Can dashboards slow down the hub?

I've been on the forums for a few years now and I keep hearing here and there that one shouldn't add all devices to a dashboard because it could then grind your hub to a halt. For those of you that may be a little bit closer to this issue than I am, can you help me out understanding this issue a little bit more?

Are the dashboards constantly processing changes to sensors that are added to that dashboard, even if no one is accessing any dashboards? Or is it only crunching the numbers when someone is accessing the dashboard?

I'm tossing up between basically adding every sensor I've got to a huge dashboard, or creating a new virtual sensor, which contains the attributes I want to see on this dashboard, attributes which are populated by the use of RM rules mirroring these from the real devices and doing some logic checks.

The benefit I see with using a virtual "Room Status" device is that this is the only device I have to add to the dashboard for each room. That would then lessen the load on the dashboard as there wouldn't be so many devices added to it and these devices would only produce events that the dashboard actually cares about. However, doing it this way means that there will be an RM rule for each room triggering every time there is an event on any of the monitored devices. So that will put a constant load on the hub, whether or not I use the dashboard.

So the question I guess is really - Will I be stressing the hub more by adding all devices to the dashboard or by creating the virtual devices that require RM rules to constantly trigger to update the attributes of the virtual devices?

For those that are still reading, I thought I'd just add what it is I am looking at achieving with my dashboard. I am creating an overview of my whole house, where I at a glance can tell if any of the lights in the room are on, if there is motion detected from any number of motion sensors in the room, what the temperature is and if any of the contact sensors in the room are open. With some playing around with CSS and stacked tiles I've come up with the following proof of concept:

Room with no activity and no lights on:

A light turned on:

Motion active:

External door open (I don't like the look of this one so much, but it serves its purpose of alerting me to something I should be aware of):

All at once (I'm going to move the temperature tile higher up in the stack so the temperature will show neatly over the contact tile):

And at the very top of the stacked tiles for the room is a dashboard link tile, so if I click on the room it brings me to the dashboard for that particular room, with all of its devices.

Now imagine that overview of every room on one dashboard! Lots of work to get there, but I reckon it will be worth it. However, I also need to make sure that my hub isn't brought to its knees in the process.

(I've only got the proof of concept running on three rooms so far. The screenshot covers about a fifth of the house.)

Truthfully, I didn't read your entire message. I quit when I saw the very first paragraph. It may not totally answer your question, but here goes.

The reason it is recommended to not add all devices to a dashboard is because when you do that, every device that is listed in the devices to add to the dashboard is queried to see its status, even if it is not displayed on the dashboard. That takes resources, thus 'slowing' thing down. It isn't horrible, just not a best practice thing.



And also

Yes, they are. The underlying app that displays the dashboard webpage subscribes to all the events created by the devices that you “add” to that dashboard app. Even if the device is not actually displayed on the dashboard webpage, and regardless of whether or not someone has the dashboard webpage open in a browser window.

So the dashboard is constantly processing changes in the background. That can (and has) caused issues for people when too many devices are included on one dashboard.


As a practical experience: I have 16 different dashboards with 5 of them in use 24/7. The % of total time use for them is 0.028% of uptime. I don't think the number of dashboards or the number of devices are as important as the number of event those devices will generate.


I tried and didn't like the native dashboards, plus the slowdowns. I resorted to using makerapi, don't know 100% that it is less CPU intensive, but suspect it is.
Try Hubitat Dashboard (Joe Page Software on Google Playstore) It loads so fast, while the native HE dashboard is still contemplating or whatever it need to do to be functional, I've already completed the command in the app.

1 Like

I don’t think OP mentioned whether he is using android or iOS.

For those that prefer not to use the built-in dashboards, there are other solutions that are supported on both iOS and Android devices.

I am aware of the Hubitat Android Dashboard and I’m using it extensively on my wall mounted dashboard tablets. However, that’s only on Android and doesn’t help me on my iOS devices or my PCs. I really want to get the native dashboards working as I can use them on any OS. I’ve had several attempts to start using the native dashboards over the years and I’ve always given up. But I think I am on to a winner now :slight_smile:

With dashboards always processing in the background, it doesn’t really make a difference then if you use only one big dashboard or several smaller ones, if the same sensors are being used across these dashboards. There will still be the same amount of sensor events for the dashboard app to process.
I’ve often heard people recommending to break up the dashboards into smaller dashboards, but I don’t see how that would help then.

Okay, then please try the hubiVue app by @gslender . Fairly certain it runs on IOS, plus there is a web version of the app, so OS independent.
This is something I've been participating in and really loving it

Download the Hubitat app