I frequently read about node red. I understand what if does, but if I only have a handful of RM rules is it even worth giving it a go?
Yes, because you'll be able to create so many more without the clunkiness of using RM. Node-red really unlocks a lot of possibilities.
The question is, are you happy with what you have? If the answer is no and you plan on expanding on your automation, then yes. You should look at it. You WILL hate it, but you will keep going back for the beating. Just like you did with everything else home automation. It is powerful, but I find it very difficult to learn. Sometimes the documentation is nonexistent.
Well at present all the automation I have works and I would say I am happy with how it works albeit some of it is fudged.
so you're not going to expand on what you have? You're done?
Then you're not missing out. Or are you?
I try to get each rule working perfectly before moving on, so I will add more and more rules/devices. I am limited by the WAF, it has to work and work well otherwise I get stick or the rolling eyes
Node red is just a way to pull the processing from the hub.
Well, that's ONE of the things it does.... It is also a visual flow based rule set instead of structured textual GUI. Can copy/paste rules at all. Edit rules as needed without them going BOOM. Etc.
But I completely agree that if someone is happy keeping everything in RM - which is a FINE thing to do - then there isn't really any reason to change to something else. There are definite PROs to keeping everything in the hub, and not adding another PC/3rd party software in the mix.
I will second all @april.brandt has written here.
I used it long enough to move all my processing from HE to NR, which wasn't very much as I have a total of 30 devices. NR was on a plenty beefy box and motion lighting took over twice as long to execute. NR is gone now.
I don't see it happening but if I find myself needing more (?) than HE can handle I'd employ NR for just those extras...and hate it still.
See? Each to their own.
My logic is considerably faster in node-red than it was in RM (measurably, and I have the data that shows it). So I'm the opposite - I would give up Hubitat before I'd give up Node-RED. (luckily I don't have to give up either)
But everyone's uses and situation is completely unique. And I do kind of miss having everything in the hub from a simplicity standpoint.
It does depend on what you run it on. I have mine on a pi. I know that some run a sandbox. Definitely not something I'd want to take on. Mine is considerably faster and I like the challenge. I also like that I can back up my work and share/borrow it easily enough from other users.
Honestly, this is the first time I've heard someone say this.
The vast majority, myself included, have seen a measurable increase in speed when moving automations off the hub.
The only issue I've run into are swamping the HE with maker requests.. other than that I agree with @MRobi it's really fast.
I find the programming to be very intuitive and as @JasonJoel says it's very easy to edit/duplicate rules etc.
It's an extremely powerful and flexible opensource system that you control. For my use-case it's the way to go.
Ok, a couple of significant differences I'd forgot about...
I'm not using RM.
Mine was on a server kinda box BUT I had to run NR in a docker container. I was also using MQTT, not maker, and that involves a broker (on the same box as NR) so I retract my commentary as it's not a fair (to NR) comparison at least as far as processing time goes.
I still don't like the NR implementation so I'll only use it under the aforementioned circumstances. I do LOVE the concept behind NR and I'm hoping a cleaner implementation comes along.
When you stop and think about it for a second, everyone using NR with HE right now are using it on a secondary system, either a pi, server, linux machine, etc... The difference is most send the commands through a webhook with makerapi.
It's actually mind-blowing to think that most are seeing measurable speed increases when adding additional steps to the process vs the built-in rule processing system. If anything the trade-off for the ease of use of NR vs RM should be that it's slightly slower.... but it's really not.
Well I can't wrap my head around the "new way" of doing things in RM. So I'm gonna shoot myself in the foot when doing complicated things. So pulling the processing from the hub has been good for me because I need visual. And, NR offers me that logic. I write a lot on paper before trying to implement it, but I tend to be thinking about the foot before the bullet. So I get way ahead of myself and end up taking the least direct path to something via 7 bridges.
Pi is great. And not just for dessert. Lots of things you can do with one.
Are you going to eventually break down and get one ... asking for a friend.
Here's one possibility that differs strikingly from RM. Sharing automations is easy ..... I think @MRobi and I have done that despite my using the Hubitat Node-RED integration while he uses the HA Node-RED integration.
I donβt really use node-red to replace RM, but node-red enables me to interface with other devices I.e. I have a older tv that does not have wake on lan, so I use the node-red CEC node to interface with by tv. Iβm also using the raspberry for other things like Homebridge.
Too late, I have 3 including a Rpi4.
Including the Rpi3b that my Pi-hole is running on I have 7 Linux boxes online so I don't see an advantage using a Pi when I already have a hyperthreaded quad core box with 16GB memory and raid1 disks that has lots of unused capacity. Do you?