Three Reasons Your Cloud Database Migration Will Fail
Moving relational data to cloud platforms has emerged as a “last frontier” for cloud-age computing. It is easy to consider the behemoth systems in place at large enterprises against the agile cloud-borne systems developed by startups and growth-stage companies—concluding that the enterprise systems must become more like their nimbler cousins in order to compete. It's also easy to overlook some common pitfalls of migrating data to the cloud that could compromise the success of the migration.
1. Unclear on What Success Means
There are several valid reasons to migrate data to the cloud, but I often speak to data architects who are frustrated by unclear reasoning. The story I get usually describes a directive put forth by someone with three letters for a job title along with a generally unreasonable target date and not much else.
Above all, the migration team must know why the migration is happening, and what the measurements of success are. With so many options for delivering data to and from the cloud, it is almost certain to come out wrong without a defined vision.
Let’s say it was determined that cost savings is a primary goal. This is a definite possibility, but if it is not stated clearly in the project charter, it is not likely to happen. The paths of least resistance for moving relational data to the cloud will generally not provide long-term cost savings.
If your story sounds a lot like those data architect stories I mentioned above, it is your responsibility to seek clarity. Go to whomever you need to go to and insist on a clear definition of success. This is not to say that you should argue against it. I’m not advising that you push back against the project. I am advising, instead, that you request a clear understanding of what it means to win for this project.
2. Unclear on Starting Point
If on-prem systems are performing generally well, you must know what that looks like before you migrate. It is important to document as much as you can about the makeup of the systems and their performance patterns before you migrate.
The system is made up of application and database, but what about those ETL processes populating a data warehouse? What about the job that automates telling finance when there’s a customer billing address change? What about the quarterly report that tells sales reps about support cases for their clients? Documenting database schema and lineage across the entire solution will help ensure you have captured all the pieces of the puzzle and determine how they should be assembled on the other side. SentryOne Document is designed to make this easy.
Inconsistent performance is costly. The larger the system and the more people using it, the more expensive slowing it down becomes. This is a hidden cost for cloud migration, and it is completely avoidable by fully understanding the performance footprint for the system on-prem and striving to replicate or improve on that same footprint after migration. This is done by monitoring and measuring performance before and after migration, and regularly iterating to meet or exceed your starting point. SentryOne performance monitoring is perfect for this.
3. Fumbling the Follow-Up
The migration is done, and all is right with the world. Great job! I’m sorry, but you’re not finished. You will continue to monitor, diagnose, and optimize this shiny new cloud system, but how do you measure against your goals?
Now that your system has a new home, it is important to go back and consistently document and measure again, comparing your results against what you collected prior to migration. This is how to tell whether the migration was successful.
If you used SentryOne to help prior to migration, you are in luck, because you already have what you need to follow up thoroughly. The same solutions you used on-prem will translate well to the new cloud environment. SentryOne CloudLifter combines a methodology for cloud data migration and a succinct bundle of software products to make each step easier.
Change is hard, and migrating your databases to the cloud is a big change. Getting clear on the measurements for success, understanding where you’re starting from, and vigorously following up after migration will help to secure your place in the cloud migration hall of fame.
Jason has worked in technology for over 20 years. He joined SentryOne in 2006 having held positions in network administration, database administration, and software engineering. During his tenure at SentryOne, Jason has served as senior software developer and founded both Client Services and Product Management. His diverse background with relevant technologies made him the perfect choice to build out both of these functions. As SentryOne experienced explosive growth, Jason returned to lead SentryOne Client Services, where he ensures that SentryOne customers receive the best possible end to end experience in the ever-changing world of database performance and productivity.