Automatically Switch to a Google database platform during a migration

This week we take a look at a combination of Infrastructure and Application modernization. Google has already had the necessary tooling to migrate your existing Microsoft SQL environment, your Microsoft database, to a supported Microsoft SQL environment in the Google Cloud. Today’s article focuses on the next step. How to automatically switch to a Google database platform during a migration?

Which of course will lead to the question. “How do those Google databases compare to the competition in the market?” In a second part of this post, we briefly explain how Gartner estimates the Google position.

Another puzzle piece in the Iaas story

Moving your “on prem data center” to the public cloud, it remains quite a challenge that will not take you overnight. What you expect from a company like Google is that they provide you with the necessary tooling to make such an operation as correct and safe as possible. For example, with Stratozone, Google already had a high-performance solution to analyze your existing environment, track down the different data flows and map out a migration path to identical databases in the GCP environment, but Google’s own solutions for a database migration (to run fully managed environments such as Google Cloud SQL) were not there. For this you had to fall back on solutions from third parties.

But that’s about to change with the Database Migration Service (DMS). Google wants to achieve four goals with DMS:

  • The migration should be easy to perform. In the first place, that means not a lot of research, but a full script, a real “wizard” that guides you through the entire process. Preparing the databases for replication, configuring secure connectivity to the source database, and validating migration settings is embedded in DMS, making installation simple, fast and repeatable;
  • The migration must guarantee minimal unavailability. Each migration with DMS allows data to continuously replicate from the source database to the destination without the need for cumbersome manual steps, minimizing database downtime and enabling rapid application switchover;
  • The migration must be guaranteed to produce a complete and faithful result. The migration must be reliable enough that the target database works flawlessly after the migration without any additional action. In a “like-to-like” migration between compatible source and target database environments, DMS is unique because it uses the database’s native replication capabilities to maximize process reliability;
  • The migration takes place without server installations and in a secure manner. DMS is available by default – so you don’t have to set up or configure anything extra – and the environment is perfectly scalable by itself. Your focus can therefore be entirely on the actual migration. If sensitive information is migrated, DMS supports multiple secure connection options so that your data remains protected during the migration process;

How does the DMS enviromment work in practise?

DMS thus offers completely faithful database migrations with minimal operational disruption for MySQL and PostgreSQL environment (Microsoft SQL is coming).

Google designed the platform completely cloud-native, built by and for the cloud. DMS uses the logs to send data from the source to the target database with very limited delay. Of course, it first starts from a snapshot of the data to constantly fill the target database and then once the base has been transferred based on the logs to keep the target database continuously up-to-date.

The source and destination are continuously up-to-date because they rely on the databases’ own native replication capabilities. Once all the tests have been performed and you are ready to make the transition to the new environment – to have your applications point to the Google Cloud SQL as the primary database – that switch can be realized with minimal downtime.

As already mentioned, DMS is serverless. The customer never has to worry about setting up, managing or monitoring migration-specific sources. The data, schema, as well as additional database functions from the source database (such as triggers or stored procedures) are reliably and scaled to the Cloud SQL destination replicated without any user intervention.

How to get started with DMS?

DMS, now in preview, supports migrations of self-managed MySQL databases, both on-premise and in the cloud, as well as migrations from managed MySQL databases at another public cloud player to Cloud SQL on GCP. PostgreSQL support is currently available on demand for a limited range of customers. This limited availability is currently also available for SQL Server and you can also use the previous document for this.

The DMS migration environment is available without additional costs.

More info can be found in the blog “Best practices for homogeneous database migrations” and in the available DMS documentation where you can find a series of documents to help you on your migration journey.

You can find the announcement of the DMS release itself here.
A more technical approach to the entire execution story can be found here.

And how does DMS fit into the broader cloud context?

Note that the IT consultancy Gartner predicts that by 2023, 75% of all databases will run in the cloud. And that is a very high number, especially when you look at the Belgian market. In any case, this means that we have reached a pivotal moment and the coming years will be spectacular in terms of migration.

It goes without saying that only well-oiled machines can cope with such a rhythm. The steps Google has now taken should make it possible for partners to smoothly process the influx of migrations. And the bonus is that this doesn’t just allow migrations from on prem to the cloud. But also from a competitive public cloud to GCP.

Offering an automated standard approach must above all also allow for proper focus in the digital transformation process. It now comes down to every company to evolve towards a digital offering and therefore also digital technology. To step in the direction of DevOps or even better in the direction of Site Reliability Engineering. To be able to focus mainly on an App Modernization process from now on. And that will become much easier if we get the technical aspects of the migration fully automated.

Gartner names Google a leader in Magic Quadrant for Cloud Database Management Systems

Gartner published a Magic Quadrant for Cloud Database Management Systems for the first time in 2020. And from day one, Google has been in a leading position here. Among the best three in terms of vision. And also among the best three in terms of the possibility of execution. Google owes Google’s strong position to its focus on hybrid and especially multi-cloud approach, to its flexible prices and to its broad integration with third-party solutions.

For Google, a database no longer stands alone, but is the basis of an extensive ecosystem for database service and, above all, data analysis. So that you as a customer can get the full value from your data. It makes no sense for them to evaluate analytical and operational use cases separately. And you can now see that reasoning at Gartner as they based their evaluation on the complete implementation.

From Google’s point of view, this announcement reinforces the first story. Gartner hereby confirms that Google’s database approach is visionary and that their offer is sufficiently strong and broad to support you as a company in all areas of data processing and analysis.

So you can migrate to a Google Cloud Server with full confidence. DMS can help you perfectly with this.

You can request the Gartner report here.

The Google blog is here.

Bart Gouweloose

Google evangelist

Ready to take the next steps in your Google Database migration?

Contact us, our experts are happy to help!

, , ,

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *

Fill out this field
Fill out this field
Please enter a valid email address.

This site uses Akismet to reduce spam. Learn how your comment data is processed.