Verify Redis implementation

[anjackson]

Having run a couple of heavy sets of seeds through the Redis Frontier, and after PAUSING the crawl, some queues are in qs:active and not in qs:scheduled - specifically:

0 | 1632996435109 | com,twitter,www,+1
1 | 1632996380612 | com,twitter,

and the timestamps are significantly older than those in qs:scheduled indicating that they've been forgotten. They are in qs:known so scanning that would bring it back.

• Track down if this is a problem!
• Determine how to properly evaluate the reliabilty of a whole replacement frontier!