We have some programs set up on our tanks that are supposed to be triggered by the water level sensor reaching a certain level in the tank. These were working fine up until about a week ago when it was noticed that the valves were remaining open (the program was not triggering the next step) when the water level sensors hit the proper % filled volume. Instead, the programs were running just on programed times, which are only entered as a failsafe.
With the way things are currently running, we cannot properly use the programs developed for this application.
The water level sensors are reading the proper volume, and the calibrations are consistent with the readings desired.
This ended up being a problem with remote publishing. The cards representing that tank had too many publishes attached, which caused some items to fall off the work queue.
After reducing the amount of publishes this began working as expected.
Microclimates will consider adding guardrails to the number of publishes and/or changing the publishing to group publishes to a destination vs. individual publishes for each element selected.
In the mean time, try to limit the number of publishes to a handful of necessary elements, and try to concentrate on the configuration elements vs the state elements. Configuration elements change infrequently whereas state elements can change quite frequently, each triggering a new publish.