Open ildar170975 opened 2 months ago
maybe the best workaround for now is using the "sampler" integration on sensors that update very slowly. I do that with many sensors not just for graphs (especially for reinmann sum sensors, before the update that basically does the same thing as sampler).
Thanks about info, I will read about “sampler”. But here no workaround needed - this a harmless glitch: since I know that this entity is excluded from Recorder - I will not want to see it’s graph in mini-graph-card. The main point is that old record is not supposed to be used since it is beyond a time interval.
Some sensors are EXCLUDED from Recorder:
Consider this card for the 1st sensor:
Note that:
And same card for the 2nd entity:
Note that:
About 2 weeks ago (23.08.24) while checking this issue the 1st sensor was INCLUDED in Recorder for a few hours. This is the last record in DB for this sensor: where
last_updated_ts
is1724421565.94111
(Friday, 23 August 2024 16:59:25.941 GMT+03:00).And
attributes_id
of this last record is with that "559" value.Means - this old "559" value recorded 2 weeks ago is shown in a graph - although
hours_to_show: 24
(a default value). There will be same glitch ifhours_to_show
is defined explicitly:Imho we should not use a value from a DB if this value is beyond a selected time interval. Probably, there may be some negative side effects after this will be fixed, not sure...
P.S. Browser's cache was cleared. v0.12.2-dev.2 2024.9, Chrome