Multiple connectivity issues, basically everything failing

[hansib 2]

Some time ago my well-functioning HC2 setup started failing miserably. Over night I suddenly got communications issues, transfer failed, dead devices that come to life, but are mostly dead etc all over my setup. Even a fibaro plug located 1 meter from the HC2 is losing its connection and going dead. I have, according to my HC2 51 physical devices, so I am fearing that @petergebruers  rule of thumb (50+ devices might give serious issues) might be hitting my setup. Or maybe I have some fun case of RF interference..

 

CPU is idling at 1-10%

RAM is at about 30% used

Same HC2 has been running here for about 2,5 years.

Most scenes are disabled

Almost no events in the event panel so I guess no devices are polling too much

 

This case has been trough Fibaro support for some time.

The HC2 has been on a vacation to service Poland, and no issues have been found. (I did get 5 candies though..)

We have rolled back to 4.180.

About 50/50 battery/mains operates devices, mostly Fibaro/danfoss.

Re-meshing fails

Going back to old backups does nothing.

And I am wondering if my next step is to reset everything and restart from scratch.. 

 

I read somewhere that after 4.510 (which I was on before sending HC2 to poland) a automatic re-meshing starts at 0300.. Well, I had a reboot-service starting at 0300.. could that cause some issues?

 

To summarize, I have hit the wall.. The WAF (Wife Acceptance Factor) of my smarthouse is in the red. 

 

Any ideas for next steps?

Edited October 23, 2018 by hansib
update

[kroeatschge 29]

hi @hansib,

that's quite a difficult case to diagnose. When you say your HC2 was checked by Fibaro and nothing found it points

to an issue with your physical devices. I guess you didn't send all of them with your HC2 to Poland.

 

That said, when you started getting the issues, what was the most recent device that you added to your network?

Probably you have 1 bad device in your network that behaves wrong. E.g. floods the network with communication.

Z-Wave communication is simular to the old radio communication. If one device is talking, all others can only listen.

 

Do you have polling enabled on your mains powered devices? If there is no vital need for it, disable polling for them.

I have it disabled on most of them, only some Greenwave Powernodes that do not update their state properly on the controller are polled.

 

You maybe can get a Fibaro installer from your region to check your network and devices to see what's going on.

If that's not possible or you don't want to spend money on this you might try this.

Disconnect 5 devices for 1 day and see if the missbehavour remains. If it didn't get better connect them back and

disconnect 5 other devices.

You continue until your find the 1 device thats causing the issue.

 

BR - kro

 

[petergebruers 1,268]

3 hours ago, kroeatschge said:

Probably you have 1 bad device in your network that behaves wrong. E.g. floods the network with communication.

 

I agree, that is a plausible explanation.

 

It might even be a "noisy load" connected to a device!

 

Here is a similar and somewhat "weird case". I advised the user to move his HC + a few devices to another house. They worked flawlessly. So we knew the source of interference wad in his house... But it took a while to find the "offending device"

 

Please read the whole topic, I know it is long but I really recommend a full read because it is full of important details...

 

 

[hansib 2]

ok, not too much time to test today. But I moved the HC2 to the living room. And things are looking a bit better. I mean I can control some more devices without too much delay. But some other devices are unreachable. 

 

When discovering a device that works ok, I try to re-mesh it, so it connects up to its buddies. So far so good. 

 

I think I have one or more rogue devices. So I am trying to systematically kill devices.

So far I have shut everything off in the basement with the fuses. But no change

Tried to remove batteries from battery operated devices in the basement, and in the Danfoss heating controllers. No change.

DVB-dongle in the mail..

 

Remaining:

Remove the rest of the battery operated stuff

Zone by zone-shutdown with fuses. in the first floor.

 

I have also reached out to a buddy of mine who is into RF-stuff. And he might have a scanner that would be capable of scanning around 868MHz. Isn´t that lovely..

I just happen to have an ImmersionRC RF power meter which can be set to 868MHz.. But it only displays a -dB number updated 2-3 times/sec. Should I be able to find a rogue device with that? Do zwave devices emit RF-data all the time, or just when polled? Meaning; would I be able to find a consistent power output only from a rogue device and not a normally working one?

 

 I was thinking that something capable of outputting so much interference over an extended time must be mains operated right? As a battery operated thingy would soon run out of battery..

Edited October 24, 2018 by hansib

[hansib 2]

One more thing I noticed, is that if I try to delete a device, it will not go away. I press remove device in HC2, I get:

 

"[22:16:20] Slave Device Detected",

"[22:16:20] Removing devices finished" 

 

or

 

"[22:16:19] Home Center in learning mode",

"[22:16:19] Device was deleted successfully",

 

And I am not able to press any buttons on a device to be removed, the exclusion process just stops with the 2 messages above..

[hansib 2]

Aaah, Found the bugger!

 

It was an MCO single wall switch that had not been included in HC2 for some time due to multiple failures. Disconnected it. And now everything works.

 

This bugger even caused the deletion-issues mentioned above.. Or at least that works flawless now.

 

At least my HC2 got some vacation time in Poland, and I got 5 candies from fibaro. 100Euros well spent...

 

WAF restored to normal. Im going to sleep.

 

 

[petergebruers 1,268]

Thank you for reporting back! Glad it's solved!