Skip to content
Mainframe
LOCAL SEARCH

Search Mainframe

    Start a project

    Mainframe Lab 03 · Reference architecture

    Observable Infrastructure Baseline

    Metrics, logs, alerts, and a runbook as a shared operating layer for small infrastructure stacks.

    GrafanaPrometheusLokiRunbooks

    System map

    Reference architecture flow

    Reference architecture flow
    1. 01SystemsServers, services, backups
    2. 02SignalsMetrics, logs, availability
    3. 03EvaluationThresholds, correlation, silences
    4. 04ResponseAlert, owner, runbook
    5. 05ImprovementReview, capacity, maintenance

    Starting point

    The problem this build examines

    Servers and services work until they do not. Without a shared view, users report problems first, alerts become noisy, and ownership remains unclear.

    Approach

    How the system is structured

    The reference architecture collects a small set of relevant signals, maps them to services and owners, and connects alerts to concrete runbook steps. Backups and recovery are tested as separate operating objectives.

    Evidence

    What this concept build demonstrates

    How monitoring becomes an actionable operating routine—with ownership, escalation, and documented recovery.

    Acceptance checks

    • Every alert has an owner and a next step
    • Maintenance cannot create an avoidable alert storm
    • Backup status and recovery are tested separately
    • Dashboards represent services, not only machines

    Typical handover

    • Signal and service inventory
    • Dashboards and prioritised alert rules
    • Runbooks for common incidents
    • Maintenance and handover plan