Thanks for the info, I realize now that itâs been quite a while since I last used SQLite 
Command in progress, this should take a few minutes / hours depending on the speed of my SD card.
Before: 2.8G
After: 1.5G
What a difference!!!
For information, I merged the PR that fixes the NaN issue causing the aggregation to crash and I launched a Docker build tagged dev 
Build: https://github.com/GladysAssistant/Gladys/actions/runs/1649220231
The bug fix for the NaN crash is available in Gladys Assistant v4.7.2!
Super thanks, I found the history of my temperatures 
.
I found a small bug in the display of the graphs that I am reporting here:
So far, so good. Then, we deactivate one or two sensors:
Still fine. We change the view, it works:
We reactivate the other sensor curves and boom!
The scale of these has not been adjustedâŠ
Nice, huh?
Edit: with the images in order, itâs even better!
Hello @GBoulvin!
I canât figure out whatâs wrong?
Do your « RDTransmitter » and « Room Temperature » temperature curves have data before March 12?
Indeed a small bug! Can you create a GitHub issue for us with these screenshots and a link to this post?
Hello everyone
I think I have a problem with a graph with a temperature sensor.
I switched my Xiaomi WSDCGQ01LM temperature sensor
from the Mqtt integration (managed with Node RED) to the Zigbee2Mqtt integration and since then I can no longer display the history for more than an hour.
Is this the same problem as before?
Do you have the latest version of Gladys?
Yes yes she passed by last night, but I have switched the sensor for 3 days already
@pierre-gilles had directly checked my database to understand the issue
@Psoy Indeed, itâs strange because we had fixed this issue with @lmilcent, there must be another problem. Yet, it surprises me, we had really solidified this part on the core & Zigbee2mqtt side!
Would it bother you to send me your DB @Psoy? It would allow me to directly see whatâs wrong and fix the issue.
I know itâs not very user-friendly to ask for this, but well, with this kind of self-hosted product, weâll never have too much choice, the data is yours and I donât have access to it
Hello Pierre-Gilles
can you guide me?
Yes, of course, but it does require a bit of SSH
On your Gladys instance, you can do:
docker stop gladys
To stop your Gladys instance.
Then, you need to retrieve the gladys-production.db file. This file is located in the folder where you installed Gladys. If you havenât changed anything, this folder is located by default at:
/var/lib/gladysassistant
To retrieve the file, you can use tools like FileZilla or Cyberduck.
Then, you can restart Gladys:
docker restart gladys
I donât know if this is easy for you or if itâs gibberish to you.
I specify that I am asking you these commands in a « developer » context. If you consider this too complicated for you or if you prefer not to touch anything, you can say no. I offer this to help and find a solution faster than by describing the symptoms of the problem
( Iâm saying all this because @Psoy pointed out to me in a private message yesterday that he found Gladys v4 not always as « user-friendly » as one might think, that he had issues he had to resolve with not-so-simple manipulations, and so it bothers me to ask him for this kind of manipulation right after he talks to me about it⊠Itâs very frustrating for me! )
No problem with that, with practice I understand some manipulations.
Itâs done, but how do I send it to you?
Either by private message (Dropbox link, Google Drive, or other), or by email (I sent you the details in a private message)