Remote Admin not working, error URL

I have one massive Google Keep.
I fear it may kill her or ME!
:rofl:

2 Likes

I found someone to do this at my house!
I can report the remote admin link now works but if I try it from the mobile app I still get invalid hub id.

1 Like

Can you send them over to my house to reboot mine?

3 Likes

Sure. I have a Tardis in my carry on :rofl:

1 Like

I am a little puzzled here. I get it when a controller gets lost and non responsive to WAN connection but in my case the controller is running all the local rules properly, the dashboards are working from remote operation, it is only remote access to get into the settings and other setup areas that is down.

The fact the dashboard work suggests the connection is up but only something awry in the remote access area. That seems fairly pinpointed.

I was thinking I would create a rule triggered by a virtual switch called reboot. I could put the VSwitch in one of my dashboards. The rule would detect the Vswitch set to on, then in the actions, switch the Vswitch off again, and run a reboot() command. I saw reference to a reboot() in older posts but I guess the structure has changed and now I can't find that command anywhere.

Alternatively I am wondering if if I should put a WiFi switched plug/receptacle to feed Hubitat's AC adapter, then via my iPhone app so I can "out of band" reset the hub but wonder if that would be be robust enough (fixes one problem creates others...).

I could also put a cheap AC timer to just cut power for 15 minutes each day at say 3AM but that seems crude way to work around this.

Too bad Hubitat doesn't have some sort of watchdog mechanism (built in app) that will initiate a self reboot() if not properly tended to (say a ping to an external site fails x times over y hours).

For some of us with remote homes its not a useful option to "just reboot the hub" that's a $700 plane ticket, $300 car rental, plus two travel days. I came to Hubitat from a different controller because it wasn't robust enough.

Update:

I was able to create a virtual switch that triggers a hub reboot. Simple on/off switch in the dashboard. Then a rule that triggers on the Vswitch turning on, delays a few seconds, turns off the Vswitch, then triggers a reboot.

Required the installation of Hub Information Driver v3 (actually v3.1.9). Then creating a virtual device using this driver. Once the device populates then you must turn on the preference "Allow Hub to be shutdown or rebooted" and save.

Lastly in the reboot rule create the action -->> Send POST to: http://youripaddress:8080/hub/reboot

Best if you have DHCP reservation turned on for Hubitat so its IPaddress doesn't change at the local router.

That's actually a good idea.
A virtual switch and use Hub Info app to reboot.
My Dashboards were all working just fine remotely.
That might have got me in before the beta fix.

Well I lost access via the dashboards now so the system is fully lost to me remotely. It seems to be still running rules locally, as via camera, the outside lights went on and off at scheduled times.

There really needs to be a watchdog routine in Hubitat to reboot/restart due to some sort of inactivity or simply a periodic reboot/restart as a stop gap.

I have ordered a couple of WiFi smart plugs I will install when I get there in a months time so I can power cycle outside of Hubitat but this seems a crude way to recover.

This all started after I went to 2.3.9.180. Now wished I would have skipped that upgrade. System had been rock steady for 6 months prior.

1 Like

.181 is out w/a fix, seems to be working well based on reports from folks who were
consistently losing connectivity on earlier releases.

2 Likes

Thanks. I am sending someone to the house to power cycle later today in hopes that gets .180 up long enough to migrate to .181

I still think there are not enough fail safe recovery pathways such that repairs can be done remotely. Locally initiated rebooting is not a good solution for many.

I have tested a reboot via virtual switch on my other Hubitat and that seem to work fine. Maybe I'll add a time of day trigger to it as well and just reboot Hubitat everyday at 3:30AM. These in addition to a WiFi smart plug directly controlled via iPhone app.

I dont know where I saw it, but I read they are looking into some sort of failsafe to ensure the remote admin stays operational. Restarting the cloud connection on an interval or something.

Keep in mind this has only been a widespread issue with the most recent 2.3.9 builds, and .181 removes the breaking changes which caused it. So it should go right back to how it was, which from what I understand was very reliable for nearly everyone.

You can also put the hub on a Wifi plug that you can control from a separate app. This will allow you to do a power cycle for this situation or also if the entire hub locks up (in which case remote admin will never work).

4 Likes

While this only works for active hubs you can schedule a reboot nightly in RM or WebCore with Hub Info or there is a community reboot app.
A genuine watchdog service would be the ideal solution.

You can make your own using Hub Info as well, it has a cloud check function and you could set it up so if it stays failed for X minutes reboot the hub. Many people have done that already.

Would be nice if that was built into the hub though, and instead of rebooting the hub it just restarts whatever background service is jammed up.

1 Like

I refined my reboot rule a little. Might not be needed given some of the positive changes in .181 and .182 but I am going to leave it in until I can put a WiFi plug in front of Hubitat.

The rule triggers when either a manual switch on of a virtual switch on the dashboard (reboot Request) or at 3:47AM each day.

The delay is just to eliminate a race condition where Hubitat returns from reboot but it is still the time of day to reboot.

As mentioned before this is based on having Hub Information Driver v3 installed and a virtual device created from it. Lastly the device created must have "Allow Hub to be shutdown or rebooted" enabled in preferences.

Of course this isn't a cure all solution. It assumes Hubitat is rational enough to process the rule and react to the reboot POST.

Yes, I have something like that already.
An embedded Linux version is what I was hinting at.

I just installed a C8 and a C5 at a friends house. Remote Admin into the C5 always works fine. Getting into the C8 doesn't. It is up and running and working fine, I just don't have any remote admin access. Also remote dashboard doesn't work. I use the C5 to reboot the C8 and then I have access for awhile. But a day later it's gone again. When I look at HUB DETAILS in my.hubitat.com, it shows it's active. It's on the latest .184 firmware.

How are you rebooting it? With a smart plug or some sort of local call between the two hubs?

Is the C8 on Wifi by chance? Is it possible it is getting totally disconnected from the LAN?

When you checked the hub details and it says active, was the last Checkin date recent?

I created a rule in the C5 to do a post reboot command to the C8. Since they are both on the same network it seems to work.

The C8 is on WIFI.

After the reboot it does show the Checkin time as when I did the reboot.

Check it next time it goes offline and see what it looks like.

Seems it is not totally off the LAN, or the post command would not work.

Do they have a Wifi Mesh? There have been issues with Mesh systems that do optimization and change the channel (usually overnight). Sometimes boots the hub offline.

3 Likes

Can I get the last 4 characters of the MAC or hub ID? Maybe the service on the C8 wasn't enabled automatically.

It always shows the last time it re-booted.

The re-boot works from the C5. So it does appear to be on the LAN ok.

I did a remote into my friends computer and was able to connect to the hub with no problem directly on the LAN. Then did a re-boot from there and now the Remote Admin is working. But as I said it seems to drop out again at some point.