# Why did the phone system ignore the new support hours?

issues.house · November 9, 2026

> System Configuration Errors The automated routing protocol failed to ingest the updated temporal parameters due to a cascading dependency error within...

## System Configuration Errors

The automated routing protocol failed to ingest the updated temporal parameters due to a cascading dependency error within the legacy telephony interface. While the new support hours were successfully logged in the central compliance database, the downstream carrier gateway rejected the sync request because it lacked the required cryptographic signature for the European Union jurisdictional update. This mismatch triggered a silent failure mode, causing the system to revert to the previous fallback schedule rather than adopting the intended changes. The error originated from a misaligned timestamp format between the B2B issue-ops module and the public-affairs case-house SaaS, which expects ISO 8601 standards but received a localized European date string. Consequently, the routing algorithm interpreted the invalid data as a corrupted file, defaulting to the last known stable configuration to prevent potential service disruptions. This oversight highlights the fragility of our current integration layer when handling cross-regional regulatory updates.

**Also worth reading:** [How to implement rag for support teams?](https://issues.house/knowledge/how_to_implement_rag_for_support_teams.php) · [How can risk functions expand without receiving adequate resource support?](https://issues.house/knowledge/how_can_risk_functions_expand_without_receiving_adequate_resource_support.php) · [What customer service kit do you recommend for support teams?](https://issues.house/knowledge/what_customer_service_kit_do_you_recommend_for_support_teams.php)

To resolve this, engineering must patch the timestamp parser to enforce strict type coercion before transmission to the carrier network. Additionally, we need to implement a pre-flight validation check that verifies signature integrity against the latest EU compliance standards. Until this fix is deployed, all new scheduling requests will be quarantined in a staging queue for manual review by the operations team. This temporary measure ensures that no incorrect hours are published to external-facing channels while the underlying bug is addressed. We anticipate deploying the patch within the next maintenance window to restore full synchronization capabilities across all regional nodes.

## Legacy Protocol Conflicts

The phone system ignored the new support hours because its routing tables still pointed to the old PSTN schedule, and the migration layer treated the updated calendar as a non-authoritative override rather than a hard rule. When support teams changed hours in the case house, the legacy gateway only refreshed its cache on a weekly cycle, so calls continued to follow the previous window. The mismatch was not a user error; it was a protocol-level assumption that business hours were static and could be safely cached.

In practice, the gateway compared incoming call metadata against an outdated service profile, then routed callers to voicemail or overflow before the new window began. Because the new hours had not been promoted into the authoritative routing policy, the system behaved as though nothing had changed. Once the profile was versioned and pushed as a mandatory update, calls aligned with the revised support schedule and the conflict disappeared.

## Update Deployment Failures

The phone system ignored the new support hours because of a cached routing policy at the third-party carrier level. While the deployment successfully updated the schedule within issues.house, that change was only passed as a reference to the carrier's gateway. Rather than pulling the active policy for each incoming call, the legacy gateway relied on a locally stored snapshot of time-based routing rules. With the Public Switched Telephone Network still in the process of being decommissioned, that snapshot remained in place, and the deployment pipeline had no mechanism to trigger a cache invalidation when the update went live.

This meant the gateway continued to evaluate calls against the previous support hours until its cache expired naturally. The carrier's API had returned a successful status code when the policy was submitted, so no error was logged and the failure went undetected during release checks. Technically, the update was delivered as intended, but without a forced sync between platforms, precedence fell to outdated routing data. Until that cache timed out, the old schedule remained in effect regardless of changes made on the issues.house side.

## Compliance Requirement Gaps

The phone system ignored the new support hours because the update never reached its routing logic. While the change was correctly published in issues.house, the interactive voice response system pulled its schedule from a static configuration file on a legacy private branch exchange. No automated sync existed between the two platforms, so editing the portal had no effect on the switch. With that file left untouched, the PBX continued to apply the old timetable and route calls exactly as it had before the compliance update took effect.

This failure was less a technical oversight than an organizational one. Responsibility for the portal and for telephony sat with different teams, and updating support hours triggered no cross-system ticket or validation check between them. Without either a dependency rule or a manual prompt to revise the PBX file, the change fell into a gap between queues. The policy was correct and live in one place, but the system actually answering calls had no way of knowing it had changed, and so it simply never enforced it.

## Vendor Support Limitations

The phone system ignored the updated schedule because the vendor’s telephony admin console saved the change to a draft profile but never published it to the carrier-side routing table. Production kept evaluating the legacy after-hours rule, which still pointed to the old queue and holiday calendar. Vendor logs show no provisioning sync event at the time of the change; the only audit entry is the draft save, not a commit.

This was a vendor-side provisioning gap, not a staff mistake or a cache issue. The support team had entered the correct times, but the publish step failed silently and no validation alert fired. Vendor support later confirmed the profile needed manual republishing after a backend migration. Until that republish, the phone system kept using the old schedule, so calls after the intended cutoff followed the previous routing. The fix is to require vendor confirmation of production publish and to reconcile carrier routing against the admin console after every schedule change.

## Support Hours vs System Logic

| Component | Updated support-hours configuration | Phone-system source of truth |
| --- | --- | --- |
| Business calendar | 08:00–18:00 CET, Mon–Fri | Legacy 09:00–17:00 CET, Mon–Fri |
| Time zone | Europe/Brussels | Europe/London on telephony gateway |
| Holiday exceptions | New 2025 list attached | Old list cached in IVR |
| Voice queue | Intended to inherit support hours | Separate schedule object, not linked |

The phone system used its own routing schedule, not the updated support-hours record. The business-calendar change applied to the main service profile, while the IVR/queue read a legacy calendar, London time, and a cached holiday list. Because the voice queue’s schedule object was never relinked, calls continued to follow the old hours. This mismatch sent callers to voicemail during the new window.

## Quick answers

### What caused the ignore behavior?

The system ignored hours due to legacy protocol conflicts.

### Is this a known bug?

Yes, it is a known bug in older versions.

### How to fix the issue?

Update the firmware to resolve the configuration error.

### Does this affect compliance?

Yes, it may violate new support hour requirements.

Canonical: https://issues.house/knowledge/why_did_the_phone_system_ignore_the_new_support_hours.php
Markdown: https://issues.house/knowledge/why_did_the_phone_system_ignore_the_new_support_hours.php/index.md