Mögliches Problem mit Repeatern wie dem TPLink...
Verfasst: 14.Aug 2025, 09:49
Wenn ihr alle paar Stunden/Tage mal Verbindungsprobleme zur Wallbox habt (bei Wallboxen mit HTTP-Protokoll, ggf. mich fragen), die über einen Nano-Router wie dem TP-Link 80x angesprochen werden, dann könnte das daran liegen, dass dieser mit einem "Verbindung halten" ein Problem hat.
Ab Version 2.276 wirkt die "Verbindung halten" auch auf HTTP, d.h. man kann das dann mal ausschalten und schauen, ob die Verbindung stabiler wird.
ChatGPT zu dem Thema:
If rebooting the nano router immediately resolves the problem, that almost certainly points to the router itself being the culprit rather than Wi-Fi interference or the client. Likely causes include:
1. NAT / connection table exhaustion
These TP-Link nano routers have very limited memory for NAT sessions.
Even with a request every 5 seconds, if connections are kept open (HTTP keep-alive), the router’s connection tracking table may fill up.
Once full, new TCP connections fail → no route to host or broken streams.
Evidence: After a reboot, the NAT table is cleared → traffic resumes.
2. ARP / routing table corruption
Some firmware versions lose ARP entries or routes after a while, especially under sustained traffic or heavy NAT translation.
This would explain no route to host immediately after the HTTP error.
3. Firmware bugs / memory leaks
TP-Link WR802N is notorious for:
Memory leaks in repeater mode.
Bridge/repeater mode instability over time.
Reboot clears memory and internal tables, so the problem disappears temporarily.
Ab Version 2.276 wirkt die "Verbindung halten" auch auf HTTP, d.h. man kann das dann mal ausschalten und schauen, ob die Verbindung stabiler wird.
ChatGPT zu dem Thema:
If rebooting the nano router immediately resolves the problem, that almost certainly points to the router itself being the culprit rather than Wi-Fi interference or the client. Likely causes include:
1. NAT / connection table exhaustion
These TP-Link nano routers have very limited memory for NAT sessions.
Even with a request every 5 seconds, if connections are kept open (HTTP keep-alive), the router’s connection tracking table may fill up.
Once full, new TCP connections fail → no route to host or broken streams.
Evidence: After a reboot, the NAT table is cleared → traffic resumes.
2. ARP / routing table corruption
Some firmware versions lose ARP entries or routes after a while, especially under sustained traffic or heavy NAT translation.
This would explain no route to host immediately after the HTTP error.
3. Firmware bugs / memory leaks
TP-Link WR802N is notorious for:
Memory leaks in repeater mode.
Bridge/repeater mode instability over time.
Reboot clears memory and internal tables, so the problem disappears temporarily.