Google IOT Core - COAP Messages dropped

[Jay P.]

I ported and modified the COAP google cloud example presented at Google IO 19 to a silicon labs thunderboard sense 2. I modified the application to accommodate accelerometer data. It seems like the COAP messages payload maxes out at around 400 bytes and I have had to segment the data into different messages. I understand COAP was not designed for this and now looking back was probably not the best choice for publishing the data to the cloud. However, is this a hard limit due to COAP limitation or due to a thread configuration parameter that can be modified.

Thank you,

JP

[Achim Kraus]

Hi Jay P,

are you aware of RFC7959 (CoAP blockwise)?

Just to remind: CoAP would support also 16K messages, if your
transmission layer is able to do that in a performant and reliable way.
At least with IPv6, 1k should be reachable. But again, it's more about
other stuff, if that works reliable. And maybe also about the memory
management. If your application and/or IP-stack message buffers are
small, you will not be able to send larger messages. But with a
challenging memory setup, that would also not work, if you use TCP, or?

If I remember well, that referred solution in that video sends a token
on every request. A smarter token management may reduce that and enlarge
the possible application payload. It uses an aged Californium milestone
version (it was even deprecated at that times). When I was watching that
video two years ago, my impression was, that it is a little too
optimistic. In my experience of the past years, it requires some more
technique (e.g. DTLS CID), to really use CoAP successful in the cloud.
But, unfortunately, using CoAP in the cloud seems not to be the focus of
the most. And therefore the work-progress there is not that fast.

best regards
Achim Kraus

Am 08.04.21 um 04:20 schrieb Jay P.:

> I ported and modified the COAP google cloud example presented at Google

> IO 19 <https://www.youtube.com/watch?v=MpkdokSyQIE> to a silicon labs

> thunderboard sense 2. I modified the application to accommodate
> accelerometer data. It seems like the COAP messages payload maxes out at
> around 400 bytes and I have had to segment the data into different
> messages. I understand COAP was not designed for this and now looking
> back was probably not the best choice for publishing the data to the
> cloud. However, is this a hard limit due to COAP limitation or due to a
> thread configuration parameter that can be modified.
>
> Thank you,
>
> JP
>

> --
> 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
> <mailto:openthread-use...@googlegroups.com>.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/openthread-users/90cdf978-2afc-4983-87cd-7c599d19b776n%40googlegroups.com
> <https://groups.google.com/d/msgid/openthread-users/90cdf978-2afc-4983-87cd-7c599d19b776n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Jay P.]

Hi Achim,

Thank you for you response. I am not familiar RFC7959 as I am pretty new to this whole thing, I'll have to read through the documentation. And yes seems like the video is a bit outdated and COAP is not the best choice for this particular use case. Perhaps i'll have better luck with the MQTT example.

Thank you again for your time.

JP