SOC monitoring lab built using Graylog, OpenSearch, and Ubuntu. Includes log ingestion, detection engineering, alerting, and dashboards.
This project demonstrates a small Security Operations Center (SOC) monitoring environment built using Graylog. The lab collects system authentication logs, detects suspicious activities, generates alerts, and visualizes security events on a dashboard.
The objective of this project is to simulate real-world security monitoring and detection engineering tasks performed by SOC analysts.
The SOC lab consists of a single Ubuntu system generating authentication logs which are forwarded to a Graylog server using rsyslog over Syslog TCP.
Graylog processes the logs, applies detection rules, and stores indexed log data in OpenSearch for searching and analysis. MongoDB is used by Graylog to store configuration and metadata.
When suspicious activity is detected, Graylog triggers alerts and sends email notifications to the analyst.
Log Flow:
Ubuntu System Logs ↓ rsyslog (Syslog TCP) ↓ Graylog SIEM ↓ Detection Rules / Streams ↓ OpenSearch (Log Storage) ↓ Alert Notifications (Email)
- Graylog 4.3.15
- OpenSearch 1.3.x
- MongoDB 7.0
- Ubuntu 22.04
- rsyslog
- Gmail SMTP (Email Alerts)
- Centralized log collection
- SSH authentication monitoring
- Brute force attack detection
- Privilege escalation monitoring
- Email alert notifications
- Security dashboard visualization
- SSH Brute Force Detection
- Successful SSH Login Monitoring
- Sudo Privilege Escalation Monitoring
The system generates alerts when suspicious activity is detected such as:
- Multiple failed login attempts
- Multiple failed login attempts then a successful attempt from same IP address
- Successful login after brute-force attempts
- Unauthorized sudo activity
When a detection rule is triggered, Graylog generates an alert and sends an email notification to the analyst.
In a real SOC environment, these alerts would typically create incident tickets in an incident management platform such as ServiceNow or Jira for investigation.
Typical analyst response workflow:
- Alert is received via email notification.
- Analyst reviews the logs in Graylog.
- The source IP and user activity are investigated.
- If malicious activity is confirmed, remediation actions may include:
- Blocking the attacker IP address
- Disabling compromised user accounts
- Investigating additional related logs
This project simulates the detection and alerting phase of the SOC workflow while demonstrating how incidents would be investigated and mitigated.
The dashboard provides visualization of security activity including:
- Failed SSH login attempts
- Successful logins
- Sudo command activity
- Log source distribution
Graylog-SOC-Lab │ ├── README.md ├── documentation │ ├── setup-process.md │ ├── detection-rules.md │ └── testing-scenarios.md │ └── screenshots
- Security monitoring
- Log analysis
- SIEM configuration
- Detection rule creation
- Alerting and incident detection
- SOC dashboard creation
This repository includes a Graylog content pack that allows the SOC lab configuration (streams, alerts, dashboards, and inputs) to be imported into another Graylog instance.
- Add Windows event log monitoring
- Integrate threat intelligence feeds
- Implement log pipelines
- Add correlation rules
Faith Okonoboh









