I have a C8 and randomly I see low memory.
My questions:
Is there a tools like in Windows Task Manager, to see what is consuming the memory? Or what application is consuming the memory?
If not, and without seeing any log errors, how do I know what is consuming the memory?
You could start by looking at the Device and App Stats tabs on the Logs pages... These can be useful... but there is some nuance in how to interpret some figures displayed on these pages.... But these would be one place to start.
When you say low memory, are you referring to system generated warnings? Or are you already using Hub Info? If you aren't using Hub Info I would suggest installing it. Plus some of @thebearmay 's other useful data and charting options to allow you to monitor things over time.
In terms of likely candidates for high memory usage.... Likely candidates can include custom apps or drivers that are LAN or Cloud based..... But that is not a given.... Ultimately more investigation is required.
Also, what platform version are you running?
yes using hub info, i monitor the memory and reboot if it get like 100 or less
I tend to tap out earlier, kudos for waiting so long... But I guess that's not the point of this discussion.... 
Do you do any long-ish-term capture of the memory readings? Just curious if there has been a change recently....
Also still curious what platform version you are running...
My C8 finally came back online , im running 2.3.9.166
What do you mean by "ong-ish-term capture of the memory readings"
Maybe its the" Quick Chart app" causing memory issues
Many people (not everyone) capture device events, including readings for the Hub Info device, in (most commonly) InfluxDB and are then able to produce charts in Grafana. The "-ish" part of my question was to encompass those that use other options such as local file storage and charting of recent history, particularly using solutions recently released.
I guess what I'm asking is can you see any trend (up or down) in the recent past to try and pin-point where any sort of change may have occurred in the memory usage pattern?
That would be quite ironic.... 
yes, I finally got back in and disable the "Quick Chart app" , and the memory increased instantly
That is also surprising... Hmmm....
It may help to explain how many devices are on the hub. What kind they are (ie wifi, zwave,zigbee) . What if any custom drivers and Apps you run. If you remember what were the last few things you installed. Maybe screen shots of the Apps and devices page with their stats for the top several items.
Memory is a hard one to pin down. Using the memory itself isn't a bad thing, the problem is when a load demands more then is free. The hub should go down and eventually stabilize as memory cleanup starts to occur more and more. When i was on a c8 it would drop to around 250 right after boot. Then it would stablizr around 200-180 and run there for weeks.
I'm sure @mavrrick58 has something more important and inciteful than me to contribute.... but it would be interesting to list out what you have setup in the Quick Charts App, plus anything else that may be a likely candidate...
Honestly it sounds like you all may have already nailed it down. I wonder who manages that app and if they may have insight into why it may be holding significant memory.
Is that Bryan's or Jean's? I pretty sure it's one of those gun dev's....
EDIT: Ah... Now @JustinL ?
I was only attempting to trend (1) temperature for 7 days, sampling once every 15 minutes using the Quick Chart, then it start the low memory issue, I finally got it back online after power cycling, (I had no choice) then it came back up, I disable the Quick Chart App, the memory is back to what I normally see freeMemory : 518968
This is why i like logging data for graphing to a external system like InfluxDB. It takes the potential DB load off the hub. I understand the appeal though.
Visualization and application proceesing are very differently database tasks. More specifically it is OLAP vs OLTP whish stands for Online Analytical Processing vs Online transaction Processing. Generally db folks try to keep those thongs seperate. You can run them on one box, but then you need a more powerful system then you need for one type of load.
@mavrrick58 is right to say that there are benefits to separation of logging and visualisation... Part of my day job.... But even in my day job money / resources and time can play a part in choosing the best solution for monitoring whatever it is you need to keep tabs on. So small, easy, local solutions like Quick Charts and similar options still have a place for those that don't want or need to have what are sometimes more complicated setups like we have both described.
Maybe we should return to the practical aspect of the topic... what next...?
I was going to say earlier, at least in terms of what you have described... I would not conflate the reboot and the pausing of the App, it will be important to know the impact of both independently... Letting each settle before assessing their impact individually. i.e. pausing the app shortly after rebooting may provide an inaccurate indication of the effect of pausing the App.
Yea. Preferably enough time for the hub to run and let the memory stabilze to a fairly consistent number maintained by memory cleanup
Download the Hubitat app
