Redshift, BigQuery, and Apache Spark, oh my! - Looker’s indispensable guide to help you choose the right database
Jul 17, 2017
Redshift, BigQuery, and Apache Spark, oh my!
For a generation of Database Administrators (DBAs), you could spend your whole career becoming an expert in one warehousing technology and one business intelligence tool.
Netezza, Oracle, Teradata. Cognos,Hyperion, Business Objects. Each was its own island with its own ecosystem.
Picking a toolset and becoming deeply knowledgeable in it guaranteed job security. Your enterprise had spent years planning its data warehousing solution, and millions of dollars installing it. They weren’t going to switch any time soon.
But the world has changed.
Today, even early-stage startups need a data strategy. Data warehouses have gotten so cheap, anyone can afford one. And while a DBA may be in charge of setting things up, it could now just as easily be an analyst, or a developer, or even a founder.
At the same time, the number of options has ballooned. Production replica or dedicated warehouse? Cloud or on-premise? Hosted or on-demand? Hadoop or columnar? All-in-one or modular? Real-time or batch processed?
Building a data strategy means weighing all of the options and charting a path forward.
So rather than experts who are deeply informed about a single technology, the need today is for people who are broadly knowledgeable about the relative merits and disadvantages of different solutions.
But with new options coming out every few months, how do you get up to speed and stay informed?
Well, it turns out that the analysts and engineers at Looker face the same challenge. Looker’s data platform integrates deeply with 34 dialects (and counting) of SQL, and because we have customers who use each and every one, we’ve become experts on them.
We see first hand what they do well, where they struggle, and how they perform in the real world. We understand their quirks and their jaw-dropping features, because we have to make sure Looker can deal with the former and leverage the latter.
Today, for the first time, we’ve collected that knowledge in a single place, and we hope it will become an indispensable resource for everyone who “does data stuff.” In the months we’ve spent putting this together, we haven’t found anything like it anywhere on the internet.
We’ll be adding to it over time--updating it with additional technologies, and maybe expanding it to cover other database types (graph, key/value stores, document stores) that are used for specialized analytics.
But we wanted to get what we’ve already collected out there for you to use. So check out the Pocket Guide to Databases. And if there are things that need clarification or that you’d love to see in future versions (or heavens forbid, things we got wrong), don’t hesitate to let us know at firstname.lastname@example.org.