Performance against Redpanda

[danthegoodman1]

Since both Liftbridge and Redpanda have a similar goal of simplifying the streaming experience (to different degrees), it would be very interesting to run the same set of producer, consumer, and e2e latency tests on clusters running on identical hardware to see which one can perform better in various scenarios, and any tradeoffs that might come with reduced complexity.

This would need to wait for #222 and #46 (Metrics and Consumer groups) to be a fair comparison.

[danthegoodman1]

Redpanda by vectorized: https://vectorized.io/