GDS Technology — Built, Wired and Secured podcast banner
Watch on YouTube →
Episodes Built
Episode 22

Smoke Tests: 10 Rapid Acceptance Checks to Stop Day‑One Failures

April 19, 2026
Key takeaways
  • Most day-one failures stem from predictable operational gaps such as stale credentials, untested failover, and inaccurate documentation.
  • A power-failover test should confirm that critical loads stay online on the intended backup source and circuits.
  • Credential testing must include both enrollment and revocation, with logs confirming that access changes propagate correctly.
  • Each smoke test needs a named owner, backup owner, clear pass/fail result, and lightweight evidence in the closeout log.
  • Batch the ten checks by location and responsibility so a small team can complete a coordinated acceptance sprint before handover.

Show Notes

Why Day-One Failures Are Usually Predictable

Move-in day can unravel quickly when a receptionist’s badge will not grant access and the backup link to the phone system does not take over. Calls do not ring, people cannot get in, the crew is pulled into troubleshooting, and a planned handover turns into a firefight.

This episode focuses on the small, operational gaps that can quietly disrupt a tenant’s first day: stale credentials, misrouted failover, untested alerts, mismatched labels, temporary services that are still carrying production loads, and missing evidence. The point is not to perform deep vendor configuration reviews or provide legal guidance. It is to run a practical acceptance sprint during closeout or immediately before move-in.

The proposed format is simple: ten rapid smoke tests, grouped by function, run by a small team in a coordinated 30–60 minute window. Each check has a clear pass/fail result, an accountable owner, and minimal evidence to place in the closeout log.

Group One: Power, Access, Monitoring, and Time

  • Power failover: Follow the documented procedure to remove mains power from a targeted system or flip the relevant breaker. Confirm that the generator transfer or UPS supports critical loads. A pass means labeled circuits remain powered and no unexpected equipment is shed. Capture a photo or short video of the breaker position and a timestamped status screen or closeout note.
  • Credential enroll and revoke: Use a tenant sample badge or temporary account. Enroll it, confirm it works at the expected doors or systems, then revoke it and confirm it no longer works. A pass requires both actions to work cleanly and appear in logs. Asymmetric behavior, such as a badge enrolling but not revoking, is a failure. Retain a brief log export or screenshot and note the doors tested.
  • Monitoring alert sanity: Trigger a known, low-risk alert by using a test sensor or intentionally muting a noncritical threshold. Confirm delivery through the intended channels, such as email, SMS, ticketing, or webhook, and verify the correct on-call person receives it. Save an alert or ticket screenshot, an exported log, and confirmation of the recipient.
  • Time alignment: Compare timestamps on access control, monitoring, and a core switch against a known accurate source. A discrepancy of about five seconds across systems may be acceptable, while a difference measured in minutes can make event correlation impossible. Save time-setting or NTP-status screenshots for the systems checked.

Group Two: Metering, Temporary Services, Spares, and Documentation

  • Metering sanity: Confirm that meters serving tenant or critical-area circuits show sensible readings under a light load. A tenant space with lights on should not read zero, and readings should not be wildly inconsistent with expected ranges or documented baselines. Capture a timestamped photo of the meter face and a one-line expectation note.
  • Temporary-services demobilization: Walk the tenant-space perimeter and service entrance. Identify temporary power or communications feeds, confirm their removal is scheduled, and safely demobilize one harmless temporary feed to verify that systems are operating on permanent feeds. A failure occurs when removing temporary service causes power or connectivity loss. Record a photo and closeout signoff.
  • Spare retrieval: Verify that physical spares are stored where operations and maintenance documentation says they are, that labels match equipment IDs on the diagram, and that the spare can actually be accessed within the expected service level. Take a photo of the location and add a short inventory line to closeout notes.
  • Closet label-to-diagram spot check: Pick one network or mechanical closet. Confirm that the door label matches the diagram, then verify the main breaker, core switch, and communications backhaul against the documented entry. A mismatch may not be dramatic at handover, but it can delay troubleshooting later. Preserve a door-label photo and timestamped diagram snapshot.

Group Three: Fallback Communications and Tenant Intake

  • Fallback communications bootstrap: Test a known backup path end to end. This might be a cellular hotspot, secondary PSTN number, or pre-approved messaging channel. Have someone call or send a message and confirm receipt by the duty contact. The evidence can be a one-line log confirmation plus the documented contact method.
  • Tenant signal intake trial: Have a tenant representative submit a simple issue through the standard help desk portal, dedicated email, or other intake channel. Confirm that it reaches the building workflow, receives an acknowledgement, and has an owner assigned. Keep the ticket number or acknowledgement screenshot as evidence.

How to Run the Sprint

Batch the checks by location and owner rather than asking one person to run every test. One person can handle power and metering checks in a zone, another can own credentials and tenant intake, and a third can walk closets and confirm spares. Use an ownership matrix with four fields: test, owner, backup owner, and evidence.

Schedule one coordinated window just before the move-in or closeout meeting. With ten checks, three people, and one hour, the goal is not exhaustive commissioning. It is fast verification of the operational dependencies most likely to create a visible day-one problem. If a test fails, log it, assign a triage priority, and require a retest before tenant access.

Lessons From the Field

A credential enroll-and-revoke check found a scheduled synchronization job had never been enabled, so revocations were not propagating to a door controller. The correction was small, and the move-in proceeded the next day. A metering sanity check identified a mislabeled meter during handover, preventing incorrect billing. In another near miss, a temporary power feed was still supporting a server rack when removal was planned; the demobilization check prevented an outage.

Three Actions to Take This Week

  • Run a 30–60 minute acceptance sprint using these ten checks.
  • Record at least two evidence artifacts for every passing or failing test.
  • Add one explicit operations-and-maintenance owner to the closeout packet.

Keep the process short, repeatable, and owned. Verify before handover, because systems can appear fine until the moment they fail. For the printable runbook and adaptable ownership matrix, visit the Built, Wired & Secured resource hub and download the 10 Smoke Tests checklist.

Deeper dive

Ten Rapid Smoke Tests That Help Prevent Day-One Failures

A tenant move-in can be delayed by issues that are neither exotic nor difficult to fix. A receptionist’s badge may not grant access. A backup link to the phone system may fail to take over. Calls do not ring, people cannot enter, the project crew gets pulled into urgent troubleshooting, and a routine handover becomes a firefight.

These failures matter because tenants experience them immediately. They also erode confidence at precisely the point when a new space, new system, or completed project is supposed to feel ready. The correction may be small, but the operational and reputational cost can be much larger when the problem is discovered after handover.

A practical way to reduce that risk is to run a short acceptance sprint during closeout or right before move-in. The sprint described here uses ten rapid smoke tests that can be completed in roughly 30 to 60 minutes when responsibilities are assigned in advance. Each test should have a simple pass/fail definition, a named owner, and minimal evidence recorded in the closeout log.

1. Test Power Failover

Power failover is a two-minute check with significant upside. Under the documented procedure, remove mains power from a targeted system or flip the relevant breaker. Then confirm that the backup source, whether generator transfer or UPS, supports the critical loads.

A pass means the intended systems remain on, labeled circuits stay powered, and no unexpected equipment is shed. A failure means equipment drops offline unexpectedly or the wrong circuits sustain the load. Facilities personnel or the electrician on site should generally own the check. Evidence can remain lightweight: a photo or short video of the breaker position and a timestamped status screen or brief closeout note.

2. Enroll and Revoke a Credential

Access control should be tested as a complete lifecycle, not just as an enrollment event. Use a tenant sample badge or a temporary account. Enroll it, verify that it works at the expected doors or systems, revoke it, and verify that it no longer works.

The key risk is asymmetric behavior. A badge that enrolls successfully but does not revoke is not a partial pass; it is an operational failure. Another failure is access being granted at the wrong doors. The access-control operator or IT team should own the work depending on the responsibility split. Save a brief log export or screenshot showing the enroll and revoke events, along with the doors that were tested.

3. Verify Monitoring Alerts Reach the Right People

Monitoring that exists but alerts the wrong distribution list is not useful when a real incident occurs. Trigger a known low-risk alert through a test sensor or by intentionally muting a noncritical threshold. Verify that the expected delivery channels work, including email, SMS, ticketing, or webhook, and confirm the correct on-call person receives the notification.

A monitoring administrator or MSP can own this check. Keep evidence simple: retain a screenshot of the alert or resulting ticket, an exported alert log if appropriate, and confirmation of who received the notification.

4. Spot Check Time Alignment

Time drift is a quiet operational problem. It may not stop a door, switch, or alerting platform from working, but it can make incident reconstruction and event correlation needlessly difficult. Compare the timestamps on access control, monitoring, and a core switch against a known accurate source.

The episode suggests that a difference of around five seconds across systems is usually acceptable, while minutes of drift are a problem. IT or the network team should own the check. Screenshots of time settings or NTP status pages provide enough evidence for a fast closeout record.

5. Confirm Metering Reads Sensibly

Metering sanity is a straightforward test: verify that meters serving tenant or critical-area circuits show plausible readings during a light load. A space with a handful of lights operating should not show zero consumption. Likewise, readings that are wildly out of range should be investigated.

The pass condition is a reading within an expected range or documented baseline. Obvious missing or incorrect values are failures. The electrical contractor or facilities team can own the check, documenting it with a timestamped meter photo and a one-line note of the expected range.

6. Demobilize Temporary Services Safely

Temporary power and communications services can create a hidden dependency if they remain in place after permanent services are assumed to be live. Walk the tenant-space perimeter and service entrance. Identify temporary feeds, confirm their removal is scheduled, and, where safe, demobilize one harmless temporary feed to verify systems remain healthy on permanent connections.

A pass means temporary feeds have been removed or are scheduled for removal, and the safe test leaves systems operating normally. A failure occurs when removing temporary infrastructure causes a power or connectivity loss. The general contractor, coordinating with trades, should own the process. A photo of the removed feed and a closeout signoff note are sufficient evidence.

7. Make Sure Spares Are Actually Retrievable

Spare retrieval may sound mundane, but it determines how quickly a team can respond when something fails. Confirm physical spares are stored where operations and maintenance documentation expects them to be. Verify that the spare label matches the equipment ID on the diagram and perform a conceptual pick-and-replace check: can a replacement be obtained within the required service level?

A spare that is missing, mislabeled, inaccessible, or locked behind unexpected approvals is not operationally ready. Facilities or inventory management should own this check. Record the location with a photo and add a short inventory line to closeout notes.

8. Match One Closet to Its Diagram

Teams often assume drawings match the installed environment. A quick spot check can reveal whether that assumption is safe. Select one network or mechanical closet, confirm the door label matches the diagram, and check the three highest-impact items: the main breaker, core switch, and communications backhaul.

A pass means the label and diagram agree for those items. A mismatch is a failure because it slows response when troubleshooting matters most. A commissioning agent or facilities technician can own the check. Preserve a photo of the closet label and a timestamped snapshot of the relevant diagram entry.

9. Bootstrap Fallback Communications

When primary communications fail, the backup method has to work end to end and be known to the people who need it. Test a documented fallback path such as a cellular hotspot, secondary PSTN number, or pre-approved messaging channel. Have someone send a message or place a call, then confirm the duty contact receives it.

A pass requires both delivery and a clear escalation procedure. A fallback path that exists on paper but cannot be reached, or that no one knows how to use, is a failure. The operations lead should own the test, recording a concise log confirmation and the contact method used.

10. Test Tenant Signal Intake

Tenants need a reliable way to report issues and know that they have been heard. Ask a tenant representative to submit a simple issue through the standard intake channel, such as a help desk portal or dedicated email. Confirm that the submission reaches the building workflow, receives an acknowledgement, and is assigned to an owner.

The pass condition is clear: a ticket is created, acknowledgement is sent, and ownership is visible. If the request disappears or receives no response, the intake path has failed. Tenant operations or the building-services coordinator should own the test. Retain the ticket number or acknowledgement screenshot.

Turn Ten Checks Into a Repeatable Sprint

These checks work best when batched by location and owner. One person can run power and meter checks within a zone. Another can handle credential and tenant-intake tests. A third can inspect closets and spares. Use a simple ownership matrix with the test, primary owner, backup owner, and evidence field.

Schedule a coordinated window immediately before the move-in or closeout meeting. The objective is not to replace comprehensive commissioning. It is to identify the operational gaps most likely to disrupt a tenant’s first day. When a check fails, log the issue, assign a triage priority, and require a retest before tenant access.

Short checks can prevent long, visible failures. A missing credential sync job, mislabeled meter, or temporary feed still carrying a server rack may take only minutes to identify, but each can create substantial disruption if left undiscovered. Keep the process brief, repeatable, documented, and clearly owned.

For a printable version of the ten smoke tests and an ownership matrix that can be adapted for closeout and move-in planning, listen to the full episode and visit the Built, Wired & Secured resource hub.