Build support runbook for LDR

[benbardin]

TSE would appreciate a support runbook of expected questions and how to solve them. This is distinct from the user-facing feature introduction.

Some immediate topics:

• Where could it break?
• How do we fix it when it does?
• why is replication lagging
• why did replication stop
• how do I recover in various scenarios
• How do I find which row is the problem
• How do I confirm a row was sent/committed

Jira issue: CRDB-43639

[blathers-crl[bot]]

cc @cockroachdb/disaster-recovery

[benbardin]

cc @nickelnickel @rupertharwood-crl