Upgrade from 1.0.10 → 1.3.0 (latest) caused significant slow down

[mshwery]

We upgraded from 1.0.10 to 1.3.0 due to some recent sidekiq-related patches and are seeing significant slowdowns across jobs configured to use this gem.

It seems there are more (slow) requests to Redis when enqueuing. Here's some of what we are seeing (and no increase in the number of operations where these Redis calls stem from).