PuppyIP Resource Center
Developer Tools & Networking 6 min Published 2026-11-17

Cloudflare Periodic Beaconing: Event Delays and False Positives

A Cloudflare periodic beaconing event identifies regular connections from a user to the same destination; it does not establish a malware infection. This Enterprise feature is disabled by default and has a default behavior level of Info. Events usually arrive three to six hours after a six-hour observation window closes. Check HTTP coverage, the observation period and the associated logs before deciding how to respond.

Cloudflare One Gateway HTTP Periodic Beaconing Traffic risk behaviors event investigation

Service eligibility and regional restrictions

PuppyIP serves only compliant overseas businesses and their authorized personnel. Proxy services are not available in mainland China. The service may only be used for lawful business activities outside mainland China. Use of this service within mainland China is prohibited.

Hosting a proxy IP or server overseas does not change these restrictions. The service must not be provided to end users in mainland China through relaying, forwarding, sharing or resale. Before use, read the Terms of Service.

Key Takeaways

  • Added on November 16, 2026, Periodic Beaconing Traffic is a behavioral lead. It does not use threat intelligence, categories or reputation to establish that a destination is malicious.
  • The criteria include more than 30 contacts to the same destination in six hours, steady intervals from 10 seconds to about 12 minutes, and few other organization users contacting that destination.
  • Current documentation requires Gateway HTTP policies and an Enterprise plan. DNS policies alone do not satisfy that prerequisite.
  • A continuing beacon is reported once. Events usually arrive three to six hours after the observation window closes, so they should not be interpreted as real-time alerts.
  • The detection toggle, user risk records and Access policies are separate controls. Disabling detection does not delete previous behavior records.

What does this event establish?

Cloudflare One added the Periodic Beaconing Traffic risk behavior on November 16, 2026 to identify regular contacts to the same destination in Gateway HTTP traffic. Beaconing describes a connection pattern here, rather than a final finding about the destination's purpose.

Read the published criteria together: a user contacts the same destination more than 30 times in six hours, at steady intervals from 10 seconds to about 12 minutes, while few other users in the organization contact it. Cloudflare does not publish a numeric threshold for few other users; do not invent a headcount or percentage.

Malware may periodically contact a command-and-control server, or C2 server, but legitimate software can make regular check-ins too. This detection does not use threat intelligence, categories or reputation. Investigate the originating software and business purpose instead of treating a match as sufficient grounds to label the destination malicious.

Check Enterprise access, HTTP coverage and the toggle

The announcement limits availability to Enterprise plans, and the risk score documentation lists Gateway HTTP policies as a requirement. Check the organization's plan and whether the user's relevant HTTP traffic is covered. A DNS request count is not a substitute for HTTP evidence. The announcement supplies no separate price or country list; contract terms and access in a particular account still need confirmation.

Administrators can open Zero Trust > Teams & Resources > Users, then Risk score > Risk behaviors, and look for Periodic Beaconing Traffic. The behavior is disabled by default and its default level is Info. Seeing the control does not establish that it is enabled or that a complete observation window exists for that user.

If the control is missing, record the organization plan, actual page and current documentation before checking with Cloudflare. The announcement gives no per-account rollout schedule. A missing control alone does not establish an outage, and another behavior's prerequisites should not be substituted.

Why can an event take hours and appear only once?

The behavior uses a six-hour observation window, with events usually arriving three to six hours after that window closes. Measure that delay from the relevant window's end, rather than an arbitrary HTTP request, login or toggle click. Cloudflare does not promise an event a fixed time after activation.

For illustration, if a qualifying observation window closes at 15:00, the usual timing described in the announcement would put the event around 18:00 to 21:00. This explains the time relationship; it is not a measured result, fixed processing deadline or detection guarantee. Use the same time zone when comparing logs and events.

Cloudflare reports a continuing beacon once. Continued regular traffic without repeated new events therefore does not, by itself, show that detection has stopped. Equally, no event does not establish that a device is safe. Check the toggle, traffic coverage, observation window and whether the pattern meets the criteria.

Investigate the user, time range and originating software

Open Zero Trust > Teams & Resources > Users > Risk score > Risk scoring, select the user, then select the behavior instance to view its associated detection log. Retain the event time and relevant traffic period, distinguishing when the connections occurred from when the event arrived.

Use authorized organization logs and device records to check the destination, connection pattern and originating software, then confirm the purpose with the software or business owner. Approved software may perform regular checks, for example. That is an investigation lead, not a reason to classify traffic as either legitimate or malicious from timing alone.

Record the supporting evidence and remaining unknowns before deciding on a response. When sharing investigation material, retain only the necessary period and fields. Redact tokens, complete private URLs and unrelated user details; full request bodies or complete customer records are unnecessary for an external explanation.

Separate monitoring, historical records and access restrictions

A risk behavior's toggle controls future monitoring. Cloudflare says previous detections remain in logs and associated with the user after a behavior is disabled. Resetting a user's risk score is another management action; these controls are not interchangeable ways to clear an event.

Access restrictions require a separately configured User Risk Score condition and depend on the actual policy action and other rules. Enabling this detection does not create new blocking rules for every application. If the organization already has policies that depend on risk scores, review their actual matches and impact.

The goal is to establish whether the connections are expected and retain a reviewable decision. Do not reset scores or disable all monitoring merely to make an event disappear, and do not report an informational behavior lead as a confirmed attack count.

Sources

Frequently Asked Questions

Does the default Info level mean Low in an Access policy?

Do not map them directly. Current documentation labels this behavior's default level Info, while the Access User Risk Score selector lists Low, Medium, High or Unscored. The reviewed materials provide no conversion from Info to those selector values. Check the actual user score and policy match instead of relabeling Info as Low or High.

Are Periodic Beaconing and Fast Flux the same detection?

No. Periodic beaconing examines regular contacts to one destination in Gateway HTTP traffic. The current risk behavior table's Fast Flux entry concerns rapidly changing IP addresses returned for a domain and requires Gateway DNS policies. Check their data sources, patterns and evaluation timing separately; one behavior's configuration does not establish that the other is enabled or has matched.