CanFlow Global
Microsoft Outage Cut CBSA Client Transmissions Nov 12-14
← All insights
carmcbsamicrosoft-outagecustoms-technologypars

Microsoft Outage Cut CBSA Client Transmissions Nov 12-14

A global Microsoft outage Nov 12-14 blocked client transmissions to CBSA. CBSA systems stayed up; the break was in the pipes brokers and importers use.

Microsoft Went Down, CBSA Didn’t — Your Pipeline Still Broke

Between November 12 and November 14, 2026, a global Microsoft outage blocked external clients from transmitting messages to CBSA or receiving messages back. CBSA systems stayed online the entire time. No internal CBSA downtime, no corrupted data, no delayed assessments. The problem was in the pipes connecting brokers, importers, carriers and service providers to CBSA’s electronic systems.

If you filed a PARS entry, submitted eManifest data, checked your CARM Client Portal for a Statement of Account, or waited for a release notification during those 48 hours, you either queued behind the blockage or switched to contingency procedures. CBSA facilitated movement under its fallback protocols, which typically means manual releases at the border, paper documentation and slower processing until the electronic pipeline came back.

What “External Clients Can’t Transmit” Means

CBSA’s Trade Chain Partners notice TCC26-0169 said external clients couldn’t transmit to CBSA or receive messages from CBSA. That hit different actors at different layers: customs brokers using cloud-based EDI software to submit PARS, file CADs in the CARM Client Portal, or receive Release Notification System messages; carriers filing eManifest ACI cargo data; importers logging into the CARM Client Portal to review their Statement of Account or confirm that an RPP bond transaction settled; service providers running integrations that pull CBSA data into WMS or ERP platforms for downstream shipment planning.

The common thread: if your software stack or your broker’s software stack runs on Microsoft Azure, Office 365, Dynamics or any other Microsoft cloud service, and that layer is how you reach CBSA, you stopped reaching CBSA when Microsoft went dark.

Why This Matters More Post-CARM

Pre-CARM, brokers filed B3s and you paid CBSA directly or via the broker’s trust account on a transactional basis. If PARS was down, you could usually get a manual release at the border with a paper cargo control document and sort out the accounting later.

CARM changed the architecture. The CARM Client Portal is now the system of record for your financial security (the RPP bond), your monthly Statement of Account, and the payment of duties and taxes assessed on your CADs. Release prior to payment depends on that bond being active, correctly sized, and not exhausted. If you can’t log into the CARM Client Portal because your browser session, your broker’s portal API integration, or the Microsoft identity layer behind the portal login is unavailable, you lose real-time visibility into your bond balance and your statement.

The November outage didn’t take down the CARM portal itself. CBSA systems were unaffected. But if your broker’s EDI vendor, your internal compliance team’s cloud workspace, or the SSO provider you use to access the portal all run on Microsoft infrastructure, you couldn’t get in. You couldn’t confirm whether a high-value CAD filed Friday afternoon cleared your bond headroom, whether your Statement of Account posted a correction, or whether a duty adjustment you disputed was reflected.

Contingency procedures covered release. They don’t give you live bond accounting.

Contingency Procedures Are Slow and Manual

CBSA’s fallback protocols allow goods to move when the electronic systems are unreachable, but the process reverts to paper and manual checks. At a land border crossing, that means the carrier presents a printed cargo control document, the CBSA officer manually verifies admissibility, and release is granted on minimum documentation. The CAD gets filed later when the electronic link is restored.

For marine or air cargo arriving at a port or airport, the delays are longer because eManifest data is usually due 24 hours before loading (marine). If a carrier couldn’t transmit that data because of a Microsoft outage, CBSA has no advance notice of the shipment, and the cargo gets flagged for manual review on arrival. Even if CBSA officers clear it quickly, the downstream drayage, cross-dock cutoffs and delivery appointments are all pushed.

Containers that should have cleared PARS and been available for pickup Friday morning sat in the terminal because the electronic release messages didn’t route. Drayage got delayed. Warehouses running tight cross-dock schedules missed their cutoffs. Detention clock started running. None of that was a CBSA processing delay. The delay was in the notification layer. FENGYE LOGISTICS’ Montreal sufferance warehouse saw inbound delays stack through the weekend for exactly this reason.

Redundancy Is the Practical Takeaway

Most customs brokers run on a single EDI vendor. Most importers manage CARM access through a single identity provider. That’s one failure point. When Microsoft’s global cloud infrastructure goes down, that single point fails.

The fix is boring and expensive: redundant transmission paths, offline backups of your CARM Client Portal data (monthly Statement of Account exports, bond confirmation PDFs, CAD summaries), and a tested procedure for switching to manual filing if your primary electronic channel is unavailable. Some brokers dual-path PARS through two separate EDI providers for exactly this scenario. If one vendor’s cloud stack goes down, the route switches within the hour. Not every broker does that, because the duplicate software licensing and testing cost is hard to justify when outages are rare.

But when the outage hits during a peak clearance window (Friday afternoon before a long weekend, end of month when Statements of Account are due, Q4 when every container matters), the cost of sitting dark for 48 hours is higher than the annual cost of the redundant path.

The Outage Is Resolved, the Lesson Isn’t

TCC26-0169 was marked resolved on November 14. External clients reverted to normal procedures. If you had cargo queued during the outage window, it should have cleared once the pipeline came back online. If your Statement of Account access was blocked and you need to reconcile what posted while you were offline, log into the CARM Client Portal and pull the November statement as soon as it’s available.

Microsoft outages are rare, but cloud dependencies are not. Azure, AWS, Google Cloud — every major provider has had a service disruption in the past 24 months that affected downstream users. CBSA is building toward systems that handle higher transaction volume and tighter security, which means more reliance on modern cloud authentication, more API integrations, and more layers where a third-party failure can break your access even when CBSA’s own systems are running fine.

The importers who came through the November outage without missing a cutoff were the ones who had a fallback: a secondary EDI route, a standing manual-release protocol with their broker, or at minimum a cached copy of their CARM Client Portal data so they could make bond and payment decisions without live portal access.

Test your contingency setup before the next Microsoft outage. Download your current Statement of Account from the CARM Client Portal and save a local copy, then confirm your broker runs a redundant EDI path that doesn’t depend on the same cloud provider as your primary one. Check our dual-path routing.

Source: CSCB

Talk to a broker