Increase web worker pool
Based on peaks in our monitoring data, we're hitting the limit of how many web requests we can handle at any given time.
For reference we have peak-time CPU utilization of 65% and memory utilization of 45%
We currently deploy 30 workers to 16 cpu nodes. I'm thinking a 10% increase to 36/node (20% more workers) would stay within acceptable CPU and memory utilization.