Skip to content
Work Insights
Login →
Talk to a senior advisor→
Marketing

Cloudflare One adds periodic beaconing detection: a practical investigation playbook for security teams

in X
Cloudflare One adds periodic beaconing detection: a practical investigation playbook for security teams
Marketing2026

Cloudflare One added a Periodic Beaconing Traffic risk behavior on November 16, 2026. The feature looks for repeated Gateway HTTP contacts to the same destination at unusually regular intervals. Such behavior can be consistent with malware checking in with command-and-control infrastructure, but it can also come from legitimate agents, monitoring software, and scheduled clients. The important engineering distinction is that Cloudflare describes the detection as an investigative signal, not a definitive malicious-domain classification. Teams should build a triage process before changing enforcement based on the alert.

What the detector measures

The published conditions are specific: a user contacts one destination more than thirty times within six hours, at approximately regular intervals ranging from ten seconds to twelve minutes, while relatively few other organizational users contact the same destination. The signal evaluates traffic behavior rather than reputation, threat intelligence, or domain category. It requires Gateway HTTP policies and is offered with Cloudflare One Enterprise. The predefined behavior is disabled by default and has an informational risk level when enabled. Events commonly arrive three to six hours after the observation window closes.

Check coverage and permissions before enabling

Confirm your Cloudflare plan, HTTP policy coverage, and the association between Gateway activity and user identities. An organization that sends only some endpoints through Gateway may have incomplete visibility; a quiet detector cannot prove the absence of suspicious traffic. Ensure investigators have permissions to review user risk details and associated logs. Agree on an owner, an escalation process, and retention requirements. Do not mistake this feature for blanket inspection of all network protocols or for a replacement for endpoint detection and response.

Original premium text-free rendering of evenly spaced network pulses connecting a workstation and distant endpoint under a forensic lens
Conceptual representation of periodic traffic assessment rather than a confirmed malware alert.

Enable the behavior without creating alert fatigue

1. Use the risk-behavior controls

In the Cloudflare Zero Trust dashboard, navigate to the user risk-scoring controls and open Risk score, then Risk behaviors. Enable Periodic Beaconing Traffic and leave it at its default informational severity during the initial observation period. Document when activation occurred. If your Access policies use user risk categories, review the downstream effects before raising the severity. An informational signal should not silently become a reason to deny every application session for a user with a legitimate background agent.

2. Examine the actual network context

When a result arrives, inspect the destination, timestamps, request pattern, source identity, and any corresponding endpoint or application logs your team is authorized to use. Compare the behavior with scheduled backup agents, health checks, update clients, and enterprise management tools. Determine whether the endpoint is approved, recently introduced, or interacting with a service that other users seldom need. A six-hour pattern can be operationally meaningful even when it is benign. Correlate multiple independent signals before labeling an incident or isolating a device.

3. Make the response proportional

If the pattern remains suspicious after validation, initiate the organization's established incident process, preserve relevant evidence, and assess containment based on actual risk. If the behavior is explained by an approved business service, record its owner and normal communication pattern so the next investigator has context. Resist the temptation to disable the detector for everyone after a few false positives. Instead, improve triage notes, tune related policies within the product's documented options, and review any additional device or identity signals available to your team.

Why a missing or delayed event is not a failure by itself

One or two connections will not meet the documented threshold. The detection needs regularity, a sufficient event count, a complete six-hour window, and subsequent processing time. An event may therefore reach the dashboard hours after the traffic occurred. Confirm that the Gateway HTTP telemetry is available before running any authorized simulation. Do not generate noisy test traffic in production without coordination; synthetic network patterns can trigger other defenses and contaminate operational logs. Evaluate the feature across multiple normal workdays before drawing conclusions.

Validation criteria and operational takeaway

A usable deployment has documented coverage, enabled risk behavior, understood delay, accountable analysts, and a clear decision tree for suspected and benign activity. Measure the percentage of investigated alerts that result in actionable findings, the time to triage, and recurring false-positive categories rather than just counting detections. Cloudflare's new capability can add valuable behavioral context, particularly when a destination has no known reputation signal, but investigative judgment and layered security remain essential.

Sources and further reading

Melody Bedingfield

Brightery Technical Support team member, and technical writer who's addict to new technology, robotics and automation

Editorially reviewedNovember 17, 2026

0 Responses

Share your perspective and continue the conversation.

No responses yet

Be the first to share your thoughts.

Join the conversation