Why is migrating a database to the cloud like a heart transplant?
Posted: Sat Feb 01, 2025 8:52 am
Migrating your organization’s on-premises database to a cloud database is fraught with challenges and potential benefits. Here are recommendations to ensure this ‘heart transplant’ doesn’t fail.
In recent years, many organizations have increased their cloud footprint and sought to modernize their IT infrastructure, accelerated by the catalyst of COVID impacts. But migrating databases from an on-premises deployment to a cloud-native deployment can present several challenges that can prevent organizations from pulling the trigger on this shift. However, the current era of cloud and infrastructure modernization is increasing the pressure to complete these migrations now more than ever.
If your organization is preparing for a project of this type, there are many factors to consider and a lot of planning required.
“Migrating a database is like getting a heart transplant,” said Ramke Ramakrishnan, senior director analyst at Gartner, during his keynote at the Gartner Data and Analytics Summit in Orlando this week. Databases are connected to many applications and are likely running the lifeblood of the organization through them every day.
The problem with database migration
To the uninitiated, a database migration project may seem instagram data straightforward. After all, you’re just migrating database objects from one database to another. Vendors provide migration tools to let you complete the job with just a few clicks.
But when you start doing your due diligence on the project, you’ll see what it really involves, according to Ramakrishnan. One of the biggest changes will be with the SQL semantics that are so familiar to users of traditional databases like Oracle or SQL. There are triggers, packages, and complex procedures.
“When you look at similar migrations in the cloud, many of the databases don’t support the robust SQL packages that you might have,” Ramakrishnan said. Instead, you might have to use Javascript or Spark-based processes.
"It's really like a heart transplant because you have to look at every piece," he said.
Another reason why you may not want to follow a “lift and shift” strategy for database migration from on-premises to the cloud is because the cloud provides data storage alternatives that you may not have available in your traditional database.
“You tried to solve everything in one database,” he said. But in the cloud, you can spread data across multiple data stores, like object storage or relational storage or NoSQL storage or graph storage.
Also, you may not want to migrate everything, Ramakrishnan said, noting that the process of migrating and modernizing your database can be an opportunity to do something like the Marie Kondo method of tidying up, getting rid of all the objects that don’t “spark joy.” You may not need to migrate all your data. The migration can be an opportunity to start over.
In recent years, many organizations have increased their cloud footprint and sought to modernize their IT infrastructure, accelerated by the catalyst of COVID impacts. But migrating databases from an on-premises deployment to a cloud-native deployment can present several challenges that can prevent organizations from pulling the trigger on this shift. However, the current era of cloud and infrastructure modernization is increasing the pressure to complete these migrations now more than ever.
If your organization is preparing for a project of this type, there are many factors to consider and a lot of planning required.
“Migrating a database is like getting a heart transplant,” said Ramke Ramakrishnan, senior director analyst at Gartner, during his keynote at the Gartner Data and Analytics Summit in Orlando this week. Databases are connected to many applications and are likely running the lifeblood of the organization through them every day.
The problem with database migration
To the uninitiated, a database migration project may seem instagram data straightforward. After all, you’re just migrating database objects from one database to another. Vendors provide migration tools to let you complete the job with just a few clicks.
But when you start doing your due diligence on the project, you’ll see what it really involves, according to Ramakrishnan. One of the biggest changes will be with the SQL semantics that are so familiar to users of traditional databases like Oracle or SQL. There are triggers, packages, and complex procedures.
“When you look at similar migrations in the cloud, many of the databases don’t support the robust SQL packages that you might have,” Ramakrishnan said. Instead, you might have to use Javascript or Spark-based processes.
"It's really like a heart transplant because you have to look at every piece," he said.
Another reason why you may not want to follow a “lift and shift” strategy for database migration from on-premises to the cloud is because the cloud provides data storage alternatives that you may not have available in your traditional database.
“You tried to solve everything in one database,” he said. But in the cloud, you can spread data across multiple data stores, like object storage or relational storage or NoSQL storage or graph storage.
Also, you may not want to migrate everything, Ramakrishnan said, noting that the process of migrating and modernizing your database can be an opportunity to do something like the Marie Kondo method of tidying up, getting rid of all the objects that don’t “spark joy.” You may not need to migrate all your data. The migration can be an opportunity to start over.