Centralized event log collection using a Security Information and Event Management (SIEM) system is a foundational requirement of Compliance Framework ECC‑2:2024 Control 2‑12‑2—this post explains how to design, deploy, and operate a SIEM-based logging pipeline that meets the control, with practical steps, configuration examples, and small-business scenarios you can follow today.
Implementation overview and scope (what Compliance Framework expects)
The control requires centralized collection, retention, and review of security-relevant events so you can detect, investigate, and demonstrate response to incidents. Start by defining the scope: identify critical assets (domain controllers, mail servers, VPN gateways, firewalls/UTMs, cloud accounts, critical applications, EDR/antivirus, and privileged admin workstations). For each asset class document required log types (e.g., Windows Security, Sysmon, Linux auth/syslog, firewall accept/deny, cloud audit logs, SaaS audit trails) and retention periods consistent with Compliance Framework policy.
Practical steps to implement
1) Inventory log sources and prioritize (critical first). 2) Choose a SIEM (options for small business: cloud SIEM like Microsoft Sentinel, Elastic Cloud with Wazuh, or managed Splunk/QRadar services). 3) Establish secure transport (TLS for log forwarding: syslog TLS on TCP/6514, WinRM over HTTPS 5986 for Windows Event Forwarding). 4) Deploy collectors/agents (e.g., NXLog, Winlogbeat/Elastic Agent, WEC + subscription for Windows, rsyslog/syslog-ng for Linux). 5) Normalize and parse incoming events to a consistent schema (CEF, LEEF or Elastic ECS). 6) Configure retention, access control, and immutability for evidence (WORM storage or append-only S3 buckets). 7) Build correlation rules and alerts aligned to the control’s detection objectives and run regular validation tests.
Technical configuration examples and details
Example configurations you can reuse: a simple rsyslog forwarder entry to send Linux logs over TLS: *Action(type="omfwd" Target="siem.example.local" Port="6514" Protocol="tcp" StreamDriver="gtls" StreamDriverMode="1"). For Windows, use Windows Event Forwarding (WEF) with a collector (Windows Event Collector service) and Group Policy to deploy collector subscriptions; WEF uses WinRM ports 5985 (HTTP) and 5986 (HTTPS) — configure HTTPS and certificate-based auth. For agent-based collection, NXLog or Elastic Agent can collect Application/System/Security and Sysmon logs; include Event IDs to capture (Windows: 4624/4625 sign-in, 4688 process creation, 4720 account creation, 4723 password change, 4672 privileged logon). Ensure all devices use NTP (UDP 123) to keep timestamps synchronized; inconsistent time destroys correlation.
Small-business scenario — phased, low-cost deployment
For a small business (10–50 users) with limited budget, prioritize: domain controllers, perimeter firewall/UTM, cloud logs (AWS CloudTrail / Azure Activity / Google Cloud audit logs), email (Office 365 audit logs), and EDR. Start with a cloud-managed SIEM (Sentinel or Elastic Cloud) to avoid on-prem hardware. Phase 1: forward firewall syslog and cloud audit logs; Phase 2: add Windows servers and critical endpoints with an agent; Phase 3: onboard SaaS and application logs via API. Use default parsers and a handful of high-value detection rules (multiple failed logins across sources, privilege escalation events, suspicious RDP sessions, mass data download from SaaS). Cost optimization: use tiered retention—90 days in SIEM hot storage for detection, archive 1+ year in low-cost object storage for compliance evidence.
Compliance evidence, validation, and best practices
To demonstrate Compliance Framework adherence, keep: (a) a log source inventory and onboarding checklist, (b) collection architecture diagram, (c) SIEM ingestion and retention policies, (d) examples of alerted incidents and root-cause tickets, and (e) periodic test results proving logs are collected and readable (safety checks such as sending synthetic events and verifying ingestion). Best practices include: encrypt logs in transit with TLS1.2+, enforce RBAC on SIEM consoles, enable log integrity monitoring (hash and store checksums), and implement retention policies aligned to legal or regulatory requirements. Consider using append-only storage for archived logs and maintain a change log for SIEM rule and parser updates.
Risks of not implementing centralized logging
If you fail to centralize event logs you risk delayed or missed detection of breaches, incomplete forensic evidence, inability to meet incident response SLA or Compliance Framework audits, and higher recovery costs. Specific scenarios: a compromised admin uses lateral movement and deletes local logs before detecion, or a cloud account misconfiguration generates suspicious API calls that are never correlated because cloud audit logs were not collected—both lead to longer dwell time and potential data exfiltration.
Operational tips and maintenance
Operationalize logging with routine tasks: monitor ingestion rates and storage growth (estimate: a Windows server with Sysmon can produce 500 MB–2 GB/day; typical desktop 50–200 MB/day—adjust by workload), tune log levels to reduce noise (avoid verbose app debug logs in production), run quarterly rule tuning sessions to reduce false positives, and schedule quarterly ingestion tests. Maintain runbooks for common alerts (account compromise, suspicious process creation, firewall policy changes) and map SIEM alerts to your incident response playbooks. If staffing is limited, consider an MSSP or managed SIEM to provide 24/7 monitoring and retention management.
Implementing centralized event log collection and a properly configured SIEM is achievable for organizations of any size and is a direct path to meeting Compliance Framework ECC‑2:2024 Control 2‑12‑2. By scoping critical sources, securing transport, normalizing data, enforcing retention and integrity, and validating your pipeline with tests and evidence, you reduce detection time, improve incident response, and create clear audit trails to satisfy assessors and regulators.
