MrBill

The "Already subscribed" bug is indeed a tough one to solve... for better than 6 months probably I blamed Mobilinc because it seemed to only show its ugly head AFTER mobilinc had been open. I never did track down exactly what caused it, however unrelated to that bug I attacked errors in the error log... I had pages and pages daily... I drove @Michel Kohanim nuts with a ticket... he kept telling me not to worry about this or that and I kept saying but it's an error log and I want to clean these up-- we literally drove each other crazy and we both got mad at each other. Eventually I got it down to where the error log only had about a half dozen entries of random warnings or informational entries per day.... Oddly after that, the "Already Subscribed" bug went away. I've thought about creating a very simple nodeserver to monitor the number of lines in the error log, just so that I can create a program that sends a notification to review the error log if there is a sudden growth.

The templates are linked in my post above. click the underlined words...

Keep in mind that scene's can function differently based on which Controller is used. I can't see the details of your scene from the screenshot you included in the first post, so let me pick a scene of mine for the example: There are 3 Controllers here. The first controller is the ISY itself, or the root node of the scene "Christmas-All" the other 2 controllers are the RED nodes. The same scene can be programed to behave 3 different ways. The root node of the scene "Christmas-All" is what is used when a program controls the scene. So you your case specifically what you need to do to make the motion sensor turn on the scene to 20% is add a second line to Program #1. Same settings as the first line except for the "Controller" drop down, make that the motion sensor. The existing line changes what happens when the switch is the controller, the new second line changes what happens when the motion sensor is the controller.

If there's an API then a node server can be created.... the problem is with little demand its unlikely one of the indviduals that develop most of the nodeservers will create one. That said... there are blank nodeserver templates available in the store for both python and nodejs. If you understand API's then it's likely something you can do for yourself, then if you think others might be interested you could also offer the nodeserver via the store. There's also paying someone to create the nodeserver for you... Depending on how advanced the api is (how many points or methods of control) For example, Simple On/Off for the whole permanent display it might not be too bad to pay someone to make a nodeserver with a single node. On the other hand if there are needs to be many nodes with RGBW control... that could get dicey... and expensive....

is the motion detector a member of the scene? Or is control 100% thru the program 'Half Bath Sensor'? If the motion sensor is also a scene controller, what's happening is the motion sensor is kicking it on to 100%, then the program is also kicking it on and that controller has been adjusted for 65%. If the motion sensor is a scene controller, one solution is to add a line to your first program to also change the scene when the motion detector is the controller, then for better efficiency you can actually drop the first line of the second programs THEN block.

Keep in mind that manual linking to another device is only a testing procedure, it's not something that should be done on a permanent basis. That suggestion is an attempt to determine if the switch is faulty. Another approach is to plug into the same place as the PLM... generally I use a plug strip to to this. If you can add the device that way, and then move it back to it's installed location and it again can't communicate, then you have an Insteon communication issue for which there are several wiki pages.

To better understand, Is this all nodeservers or some nodeservers? Were these nodeserver previously managed by this Polisy? What "Unamanged" typically means is that there is a different polyglot instance managing that slot. The ISY has 25 slots for nodeservers numbered 1-25. Each slot could be managed by a different instance of polyglot. Typically when we this when for example, someone has a slot(s) that is managed by cloud polyglot and they also have Polisy locally. In that case, the cloud instance would show "unmanaged" for the slots being managed by Polisy, and conversely on the local Polisy instance you would see "unmanaged" for the slots managed by cloud polyglot. If your instances were originally managed by the same polisy that is now showing them as unmanaged, you will likely need to delete the polyglot database manually and start over. In any case, if you think you need to do that you should open a support ticket by emailing support@universal-devices.com for further instruction.

truth is I can't hear Insteon switches when they beep. My wife sometimes can, but not always. My granddaughters can hear them 50 feet away. We must admit that as we age our hearing changes. As others have, I strongly disagree with this statement. Three specific areas that I believe a factory reset should always be completed 1) when the device is new, before it is installed. Time has proven well that the factory ships devices that have random trash bytes in the link space. 2) when an Insteon HUB is being removed from the system. In this case I can say "been there, done that"... I started with an Insteaon Hub I didn't have it long before I found the need to upgrade... I assumed I could have both, there's a thread around here somewhere were I was helped and told to factory reset the devices. I didn't want to, it was a PITA, I eventually did after trying everything else first. I should have just listened when first told. 3) Anytime a switch acts "weird" the first line of defense is "factory reset" and then "Restore Device" (no relinking necessary in this case.) That's because we shouldn't be using other tools.... The ISY should know every setting in the switch. Using a 3rd party tool to make a change is the equivalent of manually linking... the best advice is factory reset the device and LET THE ISY MANAGE IT... Every single setting... not just some of them.

Just to add to what's been said here already... Enable Internet Access is a poorly named feature, and it's from an era that UPnP was emerging. Since that era, better more secure methods have been developed. Your ISY in fact has internet access with Enabling Internet Access.. this was a good way at one time to allow internet access from the outside.... since then routers changed, portals emerged etc.... as @Geddy points out you already have Portal... you don't need this option.

Without knowing the contents of the programs its hard to say... Keep in mind that ISY programs RUN because conditions in the IF statement cause the program to be triggered. "Run at Startup" forces the program to run during the startup process. Not having a program "running at startup" doesn't mean the program is disabled, it just means that it is forced to run during the startup process. Also on the configuration tab there are two programs that might be effecting how your ISY operates during startup. "Catch up schedules at startup" and "Missed Schedule Grace period" on mine the first is unchecked. I'm not really sure what happens if it's checked or if this is related to your issue. I don't reboot my ISY very often and don't have anything change state when I do.

It's not.. both logs are written to the SD card, "log" and "error log" The size of the logfile is a factor in how fast the SD card fails to. Before node servers the log named "log" could have months of data in it, after node servers the "log" was getting overwritten very quickly-- less than 24 hours. I've since adjusted how often some nodesevers update to lengthen that. There's also a method via telnet to increase the size of the logfile, which I've considered but haven't implemented. Michel said somewhere a long time ago that size considerations needed to be balanced, but he never provided us with a forumula. When I purchased my ISY it had a 4GB microSD card in it... today its 16gb (yes i've had one fail), so I have plenty of space to make the log bigger, which should slow down on contant re-writing of the same blocks or sectors... but I don't know what the other half of the equations for balancing the size is.

There's a number of temp files that are created, i'm not sure which does what, or where they are created when running under Ubuntu, but they might be in /tmp which is probably getting cleared when you reboot. the file names can be found here and all end in .state https://wiki.universal-devices.com/index.php?title=Main_Page#Admin_Console_Minimized.2FInvisible_and_Cannot_be_Restored edit: try searching for files named "udi_*.*"

I assume it's a 994i?? not a 99i.... Doesn't help your current situation since you've already gone nuclear, but I find that when Alexa forgets how to do things the best thing to do is wait it out, don't do anything.... she has her stupid days when something at AWS broke...

I got a bunch of silly error saying my stuff (i.e. the nodeserver) was creating a race condition. I didn't believe it but restarted the nodeserver anyway.... it didn't calm down so next I got up and unplugged the tag manager. most of those errors were from the wirlesstags app, but also got an email for each one so my phone wanted to vibrate off the table. On the ISY side, my tag heartbeat programs were also detecting loss of heartbeat from the tag. also heard from 2 other users asking if i was getting those errors and what they meant, that was when my suspicion was confirmed that it was a wireless tag server issue. I'll have to shut down both internet modems and see if the tags still update the nodeserver, but my memory is they don't.

To be completely clear about it tho, it's not all local. The tag manager must be connected to that wonderful Wireless tag Windows Cloud Server. So far I know of 3 major failures in 2021 of the cloud server in 2021 with the most recent being last Thursday night. I was slammed with errors in the app, via email, and all of my ISY programs errored. In general I've been pretty supportive of wireless tags, but I'm becoming much less enthralled every time we have one of these cloud sever failure "events". 2022 the top of my list is to test zigbee temperature sensors. Unfortunately that will likely mean that they report to Home Assistant since there isn't a nodeserver. Frankly, there's also a security issue that no one talks about. If the tag manager has the ability to make network calls on our local network that originate from wirelesstags cloud server, then a hacker that gains control of the wireless tags cloud server also has open access to our local networks.

I agree. I'm not even going to bother to respond to him......

with one exception iOS 15.1.1 has made ours work well. I'm not sure what the magic combination actually is tho.... and more importantly... WHY ISN'T APPLE SOLVING THIS? to @dakall stopping programs is really the "band-aid" method to "fix" this. And for "automation" why do you want to be responsible for remembering to tell Alexa what you want everyday? "Automation" should take into account all the the possibilities.

I've spent hours over the years defining terminology. You say "except what we call it matters" then proceed to violate your own rule. You were correct the first time.... "what we call it matters".

That logic doesn't make sense. I'd use a geofence. the lights don't turn on if your home, they do turn on if you leave. the only thing you might need the switch for turning them on ahead of leaving.. in which case it won't hurt that the geofence also turns them on... if they are already on, and get turned on it's not a problem.

you can do exactly that Just create a program that has no IF, then in either then Then or the Else (depending on what you want to say to ALexa) have a Stop Program NameOfProgram assuming that it's waiting 30 or whatever, that program will be stopped before the next statement. Of course you'll also have to link Alexa in the portal to run the program (then) or (else). Probably be simplest when mapping the "program" to Alexa to have it behave like a "Scene" (there's a dropdown for that). In that case you would put the Stop Program in the Else of the program, You could add a Then block that would turn on the extra lights immediately. Then if you said "Alexa, turn off extra light" she would run Else, if you said "Alexa, turn on extra light she would run then.

I had forgotten this but you're correct. HA calculates a scene as On if any of it's members are on, and Off only if All members are off. It ignores the state of wireless remote button controllers when calculating if the scene is on or off. (ignoring wireless buttons was actually a change that occurred after an issue I opened-- so did know but forgot)

My outside lights are controlled by 8 devices, switches and a double outlet. I have 4 scenes: Day, Evening, Late Night and Bright. All 4 scenes have all the same members. (day used to be called Off, but one of the outlets turns on in the day now so it was renamed "Day"). The ISY turns on Day at Sunrise, then turns on Evening at Sunset, Late Night turns on at "bedtime" or away mode is on at a random number of minutes after 9:46pm. the scene called Bright is a button on a keypad by the front door. In fact buttons E, F, G, H each hold one of those scenes, the buttons are "non-toggle On". No scene is ever turned off, but the KPL buttons do turn off so that they act as Radio Buttons. E is Evening, F is Late Night, G is Day, H is Bright. The KPL will tell you at a glance which scene is On and the buttons can be used to override the ISY. In practice I hardly touch the buttons, but occasionally we do use bright. and to Undo bright you must hit the corresponding Day, Evening, Late Night button.

This is true when not using HA too. Any ISY scene is considered to be off only when all members of the scene are off. The solution can vary, but sending a scene an ON command will turn the scene to it's predefined on state, regardless of it's current state. Sending an Off will always turn it off. It sounds like the problem is you may be trying to use toggle when you need to use on or off. Without further context it's hard to make suggestions. ISY scenes aren't rigid.... Scene 1 can have A, B, C Scene 2 can have A and C Scene 3 Might be A, B and D. So that's to say a scene member can belong to multiple scene, and just because B is in a scene that doesn't dictate any other scene members.

HA is another controller. I do have some ways to cross things back and forth, like i do have an insteon button that opens a zigbee shade. When you get home you just need to get a PI and load up HA and start playing. there's no way for anyone to explain why it's great.

do you mean 2413u or 2414u???.... 2413U will work with polisy