uvjim / linksys_velop

Home Assistant integration for the Linksys Velop
MIT License
50 stars 7 forks source link

"Failed Setup" after many months working #447

Closed jjhplus9 closed 7 months ago

jjhplus9 commented 10 months ago

Integration was working fine. Now it gives the following error:

Logger: custom_components.linksys_velop Source: custom_components/linksys_velop/init.py:326 Integration: Linksys Velop (documentation, issues) First occurred: 12:48:11 (8 occurrences) Last logged: 12:52:30

_async_get_mesh_data --> 'NoneType' object has no attribute 'startswith' image home-assistant_linksys_velop_2024-01-29T12-58-00.279Z.log

uvjim commented 9 months ago

Has anything else changed in the environment? Anything else shown in the logs? Could you turn debugging on and see what is reported in the logs?

jjhplus9 commented 9 months ago

I attached logs to the OP using the debug capture. There are unrelated errors in the logs but nothing else ref the velop integration

uvjim commented 9 months ago

Sorry. The email just kept opening the image rather than logs. I'll take a look and see if I can work out what's going on.

Nothing changed in your environment though? I'm not having the issue here, but I'm sure mine is just a simple setup compared to most.

jjhplus9 commented 9 months ago

Nothing relevant comes to mind. Although am building out my HA I don't know of a change recently. I did have the issue where the velop nodes would go "away" from HA every fifteen or twenty minutes but I learned to live with that. The complete lack of nodes is new

uvjim commented 9 months ago

This is concerning as responses go from the API. Does the Linksys app/web UI respond correctly?

2024-01-29 12:57:58.321 DEBUG (MainThread) [pyvelop.jnap.verbose] Request.execute --> action: http://linksys.com/jnap/nodes/networkconnections/GetNodesWirelessNetworkConnections --> payload: {} --> response: {'result': '_ErrorUnexpected', 'error': '/usr/local/lib/lua/5.1/nodes/util/client.lua:66: DevDb_GetDeviceByMacAddr mac = e0:4b:41:00:65:2f failed. Error code = 24. Error message = Server responded with fatal error: Cannot execute sqlite command: database disk image is malformed @ SqlHelper::read_as_value @ ClientConn::process_data.'}
jjhplus9 commented 9 months ago

I'll go check now

jjhplus9 commented 9 months ago

Screenshot 2024-01-30 at 00 05 59 OK, so although the ios APP has access, the web UI doesnt, which I am sure must be related to the error I am seeing. I'm going to raise a ticket with Linksys

I'll update here as and when I get a response

uvjim commented 7 months ago

@jjhplus9 - did you have any luck with Linksys?

jjhplus9 commented 7 months ago

@uvjim No, made me do a 'hard reset' which of course did absolutely nothing. seems to be a certificate issue, but they where zero help last recommendation was to buy a new master node from a newest generation, which I am unlikely to do. They claimed it would be compatible with the OG velop nodes, but I have found no evidence that that would be so.

uvjim commented 7 months ago

@jjhplus9 - interesting. I thought that typically the newer version would have to be the primary node. It would work with the older secondary nodes mind you. However, not true of the newest of the new. My understanding is that they use a slightly different tech and could not work with the older versions.

jjhplus9 commented 7 months ago

I asked them if I bought a velop 6e master (MX6201) would it work with the old nodes (I have OG gen one nodes) - support said yes - I asked them to produce some evidence that this would work and they could not. Conversely, I saw plenty (or at least some) reports that older nodes do not work with 6e master. Although I have lost monitoring from HA, the rest works fine for my families purpose. Don't really want to drop £150+ on a new node with all the setup time of a new master only for it to not work. Plus given the crippled UI I doubt I could re-add the nodes to the old master again so I would be snookered and would have to buy five new nodes and that's hela expensive. So I am minded to leave well enough alone. But super open to trying anything else anyone can suggest.

uvjim commented 7 months ago

It can seemingly be done...

https://www.linksys.com/support-article?articleNum=50933

Having said that... I'd guess that the Mesh wouldn't be very "cognitive" with only one node that understood what to do. I guess the firmware would make it be "intelligent" at best.

jjhplus9 commented 7 months ago

I do think that page existed when I looked or at least I didn't find it. It may be worth a try given some evidence that it could work. Amazon having a 30 day returns policy and all that. Are there any other reports of the issue?

uvjim commented 7 months ago

Are there any other reports of the issue?

I've not seen any. Doesn't mean it isn't happening though. None of my nodes seem to be exhibiting this and some of them are the really old little nodes. I even have the little plug-in node on the mesh.

uvjim commented 7 months ago

Whilst they mention the MBE7000 series (£400 for one unit 😱) it doesn't say that it'll only work with that series.

jjhplus9 commented 7 months ago

yeah, MBE7000 is not happening here...

uvjim commented 7 months ago

I'll close this off, but gimme a shout if you do get to use the integration again and anything weird happens.

jjhplus9 commented 7 months ago

Will do - and thank you