Use Case

Define Alert Rules in SQL. Detect Incidents in Milliseconds.

Replace fragile monitoring scripts with streaming SQL queries that continuously evaluate every event, log, and metric — and trigger alerts the moment conditions are met.

Live Alert Engine
Streaming
Active Rules
1,247
SQL-based alert rules
Events/sec
84,291
all evaluated in real time
Alerts Fired (24h)
18
incidents detected
Avg Detection
340ms
sub-second latency
See Live Demo

Trusted by 1,000+ Data-Driven Organizations

for Real-time Analytics

Trusted by 1,000+ Data-Driven Organizations for Real-time Analytics

The Problem

Are You Discovering Incidents From Customer Complaints?

Traditional monitoring polls metrics every 30-60 seconds and runs threshold checks in batch. Between checks, critical events go unnoticed — your customers discover outages before your team does.

With RisingWave

Evaluate Every Event Against Every Rule, Continuously

RisingWave lets you write alert rules as SQL queries that run continuously. No polling intervals, no missed events. When CPU > 95% AND error_rate > 5% for 3 consecutive minutes, you know within seconds — not after the next batch run.

Streaming Alert Rules
Define complex multi-condition alerts in SQL. Evaluate against every incoming event without polling delays.
Multi-Source Correlation
JOIN logs, metrics, and traces in real time. Catch issues that only appear when multiple signals align.
Sub-Second Detection
Detect anomalies in under 500ms. Trigger webhooks, PagerDuty, or Slack the moment conditions are met.
See RisingWave in Action: Monitoring & Alerting
See how RisingWave processes real data in real time — not a recording, not a simulation.

An ISO (Independent System Operator) manages a 45GW regional grid. Grid frequency must stay within 59.95–60.05 Hz — deviations beyond this range trigger automatic load shedding. With 12GW of intermittent solar and wind, frequency stability is a second-by-second challenge.

SCADA polls every 4 seconds. A 180ms frequency excursion to 59.91 Hz caused a 2GW automatic load shed last summer because the control room saw it 3.6 seconds too late.
LIVEpmu_readings
pmu_idsubstationfrequency_hzvoltage_kvphase_anglerocofts
PMU-4401Riverside 345kV59.93344.2-2.14-0.082024-07-18T15:42:00.016Z
PMU-4401Riverside 345kV59.91343.8-2.31-0.122024-07-18T15:42:00.033Z
PMU-2208Mesa Verde 230kV59.96229.7-1.08-0.042024-07-18T15:42:00.016Z
PMU-2208Mesa Verde 230kV59.94229.5-1.22-0.062024-07-18T15:42:00.033Z
PMU-6615Clearwater 500kV60.01499.80.420.012024-07-18T15:42:00.016Z
PMU-3309Ironwood 138kV59.97137.9-0.73-0.032024-07-18T15:42:00.016Z
Streaming SQLRunning
Detect frequency excursions
CREATE MATERIALIZED VIEW frequency_excursions AS
SELECT
  pmu_id,
  substation,
  frequency_hz,
  voltage_kv,
  rocof,
  ts,
  CASE
    WHEN frequency_hz < 59.90 OR frequency_hz > 60.10 THEN 'CRITICAL'
    WHEN frequency_hz < 59.95 OR frequency_hz > 60.05 THEN 'WARNING'
  END AS severity,
  CASE
    WHEN rocof < -0.10 THEN 'ACCELERATING'
    WHEN rocof < -0.05 THEN 'DEVELOPING'
    ELSE 'STABLE'
  END AS trend
FROM pmu_readings
WHERE (frequency_hz < 59.95 OR frequency_hz > 60.05)
  AND ABS(rocof) > 0.05;
Correlate with generation-load balance
frequency_alertsauto-updating
substationfrequency_hzrocofseverityrecommended_actiondetected_at
Riverside 345kV59.91-0.12CRITICALSolar output drop at Mojave Flats Solar (84/300 MW)2024-07-18T15:42:00.033Z
Mesa Verde 230kV59.94-0.06WARNINGSolar output drop at Mojave Flats Solar (84/300 MW)2024-07-18T15:42:00.033Z
Sun Valley Solar 230kV59.95-0.05WARNINGLoad surge in Zone 4 South (3820 MW vs 3290 MW forecast)2024-07-18T15:42:00.016Z
RisingWave processes 50,000 PMU readings/second. When frequency at substation PMU-4401 drops to 59.93 Hz with rate-of-change at -0.08 Hz/s, the system dispatches battery reserves within 200ms.
Why RisingWave

Move From Reactive to Proactive Monitoring

Use streaming SQL to continuously evaluate every event, detect incidents in milliseconds, and alert your team before users are impacted.

Catch Issues Before Users Do
Detect anomalies and threshold violations the moment they occur — not minutes later when customers start complaining.
Reduce MTTR
Correlate logs, metrics, and traces in real time to pinpoint root causes faster. Cut mean time to resolution from hours to minutes.
Eliminate Monitoring Blind Spots
Evaluate every event against every rule continuously. No polling gaps, no sampling — complete visibility across your entire stack.

Ready to Modernize Your Monitoring Stack?

Best-in-Class Event Streaming
for Agents, Apps, and Analytics
GitHubXLinkedInSlackYouTube
Sign up for our to stay updated.