VPS network reliability and incident visibility improved
Transient router API responses no longer trigger premature network failure alerts, and administrators now have a severity-based incident center.
SCComp has strengthened the VPS network reconciliation control plane after tracing repeated operator alerts to brief empty responses from the firewall API. The underlying VPS assignments were recovering on the next pass; the new implementation retries safe reads with fresh connections and requires a second full-device attempt before an assignment can be marked failed. Firewall write requests remain single-attempt to avoid ambiguous duplicate changes.
Significant operational failures are now tracked as incidents with stable deduplication and a complete lifecycle. Continuing failures update one incident instead of generating repeated emails, recovered conditions resolve automatically, and a genuine recurrence can reopen the incident and notify administrators again.
The administrator dashboard now includes a dedicated error center with active counts, severity, state and category filters, current VPS network health, acknowledgement and resolution controls, operator attribution, and resolution notes. At release, all desired network assignments were healthy and there were no active provisioning, Proxmox, network, or payment-processing incidents.
The production deployment passed both-router API checks, the complete 45-test application suite, authenticated and authorization route validation, service restart checks, and post-deployment journal review.