Mesh reconfiguration Crashes Network

[clevertech 3]

HI,

 

Just wondering if anyone has had similar issues and whether they were able to resolve. I am thinking I might have to start from scratch, but would really rather not do that as most modules are spread throughout ceiling and in inconvenient places which I would need to get to if they didn't come back themselves..

 

This has only just occurred in the last 3 weeks or so. Last night I updated to to 4.550 to see if it would resolve, but no.

 

Appears to be a broader mesh issue, as far afield devices aren't trying to hop messages, but rather trying to go direct and failing 90% of time. (Driveway lights for example are some 50metres from HC and work on occasion which is impressive, but really should be relaying to ensure performance.) Interestingly some really close stuff is relaying (left rumpus blind for example would only be 5 metres from HC), but as it works well I dont really care).

 

So to overcome this I have tried over the last few weeks to rebuild the mesh, sometimes one by one, sometimes entire network. It tends not to put any info in the log box, then all zwave comms become unusable and a restart is required. Comes back fine, less the relaying issue which I was trying to fix.Also happens intermittently by itself, which is when I assume HC is trying to analyze the mesh itself.

 

Any advice would be much appreciated.

 

Thanks

Dan

 

Network image FYI

 

 

 

 

 

 

 

 

Please login or register to see this spoiler.

 

[clevertech 3]

SO after a poor response from support I decided to remove and add all devices again. Same outcome - but slight ridiculous relay map-- with devices at opposite physical ends of the network relaying back to the HC2 which sits in between them. i.e. it is making a 30 relay, rather than a 15 m doing metre direct connection.

 

So with this, I thought, given have been in the roof once might as well do it again..... this time set the network up with an old vera I had laying around....

all

 

Guess what - it all works , every device polls and responds/operates as expected. There is something really wrong with the way the HC2 is managing the mesh/neighbours.

[Flash 103]

Same here, strange routing.

Route_01 device 350.1 hops to neighbour 400    then to the HCL.

Route_02 device 355.1 hops to neighbour 350.1 then to the HCL.

 

Why does device 350.1 hops anyway, there is direct communication possible between device 350.1 and the HCL.

This should be automaticly fixed by de mesh software. Optimum routing.

This routing excist already a few weeks. did no mesh reconfig. this last weeks.

 

Route_01

Please login or register to see this spoiler.

 

 

Software version 4.550

55 Devices.

Struggling with stability and scenes that wil not run.

 

 

Edited August 2, 2019 by Flash

[Stiven 53]

May I ask, how did you get the mesh configuration preview?

[Flash 103]

 

 

Hello

Somewhere in Forum was a topic.

 

 

I followed the tutorial below, worked for me.

Good luck

 

Please login or register to see this spoiler.

 

 

 

Edited August 2, 2019 by Flash

[clevertech 3]

The whole thing has become a joke... fortunately it is at my house rather than a clients....

 

and to make matters worse is fibaro suport response..in short telling me 50 metres direct connection is impossible and even relaying 50 metres is pushing it. And that I couldnt possibly know if it was connecting directly or relaying as I dont have access to the "highly restricted" support tools. 

 

A few comments on that:

 

1. If 50 metres is "impossible" then every piece of marketing is in violation of product law in every country I know of... You cant just run around making claims (the manual clearly states 50 metres) that you wont support

2. The end point is showing it was connecting directly to the HC2 using the inbuilt range tester...so despite having no access to the "highly restricted" tools I can test this as reported by the mesh tool created 10der. 

3Why are these "highly restricted" tools not available to users to troubleshoot what is an increasingly problematic platform. I can not confidently recommend HC2 to clients anymore, despite 80+ installs using HC2.

 

I have swapped the problem endpoints with quibino ones, still direct connection, but slightly better transfer success rate then the fibaro endpoints.