Upstream timed out errors


#1

in our nginx setup we are seeing a lot of
upstream timed out errors

The app is doing only about half a million pageviews a day

the app has 8 backends in 3 servers ie 24 backends served from one nginx

the cpu utilization of each backend is at 70% most of the time

Any ideas on how to fix the timeout errors, should we add more backends
thanks


#2

You need to determine where the bottleneck is. Often, it is the
database layer and lack of caching.