Behavior of Sentry release health when a deployment yields no error events

When a deployment produces no error events, Sentry will mark the release healthy as long as the number of sessions (or error events) seen in the default 24‑hour rolling window stays below the internal minimum threshold (approximately 100 sessions). If the window later accumulates enough sessions, the health status can change to degraded or crashy based on the actual error rate.

Confirmed facts

• Release health is only available on paid Sentry plans (Team, Business, Enterprise).
• The algorithm evaluates a fixed 24‑hour rolling window per release.
• A minimum session/event count must be observed before a status other than “healthy” can be assigned; below that threshold the release defaults to healthy.
• When no events are received, the lack of data is treated as zero errors, not as missing information.

Likely explanation

Because the health calculation requires a baseline volume of data, a zero‑event deployment never reaches that baseline. The algorithm therefore interprets the absence of events as a perfect error‑free window and labels the release healthy until sufficient traffic arrives.

Steps to verify or change the outcome

1. Open the release in Sentry → Releases → → Health and note the “Sessions in last 24 h” (or error events) count.
2. If the count is below the internal minimum (~100), the healthy label is expected.
3. To force a reevaluation, generate synthetic sessions or error events (e.g., via a test transaction or a lightweight ping) until the count exceeds the threshold.
4. Observe whether the health status changes to degraded or crashy once the threshold is met.

Missing diagnostic detail

What is the observed session or error‑event count for this release in the last 24 hours? Knowing whether the count is above or below the internal minimum determines whether the healthy label is expected or whether another issue should be investigated.