Temporarily Enable binary logging on production Aurora database cluster

[sureshc]

Reverts #43743

Temporarily re-enable MySQL binary logging on the production Aurora database cluster to support modifying the datatype of the level_sources.id column from int to bigint using gh-ost (#41927 )

Test

$ bundle exec rake stack:data:validate RAILS_ENV=production

Pending update for stack `DATA-production`:
Modify AuroraClusterDBParameters [AWS::RDS::DBClusterParameterGroup] Properties (Parameters)
Modify AuroraReportingClusterDBParameters [AWS::RDS::DBClusterParameterGroup] Properties (Parameters)

Deployment Steps

1. Apply the update the cluster ParameterGroup bundle exec rake stack:data:start RAILS_ENV=production
   Completed 12/16/2021 16:30 PST:

Pending update for stack `DATA-production`:
Modify AuroraClusterDBParameters [AWS::RDS::DBClusterParameterGroup] Properties (Parameters)
Modify AuroraReportingClusterDBParameters [AWS::RDS::DBClusterParameterGroup] Properties (Parameters)
Proceed? [y/n]
y
 Stack update requested, waiting for provisioning to complete...
2021-12-17 00:35:10 UTC- DATA-production [UPDATE_IN_PROGRESS]: User Initiated
..2021-12-17 00:35:20 UTC- AuroraReportingClusterDBParameters [UPDATE_IN_PROGRESS]
2021-12-17 00:35:21 UTC- AuroraClusterDBParameters [UPDATE_IN_PROGRESS]
.........2021-12-17 00:36:22 UTC- AuroraReportingClusterDBParameters [UPDATE_COMPLETE]
2021-12-17 00:36:25 UTC- AuroraClusterDBParameters [UPDATE_COMPLETE]
.2021-12-17 00:36:28 UTC- DATA-production [UPDATE_COMPLETE_CLEANUP_IN_PROGRESS]

Stack update complete.
Outputs:

2. Restart the writer database instance (90 seconds of downtime) for the change to take effect
3. Verify that ProxySQL correctly identifies which database instance is the writer and which database instances are the readers