Gaps in samza data

[ngpadp12]

Some data points are not displaying a continuous graph.

Raising this issue to understand why it might be happening.
Please go through the following screenshots:

Round robin chooser:

YARN

[nmaquet]

@ngpadp12 I can try and help you troubleshoot this but I would need a bit more information. What do the exporter logs say? Is the exporter crashing and restarting? Is Prometheus properly scraping the metrics regularly? Could you maybe share the graph of scrape_duration_seconds in Prometheus to see what's happening?

[ngpadp12]

Hello @nmaquet , thanks for responding!

The only logs I see are of unrecognized metrics. But I am not sure if the unrecognized metrics can ever show up for prometheus to scrape(and hence the gaps being the times when the same metrics goes unrecognized)
Screenshot below:

We are running this as an AWS ECS service, and no, the exporter is not crashing and restarting.
Image below shows the time since when it is up and running:

The scrape duration graph of the past week is below:

Is Prometheus properly scraping the metrics regularly?
I believe yes, it is. But is there any parameter or anything in particular that you want me to check to confirm this?