I have never worked with the tools you describe, but with large data sets and charts, so I will give you one idea on how we solve a similar problem.
Suppose you get a sample of data from a temperature sensor every second for 1 year, these are 31M + records ... you donβt have βspaceβ for drawing, let's say 1 pixel for each sample. What you can do is get the average value for each pixel that will be displayed. Let's say you have a width of 800px on the screen and 31M + samples, just cut the samples 800 times and get the average (representative) values ββfor each fragment / pixel. If you need more resolution, just βscaleβ to the range and again, get the values ββ.. and so on until you have 800 samples to draw.
Hope this help and sorry for my english.
source share