markjschreiber
commented
2 months ago When incorporating the timeline.py logic we need to address:
- The timeline tool today double counts task queue time, which leads users to believe their runs take much longer than they actually do.
- add task names on left side, replace, max cpus with vCPU utilization (percentage), max memory with max memory utilization (percentage). Think about removing max memory/vCPU
We have companion visuals that users can use to visualize a run timeline, but this is provided today as a separate standalone tool (timeline.py). We should include (or migrate) the the timeline tool for in run analyzer.