1.6.0 dev - Network map weirdness

[james-fry]

Bug Report

What happened

Network map is rendering extremely strangely

I can provide logs of the network diagram log.

What did you expect to happen

Network map will look similar to how it looked on 1.5.1 (or better ;))

How to reproduce it (minimal and precise)

Unsure

Debug Info

zigbee2mqtt version:
hassio addon edge (1.6.0 dev)
CC253X firmware version:
CC2530_CC2591_SOURCE_ROUTING_20190619

[james-fry]

Refreshed again and got this:

[Koenkk]

Please provide logs.

[james-fry]

Thanks for looking

https://pastebin.com/FxHRssLP

Diagram 1 @ 21:00
Diagram 2 @ 21:21

[clockbrain]

In your first map at 21:00 the 'Ikea Router Kitchen' didn't respond to the link quality request (LQI).
In your second map at 21:21 both the 'Ikea Router Kitchen' and the 'Garden Floodlight' didn't respond.

This can be seen in the log file:

9/12/2019, 9:21:41 PM - info: Starting network scan (includeRoutes 'false')
9/12/2019, 9:21:41 PM - debug: LQI succeeded for 'Fireplace Lights'
9/12/2019, 9:21:51 PM - error: Failed to execute LQI for 'Ikea Router Kitchen'
9/12/2019, 9:21:51 PM - debug: LQI succeeded for 'CC2531 Router Kitchen'
9/12/2019, 9:21:51 PM - debug: LQI succeeded for 'Garden Switch'
9/12/2019, 9:22:01 PM - error: Failed to execute LQI for 'Garden Floodlight'
9/12/2019, 9:22:01 PM - debug: LQI succeeded for 'Ikea Tradfri WW Bulb'
9/12/2019, 9:22:01 PM - debug: LQI succeeded for 'Coordinator'
9/12/2019, 9:22:02 PM - debug: LQI succeeded for 'Kitchen Pendant'
9/12/2019, 9:22:02 PM - info: Network scan finished

and is also tagged (somewhat subtly) on the map itself:

I've no idea why those devices didn't respond - timeout/powered off/lost pairing? But that is where to start looking. Do they still respond to commands? Did the map repair itself later?

[james-fry]

Thanks @clockbrain.
Def there some issues in my network but I more raised this defect because of the rendering issue where nodes in the graph are plotted in the far extremes

[Koenkk]

@james-fry note that the rendering is not done by zigbee2mqtt.

[james-fry]

@Koenkk this I understand, but it seems there is data from z2m that is driving these strange maps.
Its not impossible that a problem with my devices suddenly occurred and made this happen, but I think it is related to 1.6.0 upgrade. If this is not correct, please feel free to close the issue.

[dreimer1986]

Some things clearly changed in that matter. See above the report I linked to. In the lovelace networkmap by azuwis there are zero links between the devices in 1.6.0 dev and the one by rgruebel looks quite funny on my side, too. Not that strange as the OP pics show, but some links are gone for sure. Both worked fine / looked different in 1.5.1:

rgruebel version (before all devices had links to almost any other routing supportive device resulting in a massive mesh):

azuwis version (before had at least links between each other):

Both no show stopper for me as 1.6.0 is more stable in pairing matters which is way more important for me, but just wanted to show my results.

log.zip

[Koenkk]

@dreimer1986 did any of the scans fail? It may that the output has been changed.

[dreimer1986]

Not really. Scans worked on 1.5.1 and do on 1.6.0 as it seems. If you dont see any strange things in that log, I must say all is fine in device using matters.

[juanjux]

My network map is also broken in 1.6-dev. Most of the routers are missing (I've more than 50, they work otherwise) and there are no lines between the others. It worked on 1.5.1:

[dreimer1986]

After 11aacb8 I am out. rgruebel version still looks different to 1.5.1 but maybe this one is the correct one. At least it looks correct connection wise. And azuwis one is fine again too.

[Gesfenlir]

I have the same issue here...

• (almost all) my router devices got a LQI and show up connected between each other on the map
• none on the endpoint devices are connected to any router (on the map)
• my Ikea Repeater does not show on the map at all (neither in the log)

Nothing special shows up in the log after I launch zigbee2mqtt_networkmap.update service.

Note that all works well, only the map has a weird rendering.

[lolorc]

same here, just a few devices connected while it was ok before upgrading to 1.6-dev.
devices work ok as well.
Is anybody running 1.6-dev having a proper map (like it was with 1.5*) ?

[james-fry]

I would be happy to close this issue now.
The network map has stabilised for me on 1.6 dev and I didnt see the really weird rendering from my 1st few posts in this issue.

This said I still have the issue with many/most end devices not showing a relationship to a router or the coordinator, and where there is often no bi-directional relationship between routers.

This is how its looking now.

BTW I have this problem with both of the following set ups:

• cc2530, Z-Stack_Home_1.2 (source_routing)
• cc1352P, Z-Stack_3.x.0

I'd really like to get to the bottom of this...

[lolorc]

all my devices used to be connected to my coordinator/router, it's not anymore. I also think this is a bug needing a fix

[james-fry]

@lolorc I dont disagree. Just not sure that this issue is the correct place.
Unfortunately I didnt have a good map before 1.6 dev either :( so for me this issue was more about the really weird rendering I saw with the very early 1.6 dev versions.

@Koenkk it seems really important to people to be able to reliably and accurately visualise their network (as evidenced by the 3+ map solutions out there). Do you prefer I close this issue and either @lolorc or me etc can create a new issue to get the missing info in the maps resolved? Or we continue to use this issue?

[lolorc]

it might be related to Koenkk/zigbee-herdsman#69

[timota]

Probably i need to put this here, not sure

I made some tests to check:

• if i add only 2 devices (end-devices) to coordinator - i can see linkquality for them and so z2mAssistant can build proper map.
• if i add 3-d device:
  • in logs i can see that device reports linkquality
zigbee2mqtt:info 2019-10-31T19:37:38: MQTT publish: topic 'zigbee2mqtt/0x00158d000274fe84', payload '{"battery":80,"voltage":2965,"contact":false,"linkquality":51,"last_seen":1572550658817,"device":{"friendlyName":"0x00158d000274fe84","ieeeAddr":"0x00158d000274fe84","networkAddress":17964,"type":"EndDevice","manufacturerID":4151,"manufacturerName":"LUMI","powerSource":"Battery","applicationVersion":3,"stackVersion":2,"zclVersion":1,"hardwareVersion":2,"dateCode":"20161128","softwareBuildID":"3000-0001"}}'

but in z2mAssistant reports all attributes except Link quality is unknown/255., so it cant build map

• if i add to the same setup additional router, and after it add 3 additional end-devices. z2mAssistant reports linkquolity for first 2 devices (including router itself) but not for 3-d device.
• same picture if i add another router and next 3 end-devices.

for me it looks like somewhere limit for 3 device (including router/coordinator) has been set.

i definitely didnt see this issue with Stable 1.6 and CC2531 as coordinator.

BTW: current coordinator is cc1352P2

[james-fry]

I think this lines up with what I am seeing, @timota.
Also cc1352p-2, but not sure that matters.

E.g. below, where I have 2 x end devices on the coord, and 2 x end devices on the router.
The rest, nothing.

[Koenkk]

Should be fixed in latest dev (should give same behaviour as 1.6.0). (reference #2307)

[axiom00]

Latest dev fixed the mapping issue. Just have to figure out why nothing is using the repeaters.

[james-fry]

Maybe the end devices can reach the coordinator directly OR via router.
Then we end up with maps like this (deconz):

I mentioned it in another thread, but I think the best solution is for some form of interactivity in the map to redraw the map in the context of a selected node.
Clearly this is much more development than what we have here...

[Koenkk]

@james-fry this is out of scope for zigbee2mqtt and should be done by the application rendering the graph.

Closing as this issue has been fixed.

[james-fry]

Don’t disagree at at all.