bpwwer

Roku node server version 2.0.5 should fix the errors

Which platform and what version of PG3? The latest versions are: eisy: PG3x 3.1.22 polisy: PG3 3.1.17 The initial release of PG3x for eisy had a bug that would prevent all node servers from starting but that has been fixed in the later releases.

I have a couple of Triathalon honeycomb shades. Initially, I had them controlled only by a Pico and that worked very well. After getting a Caseta SmartBridge and re-writing the node server for it, it is also controllable via that. However, that being said, I have them set to open at sunrise and close at sunset (via the SmartBridge) and almost never do anything else to control them. Because I have the smartbridge at the opposite end of house from the shades, every one in while, it will fail to control one of the shades and I have to hit the Pico to operate it. I'm not sure if there are limits to which shades work with which protocols as I only have Caseta compatible devices. At some point I plan to get some RA3 devices and controller, but haven't yet.

@MJsan I haven't had any problems with the MagicHome node server but I only have 2 lights set up. Node servers and PG3 both maintain log files that can provide a lot of information that will help with troubleshooting problems. I suspect that if you look in the node server log, you'll see that for the lights that aren't working, the node server says it can't connect to them. From my (granted, very limited) experience, it seems like the cloud connection is much more stable than local connections which is why the app always seems to work.

@garybixlerI'm not sure what you're asking. PG3(x) uses MQTT for communication between the PG3(x) components, not for communication with external devices. A node server is needed to create a communication bridge to external devices.

@MrBill, thanks for the correction.

It's not a problem with the credentials. If you were using the wrong credentials you'd get a "401 Unauthorized" error. The problem is that PG3x is unable to contact the UDX service. The UDX service is what does the authentication and is a separate service (that is supposed to be) running on the eisy. I'm not able to debug issues with the UDX service so your best bet is to submit a support ticket.

For the ecowitt ${sys.node.node003_wh31_3.name} Temperature: ${sys.node.node003_wh31_3.ST}F Humidity: ${sys.node.node003_wh31_3CLIHUM}

It's in release testing. The update check isn't very smart and can't tell the difference between being in release testing (which is brand new for UDI) and production. Once it passes release testing, it will be moved to production and a release announcement will be posted.

The "Unknown Error" happens PG3x is unable to connect to the UDX service to authenticate. This can happen if the UDX service is upgrading or if the UDX service isn't running.

That means that PG3 isn't running. If a reboot of the Polisy doesn't fix it, then there's something wrong on your system that is preventing PG3 from starting. You can try doing an "Upgrade Packages" again, followed by a system reboot and see if that helps. If not, you'll have to submit a support ticket.

You need to specify if you're using PG3 on a Polisy or PG3x on an eisy. There are significant differences in how each behave at startup. If you're getting a site not found error and nothing shows up, then PG3(x) isn't running. PG3(x) is the server that serves the web pages to the browser. If it isn't running then there are no pages to be served. Typical reasons why it's not running would be because it was never started (UDX is supposed to start it) or it crashed on startup or it can't find the IoX or UDX isn't running. In most cases, a restart of the device will correct any of the above. If you're getting the log in page, but it won't let you log in For PG3, that would imply the wrong credentials. The PG3 login authenticates against what's in the PG3 database. For PG3x, it could mean that UDX isn't running as it uses the IoX credentials and authenticates them using UDX. Or that the credentials being used are wrong. The error message will usually indicate which one it is.

the mongo package was causing problems for UDX and since everything depends on UDX, nothing would work.

PG3/PG3x does not have any dependency on MongoDB or the mongo-c-driver package. PG3x does have a dependency on UDX and IoX IoX has a dependency on UDX

That's what the upgrade was supposed to fix. It's not a bug in PG3x but in one of the FreeBSD packages that it depends on. The "upgrade" was suppose to downgrade that package.

That is normal. It is not supposed to automatically update. You have to manually use the Upgrade Packages button on the configuration tab of the admin console. BUT don't do this now!! something is broken and you'll likely end up with a system that doesn't work. I don't have a ETA for when this will be fixed.

I can't really add much. I tried upgrading yesterday and same thing. I believe it is UDX that is hanging while doing the upgrade and it never finishes, leaving things in a state where IoX and PG3 don't work. My development system is currently in this state. I poked around a bit but I don't know how all of the upgrade process works and wasn't able to find any quick work-a-around.

I don't believe any of the UDI components were updated. Some of the packages were updated, but the underlying components themselves were not. (Although maybe UDX was changed slightly) For example package pg3x-3.1.22_3 is simply a new package build of pg3x version 3.1.22, there aren't any changes to PG3x. The problem with PG3x wasn't actually PG3x. It was/is a bug in the language PG3x is written in -- Node.js. A new version of Node.js was released and a bug in that is what caused PG3x to fail to start random node servers. The latest updates from UDI are supposed to force PG3 and PG3x to use the older version of Node.js that doesn't have the bug.

There isn't any functional difference between 3.1.16 and 3.1.17. One of the library components was upgraded so that PG3 would continue to work with other updates to the OS. The same for PG3x 3.1.21 and 3.1.22. I have not made a new release with feature/bug-fix additions for either PG3 or PG3x yet.

You need to reload the browser. The browser loads the UI components from PG3 and your browser is currently running the UI component for 6.0.63 which is not fully compatible with PG3 3.1.17. Simply reloading/refreshing the browser will cause it to load the updated UI components.

I believe there is a bug in PG3x that breaks automatic updates. It will be fixed in the next release.

Error querying station 77677 information: NOT FOUND I would guess that means there is no station ID 77677 in WeatherFlow servers.

I don't really know what triggers the device to be locally online vs. offline. But if it's not locally online, the node server can't discover it.

I don't know why. The status status should be 0 but it's not. There's a bug in the error message as it should show the reason the status is not 0, but it doesn't. I've fixed that in version 3.0.27 which I just pushed to the store. Update to 3.0.27 and restart it. The message "weatherflow:query_station_uom: Error querying station 77677 information:" should now show the reason it failed. The WeatherFlow API documents don't document what the various failures are or what they mean. So if the message isn't obvious, you'll have to contact WeatherFlow to get more details.

It is a setting in the MagicHome app. Somewhere under the device settings I believe.