Trace explorer limits number of traces

[nathansamson]

Example When I run traces for the past 12h the trace with the max duration is 27s. When I run traces for the past 6h I get a trace with max duration 70s. (and it did happen a few hours ago, not like in the last minute)

It seems only 1000 traces are being used

What I would be able to do is check every day for the past 1 day, check ALL traces that are taking longer than a given threshold so I can optimize them. Now they are lost, if the first few traces of the day are quick...

I understand this is being done out of performance reasons

Possibe solutions:

• Let me prefilter on some variables (context) (eg response time) so I can filter intresting traces myself (eg slow ones, from a specific host, ...). I am happy to enter values myself, so no need to "loop" over all the traces to give me possible options
• Show a random sample (but this will give impredictable results. Reloading the page might give a different set)
• Show only the slowest traces. Which will solve my use case, but might give a wrong idea

[itsderek23]

What I'm thinking:

• Add a "Dataset" pulldown
• Contains 2 options to start:
  • Random (more representative)
  • Slowest

[itsderek23]

@nathansamson - released!

The dataset option you select is maintained via a cookie so you don't have to toggle it every time.

[nathansamson]

I don't see this option?

[itsderek23]

I don't see this option?

Apologies - try now.