Promethus - custom metrics at job level

[Averell]

Hi everyone,

I am trying to publish some counters and meters from my Flink job, to be
scraped by a Prometheus server. It seems to me that all the metrics that I
am publishing are done at the task level, so that my Prometheus server needs
to be configured to scrape from many targets (the number equivalent to my
max parallelism). And after that, I need to do aggregation at the Prometheus
server to get the numbers for my whole job.

My question is: is it possible to have metrics at the job level? And can I
have one single Prometheus target to scrape the data from?

I found this
http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/Service-discovery-for-Prometheus-on-YARN-td21988.html,
which looks like a "no" answer to my question. However, I still hope for
some easy to use solution.

Thanks and best regards,
Averell



--
Sent from: http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/

[Reza Samee]

Hello Averell

Based on my experience, using out-of-the-box reporters & collectors need a little more effort!
Of course I hadn't experienced all of them, but after reviewing some of them I tried my way:

Writing custom reporters to push metrics in ElasticSearch (the available component
in our project & very flexible). The custom reporters are able to group metrics with
some configurable/dynamic parameters (in my case based on a defined metric-name
and jobs): https://github.com/reza-sameei/elasticreporter

I think that maybe writing a simple custom reporter will help you :)

On Mon, Sep 3, 2018 at 10:52 AM Averell <[hidden email]> wrote:

Hi everyone,

I am trying to publish some counters and meters from my Flink job, to be
scraped by a Prometheus server. It seems to me that all the metrics that I
am publishing are done at the task level, so that my Prometheus server needs
to be configured to scrape from many targets (the number equivalent to my
max parallelism). And after that, I need to do aggregation at the Prometheus
server to get the numbers for my whole job.

My question is: is it possible to have metrics at the job level? And can I
have one single Prometheus target to scrape the data from?

I found this
http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/Service-discovery-for-Prometheus-on-YARN-td21988.html,
which looks like a "no" answer to my question. However, I still hope for
some easy to use solution.

Thanks and best regards,
Averell



--
Sent from: http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/

--

[Averell]

Thank you Reza. I will try your repo first :)

Regards,
Averell



--
Sent from: http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/