Outage report · Major outage

Shepherd outage on Mar 2, 2026

Shepherd Application System Outage

SeverityMajor outage

StartedMar 2, 2026, 09:31UTC

ResolvedMar 2, 2026, 21:49UTC

Duration12h 18m

What happened

Shepherd reported a major outage starting Mar 2, 2026, 09:31 UTC, affecting Shepherd Application, SMS (Texting Services), AI Services (TranscribeAI, SummarizeAI, DiagnoseAI), OpenAPI (Clinic Connect and Integrations). It was marked resolved at Mar 2, 2026, 21:49 UTC.

The incident lasted 12h 18m from first report to resolution.
It was the 2nd longest of 4 significant outages Shepherd reported in the 12 months up to it.
Shepherd had reported 2 other partial or major outages in the 90 days before it.
12 other vendors we track reported a significant incident within 12 hours of it.

Latest update from Shepherd

## Summary On March 2 at approximately 4:31pm ET, Shepherd experienced a production outage affecting all users for about 18 minutes. The system stabilized at 4:49pm ET. This incident was related to the same broader resource-constraint pattern identified last week. ## What Happened Following last week’s outages, we increased application thread capacity over the weekend to prevent thread starvation under load. On March 2, a different but related constraint surfaced. A brief disruption in our cache layer triggered a surge of reconnections and data reloads. Under peak traffic, this pushed the cache infrastructure past its available bandwidth. When that occurred: * Cache requests began timing out * Application threads blocked waiting on those responses * Database connections backed up * Requests stalled across the system The system recovered once the surge subsided. ## Technical Cause Last week’s incidents exposed limits in application thread and database pooling configuration, wh

Read Shepherd's official incident report ↗