Quick question, as my setup continues to grow, and my rules continue to expand. I'm curious what the maximum number of automations (rules/apps) are allowed before I encounter negative performance.
Thank you
I can't imagine it is a limit anyone could possibly hit. However, a single horribly written rule could have an effect.
5 Likes
Don’t believe there is an absolute number, it comes down to how much load your mix of devices and apps place on the hub. Some on here have 100+ devices, a large number of apps and automations and have no issues. Best advice I can give is to watch your CPU load and the rate your memory drops, when they suddenly spike, back off a bit and consider whether it is time to add a new hub or change your implementations.
5 Likes
I doubt anyone would hit a limit. I would not worry about it.
3 Likes
In my experience, it usually seems to be badly behaved devices or apps that impact performance. For example, a number of power monitoring (especially z-wave) devices flooding the mesh with reports, or a device that has malfunctioned and is flooding the mesh. A weak mesh can also cause slow response due to it taking longer to find a route to the device.
@thebearmay has a great driver for monitoring hub information (hubinfo) that you can download from HPM if you really want to keep an eye on memory, cpu time, etc.
1 Like
As others have said, your Performance is dictated by the Rules, not by quantity. Let's pretend you have 50,000 Rules. You're taking up 50,000 slots in the DB. That's a memory issue. The hub has a finite memory, not expandable, so monitoring available memory tells you when your DB is getting too large. Too large and you risk burning through CPU just finding the entry in the DB.
So you have 50,000 Rules, how many RUN at once? One usually, they are tiny and quick. Which is why the advice here is that you need to watch for badly written rules that consume CPU for no value.
Apps are under the same constraints and are usually bigger than Rules. There are more than a few Apps that are 3000+ lines. Big DB entry that has to be retrieved then run.
So.. watch the indicators available (most are tabs within the Logs menu,) and you'll be ok. 
2 Likes
This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.
