Few serverless database services have emerged in the last few years. These systems are different than traditional RDBMS and traditional database as a service model from the point of billing. First, they eliminate the need for provision on a database on a server, second, their billing is usage dependent. The traditional database as a service model has fixed pricing. As their billing system is not what we commonly used with, the topic does need extensive research before coding! Normally we do not care that much about saving read-write. This system of serverless computing is completely different than any existing models of cloud computing including PaaS. The Serverless Database model is built on the separation of storage and processing.
Using a fixed number of database servers usually tend to underutilization sometimes resulting expensive bill of materials than using a serverless database. Depending upon the use-case, a serverless database can be cost-effective. There is no operating system costs, no installations, no maintenance, and no patching. Its description points towards an absolute pay-as-you-go computing model. Commonly, with the below type of illustrations we show how the functions work :
Usually we do not talk around “serverless database”.
---
Unfortunately, serverless databases are not free from disadvantages. A virtual server also suffers from a kind of “inertia”. Not using the resource very often may cause to suffer from more significant response latency compared to a resource that is actively running. It is not completely natural but because the virtualization software spins down if the resources are not used. These kinds of factors make the serverless databases unsuitable for some computing workloads like high-performance computing. You can safely create a WordPress Widget plugin with serverless database backend but you can not rely a serverless database to get a low TTFB. Serverless databases are not more secure than traditional databases. A physical server with root access is most secure.
Amazon Aurora has a serverless version which based on MySQL and PostgreSQL. Azure Data Lake is a highly scalable data storage and analytics service. Firebase is a serverless database. Nutanix Era is a software suite which converts the database management towards the serverless database.
