Tracing aggregates

[okohll]

Hi, I've got tracing working and can drill into individual traces. I'd like to see some aggregate statistics, like what function being traced has taken the most time overall to process etc. However I'm getting some confusing results, I think there's some fundamental concept I've misunderstood.

In a dashboard there is a table with a few columns (see attached screenshot)

• name
• max(duration_nano) group by name
• avg(duration_nano) group by name
• count(span_id) group by name

However, there are many spans with some columns missing (see results screenshot), e.g. either a max. with no average, an average with no count etc. and I can't see how that can be. How can you know the average without knowing the count for example. I'm sure I must be generating the query wrong.

[srikanthccv]

Let's continue in #8268