Upstream timed out errors


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


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