Show log line count in each facet for exploratory analysis

[pranay01]

Is your feature request related to a problem?

No. This is a enhancement request

From reddit

It's more about being able to explore and visualize the data if you don't know exactly what you're looking for.

Imagine you get reports that some users are getting errors, but you don't have a really good error report -- just the helpdesk saying "hey, FYI, we've had 40 calls today about errors in ABC app, but we haven't been able to replicate it". Super vague info, but you decide to take a look.

In Datadog, you get the little sidebar that shows all the available facets, with occurrence count for each value of that facet. So you set your timeframe to "today", select error level "error", then start scrolling down the list of facets. Oh, look, 80% of errors today were in the "user profile" and "email" services. Click to apply a filter to only include those services. Now the histogram shows that almost all started at 13:25 today. Refine the time filter. Scroll through the facets again... Oh, look, 99% of the errors were in the prod-03a cluster. Click, filter for that. Oh, look, 49% of the errors are in a specific "user profile" pod and another 49% are in a specific "email" pod. Click; apply filter to include only those pod instances. Oh, OK, weird, that only leaves 1 node in the node list. You check that node and see that it was provisioned at 13:25 today, matching the start of the errors. Conclusion: bad node was provisioned and both pods got scheduled there.

Describe the solution you'd like

Enable counts in facets

Describe alternatives you've considered

Creating dashboards for above, but that doesn't enable exploratory workflows

Additional context

Add any other context or screenshots about the feature request here.

From reddit

Thank you for your feature request – we love each and every one!