update replication playbook to include Solr data

[rlskoeser]

PUL devs did something similar to what we want to do, but it's implemented as a rake task - see https://github.com/pulibrary/dpul/blob/3fc9512a5d50b25c971a3cf0ca6b17a91480457b/lib/tasks/replicate.rake#L17

they are using a backup to generate an export of production solr data, and then restoring that backup to the qa solr instance

relevant solr admin documentation: https://solr.apache.org/guide/6_6/making-and-restoring-backups.html

[rlskoeser]

Implemented in Princeton-CDH/cdh-ansible#193 and awaiting code review from PUL devops team

I tested the solr replication by clearing the index in staging, confirming in the search on the staging site that there was no content, then ran the replication and confirmed that the search results showed up again as expected.