2026 - Root Cause Analysis (RCA)

Date:
 May 1, 2026
Incident Window:
 07:30 - 10:25 AEST
Systems Affected:
 VisualCare web application (app.visualcare.com.au)
Summary
At approximately 7:30 AEST, some providers experienced intermittent issues connecting to the Visualcare web application. Worker and Participant access was not affected through Visualcare mobile applications, as those users connect through a separate domain. 
Investigation identified that affected administrators’ browsers were holding outdated connection information that pointed to a web server which was no longer available. We adjusted the underlying configuration to prevent recurrence and worked with our customer teams to guide impacted administrators through a quick browser refresh to restore access.
Impact
Administrators across some providers were intermittently unable to access the Visualcare web application (app.visualcare.com.au) during the incident window. Worker and Participant access was unaffected, and no data was lost or compromised. Once administrators completed a browser refresh, access was fully restored.  
Timeline
07:30: First reports received by the Engineering Team
07:30 - 09:30: Triage and investigation
09:30: Resolution guidance was shared with customer teams
09:30 - 10:25: Customer teams supported administrators through the resolution steps as tickets came in
10:25: Email notification sent out to all customers with resolution guidance
Root Cause
The Visualcare web application uses load balancers to distribute administrator traffic across multiple web servers. To keep each administrator’s session consistent, browsers temporarily remember which web server they’re connected to.
Overnight, one of those web servers became unavailable. Because the browser-side connection information was set to refresh once per day rather than more frequently, some administrators’ browsers continued attempting to reach the unavailable server instead of being routed to a healthy one. This is what caused the intermittent connection failures.
We are taking action to address this with improved infrastructure. (See Follow-Up Actions).
Resolution
To restore access, affected administrators needed to refresh their browser’s connection information. Depending on the browser, this was achieved by a hard refresh, clearing the browser cache, or opening the application in a private/incognito window.
To prevent recurrence with the existing infrastructure, we shortened the refresh interval on the underlying configuration so that any future change to a web server is picked up by browsers within minutes rather than up to a day.
Follow-Up Actions
Short Term
Audit related configuration to confirm appropriate refresh intervals are in place across the platform.
Medium Term
Containerise the Visualcare web application and API into a new production environment (already scheduled for next week). This work removes the dependency on the load balancing pattern that caused this incident, meaning this class of issue cannot occur in the new environment.
Long Term
Continue investing in platform monitoring and resilience to detect and prevent issues of this nature earlier.