Skip to content
Announcing BigAnimal: Fully managed PostgreSQL in the Cloud
Register for Postgres Vision • June 14-15 • Online

PostgresBuild 2021 Session Highlights: Marc Linster on PostgreSQL Cloud Strategies

EDB Team1/6/2022
EventsKubernetes

It’s been almost a month since EDB’s Postgres Build 2021 virtual event, and we’re still consuming all the value that our speakers shared. With the Database Management System (DBMS) landscape evolving ever more rapidly, there was a great deal of ground to cover, as we discussed in our highlights blog from early December. 

Kicking off the new year, we wanted to dive deeper into some of the key presentations from both the EDB team and attendees, highlighting crucial takeaways, and giving you more to consider as you look to the future of your business’ digital transformation strategy. To dive right to the heart of it, here are highlights from  the event’s centerpiece: EDB CTO Marc Linster’s keynote address from the second day of the conference. Marc addressed key questions about why and how businesses are taking advantage of PostgreSQL in the cloud.

 

The power and purpose of PostgreSQL in the cloud

Entitled Quo Vadis Postgres in the Cloud? DBaaS or Kubernetes, Marc Linster’s session explored just where organizations can put PostgreSQL in their cloud architectures. But, as Marc ultimately discussed, there’s much more to it than just that. In order to understand where your cloud PostgreSQL will live, it’s essential to touch on the reasons for the movement, and the journey they entail. It’s fitting then, for those Latin lovers in attendance, that Marc chose the famous question: “Where are you marching?”, because a move to the cloud entails major decisions, and a poorly managed move to the cloud can be arduous and aimless. 

Why are businesses moving to the cloud?

The first piece of information to understand is just why businesses are moving to the cloud in the first place. As Marc explained, the main reasons are speed and agility, especially where provisioning new databases is concerned. While in the past, provisioning a new database could take weeks or months, our customers in the cloud now need to accomplish this essential task in hours, or even minutes.

Alongside agility, other reasons frequently cited include:

  • Innovation - The cloud offers a wider array of new services and capabilities that empower evolution.
  • Data center closures - Data centers are a major investment, and less appealing in a world where ‘on-demand’ is gaining traction.
  • Global markets - Covering global markets and opening new markets on the backs of in-house data centers is prohibitively costly…if it’s even possible at all.
  • Resource refocusing - The time and money businesses devote to in-house data centers is immense, and there are more exciting, growth-focused areas where that could be spent.

Legacy databases lagging behind PostgreSQL

These reasons also play a big role in answering why these same companies are simultaneously rejecting legacy databases. Again, agility is one of the key factors, with price right new to it. Not only are legacy databases costly, but they’re restrictive, hamstringing you via contracts, and limiting your deployment and integration options. Especially as more and more organizations want to focus on consolidation, the limits of legacy databases are looking less and less tolerable, as a way forward—especially as open source solutions such as PostgreSQL prove their worth across a wider variety of workloads.

And it’s not just EDB that feels this way. In his presentation, Marc Linster cites three major industry studies from StackOverFlow, Datadog, and CNCF, each of which ranks PostgreSQL more highly than the majority of its competitors, highlighting how dynamic, flexible and agile it is.

 

Choosing your ideal destination

Now that we understand why so many enterprises are moving their databases to the cloud and why they’re trusting PostgreSQL specifically, we need to ask how we’re going to bring PostgreSQL to the cloud.

We have three main options.

1. Virtual machines on IaaS

As it stands, using virtual machines requires the most internal management of the three options, demanding both an infrastructure architect and an infrastructure DBA (database administrator). While for some, this sounds daunting, the VM approach is well-regarded by those who want the most control possible over their architecture, as this leaves infrastructure essentially entirely in your hands.

Another benefit of this approach is that your architecture does not change from a traditional on-prem architecture, which makes it easier to migrate applications and make use of critical solutions without much of a learning curve.

2. Containers and Kubernetes

While Kubernetes still requires your business to have an infrastructure architect, the need for a DBA is greatly reduced, because your Kubernetes system is managed by an operator, freeing you from much of the responsibility that comes with VMs. This does limit architectural flexibility slightly, which can be frustrating for those who abandoned a legacy database specifically because of the restrictions on their business. Additionally, in order to maintain your architecture in its on-prem state, you’ll need to invest in a microservice transformation, which can be a slight headache.

3. Database-as-a-service (DBaaS)

Finally, you have the DBaaS approach, exemplified by EDB’s BigAnimal. In this case, you need neither an infrastructure architect nor a DBA, as all infrastructure is the responsibility of the cloud services provider (CSP). While this sacrifices some flexibility, the DBaaS strategy boasts much more elasticity and on-demand capabilities than either other option. Additionally, any changes to your architecture from its on-prem form are minimal, and you aren’t tasked with ensuring architectural continuity.

For businesses who are less concerned about being able to tailor every aspect of their cloud architecture, the DBaaS approach is incredibly popular, and also makes it much easier to channel resources that might have been spent on infrastructure management to innovation and consolidation.

You can also find a more in-depth comparison from Marc’s presentation!

 

Kubernetes vs. DBaaS

For many of our customers who are looking for a true cloud experience, the decision ultimately comes down to Kubernetes or DBaaS. While the virtual machines on IaaS approach resembles a cloud, and might be cloud-adjacent, it ultimately doesn’t meet many of the most conventionally accepted criteria for what actually defines a cloud. For instance, things like provisioning scripts or infrastructure as code can serve similar purposes to but do not actually constitute a control plane, which is often considered a vital factor for being a cloud solution. Of course, VMs work very well for many organizations; but, for those who are looking to the cloud, it just simply doesn’t earn the designation.

Of the two remaining approaches, DBaaS is closer to the mainstream definition of a cloud solution, bolstered by its self-service structure, its measured options, its on-demand capabilities, and its elasticity bona fides. Still, Kubernetes is getting much closer to meeting the total definition of cloud, and its advanced operators will provide your business with a powerful foundation for operating a cloud PostgreSQL database. In fact, EDB’s own BigAnimal is built atop Kubernetes!

 

Discovering your ideal cloud future

If we’ve learned anything from helping enterprises move and operate their databases and applications in the cloud, it’s that everyone’s needs and challenges are different, but the impact of achieving success can be universally transformative.

To learn about how EDB’s BigAnimal is helping businesses to best move their databases to the cloud, visit us here; and if you want to watch Marc Linster’s keynote in full, you can check out his and all PostgresBuild on-demand sessions here!