Add large traces to k6.io tests

[joe-elliott]

Is your feature request related to a problem? Please describe.
We know that tempo has some performance issues returning extremely large traces. Let's add some benchmark testing of this to the k6.io scripts added here: #392 to help us track this issue and improve on it.

[dgzlopes]

We should create traces with tons of spans? Or with huge spans (size-wise)? Or both?

[joe-elliott]

We have tracked this down to being almost entirely an issue with marshalling/unmarshalling. Removing the jaeger-query dependency and allowing Grafana to query Tempo directly will be the best thing we can do. Also, for clarity, "large traces" are in the couple hundred thousand span range.

I think we can just close this tbh.

[dgzlopes]

Oh, if that's the case, I agree! We can close this one.

Thanks for the clarification anyway :P