Separate guest availability, networking and application health when diagnosing a failure.
Before you start
Use the access and features actually assigned to your account. Keep a rollback or recovery route before changing existing service settings.
Steps
- Check the public incident status and the assigned instance state.
- Use the console or approved diagnostics to check free disk, memory and essential service status.
- Record the first failing boundary and recent changes without collecting secrets.
- Apply one reversible repair, retest the original symptom and keep a rollback path.
Verify the result
The original application action succeeds. An SSH connection alone does not prove the service is healthy.
Help and related information
Contact support, service status, and Customer Access. Never send passwords or recovery codes.