I’m not sure if I should post this topic here or in the rules section but I’ll try here first to get some Insteon perspective. I have 18 Insteon devices on my network, 6 of which are in my living room. I’ve added them all to my Items list with the binding and added them all to the sitemap as well. All works great, the response is quick when I control each device independently. The issue I run into is when I execute a rule (like Living Room Lights On) it will run the rule, and I watch it log to the log file so I know it’s executing immediately, and I can see the Insteon binding get the command immediately as well, but it takes a good 3 to 4 seconds for any of the devices to respond.
I’m using the 2014 hub and I’m not sure if that’s the right way to be going. Has anyone run into this with Insteon before? Is it the fact that I’m sending commands for multiple devices at the same time and the modem is holding them until the commands stop? I watch network traffic and there doesn’t seem to be a backlog on my network. Would I be better off using the PLM modem instead of the hub? I’m willing to test and experiment so whatever you can suggest I’m open to trying.
Can you switch on the debug log in the insteonplm binding and see exactly when the messages go out? There is a log entry for each message when it actually goes out to the modem (or hub in your case). That at least would tell us if it’s the hub or the binding that is delaying. But in a sense it is mood because it won’t fix the problem.
I no longer remember the details, but I thought at some point a feature was added to the insteonplm binding to add messages to a whole group. This could be used to switch on or off all devices that are responding to modem messages to that group. See the section “direct sending of group broadcasts” in the wiki. How to link the devices such that they listen to that particular group is another matter. It can probably be done with the insteonterminal tool.
I have been watching the InsteonPLM debug log file and the messages get sent as soon as the rule executes, and the devices report that they responded, but they don’t actually do anything for a few seconds. I know I can send messages to groups created in Insteon, or basically control the scenes, but I was avoiding this because it reduces flexibility in my control.
One thing I like about OH is that I can create groups of items and send commands to the groups or cycle through the items in a group to perform functions and set levels making the software very flexible. If I go to creating the groups on the Insteon hub then I either lose the flexibility of being able to dynamically modify groups and levels of items in the groups or I have to create a ton of different scenes in the Insteon HUB to try to account for all possible settings, which defeats the point of using OH.
So I was hoping someone has had some experience with the HUB who can tell me if this delay is a result of using the HUB or if it’s an issue with Insteon in general. Or even if it’s an issue with OH. I’m still using 1.8.3 and would be open to migrating to OH2 if that seems any better. I’d also be willing to change to the Insteon PLM device instead of the HUB if anyone thinks that would make a difference. But I’d like to not have to go to creating scenes in the Insteon devices as this really kills the flexibility of the OH software.
I’m pretty sure you’ll get better results with a PLM since it communicates over a serial port vs polling that is done with the hub. Also, I don’t think that the hub works with OH 2, whereas I was able to get the PLM to work with OH 2.
Sorry I couldn’t reply before you went out and bought the USB PLM. Because I have a USB PLM, and I also experience lags that feel like up to 3s before a device responds (usually it’s within about 1sec). That’s definitely longer than the response times you get when using broadcast group based messaging. For some reason it takes longer when the device gets a direct message.
Either way, from a technical point the PLM modem is the more solid solution, in particular because it avoids polling.
Do you have delay issues when sending messages to multiple devices from a group? That’s the biggest thing I’m trying to avoid, that 3 second delay between pressing the button and the lights responding.
Yes, using insteon groups takes a lot of flexibility: you have to go through some painful low-level process to group devices by function. I was very disappointed by the speed as well. The feature to to send group messages from the modem was added later due to user demand: I myself don’t use it.
Eventually I decided to keep home automation to the bare minimum: my lights and bathroom fans switch off automatically based on a timer. If a fan switches off after 10mins + 3 sec instead of after 10mins, who cares? My lights are only switched off if no motion has been detected for 20mins. Those motion detectors belong to the alarm system and are non-insteon, so there is some “experience the power of openhab” moment there.
I want to chime in here because I find the USB PLM superior to the HUB for COMs speed. I’m not saying USB is perfect but if I make the USB PLM a controller/responder to everything (all devices) I get WAY better performance than when using my HUB2014.
@Bernd_Pfrommer knows way more than I on this topic so take my experience with a grain of salt but I moved from HUB2014 to USB for 2 reasons: Performance and OpenHAB2
This also allowed me to use dedicate my HUB2014 for use with standard insteon app and my USB for everything else.
It can be a little confusing when you get a new device and only pair it with one (hub or usb) not both. You must pair it with both all the time.
Beyond that - it’s the way to go - separating your “Sanctioned Insteon” (HUB) and your “do it yourself” USB.
That’s great information to have. I hear what you’re saying @Bernd_Pfrommer, there’s probably always going to be a delay there, but I might be able to live with a delay now and then to gain the flexibility. Plus I’d really like to migrate to OH2, so the USB modem for me seems the logical next step to test. If it doesn’t go as well as I’d like then I’ll go from there but I feel like it’s worth trying to see what I can make work.
On another note I will mention that having begun the migration process to OH2 I’ve noticed that OH2 seems to be much quicker in response and code execution. I know it’s still in beta and there are lots of things still needing some work but I’m enjoying the way the new system configures and works together.
If you are using or planning on using a raspberry pi with openhab 2 I HIGHLY recommend using OpenHABIAN:
The author @ThomDietrich has applied so many fixes (linux securiny linux USB configuration, SMB shares for remote configuration/file sharing, on and on) to the RPI install/config “gotchas” that it give you back 90% of your weekends.
I was considering throwing in the OpenHAB towel until I found OpenHABian.