11-lorawan: add RIOTCtrl tests for GNRC LoRaWAN

[jia200x]

Contribution description

This PR adds the RIOTCtrl counterpart for #199. This test is using a TTN account and MQTT in order to check uplink and downlink messages, besides activations from the Network Server.
I plan to migrate the existing semtech LoRaMAC tests as soon as possible in order to also use these mechanisms, but I gave priority to GNRC LoRaWAN because it had no autotests.

Test procedure

Check github workflow output

Related issues/PRs

#199

[jia200x]

since RIOT-OS/RIOT#15774 got merged, this one can be merged after the Hard Feature Freeze, since it won't affect the RIOT repo

[jia200x]

well, now with RIOT-OS/RIOT#15775 I can get rid of the hardcoded parameters here :)

[miri64]

this one can be merged after the Hard Feature Freeze, since it won't affect the RIOT repo

The drawback is that these won't run automatically in RC1 then.

[jia200x]

The drawback is that these won't run automatically in RC1 then.

I'm already on it. Let's see if we sneak it in ;)

[jia200x]

I implemented and amended directly both suggestions:

--------------------------------------------------------------------
Your code has been rated at 10.00/10 (previous run: 10.00/10, +0.00)

______________________________________________________________________________________ summary ______________________________________________________________________________________
  test: commands succeeded
  flake8: commands succeeded
  pylint: commands succeeded
  congratulations :)

[miri64]

Huh??! the self tests are not executed, I just realised oO

[miri64]

Ah, #200 fixes it.

[miri64]

Please rebase to include #200.

[miri64]

Some non-compulsory improvement suggestions.

[jia200x]

I will create the release branch now. For some reason LORAWAN_NWK_SKEY is not picked up (see https://github.com/RIOT-OS/RIOT/runs/1705350413?check_suite_focus=true).

Anyway, we can still run this test from our local setup. If we make it run, in worst case we create an RC2.

[jia200x]

should I squash? :)

[miri64]

should I squash? :)

Yes please :-)

[miri64]

Still needs tests for the new functionality added to testutils though.

[fjmolinas]

@fjmolinas fjmolinas force-pushed the pr/riotctrl/11-lorawan branch from 9e2fb6d to b1e85cb 38 minutes ago

Yes, they were. According to that, it was pushed from "9e2fb6d". This is the HEAD in #201 (comment)

Then maybe they just didn't show up on github? And so I though there where none?

[jia200x]

Feel free to force push and remove my changes if you prefer, I don't care/mind, just wanted to get this in

I'm also OK, if your test are passing let's just keep those

[fjmolinas]

@fjmolinas fjmolinas force-pushed the pr/riotctrl/11-lorawan branch from 9e2fb6d to b1e85cb 38 minutes ago

Yes, they were. According to that, it was pushed from "9e2fb6d". This is the HEAD in #201 (comment)

Then maybe they just didn't show up on github? And so I though there where none?

But that's weird because I fetched the branch just an hour ago...

[fjmolinas]

@fjmolinas fjmolinas force-pushed the pr/riotctrl/11-lorawan branch from 9e2fb6d to b1e85cb 38 minutes ago

Yes, they were. According to that, it was pushed from "9e2fb6d". This is the HEAD in #201 (comment)

Well... then I do not understand what happened...

[fjmolinas]

@fjmolinas fjmolinas force-pushed the pr/riotctrl/11-lorawan branch from 9e2fb6d to b1e85cb 38 minutes ago

Yes, they were. According to that, it was pushed from "9e2fb6d". This is the HEAD in #201 (comment)

Well... then I do not understand what happened...

But sorry for overstepping then.

[jia200x]

But sorry for overstepping then.

No problem! I think it was a GH issue.
Anyway, let's take your approach. I find it cleaner.

I will simply squash then

[jia200x]

done!

[fjmolinas]

@jia200x do we need to set environment variables for the GitHub actions to lunch this properly?

[fjmolinas]

@jia200x do you think you could setup a README for this as well? For what needs to be setup for me to run this locally?

[jia200x]

I pushed some commits that depend on #216. Using the new .devdata.env file:

tox -- -k "spec11" --non-RC

11-lorawan/test_spec11.py::test_task05[nodes0-otaa] PASSED                                                                                                                                   [ 50%]
11-lorawan/test_spec11.py::test_task06[nodes0-abp] PASSED                                                                                                                                    [100%]

RIOTBASE and friends are read now from the env file.

[fjmolinas]

I pushed some commits that depend on #216. Using the new .devdata.env file:

tox -- -k "spec11" --non-RC

11-lorawan/test_spec11.py::test_task05[nodes0-otaa] PASSED                                                                                                                                   [ 50%]
11-lorawan/test_spec11.py::test_task06[nodes0-abp] PASSED                                                                                                                                    [100%]

RIOTBASE and friends are read now from the env file.

Either you did not push, or github is acting funny again.

[fjmolinas]

I pushed some commits that depend on #216. Using the new .devdata.env file:

tox -- -k "spec11" --non-RC

11-lorawan/test_spec11.py::test_task05[nodes0-otaa] PASSED                                                                                                                                   [ 50%]
11-lorawan/test_spec11.py::test_task06[nodes0-abp] PASSED                                                                                                                                    [100%]

RIOTBASE and friends are read now from the env file.

Won't we need to change the gitub actions with this?

[jia200x]

I think GH is acting funny again :(

commit deb67ee0da0de8f39f9057fcea17a68f7dd9b447 (HEAD -> pr/riotctrl/11-lorawan, origin/pr/riotctrl/11-lorawan)
Author: Jose Alamos <[email protected]>
Date:   Tue Apr 27 14:31:58 2021 +0200

    fixup! pytest/conftest: add fixtures for GNRC LoRaWAN tests

[jia200x]

Won't we need to change the gitub actions with this?

We shouldn't, since GH actions doesn't pass the environment variables using this file