Add a timeout to the Ubidots delivery request

One of the instances has started suffering hung requests to the Ubidots API. Restarting the container makes the problem go away for some minutes(!) The problem doesn't seem to be Ubidots because the other instance is not having the same problem.

More investigation is required to determine if the hang is the actual HTTPS call to the API or before that. It is already known it isn't a hang-up in receiving messages from RabbitMQ. It could be something in the Ubidots module before the Requests call.

The problem could also be the cloud env - it is undergoing a lot of changes at present and many services are suffering.

In any case, adding a timeout to the request would be a good idea because it shouldn't hang.

Start in the Ubidots delivery module and trace the request through the Requests call, checking whether the delay occurs before or during the HTTPS request. Review how RabbitMQ delivery is excluded from the hang, then add a bounded timeout to the Ubidots request. Done means the delivery cannot wait indefinitely and the affected request path is covered by an appropriate test.

Written by the indexing model from the issue text.