I thought you had it, that’s why I mentioned it to you.
@Tlse-vins Oh I see, I made the connection, it’s the same issue that @spenceur is experiencing on the homepage, in fact you must have quite loaded devices and on the Gladys Plus side there is a limit on the size of messages! (see [RESOLU] Probleme creation de graphique sur Gladys+ - #24 par pierre-gilles )
I created a GitHub issue, and I will investigate on the Gladys Plus side:
To solve this problem, 3 ways:
-
If the size of the Gladys Plus encrypted messages can be increased, I would increase it, but I think it’s already quite large, so I doubt it’s the problem
-
Otherwise, it will be necessary to reduce the size of the API call (remove unused attributes)
-
Or split the call into several API calls
Do I have a scene with 9 devices, is that a lot for G+?
Can you check the local response size of this API call (GET /api/v1/device)?
9 devices can contain a lot of features/a lot of parameters
(In the browser => Inspector => Network)
ok this evening when I get home, via ssh, how?
they are opening sensors and I’m just asking if it is open or not.
No, but I don’t have a big scene
Oh, the dummy! Yes, you can say that.
I don’t know all the commands you can use when you connect via SSH to the Pi.
I’m starting to learn them, but they’re the most basic ones.
So, the command to copy a file from a directory to paste it on my computer is completely unknown to me.
So, I use Filezilla. I was able to download the database that Pierre Gille asked me for and I sent it to him.
He sent it back to me by email.
So, I went into Filezilla to download it into the correct folder.
You’re starting to see it? Yes, yes.
I said to myself, OK, it’s done, I replaced it.
But no, not at all. I hadn’t noticed the error code in RED, which told me I didn’t have the rights.
Sorry Pierre Gilles for making you struggle like this.
I don’t know how to download the new database with Filezilla because I don’t know how to become an admin and even less via SSH because I don’t know the command.
You can, for example, send the file via FTP to your Pi in a folder where you have the rights, then do a local mv via SSH for example.
But I don’t think this will fix this bug, now that you’ve given me more information, it’s not related.
After that, it’s still a good idea to replace your corrupted DB with the clean DB that I fixed on my side.
I tried to reproduce the issue, but I couldn’t
I wonder if it’s not a concurrency problem, that is, all requests are sent at the same time and your Pi struggles to respond and trips over itself..
Because locally, the browser has a natural concurrency of a maximum of 5 requests in parallel per domain (Browsers generally have a concurrency of 5 HTTP requests per domain), whereas on the Gladys Plus side, there is no concurrency on websocket calls, and therefore this could saturate your Pi (it’s a lead)
I took some time to realize that G+ wasn’t « Google + » 
It hasn’t arrived on my side yet!
@Tlse-vins I made a special Gladys Plus deployment to test if it’s a concurrency issue!
Could you connect to Gladys Plus at this URL and check if you still have the bug on your problematic scene or not: https://622f6796b80f670ca367960c--distracted-almeida-5007a2.netlify.app/login?return_url=%2Fdashboard
(Yes, the URL is a special URL, it’s a custom deployment to test if that’s the problem. I will deploy for everyone later)