Sebastián Estévez

Tested this locally and it looks good. @ishaan-jaff from what I can tell this is affecting completions that use `tools` for every model including `openai` even without `litellm.add_function_to_prompt=True` Can we...

> E Failed: Error occurred: dictionary changed size during iteration

I'll try to repro in a unit test

Ran out of time yesterday, I'll look into adding another test in a new PR when I get a chance.

> I don't think tuning splits would make a big difference It does. This is how I've had to do things many times when dsbulk unload fails. The reason is...

Good idea. Along the same vein, the java driver does expose some stats already that might be useful. https://docs.datastax.com/en/developer/java-driver/4.2/manual/core/metrics/

If you export the cassandra-stress.yaml you essentially have all the information about the schema and the field distributions. Does that satisfy this requirement? @jeromatron ?

Yeah let's re-focus this issue to address import export. I'll have to add a js yaml parser but it could work nicely.

I'm very much +1 on this. The regex approach is failure prone.

Does not break on collections now, but still need to be added to Viz.