prometheus-tutorial

Taller dirigido a programadores.

Para iniciar clona el proyecto. Ejecuta en tu terminal:

vagrant up
vagrant ssh grafana
sudo service grafana-server start
cd node_exporter
./node_exporter &
exit

vagrant ssh node_exporter
cd node_exporter
./node_exporter &
exit

vagrant ssh prometheus
cd prometheus
./prometheus &
exit

Grafana

Create a New Dashboard Then clic on graph And add

Grafana node_exporter

DISK

Filesystem tracking

The overall space remaining on various filesystems.

node_filesystem_avail_bytes node_filesystem_size_bytes

Grafana Panel: Bar Gauge

PromQL: (1 - node_filesystem_avail_bytes / node_filesystem_size_bytes) * 100 Query Config:

Legend: {{ device }}

Resolution: 1/1

Format: Time series

Instant: on

Visualization Config:

Display:

Show: Calculation

Calc: Mean

Orientation: Vertical

Mode: Gradient

Field:

Unit: percent(0-100)

Min: 0

Max: 100

Read & Write Latencies

The read and write latencies on our disks.

For reading node_disk_read_time_seconds_total node_disk_reads_completed_total

For writing

node_disk_write_time_seconds_total node_disk_writes_completed_total

Grafana Panel: Graph

A

PromQL: rate(node_disk_read_time_seconds_total[5s]) / rate(node_disk_reads_completed_total[5s]) * 100

Query Config: Legend: {{ device }} disk read latency

Resolution: 1/1

Format: Time series

Instant: off

B

PromQL: rate(node_disk_write_time_seconds_total[5s]) / rate(node_disk_writes_completed_total[5s]) * 100 Query Config: Legend: {{ device }} disk write latency Resolution: 1/1 Format: Time series Instant: off

Visualization Config: Display Modes: Bars: On Lines: off Points: off Mode Options: Fill: 1 Line Width: 1 Hover Tooltip: Mode: All series Start order: None Stacking & Null value: Stack: off Null value: null

Number of inodes on our system

The number of inodes still available on your system.

Metrics:

node_filesystem_files_free node_filesystem_files

Grafana Panel: Bar Gauge PromQL: node_filesystem_files_free / node_filesystem_files * 100 Query Config: Legend: {{ mountpoint }} Resolution: 1/1 Format: Time series Instant: on

Visualization Config: Display:

Show: Calculation

Calc: Mean

Orientation: Horizontal

Mode: Gradient

Field:

Unit: percent(0-100)

Min: 0

Max: 100

Overall I/O load on your instance

node_disk_io_now

Grafana Panel: Graph A

PromQL: rate(node_disk_io_now)

Query Config: Legend: {{ device }} global I/O Resolution: 1/1 Format: Time series Instant: off

Visualization Config:

Display Modes:

Bars: Off

Lines: off

Points: on

Mode Options:

Fill: 1

Line Width: 1

Point Radius: 3

Hover Tooltip:

Mode: All series

Start order: None

Stacking & Null value:

Stack: off

Null value: null

Memory usage

(1 - (node_memory_MemAvailable_bytes{instance=~\"$node\"} / (node_memory_MemTotal_bytes{instance=~\"$node\"})))* 100 Singlestat

Python Client

In one vagrant machine called python we need to install

sudo apt-get install python3-pip
sudo bash -c 'cat << EOF >> /etc/environment
LC_ALL="en_US.UTF-8"
LC_CTYPE="en_US.UTF-8"
EOF'
pip3 install prometheus_client

We need to create a file called server.py

import http.server
from prometheus_client import start_http_server

class MyHandler(http.server.BaseHTTPRequestHandler):
        def do_GET(self):
                self.send_response(200)
                self.end_headers()
                self.wfile.write(b"Hello World")

if __name__ == "__main__":
        start_http_server(8000)
        server = http.server.HTTPServer(('localhost',8001), MyHandler)
        server.serve_forever()