Recently has a business scenario is about dynamic early warning and monitoring data, the regular task to monitor the effect not beautiful, feeling can turn Flink, but not to use this project before, so some of the scenes principle, every brother please grant instruction,

Business scenarios for 20/5 s do alarm threshold, using regular tasks will appear the problem of fault monitoring data interval (below the red line range) to monitor the result does not meet the alarm conditions without warning, and implement the data does have a monitoring result is satisfy the alarm conditions, but this piece of data across the timing task two monitoring cycle (below) of purple line interval, excuse me this kind of scenario, if we switch to the Flink real-time monitoring, can realize per second for data collection within 5 s cycle before the alarm? Such as: 1 s to 5 s, 2 s, 6 s, 3 s - 7 s, 4 s to 8 s, etc.,

Thank you all for bosses to give directions,

CodePudding user response:

CodePudding user response:

No bosses will teach small white?