Connection handle 16 and 19

RE: Connection handle 16 and 19

Daniel Reisfeld — Thu, 27 May 2021 09:35:25 GMT

Hung,

I think I understand it now - thanks

Daniel

RE: Connection handle 16 and 19

Daniel Reisfeld — Wed, 26 May 2021 15:45:16 GMT

Thanks Hung,

I'll check it first myself.

Best,

Daniel

RE: Connection handle 16 and 19

Hung Bui — Wed, 26 May 2021 15:08:29 GMT

Please provide the service UUID (0x2902 is only the UUID of the CCCD, not the service UUID).

If you don't want to send the sniffer trace, you can use nRF Connect app on PC and do a service discovery of the attribute table, something like this:

If you want you can convert this case to private (GO PRIVATE) if you have confidential information.

RE: Connection handle 16 and 19

Daniel Reisfeld — Wed, 26 May 2021 14:56:39 GMT

Hung,

The nrf52840 is the peripheral and server. The phone (using Web bluetooth) writes the 0x2902 upon connection and before disconnecting. I haven't defined explicitly this service. I wonder which services are implicitly defined by the Softdevice.

Thanks,

Daniel

RE: Connection handle 16 and 19

Hung Bui — Wed, 26 May 2021 14:22:58 GMT

Hi Daniel,
We can't help you if you don't provide enough information.
I don't know if the softdevice is acting as a central or a peripheral, if it has gatt server or gatt client. We can't "guess" what you want to know without proper context (or a sniffer trace).

Please provide information such as:

- Which device is central which is peripheral

- Which device is GATT client and which is GATT server

- Which device send the write request /write command

- Which service it's writing to.

If you want to know what's 0x2902 is you can have a look at the list here (16-bit UUID Numbers)

It can be the CCCD for the service changed characteristic.

RE: Connection handle 16 and 19

Daniel Reisfeld — Wed, 26 May 2021 13:52:24 GMT

Hung hi,

I specifically described the following:

1. I would like to understand which services that are not explicitly defined are being communicated by the Softdevice (for example - I see a service 0x2902 which I don't see anywhere in my code.

2. The documentation is cryptic. There is absolutely no way to understand what is going on under the hood.

3. The Softdevice code is not available, therefore, there is no way to overcome the poor documentation by reading the code itself.

You ask for a snifet and I don't wish to go this way because I'm not trying to fix a bug. I need to know if my program is robust. It is working well but I don't understand the source of the extra services I haven't explicitly defined.

Is there a way to get an explanation of the services Softdevice is implicitly defining? Do you acknowledge that such services exists?

Daniel

RE: Connection handle 16 and 19

Hung Bui — Wed, 26 May 2021 10:13:53 GMT

I would strongly suggest to capture a sniffer trace to fully understand what happened over the air.

We don't provide source code of the softdevice but what you can do is to look at the message sequence charts here or here . They would explain most of the behaviors of the softdevice and how you should react to the event from the softdevice.

Connection handle is the handle that identify the connection. Usually when you have only one connection with a single peer the connection handle should be 0.

RE: Connection handle 16 and 19

Daniel Reisfeld — Wed, 26 May 2021 09:02:05 GMT

Hung hi,

I'm using Web Bluetooth (https://webbluetoothcg.github.io/web-bluetooth/). I also use Softdevice directly without additional libraries since I couldn't get consistent behaviour across devices that I could understand using the libraries that are in the examples. This has been a frustrating experience since examples are a poor substitute to a good documentation. My program works fine and consistently now but I obviously haven't tested it on all the phones out there. Since I don't have access to the source code of the Softdevice, there are still some things I don't understand and I cannot resolve myself by reading your code. I would like to understand which services that I haven't explicitly defined are being communicated by the Softdevice.

Thanks,

Daniel

RE: Connection handle 16 and 19

Hung Bui — Wed, 26 May 2021 08:19:53 GMT

Hi Daniel,

Please give more information. What's the application is about, how they are supposed to work ?

It's normal that you get a BLE_GATTS_EVT_WRITE before you can send notification or indication. It's because the client need to enable notification/indication by a write request or write command.

What strange here is what you described as you received connection handle 16 or 19.

Please try capturing a sniffer trace.

RE: Connection handle 16 and 19

Daniel Reisfeld — Tue, 25 May 2021 17:02:53 GMT

Hung hi,

I'm using the softdevice directly and using Web Bluetooth on the phone. When calling sd_ble_evt_get I always get a BLE_GATTS_EVT_WRITE as gatts_evt right after connections and disconnection. This is ok but it is a phenomena that is under the hood for me and I would like to understand it rather than just observing it. I noticed that I must wait for it to happen before further communicating (e.g. indicating or notifying anything).

Best, Daniel

RE: Connection handle 16 and 19

Hung Bui — Tue, 25 May 2021 13:43:24 GMT

Hi Daniel,

Which example are you testing ?

Are you asking about connection handles or attribute handles ?

Could you show some log or sniffer trace ?