Reboot, then a soft rest
I’ m running a couple of scenarios to see if I can pinpoint your issue. My current working theory is that MakerAPI may not be passing along an event on temperature change (I may need to add another capability for it, i.e. Sensor) as I can get the current temperature in Node-Red if I do a redeploy but Node-Red doesn’t seem to change otherwise (could also be that I’m not familar with Node-Red to debug that side properly too.)
1 Like
I am getting an "uptime" event in NR via MakerAPI - just nothing on temeprature/freeMemory. I have it set to update every 15 minutes.
UPDATE:
That's what I found as well. Neither temperature nor freeMemory values were changing when reading from the NR device node until a "deploy". I believe the NR/HE node refreshes values when deployed but otherwise caches them until something changes. At least that's what I remember from a previous thread (which, of course, I can't find now)
Looks like it needed the Sensor capability... I'll push out a quick update in the next few minutes.
Edit: v1.6.2 should be available now
2 Likes
Anyone else's C-4 temps in a somewhat normal range? Mind's dropped from a norm of 70C to 35-40C.
Mine is consistently around 58C haven't noticed a change or anything. This is on a hub with the (US) stick removed so no Zigbee or Z-Wave.
1 Like
My other 3 hubs are C-5's and another C-7. They are: 42C, 44C and 38C at this time.
2 Likes
Confirming that uptime, temperature and freeMemory now generate an event in NR when the driver runs the scheduled job. Thanks for the quick response.
2 Likes
Not to worry, my C-4 soon warmed up. I had recently rebooted it, but thought it would have been up to temperature when I looked at it last night.
1 Like
Added the HubConnect stub driver to HPM and also placed a link to it on the first post.
2 Likes
Added Total JVM memory, Free JVM Memory, and Free JVM %
5 Likes
OK so for the stupid 
, what's the difference in memory??
Also HPM is locking up when trying to match and when trying an install during install it says I haven't the right hub firmware, but on latest..
I may need to regenerate the manifest for the package, but to answer your question. You have the total hub memory, then inside it you have the Java Virtual Machine (JVM) where all the code runs. The JVM sets aside a portion of the hub memory for its usage, thus technically you can run out of memory to run an application but still have memory available on the hub.
1 Like
You might also want to put those into the HTML attribute. That’s the one for a Dashboard tile. Individual attributes are good for graphing.
It worked fine for me. C-7, 2.2.5.131
Good thought, give me a little bit ....
1 Like
I'm on 2.2.5.130 was on 2.2.6.** but had to r9le back and that was only option, should 130 v 131 make a difference?
It shouldn't, just haven't tested it there... I'll drop it to 2.2.5.130
Well, 2.2.6 is in Beta. Perhaps something in the Beta has an issue. I’m not in the Beta program, can’t test that. But you might want to report it as a possible issue. Perhaps 2.2.6 removed some of the API calls.
You may know this, but in 2.2.6 the format of http://hubip/hub/advanced/freeOSMemoryHistory changed (5m CPU avg got added as a column [even though it isn't a memory stat 
]).
And there is a new URL available that only returns the latest values (instead of all the history of values): http://hubip/hub/advanced/freeOSMemoryLast
2 Likes
