Database as a Service (DBaaS) is a category of cloud-based computing managed services that provide access to a database without the need to establish physical hardware, install software, or configure the database. Instead, the service provider handles most database administration and maintenance tasks. Users can quickly start up a database and load and analyze data, typically with little or no IT intervention.
DBaaS is important to the larger corporate initiative of digital transformation in which companies fundamentally change how they operate and deliver value to customers. Specifically, companies striving for greater data democratization might choose DBaaS as a way to deliver on an any analytics, any time strategy. Also, organizations concerned with ESG (Environmental, Social, and Governance) advances can leverage the DBaaS ability to spin up and down servers at will as a way to save resources.
Advantages of DBaaS
DBaaS offers some advantages over traditional methods to deploy database systems, including the following:
Disadvantages of DBaaS
There are also potential disadvantages with DBaaS compared to on-premies databases.
There are a number of types of DBaaS providers, making for an extraordinary and diverse range of options in the DBaaS space.
Public cloud providers
Many cloud vendors like Google, Azure and Alibaba have their own DBaaS offerings. Users can leverage the same account they may use for computing and storage to instantiate databases.
Proprietary cloud vendors
Vendors like Snowflake, Firebolt, and others offer a proprietary cloud, where users pay the vendor for both the use of the database and use of the cloud services. The database and servers are provided by the vendor, although the servers and storage are outsourced to a public cloud provided through the vendor contract.
Partner DBaaS companies
Many amazing options for DBaaS also appear in the partner marketplace of the public clouds. Users contract with the DBaaS partner and the cloud provider separately. These vendors, including Vertica, often offer options.
Beyond the types of DBaaS, there are other major differences between DBaaS providers, including:
DBaaS deployment options
Does your DBaaS also offer non-SaaS deployment? Some DBaaS vendors require you to lock into a specific storage place in one particular cloud. This locks the customer into one cloud, not allowing the freedom to move to a different cloud easily or take advantage of lower-cost cloud computing when available. Some vendors offer no solution for on-premises analytics or deploying in Kubernetes. Weed out vendors who don’t support all of your deployment needs.
Does your solution offer a license that allows you to easily move between multiple clouds or on-premises, or are separate licenses required for each deployment? What are the costs to maintain DEV, TEST, BACKUP and PRODUCTION? Take a look at total costs to understand which vendors will meet your needs.
Data lake capability
Do you often have locally-stored Parquet, Orc, AVRO, JSON or TEXT files that you need to incorporate into your analytics? When choosing your DBaaS vendor, explore how it can bring together a data lake’s scale and economics with the predictability and reproducibility as a data warehouse. In addition, consider how well your solution understands external table workloads and how much data movement is required.
Ability to optimize
Does your DBaaS operate in a limited compute package? All analytics is not the same, nor should it be considered the same. Make sure that the database that you select has options to properly manage all types of workloads and service level expectations. Solutions that do node-based optimization (simply adding generic nodes when your workload calls for it) may cause you to miss out on methods to keep your cloud costs lower while improving query performance at the same time. The capability to use specialized nodes, and the ability to tune slow queries is paramount.
Depth of analytics
Can you leverage your DBaaS for more than just descriptive analytics? Today’s data-centric companies have analytical needs that reach beyond standard SQL databases. For example, some workloads call for advanced analytics like geospatial, or time-series function. Predictive analytics is becoming increasingly imperative to data science teams, so consider how machine learning is supported. Consider how your solution can support a wide range of analytical use cases and a wider team of professionals as your cloud database gains success in your organization.
Vertica Accelerator is Vertica-as-a-Service (DBaaS) that delivers a unified, high-performance advanced analytics and machine learning platform with automated cloud setup and help with onboarding. It runs in your own AWS cloud account, with automation from the Vertica management plane. Vertica Accelerator is one of the deployment methods offered by the Vertica analytical database. Vertica also offers on-premises deployment, Kubernetes deployment, and more.
Vertica provides the flexibility of private and public cloud deployment – not just a proprietary cloud, but any cloud. Our database seamlessly connects on-premises environments to public clouds for a hybrid data cloud experience. By implementing hybrid cloud, you can increase flexibility, performance, and scalability. It offers you a way to maintain complete control of your data while leveraging modern cloud technologies.
Vertica Accelerator helps you create a strategy for more predictable pricing with our flexible deploy-anywhere license. It’s the best way to place workloads in the right place for price/performance and avoid single-point-of-failure scenarios.
With Vertica Accelerator, you can finally get machine learning into production. Vertica supports cluster-optimized ML algorithms, R, and Python. Data scientists and analysts can build their models using their preferred tools and languages, then leverage Vertica to power them on bigger data sets. In-database machine learning addresses every step in the ML process.