Goose66

I had a friend that had this same issue. He tried everything: Mac, PC, drivers, different cables. Then I took my laptop and cable that I used to configure mine over to his house and tried and got the same thing. Just a bad device, I presume.

That makes sense. I’ll have to think about how we should be handling that, but I will log a change for the next version.

But… thinking about this, when it comes online it won’t repost its state values (the ratgdo uses retained messages) and there is no periodic query of state, so if it does go offline and the states are set to unknown, they will remain unknown until they change (assuming the ratgdo is back online when the change occurs).

Should only happen if the ratgdo devices go offline.

I've had put in a note to fix it in next release. But as I mentioned, it is a non-fatal connection error very infrequently returned from the Schlage API the plugin uses, and the data is updated on the next poll. But it adds the notification when the error occurs. So even though it's logged as a warning and the plugin recovers on the next poll, the notification just hangs out until you clear it. Shouldn't cause any issues in the operation of the plugin.

If it is “built upon … ratgdo software libraries” like it claims, it may be easy to work it into the ratgdo plugin.

I was referring to some retrofit lighting added to the house and pool house that was totally implemented with Insteon bulbs, FanLinvs, InlineLincs, and micro dimmers. Had the linked KeypadLincs or slave SwitchLincs been removed or the Insteon links reset, this lighting would have become non-functional.

This is one major advantage of Insteon: much of this automation works completely without controllers and programs and cloud services and such. When I sold my house a couple of years ago, the new buyers wanted the “automation” with it. While I had some pool control from Keypadlincs, voice commands via Alexa, and schedules facilitated through ISY, most of the “automation” was Insteon links and scenes. In fact, contrary to @jec6613, if I had removed many of the Insteon Keypadlincs and SwitchLincs (and specifically their scene configurations) most of my floodlights, driveway entry lights, landscape and pool deck lighting, some sconces in the master bedroom, grill lights, pool house lighting, etc. would have all stopped working (or at least not been controllable). I also had an email address for the home, and, because we had lots of guests whether we were there or not, I had a “book of the house,” that I maintained on a regular basis that had operating instructions and technical details included (along with maintenance schedules and tips, paint colors, and the like). I left the ISY994i and the book of the house with the new owners - they were much appreciative.

If we could get the query to work in the ratgdo firmware, that would help too.

Resetting the status values to “Unknown” when the device goes off line is something I’m doing. I don’t really use it in programming though, so if it’s hanging other people up, I can remove it.

No worries. And I did take away your vote for option 2 (sans the separate node 😜). But now that I think about it, there's absolutely no reason not to do both option 1 and option 2 at the same time, so that's probably what I'll go with.

I addressed the idea of the “motion” node in the text above. Would just be an extra node that always said “online” but otherwise never generated a node event for the majority of users.

Selecting "Chat with a specialist" at the bottom of this page: https://www.resideo.com/us/en/support/

I chatted with Resideo support. They can see my thermostats and the "ISY Node server HomeAutomation" app or integration. This particular support rep doesn't seem to think anything has changed on their end. I wonder if the current problem lies with the authorization step. It asks you to specify "https://udi-honeywellhome-auth.azurewebsites.net/auth" as the authorization callback. This appears to be a website created by the original developer dbarentine 5 years ago. I wonder if it still functions? Sounds like this plugin needs to be brought up to newer standards and more tightly integrated with UD's oAuth and websocket infrastructures. Still doesn't explain why resideo support is not responding to password reset requests, however.

I believe this plugin was ported from PG2 by @bpwwer several years ago. I don't believe the original developer is involved at this point.

So you can no longer log into https://developer.honeywellhome.com/?

Looks like the ratgdo developer is not interested in changing (or discussing) the fact that motion will only and always report "detected" status. He implies it is intended to be just an "event" -- not a state -- suggesting this is how other motion sensors work. Of course, this is clearly not how it is documented. Further, since the MQTT motion messages are retained, then every time anybody initially subscribes or reconnects, it will look as if a new motion "event" has occurred. So we can do one of two things: 1) Keep motion as a state and have the plugin set it "cleared" 60 or so seconds after the last "detected" event; or 2) send a command (e.g., DON3) from the GDO node every time the "detected" event is posted and remove the motion state from the node. I can try and deal with the initial motion "event" on plugin restart/reconnect. I don't really want to add a "Motion" node, because there is really no way to know if the installation has a motion detector and so it would just be an extra node with "Unknown" state for the majority of users. Let me know what folks think.

Appears to be periodic REST call failure that I’m not handling gracefully. Since it’s an authorization error it’s also posting the notification that’s never getting cleared. if you can turn on debug logging and capture more info, I can release a new version that ignores the periodic error.

Is it otherwise working?

Followed the procedure for PG3x migration from here: https://forum.universal-devices.com/topic/42072-pg3-to-pg3x-migration/#comment-372757 and got stuck on the same screen as the third post in this thread: "Not connected to server... Try refreshing the page. This will disappear once the connection is re-established." Multiple reboots with hours in between over a 24+ hour period didn't clear it. Solution: Required that I click "Logout" in PG3x, and the log back in with the default (admin/admin) credentials. Unfortunately only figured this out several minutes after posting here. Just to add additional information for others, the reboots were of my Polisy. I didn't reboot my computer (or even shutdown and restart the browser, evidently).

Unlike the other status, the ratgdo firmware only appears to write out the motion status if the status read from the MyQ controller is 1 ("motion"). Specifically: if(motionState == 1){ sendMotionStatus(); motionState = 0; } All of the other states are like "if(doorState != previousDoorState) sendDoorStatus();". I can log an issue for this on ratgdo Github page. In addition, the state is retained, so it won't ever go away. One possibility is just to generate some command everything a new motion "detected" status message is posted, as opposed to having a "Motion" state value. We could generate a, e.g., DON3 command and call it "Motion" for programming purposes.

Looks like it increased sensor debounce delay in "dry contact" mode and added something having to do with reading light and lock values from sec+1. I don't see anything directly affecting the MQTT interface with the plugin, so if yours ain't broke, then... https://github.com/ratgdo/mqtt-ratgdo/compare/2.56...2.57#diff-671002be345249c9b06f425aa29425086964ffd1e884c80db89c392089a3b21bL69

Looking at the section of the log with Debug logging mode turned on, I can see that the plugin and PG3 appear to be operating correctly and fairly fast - for example, it takes 4ms after PG3 receives an Open (DON) command from IoX for ratgdo-12345 until the plugin publishes the "Open" command to MQTT broker (debug logging was not on when the plugin was started but I am assuming the local MQTT broker on your Polisy/eISY). However, a corresponding state change message is not received from the ratgdo device for another 13 seconds. That means it's taking 13 seconds for the MQTT broker to get the command to the ratgdo-12345 device, the device to start moving the door, and the device to publish a status change message back to the MQTT broker reflecting the movement. By contrast, in my network, that delay is 536 milliseconds. This points to the problem being in your network (or possibly your local MQTT broker on your Polisy/eISY). Addressing some of your specific points/questions: 1. The messages under the "homeassistant/#" topics are for discovery. They aren't actual status messages. 2. The "command" topics reflect the command message posted by the ratgdo plugins. These should show up almost immediately after you press a command button in IoX Admin Console or UD Mobile. These messages are not retained, so if you disconnect MQTT Explorer and reconnect to the broker, they will be gone. 3. AFAIK, having MyQ connected shouldn't affect anything, but I don't know the details. My MyQ continues to operate with the ratgdo devices in control, but I don't have Chamberlain MyQ door openers - I have Genie with the MyQ Universal Controllers. However, watching the doors open and close on MyQ app won't give you a very good indication of timing, because the MyQ app has tremendous latency (and variation in latency) in reflecting changes since it has to travel roundtrip to Chamberlains cloud servers (including their 3rd Party web servers and security servers). 4. Yes, hypertapping may take you to the killscreen in Tetris (evidently), but it will certainly break things in Admin Console and UD Mobile 😁. Especially when testing and debugging problems, I counsel patience so you can see what button presses causes what issues and is associated with which log entries. 5. If it is a network issue with your ratgdo devices, I won't be able to help you. If it winds up looking like a problem with the MQTT broker on your Polisy/eISY, you will need to open a support ticket with UD.

It would be helpful to change the logging level to debug and retry Open and Close commands, then resend the log. I will say, however, from the logs you sent, this looks a lot like network issues. Remember, the plugin doesn't talk to the ratgdo device. The plugin only talks to the MQTT broker, which should be the local MQTT broker on your eISY/Polisy. It seems the commands are being published successfully (i.e., no warnings or errors logged) to the MQTT broker for, e.g., "Turn Off" for light on ratgdo1234_lt and "Open" for gdo on ratgdo67899, but given the lack of change in the corresponding status values being published back from the ratgdo device, I suspect the commands aren't making it from the broker to these two ratgdo devices. For ratgdo41031, it appears to respond to commands, but takes a LONG TIME to update status: 12 seconds to respond to the "Open" command and 17 seconds to respond to the close command. You mentioned using MQTT Explorer before. If MQTT Explorer has a robust connection to the MQTT broker on your Polisy/eISY, then you should be able to monitor all this in real time and determine where the network problems are. In addition, try publishing Open and Close commands directly to the ratgdo device(s) via MQTT explorer and see how they respond. It should be virtually instantaneous.

Can you DM your Plugin log to me?