Skip to content

Database Monitoring

Beacon’s Database Monitoring module provides continuous, real-time visibility into the operational health of every registered database instance. It captures the vital signs of a database — active sessions, transaction behavior, locking activity, resource consumption, and replication status — and presents them through live dashboards and historical timelines.

This module is the foundation of Beacon’s observability model. It answers the fundamental question: “Is my database healthy right now, and how has it been behaving over time?”

Beacon tracks the availability and responsiveness of every registered database instance.

Signal Description
Online / Offline State Real-time indicator of whether an instance is reachable.
Connection Latency Time taken to establish a connection during health checks.
Uptime Tracking Duration since the instance was last observed as available.
Last Checked Timestamp When the instance was most recently probed.
Connection Errors Recorded failures with error context for troubleshooting.

Operational value: Immediately distinguishes between a database that is down, slow to respond, or unreachable due to network/configuration issues.


Beacon provides a live view of all sessions connected to a database instance.

Signal Description
Active Sessions Sessions currently executing a statement.
Idle Sessions Connected sessions not currently executing work.
Idle-in-Transaction Sessions Sessions holding an open transaction while idle — a common source of contention.
Session Duration How long each session has been active or idle.
Session Origin Application, user, or host associated with the session.
Current Query The statement currently being executed by an active session.
Wait Events What a session is currently waiting on (locks, I/O, etc.).

Operational value: Reveals session buildup, abandoned transactions, and long-running activity that may be consuming resources or blocking others.


Beacon tracks transaction throughput and behavior across the instance.

Signal Description
Commit Rate Transactions successfully committed over time.
Rollback Rate Transactions rolled back, indicating errors or application issues.
Long-Running Transactions Transactions exceeding configurable duration thresholds.
Transaction Age Elapsed time since each open transaction began.
Transaction State Active, idle-in-transaction, or completed.

Operational value: Long-running or frequently rolled-back transactions are leading indicators of locking pressure, bloat, and application-level defects.


Beacon surfaces lock contention and blocking chains in real time.

Signal Description
Active Locks All locks currently held across the instance.
Blocking Chains Which sessions are blocking which others, and on what resource.
Lock Wait Duration How long a session has been waiting for a lock.
Lock Type Table-level, row-level, advisory, or other lock categories.
Deadlock Events Recorded deadlocks with involved sessions and queries.
Contended Objects Tables or rows most frequently involved in contention.

Operational value: Blocking chains are a primary cause of sudden application stalls. Beacon makes them visible and traceable before they cascade.


Beacon monitors how each database instance consumes system resources.

Signal Description
CPU Utilization Processor load attributable to database activity.
Memory Usage Working memory, cache usage, and buffer activity.
Disk I/O Read/write throughput and latency.
Storage Consumption Database, table, and index size growth over time.
Cache Hit Ratio Proportion of reads served from memory vs. disk.
Temporary File Usage Spill-to-disk activity from large sorts or joins.

Operational value: Identifies resource saturation, inefficient caching, and workload patterns that drive infrastructure cost.


For instances participating in replication, Beacon tracks replication health.

Signal Description
Replica Connectivity Whether replicas are connected and streaming.
Replication Lag Delay between primary and replica.
Replication State Streaming, caught-up, or stalled.
Replica List All known replicas and their individual statuses.

Operational value: Replication lag directly affects read scaling, failover readiness, and data freshness guarantees.


Beacon aggregates monitoring signals into a consolidated health view per instance.

Indicator Description
Health Score A composite indicator derived from key monitoring signals.
Active Alerts Any threshold-based warnings currently triggered.
Critical Events Recent high-severity occurrences (deadlocks, outages, etc.).
Trend Direction Whether health is improving, stable, or degrading.

Operational value: Provides an at-a-glance status for each instance without requiring manual inspection of every metric.


Beacon presents monitoring data across three complementary views:

Live snapshots refreshed continuously — ideal for active incident response.

Historical plotting of any monitored signal — ideal for spotting patterns and regressions.

Side-by-side comparison of multiple instances — ideal for fleet-wide oversight.


Option Purpose
Collection Interval How frequently metrics are gathered.
Retention Window How long historical monitoring data is retained.
Per-Instance Thresholds Custom alert thresholds for specific databases.
Signal Enablement Toggle which metrics are collected per instance.
Maintenance Windows Suppress alerts during planned maintenance.

Scenario 1 — Sudden application slowdown A developer notices rising response times. The monitoring dashboard shows a growing blocking chain on a heavily updated table. The blocking session is an idle-in-transaction connection, which is terminated, restoring normal operation.

Scenario 2 — Overnight batch contention A scheduled batch job runs nightly. Beacon’s timeline view reveals recurring transaction spikes and lock waits during the batch window, prompting rescheduling or query optimization.

Scenario 3 — Storage growth planning The storage consumption chart shows steady growth over months, giving teams lead time to plan capacity expansion before thresholds are reached.


Beacon’s Database Monitoring module delivers continuous, contextual awareness of database health. By unifying sessions, transactions, locks, resources, and replication into a single observable surface, it enables teams to detect issues early, diagnose them quickly, and maintain reliable database operations.