Aggregate Logs
BSSC Node Operation - Audit Framework
Domain: Security and Access Control |
Level: SHOULD | Type: Detective
Statement of Requirement:
Node logs SHOULD be aggregated where feasible.
See it in the specification
Purpose
Ensure that all logged information is readily available from a single
source. This facilitates detecting problems in a timely manner.
Audit Procedures
Verify that logs are aggregated in ways that facilitate search, and
that security personnel can use aggregated logs
Key Evidence
- Log aggregation architecture document;
- Log inventory
- Cross-node query demonstration
- Logging configurations deployed on nodes
- Aggregation system dashboards showing ingestion rates and
coverage
Success Criteria
- Log aggregation system is deployed and operational
- Logs from node clients, system/OS, and access events across the
fleet are aggregated
- Aggregation is near-real-time with no significant gaps
- Operator can demonstrate using aggregation system for investigation
or monitoring
Best Practices
Near-real-time ingestion with no significant gaps.
Failure Criteria
Any of these circumstances is likely to mean the Requirement is not
met.
- No log aggregation system exists; logs only on individual hosts
- Aggregation system exists but critical log sources excluded (access
logs, node client logs)
- Aggregation system deployed but not functional (significant lag,
missing data, not queryable)
Justifiable Exceptions
- Separating Personally Identifiable Information (PII) or from normal logs. See also Do Not Log Sensitive Information
- Enabling fine-grained access control, such as RBAC, for log information
Common Failure Patterns
- Aggregation is configured for node client logs but system-level
(e.g. auth.log, syslog) and access logs are not included;
- The aggregation pipeline has silent failures with no alerting;
- Aggregation covers production nodes but not supporting
infrastructure.
Other Relevant Standards
A.8.15, A.8.16, A.8.17
CC7.2, CC7.3
10.3.3, 10.2
DE.CM-01, DE.AE-02, RS.AN-03