The validator has a the capability to run tests against supervisor.
However there only a few test included so far, convering connection sequence and aggregated status. The reason is that when testing a supervisor, the validator acts as a site. In RSMP the supervisor initiates all commands, status request, etc. so there's not much the validador can initiate to perform testing. For example, we don't have a way to make the supervisor send commands.
There might be a few more things we can do, e.g raising alarms.
But to test commands and build a more comprehensive test suite, we would need a way to trigger specific actions in the supervisor, e.g. sending commands.
I don't think this is possible with the current version of RMSP core directly.
We we could perhaps use (abuse?) a status message to indicate to the supervisor that it should send specific commands, so we can test the flow. But it would be a bit of a hack.
Another option would be to add new message types to core for supporting testing and debugging.
RSMP 4 is expected to be build in top of MQTT, which has no inherent concept of server/client. Everything connects to the central broker. Current idea is to allow any site to implement both sites and SXL (modules). So a supervisor system could implement command interface for triggering test flows. See https://github.com/rsmp-nordic/rsmp_core_v4 for more on RMSP 4.
The validator has a the capability to run tests against supervisor.
However there only a few test included so far, convering connection sequence and aggregated status. The reason is that when testing a supervisor, the validator acts as a site. In RSMP the supervisor initiates all commands, status request, etc. so there's not much the validador can initiate to perform testing. For example, we don't have a way to make the supervisor send commands.
There might be a few more things we can do, e.g raising alarms.
But to test commands and build a more comprehensive test suite, we would need a way to trigger specific actions in the supervisor, e.g. sending commands.
I don't think this is possible with the current version of RMSP core directly.
We we could perhaps use (abuse?) a status message to indicate to the supervisor that it should send specific commands, so we can test the flow. But it would be a bit of a hack.
Another option would be to add new message types to core for supporting testing and debugging.
RSMP 4 is expected to be build in top of MQTT, which has no inherent concept of server/client. Everything connects to the central broker. Current idea is to allow any site to implement both sites and SXL (modules). So a supervisor system could implement command interface for triggering test flows. See https://github.com/rsmp-nordic/rsmp_core_v4 for more on RMSP 4.