Problem
An international e-commerce optimization company needed additional storage to accommodate its growing amount of data. When they attempted to purchase more storage capacity, they were faced with licensure issues due to hosting key functionalities on two separate infrastructures – Oracle and SQL Server. Solvaria worked with the business and IT to perform an analysis of the current space requirements and future growth expectations. It was determined that the best solution for the business was to utilize the scalability and flexibility of the cloud.
Solution
Solvaria worked with the customer to develop a strategy to migrate from an Oracle database backend to Azure SQL Database. Elastic Pools were employed where possible to limit DTU costs. Solvaria used a combination of SQL Server Data Migration Assistant (DMA), custom scripts and SSIS packages to perform the migration. DMA was used to migrate the schema, functions, views, and stored procedures. SSIS and custom scripts were used to migrate the data. Index compression was enabled for large tables to conserve space and limit DTU consumption.
The company’s Enterprise Data Warehouse was an on-premise SQL Server. Solvaria migrated these components to Azure SQL Server VM. Native SQL Server backups were copied from on-prem file shares to Azure blob container and copied to a local disk on the Azure SQL VM from which they were restored to SQL Server. This is the quickest approach for restoring large databases to an Azure SQL VM.
Solvaria set up monitoring for the database environments and built Azure Monitor Dashboards to track and display infrastructure performance trends. Azure Monitor Alerts were created to notify DBAs and server admins of key performance conditions. Automation runbooks were coded and scheduled to perform checks for data-specific conditions and perform maintenance tasks (data purging, index maintenance). Auditing was enabled on the Logical SQL Server to capture user activity in certain environments.
Results
This effort merged the company’s databases to a single Relational Database platform and helped to greatly reduce license cost and increase the business flexibility with storage options.