Extend old events pruner to prevent data loss
In #24538 (closed), we've discussed a partitioning strategy that we need to undertake in order to allow the
events table to scale. This work has yet to be tackled and we're scaling the Database team in order to support this and other critical initiatives.
In the meanwhile, we should extend the pruner again to prevent data loss while we catch up on the work to allow this table to continue to scale.
prune_old_events_worker.rbfrom 2 years to 3 years.