OBSERVABILITY TASK Optimization dubbo meter pull delay

[songxiaosheng]

• [x] I have searched the issues of this repository and believe that this is not a duplicate.

Describe the proposal

EN : After the dubbo indicator is generated, it needs to be converted into Prometheus format through the regular task. The regular task periodically converts the data, causing a certain delay in the indicator data. Try to refer to the indicator conversion method of micrometer to optimize the data indicators to take effect in real time。

中文:可观测性任务 dubbo指标生成之后要经过定时任务转换为普罗米修斯的格式，定时任务周期性的转换数据，让指标数据产生了一定的延迟，尝试参考micrometer的指标转换方式，将数据指标优化到实时生效，