and install the node_exporter tarball or usage listed below wget to regulate to downlond it from command line,

[[email protected] |] # wget

Upon downloading and install the node_exporter data, continue and also remove it as revealed

[[email protected] |] # tar -zxpvf node_exporter-0.181. linux-amd64 tar.gz

You can examine the materials of the drawn out folder making use of the tree command as revealed

[[email protected] |] # tree node_exporter-0.181. linux-amd64


Following, duplicate the binary data called node_exporter to/ usr/local/bin course

[[email protected] |] # cp node_exporter-0.181. linux-amd64/ node_exporter/ usr/local/bin

Following, established the data consents of the node_exporter data which has actually been replicated as revealed

[[email protected] |] # chown node_exporter: node_exporter/ usr/local/bin/ node_exporter

Following, we require to set up node_exporter to run as a solution. So, continue and also develop a systemd solution data as revealed

[[email protected] |] # vi/ etc/systemd/system/ node_exporter. solution

After that paste the arrangement revealed listed below and also conserve the data


Summary= Prometheus Node Merchant
Needs= network-online. target
After= network-online. target

Customer= node_exporter
Team= node_exporter
Kind= simple
ExecStart=/ usr/local/bin/ node_exporter

WantedBy= multi-user. target

For the modifications to work, reload systemd supervisor with the command:

[[email protected] |] # systemctl daemon-reload

Following, begin and also make it possible for node_exporter solution

[[email protected] |] # systemctl begin node_exporter
[[email protected] |] # systemctl make it possible for node_exporter

Simply to make sure that the solution is running, perform:

[[email protected] |] # systemctl standing node_exporter


Simply to make sure that the solution is running, utilize the netstat energy to examine if it’s paying attention on port 9100 as is anticipated by default.

[[email protected] |] # netstat -pnltu|grep 9100
tcp6 0 0::: 9100::: * LISTEN 3472/ node_exporter
[[email protected] |] #

Perfect! Node_exporter solution is running as anticipated.

Following, open port 9100 in firewall software as revealed

[[email protected] |] # firewall-cmd – add-port =9100/ tcp – long-term
[[email protected] |] # firewall-cmd – reload
[[email protected] |] #

Repeat the actions laid out for the remote CentOS 8 Linux system too.

Lastly, you require to include the node_exporter target to the prometheus.yml data. Add the lines listed below to specify node_exporter for the Prometheus web server

[[email protected] |] # vi/ etc/prometheus/prometheus. yml
— — —
– job_name: ‘ node_exporter ’
– targets: [‘localhost:9100’]


Restart Prometheus solution

[[email protected] |] # systemctl reboot prometheus

Once More, go out to your internet browser, click the ‘Status’ tab and also click ‘Targets’

Make sure to observe a brand-new endpoint on the internet browser called node_exporter for the Prometheus web server


To include an endpoint for the remote Linux system head back to the prometheus.yml data and also add the line listed below

— targets: [‘’]

The node_exporter area must currently appear like this

– job_name: ‘ node_exporter ’
– targets: [‘localhost:9100’]
– targets: [‘’]

Conserve the modifications and also reboot Prometheus solution

[[email protected] |] # systemctl reboot prometheus

Freshen the internet browser and also observe the 2nd endpoint that has actually been included for the remote CentOS Linux system


To make sure that you are obtaining metrics, from your set up nodes. merely utilize the crinkle command as adheres to:

# crinkle http://node-ip:9100/metrics

For instance, to present metrics from the Prometheus web server run:

[[email protected] |] # crinkle http://localhost:9100/metrics


For the remote CentOS 8 host, I performed the command:

[[email protected] |] # crinkle


This can additionally be accomplished by opening up an internet browser and also surfing the LINK


You can additionally pick to chart the metrics that you desire. Just most likely to the web page pf the Prometheus web server, and also click the drop-down food selection classified ‘insert metric at cursor’.


Select the statistics you wish to chart,


Click the ‘Execute’ switch and also click the ‘graph’ tab simply listed below to expose the chart


And also this brings us throughout of this subject. You have actually efficiently mounted and also set up Prometheus to check system metrics on both your web server and also remote host. In our following overview, we will certainly incorporate Prometheus with Grafana for much better visualization and also evaluation of metrics. Do not hesitate to share your comments with us and also share the post with a pal.