Hypercare Status

Postmortem: What Happened?

At 20:10 EST, the Customer Support and Engineering teams were alerted to ongoing delays in push notifications specifically affecting users in Canada. The incident coincided with a non-routine manual backup of Canadian production databases, which was being performed in preparation for upcoming deployments.

The manual backup process consumed significant system resources, creating a bottleneck that caused significant latency in our notification services. Once a potential correlation between the backup and the latency was identified, the team immediately halted all manual backup operations. Following the termination of these tasks, system performance stabilized, and services returned to normal operations by 20:35 EST.

‌

Impact

Users located in Canada likely experienced significant latency in notification delivery across several Hypercare services between 19:00 and 20:30 EST. The scope of the impact included:

• Hypercare messaging: Delays in receiving push notifications for Hypercare secure messages.
• SMS services: Latency in SMS messages sent from the Hypercare platform.
• Escalation phone calls: Delays in escalation phone calls triggered by code team activations.
• Virtual paging: Interruptions in virtual paging services.

‌

Resolution and Next Steps

The immediate resolution was achieved by stopping the manual database backup. As this was a non-routine event, existing automated monitors did not initially flag the resource exhaustion as a service-level threat. To prevent a recurrence and improve detection, the following actions have been implemented:

• Enhanced monitoring: New, granular alerts have been configured to identify server-specific bottlenecks and resource spikes at the database level, providing an earlier warning system for manual interventions.
• Alternative non-routine database backup procedures: We have identified and adopted an alternative methodology for manual production database backups that offloads the process from the primary live environment to avoid impacting real-time traffic.

Resolved: This incident has been resolved. A post mortem will be posted shortly.

Monitoring: We are continuing to monitor for any further issues.

Monitoring: A fix has been implemented and push notifications are now being delivered. We will continue monitoring to ensure a full resolution.

Investigating: We are currently investigating alerts that push notifications are not being delivered to mobile app users (iOS and Android). This is only impacting users in Canada. Our engineering team is looking into the root cause, and we will provide an update as soon as we have more information.