Beats (Filebeat, Metricbeat) in Elasticsearch

Introduction

Beats are small programs that collect data from your computers and send it to Elasticsearch. Filebeat collects log files, and Metricbeat collects system metrics like CPU and memory.

You want to watch log files from your web server to find errors.

You need to track how much CPU and memory your servers use over time.

You want to send data from many computers to one place for easy searching.

You want to monitor your applications and systems without installing heavy software.

You want to quickly set up data collection for your Elasticsearch dashboards.

Syntax

Elasticsearch

filebeat setup
metricbeat setup
filebeat -e
metricbeat -e

filebeat setup prepares dashboards and index templates in Elasticsearch.

metricbeat setup does the same for Metricbeat data.

Examples

This sets up Filebeat and then starts it in the foreground to send logs.

Elasticsearch

filebeat setup
filebeat -e

This sets up Metricbeat and then starts it to send system metrics.

Elasticsearch

metricbeat setup
metricbeat -e

Example Filebeat config to read all logs in /var/log and send to Elasticsearch on localhost.

Elasticsearch

filebeat.yml
# Configure paths to log files
filebeat.inputs:
- type: log
  paths:
    - /var/log/*.log
output.elasticsearch:
  hosts: ["localhost:9200"]

Example Metricbeat config to collect CPU and memory metrics every 10 seconds.

Elasticsearch

metricbeat.yml
metricbeat.modules:
- module: system
  metricsets:
    - cpu
    - memory
  period: 10s
output.elasticsearch:
  hosts: ["localhost:9200"]

Sample Program

This example shows a basic Filebeat setup to send system logs to Elasticsearch. The comments explain how to run it and what output to expect.

Elasticsearch

# This is a simple Filebeat config example
filebeat.inputs:
- type: log
  enabled: true
  paths:
    - /var/log/syslog
output.elasticsearch:
  hosts: ["localhost:9200"]

# To run Filebeat:
# 1. Save this as filebeat.yml
# 2. Run 'filebeat setup' to prepare Elasticsearch
# 3. Run 'filebeat -e' to start sending logs

# Output example when running 'filebeat -e':
# 2024-06-01T12:00:00.000Z INFO  instance/beat.go:123 Setup Beat: filebeat; Version: 8.0.0
# 2024-06-01T12:00:01.000Z INFO  crawler/crawler.go:75 Loading and starting input: log
# 2024-06-01T12:00:02.000Z INFO  pipeline/output.go:95 Connecting to backoff(elasticsearch(http://localhost:9200))
# 2024-06-01T12:00:03.000Z INFO  pipeline/output.go:105 Connection to backoff(elasticsearch(http://localhost:9200)) established
# 2024-06-01T12:00:04.000Z INFO  crawler/crawler.go:95 Input started: log

Important Notes

Beats are lightweight and easy to install on many machines.

Always run setup commands before starting Beats to prepare Elasticsearch.

Use the -e flag to run Beats in the foreground and see logs directly.

Summary

Beats collect data like logs and metrics from your computers.

Filebeat is for logs; Metricbeat is for system metrics.

They send data to Elasticsearch for easy searching and monitoring.

Practice

1. What is the primary purpose of Filebeat in the Elastic Stack?

easy

A. To visualize data in dashboards

B. To collect system metrics like CPU and memory usage

C. To collect and ship log files from servers

D. To store data in Elasticsearch

Solution

1. Step 1: Understand Filebeat's role

   Filebeat is designed to collect log files from servers and forward them.

2. Step 2: Differentiate from Metricbeat

   Metricbeat collects system metrics, not logs.

3. Final Answer:

   To collect and ship log files from servers -> Option C

4. Quick Check:

   Filebeat = log collection [OK]

Hint: Filebeat = logs, Metricbeat = metrics [OK]

Common Mistakes:

• Confusing Filebeat with Metricbeat
• Thinking Filebeat visualizes data
• Assuming Filebeat stores data

2. Which of the following is the correct syntax to start Filebeat on a Linux system?

easy

A. sudo systemctl start filebeat

B. filebeat run start

C. start filebeat service

D. service filebeat enable

Solution

1. Step 1: Identify standard Linux service start command

   On modern Linux systems, services are started with 'sudo systemctl start servicename'.

2. Step 2: Match with Filebeat service

   Filebeat service is named 'filebeat', so the command is 'sudo systemctl start filebeat'.

3. Final Answer:

   sudo systemctl start filebeat -> Option A

4. Quick Check:

   systemctl start filebeat = correct syntax [OK]

Hint: Use systemctl to manage services on Linux [OK]

Common Mistakes:

• Using incorrect command order
• Confusing 'start' with 'enable'
• Using non-standard service commands

3. Given this Filebeat configuration snippet:

filebeat.inputs:
- type: log
  paths:
    - /var/log/syslog
output.elasticsearch:
  hosts: ["localhost:9200"]

What will Filebeat do when started?

medium

A. Store logs locally without sending

B. Collect system metrics and send to Elasticsearch

C. Send logs to a remote Logstash server

D. Collect logs from /var/log/syslog and send to Elasticsearch at localhost:9200

Solution

1. Step 1: Analyze input configuration

   The input type is 'log' and path is '/var/log/syslog', so Filebeat collects logs from this file.

2. Step 2: Analyze output configuration

   Output is set to Elasticsearch at 'localhost:9200', so logs are sent there.

3. Final Answer:

   Collect logs from /var/log/syslog and send to Elasticsearch at localhost:9200 -> Option D

4. Quick Check:

   Input logs + output Elasticsearch = Collect logs from /var/log/syslog and send to Elasticsearch at localhost:9200 [OK]

Hint: Check input type and output host to predict behavior [OK]

Common Mistakes:

• Confusing logs with metrics
• Assuming output is Logstash
• Ignoring the paths setting

4. You configured Metricbeat to collect CPU metrics but see no data in Elasticsearch. Which of these is a likely cause?

medium

A. Metricbeat service is not running

B. Filebeat is collecting logs instead of metrics

C. Elasticsearch is not installed

D. Metricbeat configuration file is empty

Solution

1. Step 1: Check Metricbeat service status

   If Metricbeat is not running, it cannot collect or send metrics.

2. Step 2: Consider other options

   Filebeat collecting logs is unrelated; Elasticsearch not installed would cause errors; empty config would cause no collection.

3. Final Answer:

   Metricbeat service is not running -> Option A

4. Quick Check:

   Service running = data sent [OK]

Hint: Always verify Metricbeat service is active [OK]

Common Mistakes:

• Mixing Filebeat and Metricbeat roles
• Ignoring service status
• Assuming Elasticsearch absence without error

5. You want to monitor both system logs and CPU metrics on a server using Beats. Which setup is best?

hard

A. Run Filebeat and Metricbeat but send data to different Elasticsearch clusters

B. Run Filebeat for logs and Metricbeat for CPU metrics, both sending to Elasticsearch

C. Run Metricbeat for logs and Filebeat for metrics

D. Run only Filebeat configured for logs and metrics

Solution

1. Step 1: Identify Beats roles

   Filebeat collects logs; Metricbeat collects system metrics like CPU.

2. Step 2: Choose correct combination

   Running both Beats with their proper roles and sending data to the same Elasticsearch cluster is best for unified monitoring.

3. Final Answer:

   Run Filebeat for logs and Metricbeat for CPU metrics, both sending to Elasticsearch -> Option B

4. Quick Check:

   Filebeat = logs + Metricbeat = metrics [OK]

Hint: Use Filebeat for logs and Metricbeat for metrics together [OK]

Common Mistakes:

• Mixing roles of Filebeat and Metricbeat
• Sending data to separate clusters unnecessarily
• Assuming one Beat can do both well