• State Not Answered
  • Replies 3 replies
  • Subscribers 69 subscribers
  • Views 96 views
  • Users 0 members are here
Attachments (0)
Nordic Case Info
  • Case ID: 338433
Options

Ignoring old SeqAuth Issue

Shahin97

Hello, 

I am in the process of adding LPN nodes to my network, Unfortunately I am getting a lot of "Ignoring old SeqAuth" Warnings and looking at the LPN friendship call-backs, the device is polling the friend a lot in bursts. I am assuming this is a timing issue with the LPN waking and communicating, what steps would I go to debug this?

Thank you,

Shahin

Parents
  • 0

    Hello Shahin,

    The warning isn't necessarily an issue. It comes from when a segmented packet of the same original address and destination address arrived with an old SeqAuth. With many nodes and a few retransmits this can happen.

    Is this warning the only real issue you are seeing? Otherwise communication with both the LPN and Friend works fine?

    Regards,

    Elfving

  • 0 in reply to Elfving

     
    There is a few issues with my LPN/Friendship setup. With an established friendship, the friend queues up messages for the lpn node. However they are not always sent to the lpn node. On the LPN side, I do get received cancelled SDU I believe and or duplicate SDU messages so I do believe there is a timing issue. 

    A key note to mention, it is more frequent with larger payloads. 200-300 bytes though a vendor model, within the max about of possible segments. 

Reply
  • 0 in reply to Elfving

     
    There is a few issues with my LPN/Friendship setup. With an established friendship, the friend queues up messages for the lpn node. However they are not always sent to the lpn node. On the LPN side, I do get received cancelled SDU I believe and or duplicate SDU messages so I do believe there is a timing issue. 

    A key note to mention, it is more frequent with larger payloads. 200-300 bytes though a vendor model, within the max about of possible segments. 

Children
  • 0 in reply to Shahin97
    Shahin97 said:
    A key note to mention, it is more frequent with larger payloads. 200-300 bytes though a vendor model, within the max about of possible segments. 

    Well if having larger messages seem to make it happen more frequently then maybe there is something there related to timing. At those payload sizes the packet would be split into multiple lower transport PDUs, which would need the Friend and LPN to be sending for a longer period of time. 

    You could try modifying the parameters mentioned in this case or here and see if it helps. Maybe CONFIG_BT_MESH_FRIEND_RECV_WIN would be the most interesting, but CONFIG_BT_MESH_LPN_RETRY_TIMEOUT, CONFIG_BT_MESH_LPN_SCAN_LATENCY, CONFIG_BT_MESH_LPN_POLL_TIMEOUT, and CONFIG_BT_MESH_LPN_RECV_DELAY would be worth a look as well. 

    Since you do have the logs, you do not see anything about the friendship being terminated?

    Regards,

    Elfving

Related