Losing thread devices

[bkoblenz]

I have developed my own hub that connects with matter devices. I typically have 6 devices connected to my hub.

In general, things work well. I do have devices come and go and sometimes it takes a few tries to successfully resubscribe due to timeouts or unexpected states. Sometimes I get node_updated events even though the node is currently trying to resubscribe.

But, my real issue is sometimes I get mdns timeouts for devices and once they fail to subscribe the node is shutdown and never comes back. However, upon a reboot, the node will again be part of the system.

So, I suspect I have something incorrectly configured in terms of the ipv6 networking parameters on my Debian Bookworm hub. I have read the documentation and set the following net6 values:

net.ipv6.conf.wlan0.accept_ra = 1
net.ipv6.conf.wlan0.accept_ra_rt_info_max_plen = 64
net.ipv6.conf.all.forwarding = 0

My wpan0 interface is (and I am using otbr-agent):
wpan0: flags=4305<UP,POINTOPOINT,RUNNING,NOARP,MULTICAST> mtu 1280
inet6 fdd9:c754:f9a6:740f:0:ff:fe00:fc10 prefixlen 64 scopeid 0x0
inet6 fdd9:c754:f9a6:740f:0:ff:fe00:fc11 prefixlen 64 scopeid 0x0
inet6 fdd9:c754:f9a6:740f:0:ff:fe00:fc38 prefixlen 64 scopeid 0x0
inet6 fdd9:c754:f9a6:740f:0:ff:fe00:fc00 prefixlen 64 scopeid 0x0
inet6 fdd9:c754:f9a6:740f:0:ff:fe00:dc00 prefixlen 64 scopeid 0x0
inet6 fdd9:c754:f9a6:740f:39db:bc6b:a1a2:d3a5 prefixlen 64 scopeid 0x0
inet6 fd29:5c60:27e:1:733e:c90e:9e76:5007 prefixlen 64 scopeid 0x0
inet6 fe80::78f2:ac14:b8eb:d8ef prefixlen 64 scopeid 0x20
unspec 00-00-00-00-00-00-00-00-00-00-00-00-00-00-00-00 txqueuelen 500 (UNSPEC)
RX packets 312 bytes 120984 (118.1 KiB)
RX errors 0 dropped 1 overruns 0 frame 0
TX packets 228 bytes 25596 (24.9 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0

Did I miss something else? Can someone guide me as to where to look? I am not too familiar with mdns or ipv6 debugging.

thanks
-brian