Battery powered Z-wave devices do not maintain a constant connection to the mesh, primarily to prolong the life of said batteries. This is normal.
Just chiming in so everyone is aware - in reality (more so with zwave+ devices) this graph is only showing around 1/3 of the picture as far as routing goes in a network. There are several features of the network that Zwave does not expose (or its exposed in other ways than just the neighbor list) - nodes may communicate via paths that are not present and having or not having a link between nodes and a controller in reality means nothing.
When HA can start using OZW 1.6 you will start seeing “extended” statistics that describe how a “packet” went from node to node. (But it requires a “recent” firmware on your controller) and from that you can really diagnose Zwave RF issues.
A general rule of thumb I have right now - anything with a RTT above say 150ms has issues with placement or RF interference. Looking at the neighbor list is not conclusive to why tho.
This is great information. Thanks for sharing.
Anyone know where I can find an updated firmware for my Aeotec Gen 5 Z-Stick?
According to this, there hasn’t been a firmware update for over a year
Great info! Looking forward to 1.6!
I read that to mean there hadn’t ever been a firmware update. It’s ironic that a standard which touts “firmware updates” as a selling point doesn’t actually issue any updates.
On the flip side… if it ain’t broke don’t fix it
You beat me to it! I was thinking the same thing. My Aeotec Gen 5 stick has been working perfectly for about 2 years now without any firmware update. I guess it doesn’t need one
Fibaro do consistent firmare updates… unfortunately, you can only update the firmware if you use their controllers.
Thanks for the warning, I’ll stay away from their devices then
I must disagree about Fibaro. I’ve got loads of Fibaro devices, recently I bought Fibaro HC just to update all my devices, only to find that only a couple had updates available.
Really awesome custom panel, thanks for this! I haven’t had time to check the source fully but was wondering if it would be possible to use the home assistant areas instead of the zwave (zwcfg) based locations?
Has anyone else the problem that when you click on nodes the node information for the wrong node appears? The tooltip (hovering over the node for a few seconds) is right. Not 100% certain, but it appears that only battery powered nodes are affected.
I have this issue as well, it affects the 2 nodes next to the battery node (they get jumbled), but the rest are fine.
I have finally experienced this problem and managed to fix it locally. I’ll try to update the gist tomorrow.
Updated the Gist… forgot to post here… sorry
Version 2.3: (03 February 2020)
- Graph background reflects theme background color after page reload (F5)
- Fixed problem where some removed nodes lingering in the device registry could cause wrong node info card to be displayed after clicking on nodes with higher ids
I ran through this thread, but have not really find a proper fix for the situation when nodes are not seemingly connected (eventhough they are)
It looks like this in my HA
Was/is there any proper fix to get the nodes to be connected again? I have been running this panel for some months, and the situation has not really changed.
Thanks in advance!
This can be the case right after the HA restart, but they should light up after some time.
In any case, don’t worry, it has been stated before that this graph only reflects part of the net that is exposed by HA, but the net has internal operation that is not necessarily available to HA.
If the mesh works, it works. Ignore the graph.
I am not worried, and indeed all nodes are working. But yea, it would be nice if they’d show correctly (the point of having this panel…) in the interface
If there is no fix, then too bad
The graph only has as much information as HA exposes. I have this problem with 3 Fibaro motion sensors in my mesh, they work OK, but don’t share their neighbors. Oddly enough other Fibaro motion sensors show up OK.