This work introduces and aim to overcome the potential challenges while deploying automated tuning of relational database as a service for a Platform as a Service (PaaS) provider. Some of the major challenges identified in this work include (i) automated detection of performance throttling (figure out when the performance of the system is affected due to incorrect configurations of knobs) of a database and identify potential points where a database requires a tuning, (ii) scalability and accuracy of tuning service and (iii) applying the recommendations obtained from tuning services wherein applying an obtained recommendation might require a database restart. In this work, we present a generic tuning service architecture for PaaS providers. To deal with the above challenges, we introduce performance throttling engine which is responsible to detect potential points when a relational database actually needs a knob tuning, which helps in increasing the scalability and accuracy of the tuner deployments (responsible for tuning production landscapes). This work also proposes approaches that facilitate efficiently applying the recommendations without causing much disruption in Quality of Service (QoS) of the underlying database system. Lastly, the results are obtained by evaluation of the proposed methods and modules on multiple cloud native provisioners against various set of metrics.