On Sunday, 16 March 2025 11:07:17 GMT+1 Jingwei LI wrote: > ``` > Router A Router B Router C > _____________ ___________________ ________________ > | br-lan | | br-lan | | br-lan | > |____________| |__________________| |________________| > |AP| bat0 | |AP| bat0 | |AP| bat0 | > |___|________| |___|_______________| |___|____________| > |lan|BHAP| |STA| lan | BHAP| |STA| BHAP | > |___|______| |____|_____|______| |____|_______| > ^ ^----WiFi--^ ^ ^-----WiFi---^ > |______Eth_Cable_____| > ```
This is completely distorted in a fixed width character (editor). Is router C
required to see the problem? Because only router A and B are mentioned in the
problem description.
Is the ethernet cable connection required to trigger the problem?
> Problem description:
>
> Router A (192.168.6.1) and Router B (192.168.6.2) can ping each other. The
> client (192.168.6.192) associating with Router B cannot ping Router A and its
> associating client.
> I also wrote some simple rules to trace the data frames in each router, below
> is the output for a ping from client associating with Router B (192.168.6.2)
> to Router A (192.168.6.1):
At which point are the packets dropped from client to client? The "trace" you
showed me looks like output from iptables (or actually its kernel part
"netfilter") - which is above bat0 (and therefore not in control of batman-
adv). If you want to figure out something on layer 2, please use tcpdump to
capture packets on the various interfaces involved here (bridge, ap, bat0,
lan, BHAP, ...) and then check what happens. You can analyze pcaps recorded by
tcpdump using a somewhat recent wireshark version
How does you bridge loop avoidance configuration look like and what is the
backbone and claim (table) state?
What is the content of the translation tables?
Kind regards,
Sven
signature.asc
Description: This is a digitally signed message part.
