tlp-stress v3 released

[Jon Haddad]

Just a quick note that tlp-stress 3.0 has been released.

What's changed, you ask?

* A new workload to test the performance of using ALLOW FILTERING #106

There was a user ML question about the performance of using allow filtering.  It sounded like a great opportunity for a new workload, so there it is.

* We now support compression for the --csv file flag.  If the file ends in .gz, it'll automatically be compressed.  

* Random and Normal field functions support human readable values

If you were curious to see how well (or poorly) Cassandra works with large payloads, 1-3MB for example, you can now do this: --field.table.field='random(1m, 3m)'

* We're now publishing docker containers

Grab it at https://hub.docker.com/r/thelastpickle/tlp-stress if that's your preference.  John Sanda has been working on a Kubernetes operator that will use the docker container, so if you want to run 30 stress instances against a 100 node cluster, this might be of interest.

* A couple other small things: https://github.com/thelastpickle/tlp-stress/milestone/3?closed=1

Happy benchmarking,

Jon