Discover connected nodes via OTBR

[Jonathan Vervaeke]

My setup is the RPi OTBR and 2 thread nodes.

How can I discover the thread nodes from the borderrouter?

Currently I'm using 

ping6 -c2 ff02::1%wpan0

Which gives me all Thread nodes that can be reached by a single transmission (hence it returns fe80:: addresses), but what if a node is not reachable via a single transmission and needs another node inbetween?

Then I would need 'Mesh local address' discovery.

Is there another way to discover the devices?

Since I've used the OTBR to provide unique local addresses with

sudo wpanctl add-prefix --stable --on-mesh --default-route --slaac --preferred --priority 0 fd11:22:db8:2::

Is it possible to see what adresses are in use? (maybe in a log?) 

Thanks!

[Yakun Xu]

To discovery all devices, you may try using the Realm-Local All Thread Nodes multicast address.

Here's an example to get  Realm-Local All Thread Nodes multicast address.

Mesh Local Address: fdde:ad00:beef:0:0:ff:fe00:bc00

->

Realm-Local All Thread Nodes multicast address: ff33:40:fdde:ad00:beef:0:0:1

To get what addresses are in use, you may use the network diagnostic commands. However, it's not easy for now, we are enhancing this in our border router. 

[Jonathan Vervaeke]

Thanks for pointing that out.

Pinging the Realm local all Thread Nodes (ff33)  on a OT node works, but does not on the OTBR.

Pinging to the Link local all Thread Nodes (ff32) works fine on both.

Ping result from OT node (KW41Z):

  

> ping ff32:40:fdde:ad11:11de::1
> 16 bytes from fe80:0:0:0:f812:e0f5:debb:c583: icmp_seq=2 hlim=64 time=13ms
16 bytes from fe80:0:0:0:7871:37b3:26e5:9674: icmp_seq=2 hlim=64 time=16ms
16 bytes from fe80:0:0:0:344a:548f:f779:1042: icmp_seq=2 hlim=64 time=32ms

> ping ff33:40:fdde:ad11:11de::1
> 16 bytes from fdde:ad11:11de:0:0:ff:fe00:9c00: icmp_seq=1 hlim=63 time=24ms
16 bytes from fdde:ad11:11de:0:54a2:27d7:912d:fa3f: icmp_seq=1 hlim=64 time=37ms
16 bytes from fdde:ad11:11de:0:0:ff:fe00:0: icmp_seq=1 hlim=63 time=56ms 

Ping result from OTBR:

pi@raspberrypi:~ $ ping6 ff32:40:fdde:ad11:11de::1%wpan0
PING ff32:40:fdde:ad11:11de::1%wpan0 (ff32:40:fdde:ad11:11de::1%wpan0): 56 data bytes
64 bytes from fe80::344a:548f:f779:1042%wpan0: icmp_seq=0 ttl=64 time=0.681 ms
64 bytes from fe80::b1ff:39d6:2fbd:b67%wpan0: icmp_seq=0 ttl=64 time=46.889 ms (DUP!)
64 bytes from fe80::7871:37b3:26e5:9674%wpan0: icmp_seq=0 ttl=1 time=61.115 ms (DUP!)
64 bytes from fe80::f812:e0f5:debb:c583%wpan0: icmp_seq=0 ttl=1 time=71.377 ms (DUP!)

 

pi@raspberrypi:~/coap $ ping6 ff33:40:fdde:ad11:11de::1%wpan0
ping6: unknown host ff33:40:fdde:ad11:11de::1%wpan0

This confuses me! Do you have some insight into this?

[Yakun Xu]

Could you try?

ping6 -I wpan0 ff33:40:fdde:ad11:11de::1

[Jonathan Hui]

I believe Raspbian has a minimal implementation of ping6 that does not include the `-I <interface>` command-line argument.

I can confirm that any non-link-local scope IPv6 multicast (e.g. ff03::1 and ff05::1) results in an "unknown host" error, even specifying the interface using the `%<interface>` notation.

I'm not yet sure if Raspbian's ping6 implementation supports non-link-local scope IPv6 multicast destinations.

--

Jonathan Hui

--
You received this message because you are subscribed to the Google Groups "openthread-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openthread-use...@googlegroups.com.
To post to this group, send email to openthre...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openthread-users/0cc0d618-8d0c-44f6-bce9-e045c23ad872%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Yakun Xu]

Yes, there are two packages provides command ping6. If `-I` is not supported, try installing

sudo apt-get install -y iputils-ping

[Jonathan Vervaeke]

That actually works!

Thanks!