Runtime Configuration
OverviewAI ConfigsConfigsFeature FlagsExperimentation

Release Monitoring

Protect every rollout

Catch flag-level performance and behavioral issues from the first percentage of your rollout, so you can triage faster and ship with confidence.

Start for FreeSee a Demo
AI-POWERED DETECTION

Triage and resolve release issues in seconds

Instantly identify which feature flag caused performance degradation or errors, then deactivate in under 5 seconds. No more digging through logs or stressful incident channels.

Pinpoint the flag in seconds. Know exactly which feature caused the issue without digging through logs or alerts.

Kill the flag and avoid rollbacks. Deactivate a single flag instantly instead of rolling back your entire deployment.

Auto-alert before users notice. Get notified the moment a metric degrades, so your team responds before support tickets pile up.

Root cause at 1% rollout. Catch issues when only a fraction of users are affected, not after the damage is done.

How it works

Automatic monitoring at every stage

Release Monitoring connects to your feature flags automatically. From the first user to the last, every rollout stage is watched so issues surface early and stay contained.

1

Release to any subset.

Roll out to 1%, 5%, or 25% of users with percentage-based targeting. Start small and expand only when metrics confirm it is safe.

2

Metrics connect automatically.

Performance and behavioral metrics link to the flag with zero manual setup. Every rollout is watched from the first request.

3

Get alerted and act immediately.

AI detects impact and alerts you the moment a metric degrades. Kill the flag or expand the rollout in seconds, no deployment needed.

Flag-level attribution

Catch issues that most monitoring tools miss

Standard APM tools struggle to detect issues when features roll out to small user percentages. Harness provides flag-level attribution to catch problems at 1% rollout, not just when 100% of users are affected.

Pinpoint the exact flag. Attribute issues to the specific flag causing them without digging through logs.

Catch problems at 1% rollout. Detect degradation before it spreads to the majority of your users.

Statistical significance built in. Analyze results with confidence. Know when a change is real, not noise.

Automated response

Automatic alerts and instant action

Teams that move fast rely on automation to do the right thing without adding toil. Harness monitors every flag automatically so you can ship with confidence. No manual setup per flag required.

Monitor every flag automatically. Guardrail metrics apply to all flags by default. No configuration needed per release.

Deactivate in under 5 seconds. Kill a problematic flag instantly from the alert. No deployment, no rollback, no waiting.

Resolve faster than any hotfix. Close incidents before support tickets pile up. Recovery is measured in seconds, not hours.

What our customers say

Teams are releasing features faster with Harness

10x

faster flag setup

<3ms

cached response time

“Before Harness it could be anywhere from 1 - 3 weeks to get your flag up and running. Now we're able to do it in 1 to 3 days … We're getting cached responses in under three milliseconds.”

— Patrick Laughlin, Senior Software Engineer, ADP

Read more
FAQ

Frequently asked questions

Release Monitoring continuously tracks key health metrics across every active feature flag rollout. It compares metric behavior between flag variants in real time and automatically alerts your team when a degradation is detected, so you can act before users are widely affected.

You define the metric thresholds that matter to your team, such as error rate, latency, or conversion rate. If a rollout breaches a threshold, you can quickly disable the flag or reduce rollout percentage without requiring manual intervention or a code deployment.

You can monitor any metric your application emits, including custom business metrics, error rates, latency percentiles, and conversion events. Harness integrates with analytics and observability providers so data flows in automatically alongside flag assignment data.

General observability tools show you what is happening across your entire system. Release Monitoring is specifically designed to attribute metric changes to the exact feature flag that caused them. This narrows the blast radius of an investigation from the whole system to a single flag.

Yes. Release Monitoring pairs naturally with canary rollouts. As you increase the percentage of traffic routed to a new variant, Harness continuously compares its health metrics to the control group and surfaces any degradation before you widen the rollout further.

Get started with Harness Release Monitoring

Detect and resolve issues in seconds, not hours.

Start for Free See a Demo