I am running several lightweight docker containers in a Ubuntu server (18.04) virtual machine image that is running on a Hyper-V host (running Windows Server 2019). The physical host has 2 Intel Xeon processors and approximately half a terrabyte of RAM. The Ubuntu server image has roughly half that (250Gb of RAM and a single 12 core processor).
The application I built doesn’t use much in the way of resources, I can run it on my desktop and it typically only consumes about 8Gb of RAM and I dont notice much of a slowdown processing-wise.
When I run it on our production server though, the CPU for a few containers is always pegged at or above 90%, and when I look at the logs I see events triggering that would result in the containers messaging each other over the network, but those messages take minutes to arrive.
Has anyone experienced this before? Would you have any ideas on things I could try to debug? I’m attaching a screenshot of the stats as well as the output from running docker info.
Any help is greatly appreciated!