Thank you @david2 for making the great driver for this device. We have an issue where the device turns off my wife's Sunday 'Beauty' Day shower for 'unusual activity' (it cannot learn Sunday = beauty day!). To fix this as fast as I can, I made a flow that monitors the valve, and if it turns "closed" on Sunday between x-y time it turns back open immediately (need to change so it goes into sleep mode during that time and then wakes up, on 'the list').
This works fine, except every time the device does it's 'leak check' in the middle of the night it throws some info at the node and NR flips out. Below are the logs (I do not have a debug node attached these just com in on their own).
Is there a way I can filter these out? I know what they are, and why, but I don't need them. They clutter up logs making it harder to follow other stuff. It is not the end of the world, just figure I would ask if anyone knew of a simple way to filter these. Thanks!
technically it is a 'string' so I'm not sure that would work, and I don't think I would want to exclude all 'strings'. I wonder if that could be used to 'catch' a certain 'string'
Maybe it would be better called an unhanded return string? If it was a normal return string it would not show up without a debug node right? Sorry not proficient in NR so my terminology is surely off.
So I may be completely missing the point (apologies!!!) - coffee is still settling in..
a particular device node generates an error at a certain time which shows up in the debug?
My thought was you would use the catch node which can trap such errors. The cool thing is you can select a specific node (it even has a click on the node to select option)... and it would end up looking something like this...
This is a quick and crude example of what I was thinking... if you employ the catch node then it will not show up in the debug. You probably don't even need to wire anything to the catch node itself..
edit: it only fires on an error on the selected node. All other potential errors are not trapped.
They do show up on the debug 'log' (right hand column) and they get in the way of my "real" logs.
also notice they have a orange side to them, that is different than any other log which is why I put "error" in quotations. I don't think it is an error as much as it saying "hey I am not sure what to do with this so here".
so from what I read 'errors' in NR should be red on the sides, this being orange, we'll call it a 'warning'
I get the 'warning' on both Flow and pressure nodes (which is why it is so annoying I get like 8 of them), so I added it to all of those nodes. It only happens at 1:50 am so we'll see tomorrow. Sorry to be skeptical, but I think this falls short of the 'error' designation so I'm not 100% sure it will work. I hope to be wrong.
@TechMedX I put an updated version of the driver with the text attributes corrected to use the string datatype to GitHub. Hopefully that solves the error in node?
You may have figured this out already, but the events (opens/closes/etc) are fetched from Moen at the interval set in the driver. The real-time events use something I'm not familiar with and haven't had time to reverse engineer -- possibly Google firebase.
The smallest interval the driver supports is currently 1 minute -- which means that on average the water would be off for 30 seconds before the "open" command would be sent. So, if you're already doing this on a schedule, I'd definitely recommend using sleep modes/buttons.
Yeah I figured that out. Which is why I need to switch to modes, which is high on the list for the weekend. I still needed to clean up the logs, so thank you for this. I'll post once I know it works, or not, in a day or so. Thanks!