ra2 studio - Fotolia

Google brings wizard approach to cloud database migration

Google Cloud's top database executive details the tech giant's new tool for enterprises to move databases to the cloud, as the need has accelerated in the coronavirus pandemic era.

Google said on Thursday that its new Database Migration Service is available in preview for users to migrate MySQL, PostgreSQL and SQL Server databases to Google's Cloud SQL platform.

The new service is the tech giant's response to the quickly accelerating move organizations of all sizes are making to the cloud to deploy and run database workloads.

Yet not all databases are born in the cloud, with many databases still running on premises. Cloud database migration is a big issue for cloud providers including Google, which operates a growing portfolio of cloud database services.

Andi Gutmans, general manager and vice president of engineering and databases at Google Cloud, leads Google's database cloud efforts. Gutmans joined Google in May 2020, after working at Google cloud rival AWS as vice president of analytics. Gutmans is also well known in the IT industry as the co-founder and CEO of PHP (hypertext preprocessor) vendor Zend, which was acquired by Rogue Wave Software in 2015.

In this Q&A, Gutmans discusses the challenges and opportunities of cloud database migration.

Why are you introducing the Database Migration Service now?

Andi GutmansAndi Gutmans

Andi Gutmans: Just as a bit of context, Gartner has estimated that by 2023, 75% of databases will be running in the cloud. That's a pretty interesting number given that we've had like 40 years of putting databases on premises. So I think that's really an indication that we've gotten to this tipping point of enterprises really mass migrating into the cloud.

Traditional kinds of cloud database migration services have been built on change data capture and aren't using the native database replication capabilities. So there were all these little nuances that could go wrong for customers, like schema changes that wouldn't get replicated in the right way. So we decided that we will take a different approach, and basically build on what the best practices are for migrations, which is using database native replication.

So we built a serverless experience. Customers don't have to provision a server; they don't have to pay for the compute. It's a wizard-like approach where it really takes you through the setup, how you connect to the source database, how you connect to the destination and it makes the whole journey easy.

Traditional kinds of cloud database migration services have been built on change data capture and aren't using the native database replication capabilities. So there were all these little nuances that could go wrong for customers, like schema changes that wouldn't get replicated in the right way.
Andi GutmansGeneral Manager and vice president of engineering and databases, Google Cloud

What do you see as barriers to adoption for organizations moving to a cloud database?

Gutmans: The challenges customers have, obviously we're trying to solve those for them. One of them we talked about is the migration. Sometimes there are also regulatory requirements and data sovereignty requirements. There may also be cultural elements in an organization where the whole DevOps culture and the skill set may not be quite there.

We put a huge focus on helping to make the journey easier, either directly or with our partners, to make sure that they have the right landing spot for different databases. Data sovereignty and regulatory compliances has been a huge focus of ours, not just from a security perspective, but also making sure that things like access control and all the required capabilities are there for customers.

So, I think, one of the reasons why Gartner has forecast that 75% of databases will be in the cloud is because a lot of these adoption blockers have basically been eliminated over the past few years.

Cloud SQL now supports PostgreSQL 13. How do you and Google see PostgreSQL fitting into the cloud database landscape?

Gutmans: In the late '90s, when I was doing a lot of the early work on PHP, PostgreSQL was out there but it was MySQL that really took off on the web. For many years I would say that PostgreSQL was a bit more of an academic kind of database and MySQL was more mass market.

In recent years, we have seen a dramatic change in the market and PostgreSQL adoption has exploded. A lot of customers are looking at PostgreSQL as an open alternative to enterprise database capabilities. PostgreSQL has a lot of those enterprise capabilities that customers coming from Oracle are looking for.

So PostgreSQL for us is a very important engine and we are super committed to supporting it well, and innovating on PostgreSQL. We want to make sure that we're actually the most current cloud when it comes to PostgreSQL and making sure we deliver these version updates as quickly as possible.

What has been the impact of the COVID-19 pandemic on cloud database adoption?

Gutmans: We're definitely seeing customers that want to accelerate their journeys. The pandemic is an aspect that has really accelerated the need to transition to the cloud. So we are seeing that the clarity and sense of urgency for cloud migration at enterprises has gone up significantly.

In the past, some of the conversations were lengthier conversations about why to migrate, total cost of ownership, security and so on. Most of the conversations now are about how and how fast they can migrate.

Editor's note: This interview was edited for clarity and conciseness.

Next Steps

Google Datastream advances change data capture in the cloud

Dig Deeper on Database management