Major incident

Cornerstone: Latency on US SL5

Affected: Response Time

Current status of Cornerstone: Operational — live Cornerstone status

Timeline

1. Aug 24, 2026, 12:49 UTC
   investigating
   The US SL5 swim lane is experiencing some latency issues. Clients with portals in this swimlane may experience delays or issues accessing certain pages in the application. This is our top priority and we are working to resolve the problem as soon as possible. Please check back periodically for additional updates, which will be posted as they become available.
2. Aug 24, 2026, 14:05 UTC
   resolved
   The CSOD Technology Team observed a performance degradation affecting the US SL5 swimlane. The problem began at 4:25 AM Pacific Time and was resolved at 6:35 AM Pacific Time. During this time, clients with portals on this swim lane may have experienced slow performance or intermittent errors while accessing the application.
3. Sep 3, 2026, 14:31 UTC
   postmortem
   **Issue Summary:** On August 24, 2026, users in the US SL5 Production environment experienced significant latency across multiple portals. The issue resulted in elevated response times and degraded portal responsiveness for affected users. **Root Cause:** The issue was caused by a degraded execution plan associated with a high-frequency database operation. The change in execution behavior caused the operation to consume significantly more database resources than expected, resulting in elevated CPU utilization and increased latency across the environment. Concurrent high-volume workloads further contributed to resource contention during the incident and amplified the overall performance impact. **Corrective Action:** Cornerstone cleared the problematic cached execution plans and recompiled the affected database procedure to generate a new execution plan. Additional workload was temporarily controlled to reduce resource consumption while the environment recovered. Following these actions, database utilization decreased, portal response times improved, and the environment returned to normal performance levels. **Preventive Measures:** To reduce the risk of a similar issue, Cornerstone has taken the following actions: * Reviewing execution plan behavior for the identified high-frequency database operation to reduce the likelihood of future plan degradation. * Reviewing high-volume workloads and their resource consumption to reduce the risk of excessive database contention during periods of elevated activity.

Updates as published by Cornerstone (original).