SQL stream processing, analytics, and management. We decouple storage and compute to offer instant failover, dynamic scaling, speedy bootstrapping, and efficient joins.
Although we've provided a rich set of metrics in the Grafana dashboard, as a system that handles variable workloads from users, it's believed that certain metrics are more intuitive and easy to understand when visualizing along with the stream fragment graphs in the Meta Dashboard. #14446 is an attempt for visualizing the back-pressure rate to identify the bottleneck of a streaming job.
There could be more metrics to visualize in the stream fragment graphs:
Busy/Idle time of each executor.
Data skew rate between different actors within a single fragment.
Although we've provided a rich set of metrics in the Grafana dashboard, as a system that handles variable workloads from users, it's believed that certain metrics are more intuitive and easy to understand when visualizing along with the stream fragment graphs in the Meta Dashboard. #14446 is an attempt for visualizing the back-pressure rate to identify the bottleneck of a streaming job.
There could be more metrics to visualize in the stream fragment graphs: