Loading HoneyMesh...
Loading HoneyMesh...
Measured proof · 29 October 2026 · 2.3.6-rc2
An operator issued a source-IP ban on node A. Two trusted peers accepted the signed signal. The banned source stopped receiving replies from all three protected lab hosts on five tested IPv4 traffic cases, while a separate unbanned source remained reachable.
3
AMD64 EC2 test hosts
5
IPv4 traffic cases per host
1
Unbanned control source
Isolated AMD64 lab · illustration of the tested path
Operator issues a MANUAL ban. Host A enforces locally and signs a mesh signal for its trusted peers.
Accepts the signal as REMOTE_MESH and enforces its own ban.
Accepts the signal as REMOTE_MESH and enforces its own ban.
Each host used a private virtual Ethernet pair and network namespace. The test source was 10.77.0.4; the unbanned control was 10.77.0.5. The arrow shows ban sharing, not packet routing.
Five traffic cases · three hosts
The same test source reached hosts A, B, and C before the operator issued a ban on A. After the signed ban reached B and C, that source received no reply in any tested case. A separate unbanned source still reached all three hosts. 3/3 means all three hosts.
“No reply” means the corrected client probe timed out after 1.5 seconds; this was not a packet-capture count. The detailed report explains the preliminary run and the fresh one-minute ban used for the final result.
A signed manual ban created on one HoneyMesh node was accepted by two authoritative peers and enforced on all three AMD64 hosts across the tested IPv4 TCP, UDP, and ICMP cases. An unbanned source stayed reachable.
Automatic detection accuracy, false positives, propagation latency, production load, peer outage and recovery, IPv6, Arm64, and other network paths were outside this test. The manual-ban result does not prove that every attack is stopped.
These selected raw test records contain the before-ban probes, final post-ban probes, and unbanned controls for each host. The detailed PDF includes the method, corrected preliminary run, package and binary hashes, and cleanup record.
Verify public download checksumsThis is the 2.3.6-rc2 AMD64 release-candidate test. The public evaluation downloads use the same release-candidate version. Pro mesh sharing requires a valid signed license.