Build upon the optimizations in the previous story, , test with increased DB connections per tenant to the max that is allowed by the database instance type to see if DI performance increases.  In Cornell clusters we are able to give it the maximum of db connections per db instance type. Therefore:

1. Perform the baseline 7.5K import with the profile "EBSCO ebooks new and updated" and the file  "20230307monosLoad.marc" (attached) with the default number of connections of mod-srm and mod-srs'  DB connections (15 each).  Not that this import job may fail due to existing records. 

2. Create a new task def with the number of connections for mod-srs and mod-srm to be 500 each. Rerun the import in step 1. Record duration and DB's CPU and memory utilization. 

3. Create a new task def with the number of connections for mod-srs and mod-srm to be more than 500 each, up to the max that the RDS instance type allows.  Rerun the import in step 1. Record duration and DB's CPU and memory utilization. 

Note that adding more connections means more DB memory usage. Observe the DB memory utilization graph and CPU grah to see if things still remain stable. 

Please include the findings in this ticket's comment section instead of creating a Confluence page for it.