Hello TP-Link Support Team,

I am reporting a firmware-level WAN reset issue on my TP-Link Aginet EB810v (Hardware V2) supplied by my ISP (StarHub Singapore).

Device Information

• Model: EB810v

• Hardware Version: V2

• Firmware Version: 0.21.0 3.0.0 v608b.0 Build 250821 Rel.52033n

• ISP / Connection Type: StarHub Singapore, IPoE (Dynamic IP / DHCP)

Problem Description

The router forcibly releases the WAN IP by itself during active use, causing a full internet disconnect for all devices.
The connection automatically reconnects after a few seconds.

This happens during sustained real-time traffic, such as online gaming (Dota 2), typically around 8–15 minutes into a session.

This is not caused by:

• DHCP lease expiry (lease time is 21600 seconds / 6 hours)

• User-initiated reconnect

• PC/network adapter disconnection

• Wi-Fi issues (issue occurs on wired LAN)

• QoS, DoS protection, or SPI firewall (all disabled)

Evidence from System Logs (Critical)

The router system log clearly shows that the router itself initiates a WAN DHCP RELEASE, followed by a full DHCP re-acquisition.

Example log entries (timestamps approximate):

DHCPC Unicasting a release of 116.89.25.236 to 116.89.0.1 DHCPC Unicasting a release of 10.176.167.38 to 10.110.36.138 DHCPC Send REQUEST DHCPC Recv OFFER DHCPC Recv ACK (lease time 21600)

This proves:

• The WAN IP is forcibly released by the router

• This is not a lease timeout

• The entire WAN stack is restarted

• LAN/Ethernet link remains up during the event

Windows Event Viewer confirms no Ethernet link down on the client side during these disconnects.

Troubleshooting Already Performed

The following have been tested and ruled out:

• QoS disabled

• DoS protection disabled

• SPI firewall disabled

• Full-cone NAT enabled

• UPnP enabled

• MTU tested (1500 / 1472)

• IPv4 & IPv6 enabled

• Unicast DHCP disabled

• Windows NIC power saving features disabled

• Tested both UDP and TCP (-tcp) game modes

Despite this, the router still releases the WAN IP, including cases where TCP is used, confirming this is not application-level or protocol-specific, but a router firmware issue.

Request

This appears to be a firmware bug or internal watchdog / WAN stack reset issue on the EB810v under sustained traffic.

I would like to request:

1. Escalation to the firmware engineering team

2. Confirmation whether this issue is known on EB810v (V2)

3. Any available firmware update, patch, or mitigation

4. Clarification if TR-069 / CWMP background processes can trigger this WAN release

This issue makes the device unsuitable for real-time applications despite correct configuration.

Thank you for your assistance.