As a start to profiling our code to find inefficiencies / bottlenecks, we should instrument all the tokio tasks in the agents codebase with tokio-metrics. There is an off-the-shelf grafana dashboard for these metrics, here. Most interestingly, we'll be able to sort by the tasks taking up the most resources:
As a start to profiling our code to find inefficiencies / bottlenecks, we should instrument all the tokio tasks in the agents codebase with tokio-metrics. There is an off-the-shelf grafana dashboard for these metrics, here. Most interestingly, we'll be able to sort by the tasks taking up the most resources:
Hopefully this will help fix https://github.com/hyperlane-xyz/hyperlane-monorepo/issues/3454