You want to watch CPU usage and active connections. Make sure your active connections stays below your max connection count.
On Wednesday, December 28, 2016 at 8:34:32 AM UTC-10, Alexander Gray II wrote:
We have 3 small docker containers running openresty. They sit behind an AWS ELB.
The size of each nginx container is 256 cpu units and 512mb of memory.
We get alerts if containers are using high cpu or high memory, and we have the ngnix status page that newrelic can consume like so:
![]()
So far, everything is working just fine. No alerts, everything seems to be working fine.
My questions are:
How do we know we should be scale up the number of nginx containers?
What metric should we be looking at to tell us when it is time to scale?
Thanks,
Alex