Base Station Cloud Status - changing every 4 hours?

[windracer]

I have an event that watches the Ring Base Station Cloud Status and sends me an alert when it changes. This way I know when the system goes "offline" (onto cellular backup) in addition to the alerts from Ring itself (and I can use that for other actions in Homeseer).

Last week, after updating to v1.0.0.33 I started getting a lot more alerts from this event than in the past. And when the alert was indicating the system was offline, there was no corresponding status in the Ring app. I was out of town for the holiday weekend, so couldn't check the base station physically, but I know I wasn't having WiFi issues since I have all sorts of other alerts that would triggered (outside of Homeseer).

Anyway, doing some digging, I found that the event was triggering about every 4 hours, starting on 5/26 (when I upgraded the plug-in):


So at 3am, 7am, 11am, 3pm, 7pm, and 11pm, the base station online status is changing (according to the plug-in) which triggers my event. Notice there are two events at each 4 hour repeat: one is the "offline" message, and the other is "online."

This wasn't happening before the latest plugin version, so did something change with how the base station status is being retrieved?

[dcorsus]

The ring servers disconnect every 4 hours, so that's what you see. The PI automatically tries to re-connect and it has a bit of a gradual back-off function if it doesn't succeed to reconnect immediately. That function was always there but I'm suspect that updating the state of that feature wasn't done right.

The latest version adds support for multiple locations, and depending on what kind of devices you have in those different locations, you can have more than one websocket connection active. This connection is typically associated with "an asset" and in your case it would be your base station (note if you have more assets in the same location, you still only have a single websocket connection). While introducing this new support, I fixed the alarming.

Perhaps what you can do is to change your trigger to "when device is in state xxxx for 1 minute". This will filter out the 4 hour disconnect and the sporadic disconnections you may have due to a brief network issue.

Thoughts?

ps the "alarm cloud feature" (don't have right syntax in front of me) on the MASTER device is deprecated, as stated you can have multiple webconnections so representing it with one feature on the master devices was wrong and I believe this was the feature that would be updated each time you lost connectivity.

[dcorsus]

On a different note, it is not because the PI loses connectivity with the Ring servers that your alarm hub hasn't lost its connectivity with its ring server. This is an alarm between the HS PC (PI) and the Ring Server. You probably want to monitor the wifi state which I noticed I don't have for the hub itself.

As I write this, a crew of movers, pretty much packed up the whole house and I'll be without my setup for the next 6~8 weeks or so, as we are moving internationally.
Maybe I can start writing the functions, but testing it will not be for weeks unless you want to be a guinea pig to test.

[dcorsus]

The Hub has a "NetworkConnection" attribute that is set to wlan0, eth0, ppp0
Maybe that can be of use?

[dcorsus]

Sorry for all the posts, but I think I made a mistake, where I lumped the connection status of the alarm hub (online offline, cell-backup) together with the state of the websocket. Going to have to fix that, sorry

[windracer]

Perhaps what you can do is to change your trigger to "when device is in state xxxx for 1 minute". This will filter out the 4 hour disconnect and the sporadic disconnections you may have due to a brief network issue.

Good idea, I'll start with that.

Thanks for the detailed explanation, and good luck with that move ... wow!

[windracer]

The Hub has a "NetworkConnection" attribute that is set to wlan0, eth0, ppp0
Maybe that can be of use?

That sounds to me like how the base station is connected: wireless, wired, or ?? (is point-to-point protocol even applicable here? weird).

[dcorsus]

That sounds to me like how the base station is connected: wireless, wired, or ?? (is point-to-point protocol even applicable here? weird).

I already fixed the code to go back to what it was. As to the how it is connected, I was assuming that if you loose say ethernet or wlan connectivity, it would go to cellular (ppp) and yes they still seem to run PPPoLTE. ..... but.... you might be right, that might also not do what you were looking for.

Sit tight, see if I can figure something out for these multiple locations and I'll release an update soon.

[dcorsus]

Just posted v34 which reverts some of the change I made, and now there will be a status feature under the master device for each location that has an active websocket.
The cloud status for alarm hubs now will reflect back the connectivity status.

[windracer]

and yes they still seem to run PPPoLTE.

Oh right, ppp0 would be the cellular.

Just posted v34 which reverts some of the change I made, and now there will be a status feature under the master device for each location that has an active websocket.

Thanks! Certainly did not expect a fix so quickly with the moving situation, so I appreciate that. I've upgraded to .34 and will see what happens.

[windracer]

After updating to .34, and with no changes to my event, the "every 4 hour" notifications seem to have stopped, thanks!

now there will be a status feature under the master device for each location that has an active websocket.

Should I see something different for this? I only have one location so not sure if I should see this new status ... don't see anything that looks new to me.

[dcorsus]

Should I see something different for this? I only have one location so not sure if I should see this new status ... don't see anything that looks new to me.

The master device should have a new feature

[windracer]

So in my case the master device is the alarm base station right? I don't see anything that looks new:



No more events overnight, so that issue is fixed. Although now I need to actually test having the base station go offline, onto cellular, etc.

[dcorsus]

So in my case the master device is the alarm base station right? I don't see anything that looks new:



No more events overnight, so that issue is fixed. Although now I need to actually test having the base station go offline, onto cellular, etc.

Nope, there is a device called "Master" for location "Ring", that sort of holds features that are common to the PI function , including its network connectivity etc.

[windracer]

Ah, I see it now. Forgot there was a separate device.



So I can delete that old Alarm Cloud Status feature, right?