After the AX86U is replaced with an optical cat, it will not automatically replay when encountering a WAN without network
Problem description: I am using Merlin 3004.388.11, and I will suddenly be unable to connect to the Internet when I use it, and I can't ping the website after connecting to the router ssh, but the online status still shows that I am online, and I need to reapply the settings to trigger the replay to restore the network, so that it can be restored.
is this problem that occurred after replacing the optical cat:
because of the upgrade package, the telecom first replaced the TEWA7501V, and when using it, it will encounter a light signal flashing light without a network, and it needs to wait for a period of time to recover.
later find the operator to come to the door again, after the light cat is replaced, the problem of flashing light is gone, the master said that the equipment on the other side has also been changed, and it has been normal for several days, and there is a probability that the network will be blocked in less than 24 hours at the beginning of these two days, the performance is that the router shows that it is online, but there is actually no way to access the Internet, and there is no trigger to re-dial the router, it must be manually restarted or triggered to dial, and if you do not manually trigger the dialing, it will always show that it is online and will not dial by itself.
1. Determine whether it is a router or an operator problem
a quick verification method
- computer direct connection to the optical cat dialing test
- disconnect the WAN port of the router, and the computer network cable is directly connected to the optical cat LAN port
- a new PPPoE dial-up number on your computer to see if you can access the Internet normally
- if you can get on →, it means that the operator's line is normal, and the problem is negotiated between the router and the PPPoE of the optical cat
- If you can't go on →, there is a problem on the carrier side (optical cat or upper device)
-
- check the status of the light cat indicator
- Observe the optical cat PON/LOS light when the is disconnected: If the PON is always on and the LOS is not flashing, it means that the optical path from the optical cat to the operator is normal
- if the computer is directly connected to the optical cat at this time, but the router does not, it is basically determined that it is a problem with the router's PPPoE heartbeat mechanism
- Observe the optical cat PON/LOS light when the
2. Merlin's firmware-specific troubleshooting points
1. Check PPPoE Dial Logs
# View the dial log cat /tmp/syslog.log | grep -i ppp # or view in real time logread -f | grep -i ppp -
No response to x echo-requests(PPPoE heartbeat unresponsive) -
Serial link appears to be disconnected -
Couldn't increase MTU(MTU negotiation issue)
2. Check the current PPPoE session status
# View PPPoE connection status cat /var/run/ppp-wan.pid 2>/dev/null && echo "PID exists" || echo "PID missing" ifconfig ppp0 # to see if the interface exists ping -I ppp0 223.5.5.5 -c 3 # Specify interface ping test ppp0 interface exists but the ping is not , the PPPoE session is dead at the bottom and the daemon is not detected. 3. Targeted solutions (by priority)
Scenario 1: Enable PPPoE heartbeat detection with automatic reconnection (most likely to resolve)
- enable the PPPoE Echo Detection if available
- adjust the LCP failure threshold : set "LCP echo failure" to 3 and "LCP echo interval" to 5 seconds
- enable the Connection Retry
# Edit PPPoE configuration nvram set wan_pppoe_options="lcp-echo-interval 5 lcp-echo-failure 3" nvram commit reboot Scenario 2: Adjust the MTU value (common compatibility issues)
# Ad hoc test ifconfig ppp0 mtu 1480 # Permanent modification (Merlin interface) Change the WAN port setting → the external network → MTU setting → to 1480 Scenario 3: Change the dial-up mode (key attempt)
- try IPv4 dial-up only : Turn off IPv6 DHCP-PD
- check the bridge mode of the optical cat: confirm that the new optical cat is indeed in bridge mode (non-routing mode), and the VLAN ID is the same as the old optical cat
- MAC address cloning : Try to clone the old optical cat WAN port MAC or computer network card MAC (some regions are bound to MACs)
Scheme 4: Compulsory preservation of scheduled tasks (bottom plan)
# Edit the scheduled task cru a wan_check "*/10 * * * * /usr/sbin/check_wan.sh" # Create a detection script cat > /usr/sbin/check_wan.sh << 'EOF' #!/bin/sh if ! ping -c 3 -W 5 223.5.5.5 >/dev/null 2>&1; then logger "WAN check failed, restarting wan..." service restart_wan fi EOF chmod +x /usr/sbin/check_wan.sh 4. Key points of communication with operators
- check the BRAS session timeout time : The new plan may have adjusted the aging time of PPPoE sessions (for example, 24-hour forced offline
- check the of the OLT template: confirm that the SN/LOID registration of the new optical cat is correct and the optical power is normal (-8 to -25dBm)
- ask to check the of the optical cat management interface: confirm that it is pure bridge mode (non-routing mode), and there is no function such as "PPPoE proxy" enabled
5. Summary suggestions
| priorities | operation | expected effect |
|---|---|---|
| 1 | LCP echo detection enabled (5 seconds/3 failures) | solve 90% of the problem of suspended animation |
| 2 | MTU changed to 1480 | solve the problem of large packages |
| 3 | the next time the network is disconnected, the computer will be directly connected to the optical cat test | clarify the responsible party |
| 4 | check if the light cat is a pure bridge | Eliminate dual NAT conflicts |
