This is arguably unnecessary as you can infer the block time if you know what other metrics to aggregrate, however given this is a super core metric for the datanode I am erring on the side of caution and explicitly adding a metric for this which captures the whole of the block processing logic and gives us a simple metric to monitor.
This is arguably unnecessary as you can infer the block time if you know what other metrics to aggregrate, however given this is a super core metric for the datanode I am erring on the side of caution and explicitly adding a metric for this which captures the whole of the block processing logic and gives us a simple metric to monitor.