Changes / Statsig / Degraded SDK initialization global success rate
Statsig: Major incident: Degraded SDK initialization global success rate
Summary
Statsig reported a major partial outage affecting SDK initialization APIs (/v1/initialize, /v1/download_config_specs, /v2/download_config_specs) in all clusters, reducing initialization success. It started at 2025-10-30 18:56 UTC and was resolved at 2025-10-30 19:20 UTC, lasting 24 min. The vendor attributed it to a networking configuration change and said it was reverting that change.
Machine-written from the diff below (Sol; type: incident outage). Check the diff before relying on it.
What changed
Before: not present
After: Degraded SDK initialization global success rate
Before: not present
After: major
Before: not present
After: resolved
Before: not present
After: 2025-10-30 18:56 UTC
Before: not present
After: 2025-10-30 19:20 UTC
Before: not present
After: Asia Southeast 1, Config Delivery API - Asia Southeast 1, Config Delivery API - Europe West 1, Config Delivery API - US East 1, Config Delivery API - US East 5, Config Delivery API - US West 1, Europe West 1, US East 1, US East 5, US West 1
Before: not present
After: Starting at 11:56AM PST, a networking configuration change caused a partial outage for SDK initialization APIs (/v1/initialize, /v1/download_config_specs, /v2/download_config_specs) in all clusters.
The team is in the process of fast reverting this change, which will restore availability shortly.
Mitigated: 10/30/2025 12:20:00 GMT-0700 (Pacific Daylight Time)
Before: not present
After: Resolved.
First version of this incident archived by Rewindex. View the live incident (opens in new tab).
History of this incident
| When (UTC) | Change | Kind |
|---|---|---|
| Major incident: Degraded SDK initialization global success rate (this change) Statsig reported a major partial outage affecting SDK initialization APIs (/v1/initialize, /v1/download_config_specs, /v2/download_config_specs) in all clusters, reducing initialization success. It started at 2025-10-30 18:56 UTC and was resolved at 2025-10-30 19:20 UTC, lasting 24 min. The vendor attributed it to a networking configuration change and said it was reverting that change. | Major incident |