We use AWS RDSProxy as a middle tier between our API and the SQL Server database.
Noticing very high waits on the following statements called from RDSProxy even though they run very quickly manually:
SELECT IIF((DATABASEPROPERTYEX('rdsadmin','Updateability')='READ_WRITE'
AND DATABASEPROPERTYEX('rdsadmin','status')='ONLINE' AND (DATABASEPROPERTYEX('rdsadmin','Collation') IS NOT NULL
OR DATABASEPROPERTYEX('rdsadmin','IsAutoClose')=1)), 1, 0), @@options & 2
, (SELECT cast(value_in_use as int) FROM sys.configurations WHERE name = 'network packet size (B)')
SELECT @@options & 2, IIF((DATABASEPROPERTYEX('rdsadmin','Updateability')='READ_WRITE'
AND DATABASEPROPERTYEX('rdsadmin','status')='ONLINE' AND (DATABASEPROPERTYEX('rdsadmin','Collation') IS NOT NULL
OR DATABASEPROPERTYEX('rdsadmin','IsAutoClose')=1)), 1, 0)
, (SELECT cast(value_in_use as int) FROM sys.configurations WHERE name = 'network packet size (B)')
Does anyone know why they calls are made and if there is a way to stop them? We don't do anything to explicitly request info from rdsadmin. The db is Multi-AZ.
From my experience, this is likely due to the RDS proxy requesting more database connections than the database can service at the moment, either due to CPU limitations, or a max connection limit.