Child ID request fails (for test 6.1.1 in GRL test harness)

54 views
Skip to first unread message

Martine Da Silva

unread,
Aug 24, 2017, 4:17:10 AM8/24/17
to openthread-users

Hello

previously the test 6.1.1 (Attaching to a router) passed with a DUT as End Device.
Now the test fails.

I noticed in Wireshark captures, that there is no more the TLV Active Timestamp. Is it normal ?
Can someone detect what is wrong with the test ?

I attach to this message, the snapshots of failing capture and correct capture.
I appreciate any help or suggestion.

I am working with GRL software release 1.1.0

Best regards,
Martine DA SILVA

6.1.1_ko.JPG
6.1.1_OK.JPG

Martine Da Silva

unread,
Aug 24, 2017, 5:37:00 AM8/24/17
to openthread-users

Additional information:

In my test 6.1.1 that is now failing, the DUT does not switch to Child mode.I set mode rsn as I did previously. But now it never reaches Child mode.  It goes into Detached mode instead.

Martine

Martine Da Silva

unread,
Aug 24, 2017, 7:07:47 AM8/24/17
to openthread-users
In traces, the error is:
"Frame rx failed. error : Duplicated"

See snapshot attached.

Martine

traces_error.jpg

Rongli

unread,
Aug 24, 2017, 10:12:27 PM8/24/17
to openthread-users
Hi Martine,

From the failure capture, there is no acknowledgement for the unicast message from parent.  are there any difference in the radio driver between the pass run and failure run?

Martine Da Silva

unread,
Aug 25, 2017, 10:42:25 AM8/25/17
to openthread-users
Thank you for your quick answer.
Indeed it seems something has changed concerning AutoAck feature in radio driver code.
I am still investigating.

Meanwhile, if 2 boards with my software (including OpenThread) communicate, there is no problem. One is Leader and the other one switches to Child role.
But if I use my board as DUT to communicate with golden devices, my board will never switch to Child (detached instead)
Are the golden devices waiting for an Ack and it never arrives ? Or any other idea to explain between my 2 boards this is ok but not with golden device ?

Thanks,
Martine

Jonathan Hui

unread,
Aug 25, 2017, 1:58:00 PM8/25/17
to Martine Da Silva, openthread-users
One common interoperability issue I've seen at the IEEE 802.15.4 layer is the timing of ACK frames.  IEEE 802.15.4-2006 specifies relatively tight timing constraints for ACKs sent in response to received data frames - the ACK transmission must start within several hundred microseconds after the end of the data frame being acked.

--
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-users+unsubscribe@googlegroups.com.
To post to this group, send email to openthread-users@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openthread-users/2d2a23a3-784a-4666-83d6-f56b95769279%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

Martine Da Silva

unread,
Sep 5, 2017, 10:46:04 AM9/5/17
to openthread-users, martine....@gmail.com, jon...@nestlabs.com
Indeed the error is inside low level driver by our side.
Thanks for your help and info.

Regards,
Martine DA SILVA

Reply all
Reply to author
Forward
0 new messages