Stream analytics receiving messages from several devices through IoT hub about the temperature of devices. I want to use streaming analytics to produce the average temperature across all of the devices. I can't use the LAG because I don't know when the device will send the information and if one device hasn't sent a message for a while I still want to account its temperature in the average.
UDA is working over a time window and to account for all the devices I will have to use a really big window which I feel will degrade the processing speed of SA if there will be hundreds of thousands of messages in each window.
Is there an SA functionality that will allow me to create a table that will be persistent over messages and can be changed dynamically (I am aware of reference input blob storage, but it's a 'cold' storage) to use as a reference for calculating the average?