Matter devices communicate primarily within the local network. They use IPv6 and multicast-based discovery so that devices, controllers, and services can find one another without requiring manual network configuration.
Normally, this multicast traffic is small and has little effect on network performance. Problems can occur when network equipment begins forwarding or duplicating multicast packets incorrectly.
This condition is commonly known as a multicast storm.
A multicast storm may be caused by:
- Incorrect router or managed-switch settings;
- Missing or improperly configured IPv6 MLD Snooping;
- An inactive MLD Querier;
- Multiple mDNS or Bonjour gateways forwarding the same packets;
- An mDNS reflector loop between VLANs;
- A physical Ethernet loop;
- Incorrect mesh or wireless bridge configuration;
- Wireless controllers replicating multicast traffic to every AP;
- APs converting each multicast packet into multiple unicast packets;
- Router, switch, AP, or controller firmware issues.
Why does the Internet still appear normal?
A multicast storm takes place inside the local network. It may mainly affect communication between local devices rather than Internet access.
As a result, users may still be able to browse websites, stream video, or run a normal speed test while smart-home devices experience:
- Offline status;
- Delayed commands;
- Missing status updates;
- Intermittent connectivity;
- Multiple devices disconnecting at the same time;
- Temporary recovery after restarting the router.
Internet speed and local-network stability are related, but they are not the same thing. A speed test alone cannot rule out a local multicast problem.
Recommended troubleshooting procedureļ¼
Step 1: Restart the network infrastructure
Restart the main router first, followed by managed switches, wireless controllers, APs, mesh nodes, and Wi-Fi extenders.
Once the network is fully available and stable, power-cycle the affected Matter device so it can reconnect cleanly.
Step 2: Simplify the network
Temporarily disconnect nonessential equipment, including:
- Secondary routers;
- Wi-Fi extenders;
- Additional APs;
- Mesh satellite nodes;
- Optional managed switches.
Test the Matter device using only the main router. If it remains stable, reconnect the other network components one at a time.
If the problem returns immediately after reconnecting one component, cable, or AP, that part of the network is likely involved.
Step 3: Check for network loops
Make sure that:
- Two switches are not connected by multiple ordinary Ethernet cables;
- Two ports on the same switch are not connected together;
- A mesh node is not using an incorrectly configured wired and wireless backhaul at the same time;
- A secondary router is not unintentionally operating as another main router;
- Locally switched and centrally switched WLAN paths are not being bridged together incorrectly.
Step 4: Update network firmware
Install a stable firmware version for the router, managed switches, wireless controller, APs, and mesh nodes. Multicast forwarding and table-management issues are sometimes corrected through firmware updates.
Step 5: Check managed-network settings
For business, campus, or WLC/AC + AP networks, ask the network administrator to verify:
- MLD Snooping is enabled for IPv6 multicast;
- A working MLD Querier exists in each relevant VLAN;
- MLD membership tables are not repeatedly expiring;
- IGMP Snooping is configured correctly;
- Unknown multicast is not being flooded unnecessarily;
- Multicast-to-unicast conversion is not creating excessive copies;
- Only the required mDNS/Bonjour gateway is active;
- mDNS relays or reflectors are not forwarding packets in a loop;
- STP/RSTP and loop-protection mechanisms are operating correctly;
- The Layer 2 broadcast domain is not unnecessarily large;
- Multicast and broadcast storm-control policies are reasonable.
Important
Do not solve the problem by completely disabling IPv6, mDNS, or multicast. Matter relies on these mechanisms for discovery and normal operation. Blocking them indiscriminately may prevent commissioning, discovery, control, or status updates.
The correct solution is to identify and stop the abnormal duplication or flooding while preserving normal Matter traffic.
