• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 55 subscribers
  • Views 441 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Reference nodes hang randomly when blind nodes start working

"Henry"
Intellectual 675 points

Other Parts Discussed in Thread: Z-STACK

I have 200+ reference nodes (2430) and 100+ blind nodes (2431) to run the locating application using z-stack 1.4.3-1.2.1 with minior modification in an indoor environment.

I found when all blind nodes are powered down, everything is well, but after these blind nodes start working, some reference nodes may hang in several days randomly. I must reboot these reference nodes, everything goes well.

I have no idea to identify and address this issue. I think it is not hardware problem. I guess something goes overflow in z-stack or wrong conflicted data received.

Please advise how to address the root cause, thanks.

Henry

  • 0
    Intellectual 645 points

    Hi,

    I had the same problem with 13 reference nodes and 53 mobile nodes. The original design had all 53 nodes reporting their position every 5 seconds. This set up, as with yours, caused reference nodes to hang a random intervals. This all took place at the the customers site and was very embarrassing. We never did resolve the problem but changed the strategy so that a central server was used to sequence location determination and reporting from each mobile node. A crap solution but when the customer is standing over you...

    I'm convinced the problem was due to network loading. In the initial system the reference nodes were responding to over 50 messages/second. I had sniffer logs containing 50,000 messages and I spent many hours trying to identify a sequence of events that lead up to failure.

    The TI blast solution generates far to much traffic and you should not think of it as "the solution". Look at getting your traffic levels down and the problem will go away.

    I'd love to be able to hit TI with some hard evidence but so far I have none. In a distributed system like Zigbee tracking down random failures like this is a nightmare.

    Steve,

     

     

     

  • 0 in reply to Steve3483
    Intellectual 675 points

    Can anyone give us help to solve this issue?

    I guess high network loading causes many simultaneous wireless signals. Sometimes it causes error signals, When the reference node receive special error data, it hangs.

    Henry

  • 0 in reply to "Henry"
    TI__Expert 4155 points

    Can both of you send me a pm so I can get your contact info? I have some suggestions for you to try.

  • 0 in reply to "Double 0"
    TI__Expert 4155 points

    Also, I was reviewing the code in RefNode.c and one needs to initialize ‘blastPtr’ to NULL in line 144. It’s used before being intialized with any value. This may cause
    a problem.

  • 0 in reply to "Double 0"
    Intellectual 675 points

    Mr "Z",

    Thanks for your help. Cannot find a way to send PM to you, just send a friend request with my email, thanks.

    Henry

  • 0 in reply to "Henry"
    Intellectual 675 points

    I added the init code for the pointer in “RefNode.c” as below, am I right?

    static blastAcc_t *blastPtr = NULL;

    Cheers,

    Henry Zhou

  • 0 in reply to "Henry"
    Intellectual 675 points

    got "Z"'s confirm. I'm correct.

  • 0 in reply to Steve3483
    Intellectual 675 points

    Hi Steve,

    Have you got any progress of the issue?

    Henry