A new report by Gartner predicts that by 2022, 75% of databases will be deployed in or migrated to the cloud. Moving on-prem databases to the cloud offers organizations flexibility and scalability, faster infrastructure configuration, consumption-based pricing, and access to a wider variety of database management systems. However, moving databases can be the trickiest part of a cloud migration, often requiring downtime, reworking of data schemas, and refactoring of applications.
Before the actual database can be moved, all of the data it contains must be migrated. Data migration is an extremely high-risk undertaking. According to Gartner, 83% of data migrations exceed budgets and timelines or fail altogether. However, with this great risk comes great rewards. Most large enterprises perform their cloud migrations iteratively. A properly orchestrated data migration minimizes the risks of moving databases and applications later on, vastly increases application performance, improves query efficiency and response time, streamlines deployment processes, hardens security, and enhances data governance, putting enterprises in a better position to comply with the GDPR, the California Consumer Privacy Act, and other data privacy legislation.
Here’s a handy checklist to ensure that your migration process goes smoothly:
1. Assess your current data environment
Before doing anything else, you need to understand what you’re working with, including any challenges or risks. Understanding obstacles will enable you to thoroughly work through them during your proof of concept. Size databases so that you know how much storage you’ll need after migration; assess applications to ensure they are cloud-compatible; determine which databases and apps are mission-critical; and determine which need to stay on-prem for compliance reasons or because it’s not feasible to move them to the cloud.
2. Assess the skills & expertise of your internal team
Determine the areas in which your team members need upskilling so that you can allow for this in your migration timetable and ensure that your team can make the most of your new cloud platform after migration. If you will be using a different database in the cloud, your team members will likely need upskilling in that area as well.
It’s also important to assess your team’s current workloads to ensure that they do not get overscheduled during the migration process, which could result in internal projects being delayed or derailed.
3. Select the right migration tools
There are a variety of tools to help new customers with database migration. For example, Google Cloud Platform (GCP) recommends Alooma, an ETL tool that simplifies database migration by automating one of the most tedious tasks in the process: transforming and normalizing data. Alooma also removes PII (Personally Identifiable Information) and detects errors while moving data to the target database. When evaluating and testing migration tools, be sure to determine which best fits your team’s skills and needs as well as your particular use case.
4. Deploy a proof of concept
Deploying a proof of concept allows you to work through the challenges and risks you identified when assessing your data environment, brings unforeseen issues to the surface, and provides assurance that your migration will be successful. For best results, don’t create an irrelevant “test” app; use a real application that would be a good fit to explore the flexibility and scalability of the cloud. That said, don’t pick an extremely difficult use case or one that involves complex security and compliance issues.
5. Review your data and reclassify it if necessary
It is imperative that your organization performs a review of all data prior to migration, reclassifying it if necessary. This maintains the integrity of existing security and compliance controls, as well as any legal requirements to hold, retain, or keep certain records on-prem. The review and reclassification process will involve multiple business units, including compliance and legal teams. After it is completed, you’ll need to update your IT governance policies to reflect any changes.
6. Cleanse your data
Migrating your databases to the cloud is a great opportunity to improve your enterprise data quality and utility by ensuring that you don’t transfer records that are old, duplicative, incomplete, or corrupted. Data cleansing is an extremely involved, error-prone process, but since it’s unlikely the data will be cleansed again for some time after the migration, it’s important to ensure that you’re moving accurate data. Enterprise ETL solutions such as Alooma automate the tedious work of data cleansing, ensuring accuracy and saving countless hours.
7. Rebuild your database schema
Before data migration commences, your team will need to convert your current data schema to one that works with your new database. This will involve creating tables and writing scripts for functions such as stored procedures and database triggers.
8. Migrate, validate, and repair your data
Depending on the size of your database, the initial data load could take days to complete. During this process, the data will be transformed, normalized, and checked for errors. Once the data is loaded, it must be checked again to ensure completeness, validity, and accuracy. Any issues must be addressed before migrating the actual database.
9. Migrate your access privileges & security settings
As evidenced by the epidemic of cloud breaches caused by misconfigurations, cybersecurity is too often an afterthought during the migration process. While your cloud provider is responsible for the security of its cloud, your company is responsible for anything it puts in it, including enterprise databases and their contents. Make sure that you duplicate all of the access privileges and security settings you had in place on-prem in your new cloud deployment.
Cloud migration is a marathon, not a sprint. Depending on your organization’s size and data environment, the entire process can take anywhere from several months to a year. The important thing is to get your databases moved efficiently, securely, and accurately, while minimizing disruption to your business and your clients.
