I'm afraid of there's no document for it. But there's a FLIP[1] define the metrics that the connector (source / sink ) should expose. And I think the offical connectors will cover most of these metrics if possbile. [1] https://cwiki.apache.org/confluence/display/FLINK/FLIP-33%3A+Standardize+Connector+Metrics
Best regards, Yuxia 发件人: "casel.chen" <casel_c...@126.com> 收件人: "yuxia" <luoyu...@alumni.sjtu.edu.cn> 抄送: "Salva Alcántara" <salcantara...@gmail.com>, "User" <user@flink.apache.org> 发送时间: 星期日, 2022年 7 月 17日 下午 12:00:11 主题: Re:Re: Metrics & Monitoring in Flink SQL How to get all metrics of those connectors shipped inline with flink release? any document? At 2022-07-13 11:05:43, "yuxia" <luoyu...@alumni.sjtu.edu.cn> wrote: With Flink SQL, You can define your own source/sink metrics [1], but you can't define the metrics for the intermediate operators. [1] https://nightlies.apache.org/flink/flink-docs-master/docs/dev/table/sourcessinks/ Best regards, Yuxia 发件人: "Salva Alcántara" <salcantara...@gmail.com> 收件人: "User" <user@flink.apache.org> 发送时间: 星期三, 2022年 7 月 13日 上午 1:33:30 主题: Metrics & Monitoring in Flink SQL I have a question regarding Flink SQL, which I'm lately getting into. So far, my experience is with the DataStream API mostly. In that context, it's easy for me to generate metrics for my operators. However, I'm just wondering which level of control there is regarding monitoring & metrics when working with Flink SQL. Is it possible to define "metrics for your queries"? Whatever that means and assuming that it makes any sense :laughing:. At least I should be able to generate typical metrics for common connectors, e.g., messages read/produced & things like accumulated lag for the case of the kafka connector, to put an example. Sorry for the vagueness, but I could not find a section for metrics & monitoring within the Flink SQL docs. Any guidance would be appreciated! Salva
