Please update the history section as it is extremely slow on the Pi…
What do you expect of 5 years old hardware which was already under powered when released? We are hoping that it will be better with the in-progress modification of the db backend.
Figured it out; it depends on sd card read/write speed ;).
Yes, on the Pi an ultra high speed card is suggested. I bought a Samsung 64GB EVO Plus Class 10 Micro SDXC - 80mb/s at the suggestion of @brusc in his video and glad that I did.
Nice, thanks!
I’ll need to try this too.
About to purchase it now - did you notice significant improvements?
Well, I really don’t have anything to compare it to as that’s what I started out with. But I just bought another one for my drone if that means anything. 
Haha, bought the 128 gig version! My PI came with a 30mb/s microsd. This should be an insane difference for me.
Oh yeah! You should definitely see a difference. 
I’m using a mySQL backend on a remote server but this only seems to be used for the initial SQL query.
I previously has a Class 4 microsd in my Pi2 and changed it out for a UHS-1 Samsung EVO.
These are the timing test results
Timing with sdcard class 4
First time call after starting up Hass
Logbook 25 seconds
History 19 seconds
Sept 12 (second test for a specific date)
Logbook 15 seconds
History 10 seconds
Timing with sdcard UHS 1 - Class 10
First time call after starting up Hass
Logbook 25 seconds
History 18 seconds
Sept 12 (second test for a specific date)
Logbook 14 seconds
History 10 seconds
Is this to be expected or it there something else I can tweak?
What kind of speed improvements would the Pi3 make?
Back in June when I first answered this thread, things were coming up fast but now - even with a super fast SD card - it seems to take forever when I open a card or look at the history/logbooks so I’m afraid I’m in the same boat.
I have the exact same issue. Judging by the CPU and memory load the problem is not in the processing power. I do have a relatively fast as card (30mb/s) so that should be capable of outputting those values in a blitz.
In my case this turned out to be the db getting large. I moved to MySQL and now it works much much faster.
Check out this thread: