Why build a “Serverless” Data Warehouse?

Posted by Rodrigo Mestres on Dec 11, 2018 10:30:21 AM

In Data Warehouse, Serverless

Is “serverless” just a temporary trend or companies are really adopting it?

What does “serverless” actually means?

“The name “serverless computing” is used because the business or person that owns the system does not have to purchase, rent or provision servers or virtual machines for the back-end code to run on.”
Source: Wikipedia

In the context of a data warehouse, serverless means being able to simply store and query your data, without having to purchase, rent, provision or manage any infrastructure or software licensing.

So, should you use a Serverless Data Warehouse?

Is your company Google or Twitter?

Unless you are a very large corporation like Twitter or Google with a lot of resources or you have a very specific necessity/use case, nowadays it doesn’t make so much sense to reinvent the wheel inside each company.

So the main reason for a “serverless” data warehouse is actually the answer to this question: “do you really want to spend your company resources and your team’s time building, managing or tuning a data warehouse?

You don’t really have to, there are plenty of data warehouse as a service platforms like SlicingDice and Google BigQuery where you simply need to send your data and later make your queries, very fast and in a cost-effective way, so why build internally or even manage servers like Amazon Redshift?

Is your team ready to build or manage it?

As the recent study from 451 Research on machine data analytics shows:

“When asked what challenges they have faced with achieving hoped-for results with analytics projects, respondents cited infrastructure requirements (cited by 36 percent), scaling challenges (cited by 33 percent), staffing requirements (cited by 33 percent), slow analytics (cited by 32 percent), expense (cited by 31 percent) and technical challenges (cited by 30 percent).”
Source: Information Management

On top that, the IT talent shortage is a global reality and will continue to be, so specialized professionals that can build or even manage a data warehouse will become increasingly difficult to find, hire and retain.

Is anybody really adopting it?

As you can see in the Google Trends graph below, the “serverless” concept is growing really fast and in parallel with serverless data warehouse solutions like Google BigQuery.

Also, as Gartner states on their latest Magic Quadrant for Data Management Solutions for Analytics (February 2017):

“More organizations are considering cloud IaaS or PaaS as the means of deploying their analytical environments. […] Cloud solutions are also challenging the traditional positioning of appliances, which is causing further disruption for traditional data warehouse vendors.”

What the future looks like?

From a technical standpoint, we believe that platforms like SlicingDice and BigQuery will continue to become more flexible in terms of data types and analysis it can support, while still getting faster, easier and cheaper.

As stated before, considering that the IT talent shortage is a global reality and will continue to be, from a market and business standpoint we believe companies from all sizes will continue to “outsource” their data warehouse operation to specialized providers like SlicingDice. That way their teams can focus entirely on extracting value and insights out of the data, instead of spending time building or managing a data warehouse technology.

Not sure if a serverless platform can support your needs?

Click here and schedule a 15-minute talk with our developers, so we can evaluate your case together or let’s chat online now.

Why SlicingDice?

Developers want to store and query their data, but they hate having to manage complex data infrastructure, so we do it for them.
All they have to do is simply store and query billions of real-time and historical data, without managing any infrastructure.

Store your data and make fast queries using our simple API or any of our clients available in Python, Ruby, JavaScript, Go, Java.NET/C#, PHP, and Arduino.

High availability?
You don’t need to worry. We always replicate and backup your data on three different datacenters, located in France, Canada and Germany.

Query speed?
If any query takes longer than 10 seconds to complete, no matter its complexity or data size, you receive a 10¢ discount on your account for each one. We believe ensuring fast queries is our responsibility, not yours.