This message was deleted.
# generals
Slackbot
01/08/2024, 3:51 PMThis message was deleted.
m
Mike Sherman
01/08/2024, 7:00 PMHey @PANKAJ KUMAR at my last company, I think we emitted all the metric values but then filtered them when they were ingested (via Kafka in our case), I know that will work.
p
PANKAJ KUMAR
01/09/2024, 5:50 AMokay, But the issue i am facing is the high throughput on kafka. and during loadtest it goes very high.
PANKAJ KUMAR
01/09/2024, 5:50 AM@Abhishek Agarwal, is there any issue in having the same metricDimensionMap config for kafka emiiter as well?
m
Mike Sherman
01/09/2024, 1:46 PMHey @PANKAJ KUMAR I wish I had saved the config for the druid cluster I was using but I didn't. We only had the consumer side. You'll have to do some experimentation. On the consumer side in the Supervisor you have for the topic, you should be able to configure more processors to deal with the throughput isue.