Рет қаралды 3,864
Squeezing Django performance for 14.9 million users on WhatsApp
by Rudi Giesler
pretalx.com/dj...
At the start of the pandemic, there was a large need for accurate information to combat misinformation. This is how we used Django as part of our South African WhatsApp service "ContactNDoH"
At the start of the pandemic, there was a large need for accurate information to combat misinformation. For this, we developed a South African WhatsApp service, "ContactNDoH", which disseminated accurate information, and later managed registration and bookings for vaccinations.
With over 14.9 million users, and over 850 million messages, we needed to squeeze as much performance from Django as we could.
This talk will go over:
how we approached profiling and the tools that we used,
what scaling issues Django limitations we ran into and how we overcame them,
and general tips, tricks, and tools that helped us.
It will focus mostly on backend performance, as the role Django played here was as a REST API.