My data killed my cloud project!

Data transfers seem to be the easiest part of cloud migration. After all, migrating applications is the biggest pain in the neck. Data replication and migration should be simple, something that’s done during the last step of the application and data migration process. Right?

Many people in IT will sit in a big circle and tell you that data is their most valuable business asset. You would never know it, based on the current state of their data. They have no single source of truth, and replication is the most common way they solve data problems. Also, the data is not properly cataloged, data integration is often missing or just adds more complexity, and database administration and security are lackluster. 

The problem now? They want to move all this jumbled data to the cloud.

News flash: The cloud fixes nothing. It’s simply another platform that will host your existing data problems. It might even make things worse, considering the ease of allocating storage and databases with a click of a mouse to drive quick fixes. Now we can do dumb things faster and cheaper in the cloud. 

Here are the core opportunities to avoid these problems:

Make data a first-class citizen. Have you ever thought, “If it works, then it does not need attention”? We often overlook poorly designed and maintained databases because they can still store and retrieve data; therefore they are doing their job and do not need to be an IT priority. 

To compound the problem, there is no immediate downside to failing to prioritize data. Data-related jobs often go on the block first if any downsizing occurs. (Remember those days?) IT leaders don’t seem to push back as much as they should, considering that data is hidden by applications and tools. 

The poor state of enterprise data is often seen as something to address at some undefined future date. Today is that day.

Fix data as it moves to the cloud. Again, the cloud won’t fix your data. A data mess on premises will become a data mess in the cloud. The best time to find and fix problems is before you relocate data to the cloud since you’ll already disrupt the use of the data during the migration. 

Those who skip this step often run into issues when they attempt to migrate and link applications in the cloud to the data. You would think that if it worked on premises that it would work in the cloud, but that’s just not true. Most find that at least some data problems need to be fixed before the applications will even function properly. This is certainly true if the applications are being modernized, such as being moved to containers or serverless. 

This problem is super easy to fix. Just pay attention to the data that has already migrated to the cloud, the data that will migrate to the cloud, and perhaps the data that will never migrate to the cloud. Make it an early priority to get applications that use data working properly in the cloud before all the data gets transferred.

If you fail to address data problems, you’ll end up with another set of patches and quick fixes that will cost more money in higher cloud bills than it would to just take the time and tools to fix the problems once and for all. 

It’s time to assign data its first-class seat at the table. Fix existing data problems before they carry over to become cloud data problems. Don’t let unaddressed data issues kill your cloud migration project.